TI'I'IIIII'IIIIIIIIIIIII' E: r - .

INN/Mi{Ill/I/U/H/WIIIHIlf'llllilllU/Il/N/i

3 12930

 

 

LIBRARY
Michigan State
University

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

MAR 0 7 2302

 

112002

 

 

 

 

 

 

 

 

 

 

 

 

woo dam-peep.“

 

LEVERAGE, POWER AND IMF CONDITIONALITY:
A STRATEGIC BARGAINING MODEL OF CONDITIONALITY AGREEMENTS
BETWEEN THE IMF AND DEVELOPING COUNTRIES
By

SEONJOU KANG

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Political Science

2000

ABSTRACT
LEVERAGE, POWER, AND IMF CONDITIONALITY: A STRATEGIC
BARGAINING MODEL OF CONDITIONALITY AGREEMENTS
BETWEEN THE IMF AND DEVELOPING COUNTRIES
By

Seonjou Kang

This dissertation analyzes the dynamics at work in negotiations for structural
adjustment lending to developing countries from multilateral donor agencies such as the
International Monetary Fund (IMF), dynamics which can account for the difference in
conditionality agreements. The disappointing results associated with structural
adjustment are more often than not identiﬁed with the onerous inﬂexibility of conditions
imposed on developing countries by the IMF, with little consideration for differences in
types of economies and level of development. I, however, ﬁnd that the terms of policy
reform varied signiﬁcantly in content, scope, and timing country by country.

In order to explain such variance, conditionality agreements are modeled as
outcomes of strategic bargaining between the IMF and developing countries, whereby the
relative bargaining strength of each player determines the nature and variance in
outcomes. The relative bargaining strength denotes leverage within a Speciﬁc issue area,
as distinguished from the conventional notion of aggregate structural power in
international relations. Strategic interaction between the two players was constructed with
a game model, which illustrates that bargaining leverage is a necessary condition for the

differences in conditionality agreements.

 

Subsequent statistical analyses concentrated on measuring the effect of leverage
relevant to the issue area of conditionality bargaining, using the number of agreed
conditions extracted from IMF Survey as dependent variable, and supported the
theoretical results of the game model. A case study of conditionality bargaining between
the Republic of Korea and the IMF also Showed that some exceptional conditions in its
conditionality agreement, such as radical liberalization of ﬁnancial market, were by and
large a result of failure to mobilize effective leverage to the negotiation with the IMF.

The above ﬁndings suggest that there are major distinctions in the instruments,
issues, and interests that developing countries could bring to their relationship with the
IMF. The ﬁndings also suggest that the sophisticated and symbiotic interplay of
interdependence between the IMF and recipient countries was not properly appraised in
past scholarly works. The results of structural adjustment seem to be inﬂuenced by the
choice of conditions through bargaining as much as they are by domestic implementation.
Studies of structural adjustment Should be reorganized so as to incorporate the
consequences of politics at the negotiation stage. That is, without a clear idea of how the
terms of conditionality were drawn and what the source of the variance in conditionality
terms was, a diagnosis of where the conditionality implementation went awry is not
reliable, and much less is a search for cure for that malady. Conditionality agreements
have been plagued by political rationality from the beginning. Therefore, what is needed
for better results from structural adjustment is an improvement in politics at the
bargaining stage in order to bring their interests gradually closer together under the

shadow of ﬁIture.

 

© Copyright by
Seonjou Kang
2000

To All my family who kept faith and hope in me.

ACKNOWLEDGEMENT

It is exactly twenty years since I ﬁrst dreamt of being a professor in political
science. The past twenty years have been a series of seeking, planning, and executing
endeavors in order to hold on to my dream, and the last seven years of them were the
ﬁnal dash to the ﬁnish line. Now that dream has become reality.

At that time I did not know how difﬁcult it would be to obtain a PhD. and what it
would take. Although this Ph.D. is to ﬁtlﬁll a dream and challenge of my choice, many
people and institutes contributed and never hesitated to help and support me as I worked
on this lifetime project. My PhD. is literally a work of collaboration and cooperation for
a personal glory.

I ﬁnish my doctorate program with a dissertation topic I love to study. I cannot
leave Michigan State University without expressing my genuine thanks to dissertation
committee, Dr. Scott Gates, Dr. Dennis Patterson, Dr. Jeffrey Riedinger, and Dr. Nicolas
van de Walle. All of them are established scholars in their specialties and maintain as
high standard as possible over academic matters. I do not believe that I met all their
demanding expectations in my dissertation, but its current quality is a glimpse of their
outstanding guidance and my tenacity to keep up with it.

Especially Dr. Gates, my advisor, deserves special thanks from me. He gave me
not only advice on dissertation substance but also on scholastic attitude. He was always

enthusiastic about guiding me and I received lots of advice that enabled me to carry on.
Furthermore, he played the most important role in making dissertation writing a pleasant

and positive experience. Now he is my role model in academe and will remain a mentor-

vi

colleague to me. I sincerely hope that I will be given an opportunity to replicate my
experience with younger graduate students in the ﬁIture.

During the last year of my dissertation writing, Hudson Institute in Indianapolis
gave me generous support through the Herman Kahn Fellowship. Although I partially
neglected my duty and responsibility there due to frequent trip to East Lansing, they
neither mentioned it nor stopped encouraging me. I am glad that I became the ﬁrst to
ﬁnish a degree during the fellowship. I am thankful to all the Hudson members.

All of my family is at the center of my heartfelt appreciation. They are the ones
who kept faith and hope in me and with whom I love to share the joy of this completion.
Among them, I am very grateful to my eldest sister, Sun-Sook. As the oldest of the four
children, she had a signiﬁcant inﬂuence on me while I was growing up and did not spare
emotional and material support during my PhD. program. I have nephews who will soon
be adolescents and my Ph.D. teaches them, I hope, that this is a way of life worth
pursuing in the future and, more importantly, that dreams will never leave us unless we
give them up ﬁrst.

I have many friends to thank for their beautiful friendship. Especially Soo-Young
Kang, Barbara Czerny and Lucinda Deason, three close friends I made at MSU. They
were understanding and supportive as allies working on their own Ph.D.’s in human
nutrition, computer science and political science, respectively. Young-Soon Kim
extended her friendship from our high school days to life of thirty something in the
United States. And for the last, but not the least, I would like to remember Sameh Kamel
for his invisible companionship and invaluable motivation for many years. My thanks are

and will always be with him.

vii

I am closing one chapter of my life and about to start a new one. This new chapter
will essentially comprise returning all of the above. The best way I can keep my solemn
promise is to be a successful scholar and a more mature individual, who they can count
on whenever they need help and support in order to carry on their dreams and meet

challenges.

viii

TABLE OF CONTENTS

LIST OF TABLES ........................................................................................................ xii
LIST OF FIGURES ...................................................................................................... xiv
CHAPTER 1
CONDITIONALITY: A PANDORA’S-BOX? .............................................................. .1
1.1 Puzzles of Conditionality ............................................................................... 1
1.2 Theoretical Overview of the Subject ............................................................ 11
1.3 Research Strategy: How to Solve Puzzles ..................................................... 31
1.4 Expected Achievements: What We Will Learn from This Study ................... 35
CHAPTER 2
PRELUDE: THEORY OF BARGAININ G
IN INTERNATIONAL RELATIONS ........................................................................... 39
2.1 Bargaining AS a Way of Resolving Differences between State ..................... 39
2.1.1 Game Theory and Modeling International Negotiations ................. 42
2.2 Factors Affecting Bargaining Outcomes ....................................................... 51
2.2.1 Bargaining Tactics:
Commitment, Linkage, and Side Payment ..................................... 54
2.2.2 Resources: National Power in International Negotiation ................ 59
2.3 Asymmetry in Resources and International Negotiation ............................... 66
CHAPTER 3
MODELING CONDITIONALITY BARGAIN ING GAME .......................................... 74
3.1 Developing a New Concept of Leverage ...................................................... 74
3.2 A Model of Conditionality Bargaining Game ............................................... 82
3.2.1 Strategic Interaction between Donor and Recipient
in Conditionality Bargaining .......................................................... 82
3.2.2 Assumptions of Conditionality Bargaining ..................................... 86
3.3 Expected Utilities and Preferences in Conditionality Bargaining ................ 103
3.3.1 Expected Utilities of Outcomes .................................................... 103

ix

3.3.2 Preference Orderings and Restrictions

in Conditionality Bargaining ........................................................ 106
3.4 Equilibrium of Conditionality Bargaining with Perfect Information ............ 1 l6
3 .5 Theoretical Propositions of Conditionality Bargaining
and Hypothesis ........................................................................................... 121
CHAPTER 4
INTERMEZZO: BRIDGING ABSTRACT TO CONCRETE ...................................... 133
4.1 Data for the Analysis of Conditionality Bargaining .................................... 133
4.1.1 Dependent Van'able: TITNESS .................................................... 134
4.1.2 Expected Utility Components ...................................................... 142
4.1.3 Independent Variable: Leverage ................................................... 150
4.1.3.1 Operationalization and Measurement
of Leverage for the IMF ................................................ 151
4.1.3.2 Operationalization and Measurement
of Leverage for the Recipient ........................................ 160
4.2 Statistical Method: Poisson Model for Count Data ..................................... 170
CHAPTER 5
EMPIRICAL TESTS OF CONDITIONALITY BARGAINING MODEL ................... 175
5.1 Controversies on Conditional Lending Practice .......................................... 175
5.2 The IMF and Leverage ............................................................................... 180
5.3 The Recipient and Leverage ....................................................................... 192
5.4 Comprehensive Speciﬁcation for Conditionality Bargaining ...................... 209
5.5 Theoretical Implications of Findings .......................................................... 219
CHAPTER 6
LEVERAGE, CAPITAL MOBILITY AND CONDITIONALITY:
INTERNATIONAL BAILOUT OF KOREA, 1997 ..................................................... 226
6.1 Summary of Events .................................................................................... 228
6.2 Reconstructing the Case to the Conditionality Bargaining .......................... 232
6.3 Implications to Debates on Conditionality .................................................. 250

 

 

r
p
I

F-"

CHAPTER 7

FINALE: CONCLUDING REMARKS AND FUTURE RESEARCH ......................... 267
Findings and Implications ................................................................................ 267
Future Research Agenda .................................................................................. 273

APPENDIX 1

METHOD OF DATA COLLECTING:

COUNTING THE NUMBER OF AGREED CONDITIONS ....................................... 278

APPENDIX 2

STATISTICAL RESULTS OF NEGATIVE BINOMIAL REGRESSION ................... 283

APPENDIX 3

TREND IN TIGHTNESS OF CONDITIONALITY .................................................... 293

REFERENCES ............................................................................................................ 294

xi

 

Table 3.
Table 4.
Table 4.
Table 4.
Table 4.
Table 4.
Table 4.
Table 4.
Table 5.
Table 5.
Table 5.
Table 5.
Table 5.
Table 5.
Table 5.
Table 5.
Table 5.
Table 5.

Table 5.
Table 5.

Table 5.

LIST OF TABLES

1 Outcomes of Conditionality Bargaining with Perfect Information ............... 118
1 Descriptive Statistic of Type of Arrangement .............................................. 140
2 Descriptive Static of Tightness of Conditionality ......................................... 140
3 Descriptive Statistic of Expected Utility Components ................................. 150
5 Frequencies of Donor's Leverage in Binary Form ........................................ 160
6 Descriptive Statistic of Recipient's Leverage ............................................... 166
7 Frequencies of Recipient's Leverage in Binary Form ................................... 169
8 Descriptive Statistic of IMF-Recipient Financial Transactions .................... 169
1 Quota Proportionality and Conditionality .................................................... 176
2 Type of Arrangement and Conditionality .................................................... 179
3 Donor's Expected Utility Components and Conditionality ........................... 181
4 IMF Institutional Setting and Conditionality ............................................... 183
5 Financial Relations and Conditionality ........................................................ 185
6 Bilateral Political Relations and Conditionality ........................................... 188
7 Bilateral Economic Relations and Conditionality ........................................ 190
8 Recipient's Expected Utility Components and Conditionality ...................... 193
9 Economic Crisis and Conditionality ............................................................ 196
10 External Debt and Conditionality .............................................................. 198
11 Economic Characteristics and Conditionality ............................................ 203
12 Global Economic Opportunities and Conditionality ................................... 205
13 Politics of Special Interest and Conditionality ........................................... 208

xii

Table 5. 14 Comprehensive Speciﬁcation for Leverage and Conditionality .................. 211

Table 5. 15 Comprehensive Speciﬁcation with Subjective Probabilities ....................... 213

Table 5. 16 Descriptive Statistic of the Alternative Dependent Variable ...................... 215

Table 5. 17 Comprehensive Speciﬁcation

with the Alternative Dependent Variable ................................................... 215

Table 5. 18 Comprehensive Speciﬁcation with the Alternative Dependent Variable

Table A2.
Table A2.
Table A2.
Table A2.
Table A2.
Table A2.
Table A2.
Table A2.
Table A2.
Table A2.
Table A2.
Table A2.
Table A2.
Table A2.

Table A2.

and Subjective Probabilities ...................................................................... 218
1 Quota Proportionality and Conditionality ................................................. 283
2 Type of Arrangement and Conditionality .................................................. 283
3 Donor’s Expected Utility Components and Conditionality ........................ 284
4 IMF Institutional Setting and Conditionality ............................................. 284
5 Financial Relations and Conditionality ..................................................... 285
6 Bilateral Political Relations and Conditionality ........................................ 285
7 Bilateral Economic Relations and Conditionality ...................................... 286
8 Recipient’s Expected Utility Components and Conditionality ................... 287
9 Economic Crisis and Conditionality ......................................................... 288
10 External Debt and Conditionality ............................................................ 288
11 Economic Characteristics and Conditionality .......................................... 289
12 Global Economic Opportunities and Conditionality ................................ 289
13 Politics of Special Interest and Conditionality ......................................... 290
14 Comprehensive Speciﬁcation for Leverage and Conditionality ............... 291
15 Comprehensive Speciﬁcation with Subjective Probability ...................... 292

xiii

LIST OF FIGURES

Figure 3. 1 Conditionality Bargaining Game .................................................................. 83
Figure A3. 1 International Context and Trend in Tightness of Conditionality ............... 293
xiv

CHAPTER 1

CONDITIONALITY: A PANDORA’S—BOX?

1.1 PUZZLES OF CONDITIONALITY

Since the early 19805, structural adjustment as condition to lending from
international ﬁnancial institutions (IFIs)—especially from the International Monetary Fund
(IMF or Fund) and the International Bank for Reconstruction and Development
(informally known as the World Bank)——has dominated economic policy agenda in the
developing world.1 In academia, parallel efforts have been made to ﬁnd the relationship
between policy choice and economic performance, what policies caused macroeconomic
imbalances compounded by national debt crises, and what corrective measures have been
successﬁIl in managing them. A general conclusion that emerged from them is that
stmctural adjustment was nothing but failure, as blatant titles such as “The Twilight of
Conditionality?” (Nelson, 1996) and “The Failure of Conditionality” (Collier, 1997)
Suggest.

While conditionality was being established as a major research theme, a growing

number Of research on it increasingly depended on a presumption that failures in structural

\

l .

TakmE a handy example, the average number of programs and committed funds indicates the degree of
acceptance of structural adjustment in developing countries. Since 1982 through April 30 of 1998, 425
mgements were contracted with the IMF in the name of structural adjustment. On average, 26.56
adJllSt“lellt programs were annually arranged during that period; funds conunitted under those
arrangements reached 142,910 SDR millions in total, of which annual average is 8931.88 SDR millions.
w” the Other hand, for the ﬁrst 30 years of HVIF history from 1952, the annual average of arrangements

as 17.37 with 1846.77 SDR millions annually committed. Source: IMF Annual Report. 1998.

ashmgton, DC: International Monetary Fund.

adjustment were caused by the application of wrong policies from the IFI’S, and that
structural adjustment would succeed if the [FPS modify their economic model so as to
apply a different set of policies.2 This presumption also assumes away universal imposition
of such policies by the IFI’s, whereby process for policy choice was precluded from its
discourse. However, reviewing closely its track record, conditionality appears to have
varied case by case. Despite the contention of strict imposition of the same wrong
economic model, some countries obtained easier terms, quantitatively and/or qualitatively,
than others in the ﬁnal agreements with the IFI’S. This conﬂicting observation inevitably
raises fundamental puzzles: what are the bases of these differences, and what made them
settle for those unpromising policies in the ﬁrst place; and can the contention of dictation
or coercion by the two representative IFI’S explain the discrepancy consistently within its
theoretical boundary?

Such observation also presents a theoretical imperative that the policy choice
process is explicitly endogenous to conditionality agreement rather than exogenous and,
thereby, unﬁilﬁlling policies prescribed for structural adjustment become an outcome of
deliberate action from both parties involved to make choices. Furthermore, the structure
of their relationship based on the ability to inﬂuence their interaction, which can be dubbed
ba’gaim‘ng strength, may determine the ﬁnal choice of conditionality agreements. It is this

missing political process of choice, speciﬁcally interaction and bargaining strength in

\

2
There has been much discussion of the success of the IMF ’s role. However, reviewing the discussion or
1':le attempting to evaluate policy advice dispensed by the IMF is beyond the scope of this dissertation.
D‘" the review of evaluations and analytical issues, see Killick, Tony. 1995. IMF Programmes in
eve/oping Countries: Design and Impact. London: Routledge. pp. 36-48; Lensink, Robert. 1996.
ggg'ur‘al Adjustment in sub-Saharan Africa. London: Longman, pp. 95-108; and Williamson, John.
a. ‘On Judging the Success of IMF Policy Advice.” in IMF Conditionality, ed. John Williamson.

Washington, DC: Institute for International Economics, pp. 129-143.

negotiation for development ﬁnance between the IFI’S and developing countries, that this
dissertation intends to address.

These seemingly simple puzzles related to policy choice process of conditionality
are something that converges to the core of debates revolving around structural
adjustment, and answers to these puzzles are likely to Shed a new light on the question of
failure of conditionality and provide clues to viable solutions. To better understand the
constellation of relational aspects of conditionality, speciﬁc issues and unsolved problems
within the subject of conditionality are reviewed in depth.

The IMF, as part of Bretton Woods post-World War II economic architecture,
was established to promote international monetary cooperation and world trade through
exchange-rate stability. In the same light, the M was designed to be a ﬁnancing
institution, providing temporary ﬁnance to member countries experiencing balance-of-
payments deﬁcits and prescribing appropriate means of adjustment to help member
countries overcome their deﬁcit positions. On the other hand, the World Bank, the other
Pillar of the Bretton Woods system, was left with development assistance that provides
long-term ﬁnancing for speciﬁc development projects—ﬁrst for post-World War II
reconstruction and development efforts in Europe, and later for development projects in
less economically advanced countries.

The role that the two Bretton Woods institutions played in the Stability and
proSperity of the post-World War economy cannot be understated in any event.
Nevertheless, events in the 1970’s, such as abandoning ﬁxed exchange rates, undermined
the glObal systemic role of the IMF—and under that circumstance, the IMF was almost

forced to accept a new and more speciﬁc role: development ﬁnance (Killick, 1995;

Vaubel, 1991).3 This role reﬂected the evolving balance-of-payments problems that
developing countries encountered in the wake of oil crises of the 1970’s and became ﬁxed
with the outbreak of debt crisis in 1982. In response to the deepening debt crises in the
developing countries and a growing sense that traditional sources of concessional lending
were insuﬂicient for the task—the IMF, as the lender of last resort,4 created two new
facilities, Structural Adjustment Facility (SAF) in 1986 and Enhanced Structural
Adjustment Facility (ESAF) in 1987, besides the existing credit under Stand-by
Agreement (SBA) and Extended Fund Facility (EFF).5 Each of these facilities involves a
condition that recipient countries Should agree to undertake speciﬁc economic policy
reforms over the ﬁxed term. The agreed policy reform is typically associated with an

orthodox IMF stabilization program, which is designed principally to reverse acute

 

3 The IMF has been mistaken as a development agency nowadays, and this triggered debates on the right
place and role of the IMF in the world economy. See Finch, David C. 1988. “Let the M be the IMF.’
International Economy (January/February); Finch, David C. 1989. The IMF: The Record and the
Prospect. Princeton Essays in International Finance, no. 175; Bergsten, C. Fred. 1990. “From Cold War
to Trade War.” International Economic Insights (July/August); and Bird, Graham. 1995. IMF Lending to
Developing Countries: Issues and Evidence. London: Routledge.

4 The usage of this term requires caution. As a matter of fact, it is not that clear whether the IMF is a full-
ﬂedged lender of last resort. There are some qualiﬁcations that economists think the IMF should meet to
be a lender of last resort in international monetary system just as central banks do in domestic monetary
matters, Discussing this issue is beyond the scope of this dissertation. The IMF itself argues for the need
for an international lender of last resort and steps up to achieve that status in the middle of debates on
reforming the IMF. The use of this term here reﬂects this. For a theoretical discussion of this, refer to
F‘SCherr Stanley. “On the Need for an International Lender of Last Resort.” paper delivered at the joint
luncheon of the American Economic Association and the American Finance Association, New York.
[Onllne] Available http://www.imforg/external.np/speeches/ 1999/010300.htm, January 3, 1999.

5
Features of SAF and ESAF are that, ﬁrst, both facilities are available only to designated low-income
comm“; and they are granted for ten years at a half percent interest rate, compared to short-tem (1-3
y ears) and higher interest rate up to 9% of SBA and EFF. For detailed description of each facility, refer to
4:23“er DepartmenL 1998. Financial Organization and Operations of the IMF. Pamphlet Series, no.
' Washington, DC: International Monetary Fund.

 

‘
I
u

n

u
n.
P
-P
‘1 a
A
O

Balance-Of-payments deﬁcits by generating large trade surpluses in relatively Short period
of time.6 Those policy measures are also a catalyst that facilitates consensus formation on
adjustment policies both within the recipient country and among international lending
communities, and to provide a focus for the mobilization of additional funds (Bird, 1995).
In the meantime, the World Bank’s development activities were also changed with
a view to enhancing the IMF’s mission of promoting balance-of—payments stability. Its
traditional project lending was Shifted toward policy-based lending under Structural
Adjustment Loan (SAL) and Sectoral Adjustment Loan (SECAL), which are based on
neoclassical economic models of development, promoting longer-term economic recovery
through reduced govemment’s intervention in economic activities and the strengthened
role of market in resource allocation. These are efforts to institutionalize elements of the
reform measures initially introduced as part of short-term IMF stabilization efforts
(Biersteker, 1993).

This practice in international development ﬁnance in the early 1980’s was the
reinstatement of an age-old principle of conditionality in relation to developing countries
(COhen, 1982). Strapped for resources and faced with potential calls on its credit of
unprecedented portions by developing countries, the Fund reasserted the framework of
conditionality. Thus, the IFI’s make loans based on the promise of borrowing countries to

pursue a speciﬁed set of policies as conditions for ﬁnancial support.7 It is in this context

;\\H
F“ an authoritative exposition and discussion of IMF adjustment model, see IMF. 1987. Theoretical
ASP“? Of the Design of Fund-Supported Adjustment Programs. IMF Occasional Paper, no. 55.
ashlngton, DC: International Monetary Fund.

7
I 9““ the Outset, this meant balance-of-payments stabilization in exchange for debt rescheduling. After
#5, uIlder Baker Plan, debt crisis would be eased through continued reliance on market mechanisms,
ch meant that medium-tenn structural adjustment in debtor country economies would be ﬁnanced by

that the IMF and the World Bank have been widely viewed as exercising tremendous
inﬂuence over the design and implementation of economic reforms in the developing
world since the early 1980’s, and much of their inﬂuence is assumed to ﬂow from the
conditionality (Nelson, 1996).

Countries in debt crisis that appealed to the Fund or the World Bank for new loans
were the ones who had already been judged to be uncreditworthy by normal market
criteria. In such a treacherous circumstance, a ﬁlndamental assumption of conditional
lending is that the IMF and the World Bank can compel, or at least induce, countries to
undertake stabilizing actions in return for loans, thereby making the loans prudent even
when the private capital markets have declared the country to be uncreditworthy (Sachs,
1989). Countries in crisis are often in poor economic shape in large part because of bad
policy choices in the past, and the Fund and the World Bank’s measures are appropriately
focused on key policy weakness. Moreover, the IMF and the World Bank have the
expertise and institutional clout to design high-conditionality programs, while the
commercial banks that used to inundate developing countries with massive capital prior to

the debt crisis do not.8

 

 

mew Fund and World Bank lending to provide the needed incentives for renewed private bank
le‘l‘hng- This was replaced with Brady Plan in 1988 that called for debt relief in exchange for structural
adJUStment of debtor countries.

8 The Fund remains in a uniquely strong position to collect and interpret information of members’
economic conditions and to act on the basis of it. Not all the information the Fund share with the member
Countries is openly available to banks, and conﬁdentiality is maintained for the purpose of keeping the
ﬂow or accurate information, which would make the capital transfer mechanism less imperfect. On the
Otherhand’ private lenders possess less information than the Fund and, therefore, tend to reach inferior
dec1310118, although the Institute for International Finance collects and processes information on behalf of
2:; banks. As indicated by their reaction to debt crisis by simply taking decisions not to lend, which was
COu°w°d by the resurgence of bank lending to some developing countries in the 1990’s, their analysis of
detain"? nsk Is neither sophisticated enough nor closely related to signiﬁcant economic factors
Vaubneljlmng the future ability to service debt (Bird, 1995). To the contrary, public choice theorists like

(1991) dismiss the superiority of the Fund in information processing on the ground that most of

Therefore, it is no wonder that the two leading IFI’S have been the main targets of
criticism for the disappointing results of structural adjustment. Some say that, to begin
with, the intellectual underpinning of structural adjustment was no more than a mixture of
ideology and common sense, with very little theory to recommend it (Buira, 1983). Also,
some believe that structural adjustment was too harsh for developing economies, and that
such harshness had usually been the result of dictation or coercion by the IMF and the
World Bank (Eckaus, 1986). The IMF and the World Bank imposed inﬂexible and
onerous conditions on individual countries, with little consideration for differences in types
of economies (Williamson, 1983b; Killick, 1990, 1995).
Nonetheless, both the role for high-conditionality lending and the inﬂuence of the
IMF and the World Bank seem to be overstated. Conditionality, in reality, has a facade
that makes the above arguments less convincing than they sound, particularly in some
aspects. Among other things, against the coercion, compliance with conditions remains
low, at around a 50 % level. Using as the indicator of performance discontinuance of a
Program before the end of its intended life—in which 20% or more of the total value of
the credit remained undrawn as well—Killick (1995) found out that just over half (53%)
of 305 Fund programs approved in the period of April 1979 through April 1993 were
uncompleted. What was worse is that such poor performance was more serious with

medium-tenn programs than Short-term and delinquency seemed to be the main cause of

failure,

\

the time the Fund’s forecast has been less accurate titan both private forecasts and forecasts by national
p “bl“: agencies.

Moreover, rather than ﬁnding the presumed strict imposition of same conditions, it
was observed that the terms of policy reforms varied signiﬁcantly country by country.
Some countries got exempted from a key component of orthodox reform (such as the
removal of a particular subsidy and devaluation), whereas others achieved qualitative
changes (i.e., the Size of cutbacks in government spending and money growth). Around
the world, large debtor countries like Mexico, Brazil, and Argentina have tended to
receive a prompt response from the IFI’S, loans far exceeding preset access limits, and
more discretion in their debt management than countries with relatively smaller debt
burden. They have also pioneered more unorthodox debt rescheduling agreements and
have been more successﬁll in securing additional ﬁnance from commercial banks (Haggard
and Kaufman, 1989; Stiles, 1991). It is indeed true that those policy reforms mentioned
above typically deﬁne the initial position of the IFI’S in confronting heavily indebted
countries. However, the precise terms of the agreements ﬁnalized by developing countries
were by no means all alike. Apparently, some developing countries managed to escape this
externally thrusting plight, which indicates that within the general guidelines of orthodox
stabilization and adjustment, there is room for considerable variation in policies to be
chosen.

This obvious discrepancy demonstrates that past studies of conditionality started
from a SUperﬁcial or, at most, a partly accurate observation, and thereby were impaired by
conceptualization problems. Conditionality intrinsically reﬂects power inequality due to
the fact that one of the parties involved is the holder of readily disposable material
resources, while the other is heavily dependent on those resources for its survival. It is this

Inequality between the two parties that makes the presumption of imposition sound

plausible. However, this is nothing but invoking to the subject of conditionality the
stereotyped notion of power in international relations that stronger actors with the greater
resources and capabilities (such as GNP and large population) will, by deﬁnition, prevail in
any encounter with the weaker. It is well compiled that such Single-minded applications of
the concept of power could not explain numerous cases where weaker states won many of
their objectives vis-a-vis stronger states. Presuming that the IFI’S strictly imposed
conditions, since there was power disparity between them and developing countries, is
committing the same mistake prevalent in international relations. To the contrary, the
variance in conditionality agreements indicates that the conceptualization of conditionality
needs to go beyond the issue of power and, furthermore, that there are major distinctions
in the instruments, issues, and interests that developing countries can bring to their relation
to the Fund and the World Bank and draw easier terms of conditionality (Biersteker,
1993).

Taking this into account, much of the criticism mounted against structural
adjustment and the role of international donors in laying down adjustment policies is
polemical with little ground ﬁ'om empirical analyses utilizing reliable models. Research in
the past, especially focusing on implementation, often proceeded without a clear
exposition of the process of conditionality agreement and, instead of seeking it eagerly, it
was blanketed altogether under the inequality in capabilities of the two actors involved.
Within the contention of coercion or dictation by the IFI’S, not only the relative autonomy
of debtor countries is ignored, but also the sophisticated and symbiotic interplay of global
and domestic forces in shaping relations of asymmetrical interdependence is not properly

appraised (Zartman, 1987a).

The presumed relationship between coercive conditionality and its unsuccessful
results of structural adjustment should be substantiated as a prerequisite to studies of
condition implementation. This is likely to be most successfully carried out by answering a

Series of questions: how are the terms of conditionality drawn, how do the two actors
involved interact with each other in that process, and what factors inﬂuence their
interaction which eventually leads to the variance in terms of agreement? These questions

ﬁtndamentally boil down to the so-called power relationship of the actors, but needs to
transcend the traditional, as mentioned above. Therefore, as an attempt to solve the
puzzles of conditionality, I propose research that lays out a framework for analyzing the
relationship between the IFI’S and debtor countries as a strategic bargaining game in
which leverage measuring relative bargaining strength largely determines the nature
and variance of conditionality agreements. In this dissertation leverage is revived as a
state’s capabilities or endowments conducive to changing the outcomes of bargaining
within a speciﬁc issue area, which otherwise would not occur. This is a noticeable shift
away from the conventional concept of power, and analytical issues related to leverage
will be explicated in Chapters 2 and 3.

The concept of leverage has a special importance in this project since the process
of negotiation is the process of creating outcomes, altering and modifying values, and
bargaining strength based on national resources creates options that are useful within the
context of speciﬁc encounters a state faces (Habeeb, 198 8). Conceptualizing conditionality
with leverage, this dissertation aims to Show the dynamics at work in negotiations for
conditionality agreements with the major IFI’s—dynamics that can account for the

difference in their outcomes and that cannot easily be explained away as the past studies

10

that presumed imposition of conditions would have us do. Thus, this research will
constitute an integral part of making the study of structural adjustment complete, covering

the Whole gamut from agreement to implementation, so that we can grope for a better

understanding of the subject.

1 .2 MORETICAL OVERVIEW OF THE SUBJECT

This section critically reviews the ﬁeld, focusing on the conceptualization problem
in conditionality. Identiﬁed failure in the past studies is the one that makes my project an
imperative for reviving the ﬁeld. Their success, on the other hand, will be absorbed,
reorganized and developed further for my project to conceptualize conditionality from the
Perspective of power relationship.

Criticisms of IMF monetarist conditionality and the IMF’S reply to them have a
relatively long history dated back to the late 1950’s and early 1960’s. Since then, this
debate has changed its outlook with new features added to the core, reﬂecting
developments in international economy.9 Conﬁning overview of the debate on
conditionality more relevantly to that of the 1980’s and 1990’s, it can be entitled IMF’S
(neo)orthodoxy vs. (neo)structuralism, which falls on two dimensions: the coherence of

StrUtztural adjustment as an economic program, and its implementation. While the ﬁrst

dimension is mainly about the formulation and design of Structural adjustment as well as

 

 

9 Historical overview of debates is well compiled in Kahler, Miles. 1990. “Orthodoxy and Its Altemative:
EXplaining Approaches to Stabilization and Adjustment.” in Economic Crisis and Policy Choice: The
Politics of A djustment in the Third World, ed. Joan M. Nelson. Princeton, NJ: Princeton Univ. Press.

11

its theoretical and ideological base, the second is concerned with the implementation of
conditionality and its effects in real world followed by policy alternatives. Most studies of
structural adjustment have concentrated on speciﬁc country experiences on the second
dirnension and took part in attacking the ontological base of conditionality at the end of
their studies on the ground of disappointing results they found. Whichever dimension their
research may stand for, however, their common starting point is that structural adjustment
failed to achieve its proclaimed goals. Disappointing outcomes in terms of economic
indicators and negative effects on human and social indicators have prevailed in the
developing economies, even where the conditions were sincerely implemented. Not all
adjusting countries had poor performance of this kind, but good performance among
adjusting countries was the exception rather than the rule (Stewart, 1995). Still, to do
justice with this debate, it should be admitted that these studies lack deﬁnite criteria to
evaluate successful implementation of conditionality,10 and this only ﬁlels the debate on
the causes of failure and viable policy alternatives between the advocating IFI’S and
critical scholars.
On the first dimension, structural adjustment is criticized for its defects; while
Stabilization is a crucial element of adjustment, ﬁscal stringency, balance of payments and

dth constraints overkill potentials for productive activity required for development.

 

 

‘0 Using the IMF’s term, this is equivalent to performance criteria. The outcomes of performance could be
mum-staged: ﬁrst, a borrowing country implements each of agreed conditions but still it either can or
Cannot meet the targets of structural adjustment. Second, meeting the targets can lead to improving the
balance of payments so that it can reimburse the Fund, repay its debts, and continue to honor its
international obligations, or it does not. Of course, there is no rigid logical connection between the success
at the ﬁrst stage and the second. The success at the second stage can occur without the ﬁrst. Thus, a

natural question is success at which stage the performance criteria should count. Neither the IMF nor
critical scholars are clear about this.

12

Moreover, the IMF insists on the same defected economic reform programs for every
country requesting ﬁnancial assistance. That is, they provide the same medicine for all the
economic ailments of the developing world (Dell, 1983), which consequentially produces
negative effects in developing countries, rather than positive ones. Basically, this line of
criticism endorses many of orthodox principles, including full recognition of the economic
costs of subsidies, overvalued exchange rates, and other price distortions, although it
questions the legitimacy of highly detailed conditionality and external intervention. It is
skeptical of certain elements of orthodox adjustment policies at the minimum and, thus,
suggests that the IMF modify its methods: more sensitivity and ﬂexible application.

An overarching assumption for policy recommendation from the above line is that
developing countries are structurally different as a group, and that each developing
country deserves independent analysis for its economic reform program. Thus, an
adjustment program tending to this need is one that gives more careﬁil attention to
sequencing, a more realistic time horizon, greater priority to growth, and short-run current
account balance and containing demand. Even before the advent of full-blown structural
adjustment in 1983, Williamson (1983b) contended that a country should be able to decide
for itself how best to adjust its balance of payments, and this Simply resonated with more
SPeciﬁc requirements by other scholars. For instance, Killick (1984) argues for extending
and consolidating supply-side measures and longer-term adjustment periods: the Fund also
0light to adapt to new problems created by the oil shocks and the world recession in a
number of ways: increasing its lending power; making increased use of the EFF; reducing
insistence on rigorous conditionality. In presenting alternatives to current structural

adjustment, Taylor (1987) treats orthodox analysis as a hypothesis and stresses the need

13

for receptiveness to the economies in question and wisdom in applying institutional

parameters to them. Related to Taylor, Rodrik (1990) says that a type of wise application
of institutional parameters is prioritizing policy sustainability in a structural adjustment
package—given the nature of the crisis—over liberalization, whereby private investment is
resuscitated and generates lasting growth. In the case of Sachs (1987)——who is skeptical
of the orthodox program but not orthodox analysis—he notes that market mechanisms,
despite considerable potential, have their limits. He also suggests, from careful
assessments of East Asian experience, a much more active role of state and more gradual
and phased reform than current neo-orthodoxy prescribes. The search for adjustment
programs that combine adjustment with growth and equity is also in this line. The SO-
called “adjustment with a human face” (Cornia et al., 1987; Stewart, 1995; Oxfam, 1995)
represents concern about the distributional consequences of orthodox adjustment.
Reporting adverse impact on particular vulnerable groups—children and the poor—and
eHVironmental damage in the wake of adjustment, they argue that alternative policies can
be implemented within the orthodox framework.

Not too far from these economically oriented debates are ideologically charged
criticisms of political function of conditionality. The IMF and the World Bank are
overpoliticized, these critics charge, in that major creditors within the two organizations
use them to increase the liberalization of international economic system, and directly to
sustain the economic viability of nations considered strategic allies to them (Anunobi,
1992; Assetto, 1988). The Fund and the World Bank together serve merely as a debt
collector and enforcer of the interests of the world capitalists, which is why and how they

impose inﬂexible and painful conditions on developing countries, regardless of differences

14

in types of economies and level of development (Helleiner, 1986). Such an orientation

within the IMF and the World Bank and their conditionality lead to fears in the developing
world that the IMF and the World Bank intentionally perpetuate economic dependency
through conditions that re-establish colonial economic patterns based almost wholly on
raw materials exports and de-industrialization (Stein 1992; Chem, 1992). In short, these
arguments claim that conditionality is designed to take care of the well-being of developed
countries, not developing countries. Thus, conditionality is based on a political rationale
rather than economic theory.

The IMF’S response to these criticisms has been that conditionality does, in fact,
vary depending on the country concerned and its economy and history doesn’t bear out
the allegation that conditionality has a negative impact on developing economies. The IMF
asserts that its institutional purpose and objectives have not changed over the years; the
IMF has maintained its primary focus on sound money, prudent ﬁscal policies, and open
markets as preconditions for macroeconomic stability (Fischer, 1998b). This means that in
Stabilization programs, many elements of the traditional orthodoxy remain central to Fund-
supported programs. However, this continuity does not necessarily imply inﬂexibility and
insensitivity, the IMF argues. While its purposes have not changed, the IMF has Shown the
ﬂexibility to adapt its approaches to problems and the instruments it employs to an ever-
Changing global environment (Fischer, 1998a). The scope of its policy concerns has
bl‘Oadened to include other elements that also contribute to economic Stability. A good
example of this is upgrading economic growth to the status of primary objective of a
Fund-supported program along with stabilization (Polak, 1991). The Fund’s traditional

position had been that its principle of political neutrality made it inappropriate to get

15

involved with domestic policy objectives that are pursued from equity, social, or political
considerations (Guitian, 1987). Thus, the distributional impact of stabilization programs is
the govemment’s concern, not the Fund’s. This position has gradually changed. The IMF
now explicitly discusses distributional effects of adjustment. For more than a decade, the
[IN/IF has sought to understand the social and human dimensions of adjustment and, as a
result, provided social safety—nets to vulnerable groups and an environmental
consciousness, which have appeared in many adjustment programs since late 19805. ”
Another area in which the IMF has been responsive to members’ needs is the
extension of the time span for stabilization. Recognizing that low-income countries’
problems are deeply rooted in structural weakness, achieving balance-of-payments viability
should take place over medium-term than short-term. The newly established SAF and
ESAF largely reﬂect this change within the IMF. These facilities are to address problems
in heavily indebted low-income countries suffering protracted payments difﬁculties, and
substantial progress is supposed to be made within a three-year period. Although it
continues to look at all its member countries through the same lens—the requirements for
ecOnomic stability—the IMF has heeded the criticisms leveled at its approaches and dealt
With distinctive problems in member countries in differentiated ways.
On the issue of negative effects of Fund programs, especially on economic growth,
the IMF dismisses such criticism on methodological grounds. Critical studies of Fund
Programs apply various methods, each of which faces some intractable obstacles, and thus

their evaluations do not give reliable results. Despite such methodological problems with

K

H Thus, Polak (1991) classifies objectives that conditionality pursues into two categories: primary and
secondary. Stabilization, growth, and price stability are in the primary objectives; poverty alleviation,
environment, military spending and political consideration are in the secondary.

16

evaluating the effects of conditionality, the IMF argues, one piece of solid evidence is that
the balance of payments and the current account of countries with Fund programs have
improved in short term. In other areas, the effects of programs are mixed and differ
country by country, depending on the initial economic conditions that adjustment
programs were introduced to and the developments after program periods (Khan, 1990;
Schadler et a1, 1995).12 What the IMF suggests is that vigorous implementation of sound

programs over medium-tenn periods will lead to more satisfactory outcomes in the long

13
run.

Regarding economic growth suppressed by stabilization, the IMF says that a
distinction should be drawn between short and long-term issues relating to the effects of
Fund-supported programs on growth.14 The decline in the growth rate is a necessary part
of adjustment to eliminate underlying imbalances in the economy. However, balance of
PaYInents recovery should not conﬂict with economic growth when the time horizon for

both objectives is properly speciﬁed to be the medium term. Indeed, adjustment and

x

12 This evaluation is also applied to ESAFs. While Khan and Schadler focus on SBA and EFF, a staff
repon only on ESAF reaches a similar conclusion. See IMF. 1997. The ESAF at Ten Years: Economic
Adjustment and Reform in Low-Income Countries, IIVIF Occasional Paper, no. 156. Washington, DC:
IMF.

‘3 A very similar argument is forwarded about World Bank-supported programs. That is, the World Bank
(1989) maintains that, on average, program countries had “moderately” better economic performance than
nOil-program countries, in terms of economic growth and both internal and external balance. The
diﬁ‘erential in performance was greater for countries that had implemented a succession of program; thus,
adjustment programs have a better chance of success in countries with an appropriate macroeconomic
POlicy framework and a high degree of borrower ownership (World Bank, 1993). However, Mosley and
his colleagues (1989, 1995) found that Bank programs had no measurable effect on real GDP, positive
effect on export growth and balance of payments and negative eﬂects on investment levels. They also
found a rather larger rate of non-implementation of program policies than 40% as claimed by the Bank.

M Lensink (1996) admits that methodological disparity hinders a generally acceptable conclusion on this
matter, too. Even if focusing only on a few macroeconomic indicators—such as GDP, exports, investment,

and savings—a unanimous answer cannot be given to the question whether the adjustment programs have
had a positive or negative effect.

17

growth are complementary: by removing distortions and impediments to eﬂiciency,
adjustment establishes necessary conditions for sound growth. In the presence of
imbalances, the choice that policymakers confront is adjustment now versus often harder
adjustment later: the notion that adjustment is inimical to growth only serves to conceal
that without adjustment, growth today is at the expense of often signiﬁcantly less growth
t-omorrow (Guitia’n, 1987). I
Note that no matter how much critics disagree with the IFI’s in economics and
politics, they do not discuss explicitly how conditionality agreements were reached.
Rather, while ignoring that issue, they imply or assume that conditionality was dictated by
the IFI’s and coercive in nature, and as a consequence, negative results ensued. However,
if donor conditionality is necessarily coercive, all of the conditions should always be
implemented and always damaging to the recipient country. In this situation, there is no
need for policy dialogue between the two parties, since the IMF and the World Bank
always win on their own terms (Mosley et al., 1995). On the contrary, however,
compliance roughly at a 50% level is almost a norm, as mentioned earlier, and has
tarnished the reputation of the Fund and the World Bank. If the IMF and the World
Bank’s conditionality were truly coercive, such an ostensive evasion would not be
allowed. This also implicates counter-evidence to the opposite case that implementation of
the conditions is in the mutual interest of the [FPS and the developing countries; the
actions recommended to those countries are desirable for them and, thus, they need not be
compelled to undertake the policy reforms. Only a lack of mutual interests on both sides,
however, can explain the fact that recipient governments did not spontaneously implement

the conditions desired by the Fund and the World Bank without any prodding from them.

18

Apparently, both interpretations, which were taken with neglect, conﬂict with readily
available evidence.

If the critics discussed above exaggerate the inﬂuence of the IFI’s on
conditionality, we should also recall that some analysts understate the inﬂuence of the
[FPS in designing conditionality. Some inﬂuential scholars who emphasize the dimension
Of politics and public policy for structural adjustment in developing countries focus on the
political factors that shape governments’ adjustment choices and the implementation of
those choices. They argue that the role of external agencies, while important to the details
of particular cases, is less important than essentially internal factors as a determinant of
adjustment choices and action (Nelson, 1990). This view seems willfully self-

contradictory: while it admits that the IFI’s have a strong belief in orthodoxy, it denies that
they are interested in imposing their belief in the programs that they ﬁnancially support. To
the contrary, the apparent dominance of the orthodox program was due in large measure
to the preferences of G—S governments, which endorsed the prescriptions being offered
and were in the course of implementing them in their own countries (Kahler, 1990). Their
preferences have been reﬂected on the policies pursued by developing countries,
particularly after the onset of the debt crisis when many developing countries undertook
programs of adjustment supported by the M or the World Bank.” Taking the inﬂuence
of external agents as irrelevant to the choice process, or attempting counterfactual
explanations—what would probably have occurred with external intervention—is likely to

make explanations incomplete and distort the reality. The inﬂuence of the IFI’s on the

 

'5 The inﬂuence of the major donor countries within the IMF was not only on the conception of structural
adjustment, but also on the changes in conditionality that the IMF went through since mid 1980’s as
discussed earlier.

19

choice of adjustment program has been there all the time, and thus deserves as much
attention as we pay to domestic political factors. It should be clearly understood that
political forces work not only at implementation, but also at choice stage.

Another stream of debates over the disappointing results of structural adjustment is
focused on poor implementation, rather than the merits of structural adjustment.
According to Sachs (1989), reneging on the agreed conditions is inherent in conditionality
due to the fact that an appropriate debt relief was excluded from its design. Without debt
relief, required structural adjustment acts as a tax on investment, which will beneﬁt foreign
creditors rather than domestic constituents. This will be a major disincentive against
investment in the debtor country that would contribute to future debt service. Since the

government will not undertake such investment, it is not likely that the debt-servicing
capacity of the country will be enhanced.'6 Adjustment is not occurring in the long run.

In this context, debt relief is a necessary, if not sufﬁcient, condition for inducing
the country to undertake needed reforms. Furthermore, debt reduction has a strategically
important connection to the Fund’s catalytic role, since the enhanced effectiveness of
adjustment through debt reduction and better economic performance of a recipient
country will draw in additional capital to the country, which will allow the burden of
balance-of-payments ﬁnancing to be spread between commercial lenders and the IMF
(Bird, 1995). Unfortunately, however, the IMF failed to recognize this, and poor

implementation continues.

 

1" A research has shown that there are actually the opposite incentive effects of debt relief and that this
line of argument could be valid only under particular circumstances. Theoretic diagnoses of debt problem
are collected in Frankel, Jacob A., Michael P. Dooley and Peter Wickham, eds. 1989. Analytical Issues in
Debt. Washington, DC: International Monetary Fund; and Wells, Robin. 1993. “Tolerance of Arrearages:
How IMF Loan Policy Can Effect Debt Reduction.” The American Economic Review 83, no. 3 2621-633.

20

Mosley (1987) puts such a situation in more generic terms. “Slippage,” or poor
implementation, is induced by unnoticed disincentives. A debtor government accepts ex
ante the need for policy adjustment as the quid pro quo for a loan, but the government has
a strong incentive to avoid the policy change after the loan is arranged. Conditionality, as
a course of future actions to which a recipient country will bind itself, makes senses today,
but will become unattractive in the future. This even involves a possibility of disguised
exploitation of the donor by the recipient: it often takes a long time for the donor to see
whether the recipient is acting in the spirit of the policy measures it wants implemented,
and the recipient who reluctantly seeks short-term conditional help from the donor can
escape unscathed from it at the end. This “time inconsistency” is the ground for that

recipient countries must be compelled to accept conditionality (Mosley, 1988).

This perspective shows that the conditional lending from the IMF and the World
Bank does have positive sides, but conditionality faces inherent constraints given the
limited enforcement power at the hands of the IMF and the World Bank. Domestic loans
are normally balanced by collateral in the form of a security, or negotiable capital asset, to
be handed over to the lender if the borrower cannot otherwise repay the loan. The
conditionality of the Fund and the World Bank is an instrument intended to maximize the
probability that the loan will be repaid as in the case of ordinary domestic loans. However,
when IFI’s lend to overseas governments, the conditions often become more complex, and
their application becomes controversial because the borrower is a sovereign body on
whose assets the lenders have no legal claim. This makes the link between instrument and
the ultimate target of economic policy change less certain (Mosley, 1987). Thus, what is

implied by such a prevalent slippage in structural adjustment is that the alleged superiority

21

of the M and the World Bank in imposing conditionality to commercial banks, is
probably correct in general but much oversold quantitatively (Sachs, 1989). The
conditions for structural adjustment, whether by the IMF or the World Bank, should not
on an a priori basis be expected to be implemented from A to Z.

This analysis of limited ability of the IMF and the World Bank to enforce contract

Vis-a-vis developing countries goes beyond just explaining low degree of implementation.
It points to the general nature of the relationship between the [FPS and developing
countries in the matter of conditionality as well. Neither actor involved in conditionality is
so predominant as to manipulate their relationship thoroughly in its preferred way. Thus,
this can be taken as a ﬁrst step for developing a new framework to illuminate the nature of
conditionality more accurately. What this perspective leaves still untouched, though, are
questions of what could determine the nature and pattern of their interaction, and what
outcome will result ﬁom it. Rephrasing these questions for conditionality, it is whether
conditionality is necessarily or inevitably coercive; why conditionality agreements are so
different case-by-case that they cannot be classiﬁed by a reasonable ﬁrst-hand principle,
such as economic difﬁculty; and what motivates the IMF, the World Bank, and developing
countries to keep contracting sub-optimal agreements on conditionality given all the
failure experienced in the past. To ﬁll in the missing link between the newly burgeoning
framework and its vital content, other facets of the relationship between the [FPS and
developing countries must be ﬁrrther scrutinized.

It is a conventional notion that the IMF and the World Bank predominate over the
developing economies. As a matter of fact, however, there are some sensible grounds to

claim that both parties have been dealing with each other with certain caution and strategy

22

in conditionality agreement. Furthermore, while developing countries appear to have
succeeded in ﬁnding chinks in the seemingly impenetrable wall of the IMF and the World
Bank, these IFI’s have been in a relatively weak bargaining position than many think.

Above all, the Fund and the World Bank are the clients of governments to whom
they lend money. Because of this formal position of the IF I’s as the clients of the member
governments, they must accord a formal equality of treatment for all member governments
with regard to conditionality negotiations (Sachs, 1989). This would render it difﬁcult for
the Fund or the World Bank to make invidious comparisons among countries concerning
the likelihood that they will actually live up to commitments to economic policy reform.
Another problem that the Fund and the World Bank face is that they have many goals
other than proﬁts, which can make them approach conditionality with a soﬁ touch. Private
capital markets operate under one principle, making proﬁts out of loans. Conditionality,
however, is expected to fulﬁll purposes other than the mere development of a healthy cash
ﬂow—such as nurturing the potential for grth and keeping political stability, which
requires the Fund to settle for a sub-optimal program (Polak, 1991). These are the raison
d ’étre of international institutions. These concerns do not always allow for a hard-boiled
judgement about the potential success or failure of a conditionality package. On this
Wiarda (1990) succinctly remarks that the IMF has not been nearly so draconian in the
implementation of its austere measures as its popular image would lead one to believe; the
IMF measures often sound ﬁerce and mean on paper but in fact the [MIPS bark is much
worse than its bite.

Finally, other than fulﬁlling their founding purposes as international institutions,

the IMF and the World Bank, as narrowly deﬁned organizations, have an imperative to

23

protect their own existence, i.e., circulating quota subscriptions from member countries.
Developing countries have tried to bypass the Fund whenever possible in their search for
ﬁnancing resources to ﬁll in their external deﬁcits. The refusal of the developing countries
to use the IMF ’s high-conditionality resources, and the substantial loss of its clientele
among the developing countries due to afﬂuent liquidity in the world during the 1970’s,
were the most dramatic threat to the Fund’s overall institutional purpose: ﬁnancing
mandate (Ferguson, 1988).17 This threat was evidently reﬂected in the establishment of
new facilities, SAF and ESAF, in mid 1980’s that departed from its traditional lending
philosophy.18 The IMF cannot afford to watch countries reform on their own because it
would risk revealing the IMF’s irrelevance (Vésquez, 1998). The most recent innovation is
the Supplemental Reserve Facility (SRF), which was created in December 1997 to assist
emerging market economies. It allows the IMF to make large short-term loans at higher
rates than it normally charges. This self-preserving imperative looks a little different within
the World Bank, where country loan officers face lending pressure to meet country
commitment target, whatever the posture may be adopted by the recipient government,
and to meet country disbursement targets, however unpromising that government’s
subsequent implementation performance appears (Mosley et al, 1995). If a program looks

good on paper, there is great pressure for the program to be approved, even if it seems

 

‘7 The inverse relationship between demand for IMF credits and world liquidity was tested by Cornelius.
See Cornelius, Peter. 1988. “Some Arguments against a Global Stabilization Policy by the IMF.”
Economic Intemazionale 41 :3-18.

'8 Furthermore, it is observed by Killick (1995) that, as the IMF established these medium-term
stabilization programs, there has been some relaxation in Fund procedures, which has given its missions
more freedom to accommodate recipient goverrunents’ views; some borrowing governments have acquired
more of voice in the process of drafting Policy Framework Papers for SAF and ESAF.

24

unlikely that the program will actually be carried out. Thereby, these powerful internal
pressures to lend not only act as a counterpoise to the more overt pressure to apply strict
conditions, but also make their conditionality less credible. Without such lending pressure
or organizational incentive for self-preservation, recidivism in conditionality agreements,
with all the poor performance and disappointing results, will remain clueless.

On the part of developing countries, their motivations toward the IMF and the
World Bank are double-edged. First, recipient countries understand that the Fund and the
Bank are ongoing institutions, and that the ripple effects of rupturing relations with these
institutions will impair their economic survival. While defaulting to some private creditors
in particular syndicates might not forestall further borrowing from new lenders elsewhere,
ruining relationship with the IMF and the World Bank by defaulting debt service to them
will presumably endanger the entire opportunities. Nelson (1990) sees that the fear that
multilateral support might be affected (not the hope for resumed long-term private ﬂows)
helped fuller default inhibited during the 19805’ debt crisis. This argument can be
substantiated by the fact that the number of countries that have incurred protracted arrears
to the IMF is much smaller than the number of countries that have failed to service their
debt to other creditors and/or have had to enter into debt-rescheduling arrangements.19
Developing countries are likely to place relatively greater emphasis on meeting their
ﬁnancial obligations to the IMF, considering the IMF’s status as a creditor and its role as a

catalyst in attracting ﬁnancing for them from other creditors (Aylward and Thome, 1998).

 

'9 The IMF regards arrears longer than six months as protracted. While 23 countries have had protracted
arrears on principal obligations to the IMF during its history, 58 countries rescheduled their official debt
through the Paris Club in the period 1980-92. Source: IMF (1983-1998) Annual Reports. Washington,
DC: International Monetary Fund; and World Bank. 1994. World Debt Table: External Finance for
Developing Countries 1994-95. Washington, DC: World Bank.

25

By this, it is suggested that the IMF and the World Bank be strategically permitted to
dictate terms to a country while the private sector cannot (Sachs, 1989).

Second, and perhaps more important, the major creditor governments have made
IMF and World Bank conditionality a precondition for a debtor country’s ﬁJture bilateral
ﬁnancial relations with them. With few exceptions, a debtor country in crisis must have a
good relationship with the IMF and the World Bank in order to qualify for rescheduling
official loans and new credits from official export credit agencies to the debtor
government. Thus, a country’s concern about foreign economic relations with the
developed countries often strengthens the hand of the IMF and the World Bank indirectly
(Cohen, 1985). By all means, it is indicated that developing countries maneuver their
relationship with the [HS under the wide shadow of ﬁiture.

These stances that each party holds in international development ﬁnance indicate
that, in spite of the apparent asymmetry in power, the relationship between the [FPS and
developing countries is interactive rather than one-directional. And all-in-all, this boils
down to the necessity to employ different analytical tools to understand their relationship
correctly, and setting up conditionality agreement as an outcome of strategic bargaining
process will ﬁnd a perfect match. Considering the above situation, the borrower and the
lender of conditional aid have both inherent conﬂicts and harmonies of interests, and the
outcome of this love and hate relationship is not so much from predetermination as from
circumstantial creation. That is, conditionality agreements are negotiated outcomes on the
continuum of absolute harmony to absolute conﬂict of interests, depending on strategy
deployed by both parties and external circumstances. In such negotiation, typically the

lending party seeks to get consent from the borrowing country to conditions that will shift

26

economic policy in the directions that they think are desirable, while the borrower seeks to
consent on conditions only palatable to it. Now the question is, what affects their abilities
to deal with each other in their preferred way? In other words, where does the variance in
conditionality agreements come from, if both sides perceive each other as qualiﬁed to
negotiate and are doing so?

The factors that differentiate actors’ negotiation ability for conditionality
agreements can be generically dubbed leverage. This subject was ﬁrst explored by
Haggard and Kaufman (1989) in their early study of debt scheduling and later by Chan
(1993). They argued that the differences in debt rescheduling is best explained by
bargaining power of developing countries represented by three factors: the size of the
country’s debt, its strategic signiﬁcance, and its access to other nonconditional resources.
Larger debtor countries in the Southern Cone in a position to threaten the international
ﬁnancial system as a whole (largely through the potential of their unilateral action) and
capable of sustaining relatively autarkic economic measures, at least for a time, have
tended to receive lower interest rates, longer maturities, and a longer grace period in their
debt rescheduling than countries with relatively smaller debt burdens. Geography, history,
ideological affinity and alliance behavior of debtor countries have also played an important

role in determining creditor countries’ consideration of debt rescheduling.20 The

 

2" The importance of those factors in economic relationship has been enlisted in various ways. According
to Haggard and Kaufman (1989), the United States has been relatively more concerned with relief for
Latin American countries, Germany for central Europe and Turkey, France for its former colonies in
Africa, and the Great Britain for the indebted members of the Commonwealth. The beginning of the debt
crisis in 1982 coincided with the height of the second cold war, and accordingly. there has been a
relatively close association historically between favorable terms in international ﬁnancial negotiations and
the strategic relationship of the debtor country with the United States or its allies. Therefore, countries
with poor reputations for meeting performance criteria, such as Zaire or the Philippines, consistently
obtained ﬁnancing from the IMF throughout this period. Biersteker (1993) says that the cold war may be
over, but alliance behavior remains important nonetheless, as indicated by the generous debt relief

27

availability of nonconditional ﬁnancial resources (i.e., petroleum revenues or proceeds
from commodity price boom) enhances a debtor country’s bargaining position and the
ﬁnal terms it is likely to receive. Those additional resources render a country less
vulnerable to external pressure from the [HS because it will probably have fewer political
incentives to make a difficult economic adjustment. The credibility of a debtor country’s
bargaining position in international ﬁnancial negotiations is inﬂuenced by various types of
resources that enable it to sustain itself without immediate IMF support.

In addition to these factors, Haggard and Kaufman outline how a country’s
internal or domestic bargaining space can be crucial to understanding both the terms it
receives in international ﬁnancial negotiations and the different degree of implementation
of such terms. This argument is more clearly explicated with what Putnam (1988) refers to
as a classic “two-level game”: a situation of simultaneously negotiating on two different
fronts. Putnam says that the size of “win-set” or room to maneuver negotiation is allowed
by domestic constituencies. The size of the win-set is a function of the distribution of
power, preferences, and possible coalitions among domestic constituents, as well as the
political institutions and strategies of international negotiators. Therefore, a number of
different domestic factors can inﬂuence the outcome of international ﬁnancial negotiations,

including the degree of internal cohesion, the salience of the issue, the potential for issue

 

provided to Egypt following its involvement in the allied coalition against Iraq during Gulf War. This is
also supported by the other studies; Timothy McKeown and his colleagues ﬁnd out that ideological
affinity, which is measured by the nature of regime in developing countries and voting at UN approaching
the US position, affects the amount of aid extended and the degree of conditionality in that aid
relationship. See McKeown, Timothy, Leona Pallansch, and Strom Thacker. 1999. “Political
Conditionality in US. Bilateral and Multilateral Foreign Assistance.” paper presented at the 40th Annual
Convention of the International Studies Association, Washington, DC, February 16-20. Even the lending
behavior of intemational banks is affected by the political relationship between banks’ home country and a
borrowing country. See Jain, Arvind and Douglas Nigh. 1989. “Politics and the International Lending
Decision of Banks,” Journal of International Business Studies 20, no. 2 (Summer):349-359.

28

linkage, or the articulation of an indigenous alternative to orthodox IMF or World Bank
conditions.

In developing countries’ political setting, Putnam’s two-level game is quite
persuasive. Economic adjustment creates a dilemma for borrowing countries because the
state as an actor must bargain with two distinct groups whose interests are often opposed
to each another. On the one hand, the state needs to satisfy key domestic constituencies in
order to maintain political stability and its hold on power. On the other hand, the state
needs to convince its international creditors of continuing to provide capital or at least to
accept the periodic rescheduling of old debt service (Lehman, 1992). The conﬂict between
these two relationships has intensiﬁed as the IFI’s have increasingly attached strings to
their ﬁnancial support on economic policy reforms that would negatively affect the
material interests of those key domestic constituencies. The more a government agrees to
creditors’ demand, the greater risk it faces of political instability—but if it continues to
meet the demands of domestic constituencies, it runs the risk of losing the ﬁnancial
support from international creditors. Agreeing upon conditions that ignore a win-set
determined by the distribution of power among domestic constituents and political
institutions already conceives of possibility of sloppy implementation. Thus, by delivering
this to the bargaining table, whether with or without sincere intention to comply with the

program, recipient countries can adjust the level of toughness in conditionality.”

 

2' In two-level game, the opposite situation is also possible. That is, negotiating developing countries
make an excuse of external pressures to domestically deliver economic policy change. Kapur (1998)
argues that, in many cases, the IMF devised loan conditions at the behest of borrowers, whether local
oﬂicials who felt powerless to sway their political leaders or politicians who used the IMF to shield
themselves from popular rejection of policies that they too recognized as essential.

29

The factors that Haggard and Kauﬁnan enumerate to explain different bargaining
outcomes in debt restructuring are good examples of leverage. Unfortunately, they do not
explicitly discuss creditors’ leverage to keep things on their side, but this imbalance seems
to be related to their purpose to stress that developing countries fare relatively well thanks
to those leverages. In any event, Haggard and Kauﬁnan’s argument should be interpreted
as that each actor’s relative, not absolute, strength calibrated by leverage plays a key role

in bargaining process, and that many national characteristics and contextual factors of
developing countries have a legitimate potential to be leverage.

Summing up the state of the ﬁeld, many areas of conditionality have been
explored, and some of their ﬁndings were convincing enough to change IFIs’ lending
policies and content of structural adjustment. Nonetheless, the agreement process of
conditionality at the very beginning still remains unexplored or poorly charted if
explored.22 Although there has been plenty of evidence against the stereotyped view of
predominance by the IFI’s, they have not been orderly put together to bring a Copernican
revolution to the study of conditionality. Based on the above theoretical grounds,
conditionality agreements are neither inevitably coercive nor invariably compromising.
Rather, each agreement is struck somewhere on the continuum of absolute coercion to
absolute compromise, and the distance from the center toward either extreme will be

critical to determining the nature of the agreement. Furthermore, the distance canied from

___

22 The word bargaining or negotiation on conditionality appears in the text more often than expected, but
it stops there. Not many scholars showed a serious interest in developing the bargaining process of
conditionality. Biersteker (1993), Mosley (1987, 1988, 1995), Stiles (1990, 1991), and van de Geest
(1994) are those scholars who attempted to explore bargaining relationship between the IFIs and
developing countries in conditionality agreement, but the analytical tools they applied didn’t match their
research purpose very well in a strict sense.

30

the center is largely a function of strategy and players’ relative strength based on
mobilized resources. Now it is only a matter of applying a matching analytical tool to get

satisfactory results.

1.3 RESEARCH STRATEGY: How TO SOLVE PUZZLES

Haggard and Kaufman’s work gets closer to showing the nature of conditionality
agreements by incorporating the notion of leverage into debt rescheduling and identifying
various sources of bargaining power. Recently, Aggarwal (1996) completed a similar
work titled “Situational Theory” using game matrix and focusing on several cases in depth
over a long period. The notion of leverage will be extensively used in this dissertation, as
well. Although these literatures contributed to opening our eyes to the possibility of
bargaining between the IF 1’s and developing countries, they are limited in showing the big
picture of the subject they are studying. These studies describe each tree in the woods, but
do not give us the equipment with which to view the whole woods. While we have
accumulated enough case studies so far, we still lack a generalized framework under
which those case studies can be incorporated.

In addition to needing a generalized framework, the manner that they chose their
cases is called into a question. To begin with, we are not informed of why particular cases
should be studied and what their selection criteria were. The problem with their selection

process is that they focus exhaustively on the occurrences of an independent variable, here

31

leverage, and select exclusively those occurrences.23 The attainment of answers to the
questions they pose is already logically precluded by their method of selecting cases. By
examining all of those occasions in which the independent variables occurred, but only
those cases, these analysts are testing the questions of sufficiency (Most and Starr, 1989).
With that method, however, they cannot draw any conclusions about whether or not those
independent variables are necessary conditions for dependent variable.

Although my research design owes a lot to Haggard and Kaufman and to
Aggarwal in terms of conceptualization of bargaining and leverage, mine differs from
theirs in four points in addressing the puzzle of coercion or compromise in conditionality
agreement. First, I will explicitly use game theory. Of the approaches to model building
that one might take—including models of evolutionary processes, cognition, structural
imperatives, temporal imperatives—models of choice processes as part of goal-oriented
behavior show the most promise for clarifying the nature of conditionality agreement.
Although I set out with a bargaining model of conditionality, it does not necessarily mean
that the outcomes produced would tilt toward a particular nature. There is no
predisposition toward the nature of interaction and its outcomes. Rather, the logic and
evidence developed in this model will show that conditionality bargaining is open to a
wide range of outcomes, not doomed to produce only one type of outcome, and all of
them are results of strategic choice made by the players.

Relying on a game theoretic model to explain conditionality also has an additional

merit of methodological inclusiveness. Assetto (1988) and Stiles (1991) set up two or

 

23 Chan’s work (1993) should be excluded from this category, however. He conducts a single-point (1986)
cross-national analysis of the relationship between leverage and debt rescheduling of 87 developing
countries, using the terms of loans—interest payment, maturity, and grace period—as dependent variable.

32

 

three different models to explain differences in conditionality agreements. This inductive
method has a problem that it has to produce new models whenever odd cases turn up. In
the worst case, we will see islands of model spawning in the same number of individual
cases, which will make generalization impossible. But a game model that will be built in
this dissertation will provide a solution to such a problem. It is deductive and free from
particular cases at its construction stage, thus it is generalized. Moreover, what Assetto
and Stiles call cases of political or ﬁinctional model can be explained as one of the
outcomes produced by the game model based on the concept of leverage. That is, the type
of mobilized leverage and the degree of leveragedness determine whether those cases were
political or ﬁmctional. In this sense, the emerging game model is comprehensive and has
more explanatory power.

Second, whereas most studies of conditionality use only one or a small number of
cases that were selected before their theorizing, or selected because they ﬁt the theory (a
common violation of the principle of scientiﬁc research), my empirical assessments will be
derived independently from the theoretical structure. By using a wider range of cases (i.e.,
all the eligible conditionality agreements), I will focus on identifying necessary, not
suﬂicient, conditions for variance in the outcomes of conditionality bargaining, which will
be led to generalize the arguments and provide an analytical framework within which case
studies can be incorporated. For this purpose I deliberately concentrate on the
conditionality agreements made only with the IMF as eligible because IMF ’s conditionality
is designed to correct disequilibrium in balance of payments within a short period of time;
therefore, it applies a well-structured and clear-cut policy measures than that of the World

Bank. While these characteristics of IMF conditionality are more apt to preserve

33

continuity and inﬂexibility in its application than change and ﬂexibility, such characteristics
at the same time make it easy to notice variance in conditionality agreements, if it occurs,
and enable us to classify them and understand their nature better. My research is matching
the cases it considers with the questions it asks and the methods that it employs.

Third, while many studies of structural adjustment cover both bargaining and
implementation as one dimension, my research is strictly conﬁned to the initial bargaining
process. One reason that previous studies were not so successful in revealing the nature of
conditionality, even though they viewed conditionality from a bargaining perspective,
seems to have to do with combining and analyzing two different processes of different
contexts at one dimension (see Mosley, 1995, p. 76). Once a conditionality agreement is
contracted, however, the arena of interaction is changed from the international for
bargaining to the domestic for implementation, where actors, expectations and strategies
are different. Unlike at the bargaining table, where one voice through a uniﬁed channel is
quite expected, there is likely to be a considerable diffusion of power within the
government, and the department within the government negotiating the conditionality
agreement may well lack the authority to implement the agreement. Therefore, the analysis
of the two processes requires different reasoning and should remain separate. Even though
this fact was perceived, it was not well incorporated in the previous works. Bargaining
conditionality is one thing, and implementing it another.

Lastly, this study will pay a special attention to testing the validity of the model.
Let alone exploring the internal validity of the model persistently, its external validity will
also be pursued through a detailed case study. For the external validity’s sake, it is not

enough to identify general characteristics of conditionality bargaining through a game

34

model and supportive cross-national statistical analysis. For more satisfactory results,
going the extra mile with a case study is essential. The case chosen for this purpose is the
international bailout of the Republic of Korea in 1997. The Korean case is interesting
enough in terms of context and outcomes of bargaining. As a model of miraculous
economic development to be ranked at the 11th largest economy in the world, the ﬁnancial
disaster that forced South Korea to rush to the IMF for rescue, and the entire process that
deﬁned how the Korean crisis was handled, show vividly the dynamics working in
conditionality agreement in the era of extreme capital mobility. The examination of South
Korea’s recent experience with the IMF will concentrate on two things. The ﬁrst is
conﬁrming the ﬁtness of the theoretical model to a speciﬁc context. The presence of each
ingredient of the game model within the case and their inﬂuence on the outcome must be
the ultimate test of the external validity of the ediﬁce as a whole. Second, the distinctive
content of the agreement between the IMF and the Korean government, which was lost
during Operationalization for data analysis, will be brought to light with details. This is to
determine the effectiveness of the leverage that the Korean government could mobilize in

the bargaining and will also be compared with general statistical results.

1.4 EXPECTED ACHIEVEMENTS: WHAT WE WILL LEARN FROM THIS STUDY

This research seeks to obtain a better understanding of structural adjustment or
conditionality by tracking down the nature of conditionality agreement between the IMF
and developing countries. A game theoretic model combined with statistical analysis and a

case study will make it possible to test theoretical outcomes and validate them against real

35

world experience. When this research is successfully conducted, three academically and
practically meaningful achievements are expected.

First, this research will provide a decisive answer to some of the controversies
revolving around conditionality, which could not form consensus on the ﬁindamental
causes of failure and their remedy due to the lack of appropriate framework and evaluation
criteria to rely on. This dissertation will provide a general idea of the domestic and
international factors that inﬂuence conditionality agreements by opening up their
bargaining process, which has been put aside as black box so far. Based on this, we will be
able not only to determine the nature of conditionality agreements but further pinpoint
factors that may lead to success or failure of structural adjustment in individual cases. In
order to put study of conditionality implementation on the right track at the next stage, it
is necessary to ﬁgure out which of those factors is likely to have what inﬂuence on
domestic implementation, helping achieve or abandon the intended goals. As a preliminary
step, examining the correlation between the types of conditionality agreements and the
degree of implementation will be helpful, which is possible by classifying the existing
conditionality agreements to the outcomes produced by the game model.

Second, in practice, ﬁndings from my research will have implications to viable
policy alternatives to current lending practice of the IFl’s. Although my research is solely
focusing on the IMF, its framework and ﬁndings from it can be applied to all international
ﬁnancial institutions owing to its high level of generalization. As suggested by the
conception of this research, economic reform package and supportive international
development ﬁnancing are highly political decisions that are produced through

coordination and coalition of various forces operating at international and domestic level

36

 

from the stage of negotiation through implementation. Therefore, the IFl’s should take
factors other than economics into account when formulating lending policies and
supervising their progress, which will make a necessary condition for increasing the
probability that the purported goals of each lending will be achieved.

Recently, the notion of ‘ownership’ of reform, for instance, is gaining momentum
in international development community. The results of this research can be used as a
guide to map out under what conditions the ownership is created and what role the IFI’s
should play to nurture it. In this regard, this research will be providing an essential part in
seeking viable solutions to the problems of current conditionality and reorient the IFIs’
lending policies. By this merit, my dissertation will be in line with establishing measures to
use limited international resources with efficiency. International resources for development
are limited in the sense that most funds operated by the [FPS come mostly ﬁom tax payers
within limited number of industrialized countries; therefore, maximizing its efficiency and
productivity should be a part of the goals that international development ﬁnance tries to
achieve altogether.

Lastly, in addition to its contribution to a speciﬁc area of structural adjustment,
this study will contribute to the study of international bargaining in general, as well. My
research is equally concerned with theoretical development of the concept of leverage in
the bargaining process: how to deﬁne and measure leverage, under what condition
leverage works, and how effective it is. This is a sophisticated conceptualization of power
that demonstrates how power is translated into action in a causal process that leads to
outcomes, with emphasis on issue-speciﬁcity in international interaction. This research is

also an advanced scientiﬁc study of bargaining, focusing on systemic factors, rather than a

37

diplomatic account based on negotiators’ skill as in the past. In this regard, it deserves to
be called a progress.24 Therefore, the theoretical ﬁndings of leverage based on the
asymmetric relationship between international organizations and developing countries can

be applied to other forms of international bargaining with subtle modiﬁcations.

We learned ﬁ'om experiences that many things in life are path-dependent: previous
choices constrain next choices available. This quasi law of life may be applicable to
conditionality; that is, the way that adjustment is chosen is likely to set courses for the
remaining parts of it. Thus, without a correct understanding of the ﬁrst process that sets
the courses, any attempts to understand the rest could be easily disoriented and leave a
hollow hope for achieving the intended goal somehow. In this sense, the study of
conditionality may be a Pandora’s Box in academia. Curiosity opened the Box and many
good things, as well as bad, popped out instantly. But all of a sudden, fear intervened and
the Box was hastily closed at the very moment that hope, one of the vital things for life to
move forward, was about to come out. Having it kept in the Box, life was nothing but
miserable. As we see the state of the ﬁeld, almost all parts of conditionality but the
bargaining process are out of the Box and ﬂoating freely around us now. It is time to
reopen the Box and set the bridled part free. When it ﬂoats free with the other goods

already out, the ﬁeld will become complete and function better.

 

2" There are researches that describe the Fund’s ﬂexibility at interpersonal level, namely personal
involvement of Managing Director and mission officials with country ofﬁcials. Putting conditionality
bargaining under judgmental approach by technocrats is more likely to produce erratic and inconsistent
explanation for variance in bargaining outcomes. Thus, my dissertation attempts to examine the outcomes
of conditionality bargaining at a systemic level. For a comparison, refer to Stile, Kendall. 1991.
Negotiating Debt: The IMF Lending Process. Boulder, CO: Westview Press; and Martin, Matthew.
1991.The Crumbling Facade of African Debt Negotiations. New York: MacMillan.

38

CHAPTER 2

PRELUDE: THEORY OF BARGAIN IN G IN INTERNATIONAL RELATIONS

The necessity of using a game theoretic model for the study of conditionality rose
to an imperative in the previous chapter, with a view to clarifying the process and nature
of conditionality agreements to a higher standard of science. Before moving on to real
modeling, it will be worthwhile to conﬁrm the appropriateness of applying a game
theoretic bargaining model to conditionality agreements by reviewing general bargaining
theory in international relations. The purpose of this chapter is to survey various
components of the existing bargaining theories and extract from them contents useﬁil to
constructing a theory of leverage and conditionality bargaining model based on it in the

next chapter.

2.1 BARGAINING AS A WAY OF RESOLVING DIFFERENCES BETWEEN STATES

Resolution of international disputes can follow two paths, war or negotiation.
Negotiation as a means of resolving disputes and managing relations, has played an
important role in interstate relations for centuries from the beginning of modern
international system, and has become a primary activity of states with the advent of
massive destruction weapons since World War II. This is more so as the current
international system is getting closer to what Keohane and Nye ( 1989) called “complex
interdependence,” which is characterized by multi-layered transnational relations rather

than military competition and confrontation. For the larger part of the world, military

39

security is not at the top of a clear hierarchy of issues for government but just one of the
multiple and complex issues. Moreover, force as a counterpart measure to negotiation is
usable only at high cost. Thus, the fact that the utility of force declines and issues become
more equal in importance leads states to rely on negotiation more than ever in order to
wield power and resolve disputes.

What do we then legitimately expect from the word of negotiation? The most
basic deﬁnition of international negotiation showed up in one of the early systematic
studies of negotiation; Fred C. Iklé (1964) deﬁned negotiation as a process in which
explicit proposals are put forward ostensibly for the purpose of reaching an agreement on
an exchange or on the realization of a common interest where conﬂicting interests are
present. This deﬁnition denotes one of the fundamental features of negotiation called
mixed motives in negotiation. Interests of two parties involved conﬂict with each other,
but they also share a common interest in solving their problem through negotiation rather
than leaving it unsolved (Brady 1991). Relations between any two parties or more may
fall on a continuum of completely identical interests to totally incompatible. At one end
of this continuum, the two parties’ interests are in harmony so that they need not
coordinate their interests through negotiation; even if each party behaves as it prefers, this
does not hinder the other from achieving its goals. At the other end of the continuum, on
the other hand, the interests of the two parties cannot be realized simultaneously, and
thus, the dispute will almost inevitably be resolved by force or coercion rather than by
diplomacy and negotiation. Virtually all of international relations fall on this continuum,

perhaps at times tending toward one or the other end of the continuum, but most times

40

somewhere in-between. It is within this middle territory that Iklé’s essential condition for
negotiation is met.

The second feature of any deﬁnition of negotiation is that it entails a situation of
interdependent decision making, where two parties make decisions and thereby the
outcome of the negotiation is not exclusively under their control. The result of their
interaction will in part be inﬂuenced and constrained by the action of the other as much
as by that of one. In such joint decision-making, it is no longer possible for any party to
optimize the outcome alone. However, it is usually possible to create conditions in which
one’s decision will produce a mutually acceptable outcome that will at least serve the
fundamental interests of all of the parties affected by the decision (Hopmann, 1996).
With two parties simultaneously participating in decision-making, negotiation becomes a
dynamic or moving event, not simply a static situation, and an event concerning the
selection of a single value out of many for implementation and action (Zartman, 1976).

One thing to be noted about the theory of negotiation is that scholars of international
relations use the terms “negotiation” and “bargaining” interchangeably without any
hesitation. To draw a ﬁne line between them, however, John Cross (1969) deﬁned them
respectively: the term bargaining will refer to the process of demand formation and revision
which provides the basic mechanism whereby the parties converge toward an agreement,
while negotiation will refer to the whole situation within which bargaining occurs. It is clear
that Cross portrays negotiation as a more comprehensive concept, within which bargaining
forms the nucleus. This dissertation is concerned not only with bargaining per se, but also
with the whole negotiation procedure by Cross’ deﬁnition. It will keep balance between the

narrow bargaining aspect of conditionality and a broader range of contextual elements

41

affecting the magnitude of the parameters of bargaining. Nevertheless, these two words will

be used interchangeably in this dissertation as well.

2.1.1 Game Theory and Modeling International Negotiations

In light of the essential nature of negotiation as a process of joint decision-making
that combines conﬂicting positions into a common outcome, game theory provides a
good logical underpinning of interdependent decision making in any situation that
involves at least some conﬂicting interests between two or more parties. The logical
structure of game theory is independent of any concrete content and helps visualize the
underlying principles without being distracted by specifics of particular situations.
Rather, it organizes them in a systematic way so as to begin with the most basic choices
that an analyst must make, namely starting with the simplest assumptions that may
provide the most general explanations, and then to move along to more reﬁned choices
and complex analyses that may offer a more differentiated explanation of the process in
particular categories of negotiations (Hopmann, 1996). Thus, the analysis of negotiation
using game theory enables us to go beyond a simple reductionist typology of negotiations
or ad hoc list of factors that might account for the behavior of actors in international
negotiations.

The basic bargaining model in game theory is a simple representation of two-

party negotiation.l These two players are self-interested and rational in the sense that they

 

' Game theoretical approaches to bargaining more or less assume an agreement given the game, and
attempt to predict the nature of that agreement. It should be noted, however, that while some sort of
agreement is inevitable in some context, there are many social contexts where reaching any agreement is
quite difficult.

42

have goals and attempt to realize those goals through their strategic actions. A utility is a
summary measure of all of the costs and beneﬁts that may be collectively associated with
a particular outcome. Actors’ preferences over the outcomes are formed in accordance
with the costs and benefits associated with each alternative and how the other player will
respond to their decision. The players respond to the circumstances by making choices of
strategies that they believe maximize their expected utility.2

The outcomes of interaction are structured in two settings, zero-sum and nonzero-
sum. In zero-sum games, players’ payoffs are always inversed of each other, and hence,
they have no interest in communicating or coordinating their strategies for mutual
beneﬁts; no negotiated solutions are possible that can beneﬁt both parties simultaneously.
The nature of this negotiation process is so competitive that each party seeks to maximize
his relative share in the context of ﬁxed sum payoffs, whatever value is being divided and
to protect themselves from being exploited by the other (Friman, 1993). The structure of
the game itself determines outcomes for the players.

On the other hand, nonzero-sum game represents a situation in which one player’s
gain is not automatically the other’s loss. The players have both competitive and
cooperative interests and may ﬁnd outcomes that are preferred by both of them to others.
Since there are generally no determinate solutions in this type of game, the players may
try to communicate joint strategies that will improve payoffs for both, compared to
unilaterally determined strategies. In this sense, nonzero-sum game captures the mixed
motives in negotiation better and highlights more cooperative efforts to enlarge the

simultaneous joint gain of both parties in otherwise conﬂictual situation (Morrow, 1994).

 

2 The reason that utility is expected, not decided in advance, is explained later in this chapter.

43

Unilateral strategies, however, are still usable even in nonzero-sum bargaining,
because even in this situation one can always compare negotiated outcomes with what
one could obtain by playing the game as if it were a zero-sum game in which one plays
one’s best unilateral strategy (Hopmann, 1996). The players will only accept resulting
agreements that leave both of them better off than they would be in the absence of an
agreement. This is due to the fact that players can act only in anticipation of their
opponent’s actions, but cannot precommit themselves to one or another course of action
—a typical situation in international relations. This inability to make precommitments
springs from a fundamental of international relations that there is no higher authority to
enforce agreements between states, although this condition does not preclude cooperation
among self-interested states (Bueno de Mesquita and Lalman, 1992).

The above structures of a game can be reorganized around two major paradigms
in international negotiation: bargaining and problem-solving. These distinctions and the
implications drawn from them are also in parallel in many important respects to realism
and liberalism that have divided the larger bodies of general theory about international
cooperation.3 Indeed, there is a strong correspondence between the central concepts of
bargaining theory and realist approach to larger international relations theory, whilst the
problem-solving paradigm is closely linked to the liberal tradition toward international
relations in basic assumptions and concepts (Hopmann, 1995). The most conspicuous
difference between them, which is relevant to bargaining, is the perception of gains,

whether they are relative or absolute: the pursuit of gains in international negotiations

 

3 For a critical review of contrasting theoretical traditions of realism and liberalism in international relations
in terms of the possibility of cooperation, see Grieco, Joseph M. 1988 “Anarchy and the Limits of
Cooperation: A Realist Critique of the Newest Liberal Institutionalism.” International Organization 42, no.
3 (Summer):485-507.

44

could be at the expense of the other in a largely zero-sum competition or joint gains
through cooperative search.

In realism, national interests are charted as ﬁxed preference and the magnitude of
demand in negotiation is endogenous to the structure of the game so that it is selected to
maximize expected utility (Bueno de Mesquita and Lalman, 1992). Therefore, states will
even forgo agreements that will produce beneﬁts greater than the status quo or their next
best alternative to an agreement, if their potential competitors were perceived to be
gaining more than they are from that agreement. It is perceived as a strategic
disadvantage that could do serious harm to one over the long run if the other gains more
relative to one, even if both are absolutely better off (Grieco, 1993).4 In realist
perspective the goal of international negotiations is winning them at the expense of the
other by remaining ﬁrm, while the other should be ﬂexible and offer compromises.

By contrast, a problem-solving approach to international negotiations is generally
associated with a more liberal or institutionalist stance on international relations theory.
Reﬂecting the changing nature of post-Cold War world, liberalists have focused on
interdependence and contributed to a substantial reconceptualization of the international
relations and negotiation process. To liberalists, the most fundamental purpose of

negotiation is solving common problems that parties face and ﬁnding solutions to those

 

4 This statement could be controversial in light of recent developments in international relations theory,
especially regarding state’s preferences over relative and absolute gains and cooperation. For example,
Snidal proved that even if states stick to relative gains, its impact on cooperation dramatically decreases as
the number of actors involved in interaction increases. On the other hand, Powell proved formally that it is
not states’ preferences over the type of gains that hinders cooperation among nations but international
environment, namely, anarchy that lacks a central authority to enforce contract. Nonetheless, this statement
is given here for the purpose of caricaturing conﬂicting theoretical traditions as reference at the expense of
details. See Snidal, Duncan. 1991. “Relative Gains and the Pattern of International Cooperation.” American
Political Science Review 85, no. 3 (September):701-726; and Powell, Robert. 1991. “Absolute and Relative
Gains in International Relations Theory.” American Political Science Review 85, no. 4 (December): 1303-
1320.

45

problems that will optimize mutual gains for both parties (Hopmann, 1995). It also
assumes that a state can accept and implement agreements that serve its interests, even if
they also serve the interests of other states at the same time. This is thanks to
international institutions, formal and informal, that create a set of norms and rules within
which negotiations take place and facilitate cooperation for joint gains. Keohane (1983;
1984) argues for the positive role of international regimes, saying that their most
important ﬁinction is to facilitate negotiations leading to mutually beneﬁcial agreements
among self-interested states by reducing uncertainty and transaction costs and allowing
states to redeﬁne their national interests.

The choice of bargaining tactics, too, is shaped in accordance with players’
selection of a bargaining framework. Given the emphasis on power within its tradition,
the realist prescription for negotiation has tended to emphasize the tactics of manipulation
such as commitments, threats, and promises, which are likely to win a larger share of
good for one party (Hopmann, 1995). States make commitments to reinforce their offers
and demands; they manipulate information to disguise their true preferences and their
actual alternatives to agreements in order to gain settlements that are unilaterally
favorable to their own interests; they issue threats and promises concerning punishments
and rewards, which are in turn made credible by demonstrating that the states have
sufficient (military) capabilities to carry them out. However, this tactic concomitantly
entails a risk of stalemate or sub-optimal solution, if the opponent is similarly motivated
and then both parties become engaged in a contest of wills. This, Grieco argues (1993), is
the risk inherent in avoiding exploitation at the hands of the other by the search to expand

benefits with a potential competitor.

46

On the other hand, to the liberalists, negotiation is increasingly viewed as a tool to
resolve conﬂicts in such a way that mutual beneﬁts for the parties are produced rather
than exclusive beneﬁts for one at the expense of the other. Therefore, as long as a
mutually satisfactory agreement is possible, both parties have an interest in avoiding
tactics or other mistakes during the bargaining process that will prevent them from
achieving an agreement (Walton and McKersie, 1991). Moreover, the players will apply
tactics that will facilitate reaching an agreement such as issue linkage and side-payment.
For them, it is not just the outcome of the negotiation that is important but also the
improved nature of the long-term relationship between the parties that results from
achieving mutually beneﬁcial agreements (Fisher, 1987). It serves their long-term
interests because agreements based on mutual beneﬁts have more potential to endure,
while agreements that may undermine the fundamental interests of at least one of the
parties are likely to be violated and eventually dissolve.

With the above bargaining strategies in mind, how do the parties then reach an
agreement as outcome? The primary problem for negotiating states is to identify the
approximate range of bargaining space within which mutually acceptable agreement can
be realized. Each player may enter into a negotiation with an absolute bottom line set
according to some abstract criteria of gains and losses. This bottom line is not set
arbitrarily, but rather determined by identifying a point at which agreement would no
longer be beneﬁcial for the party. Fisher and Ury (1991) refer to this as a resistance
(security) point or BATNA—the Best Alternative To a Negotiated Agreement—that
represents the point of minimum acceptable agreement. The BATNA is a minimum both

for the amount of utility to be satisﬁed and the degree of success to be individually

47

judged in negotiation. In game theory players can determine their BATNA by playing the
game as if it were a zero-sum game and by calculating their minmax strategy. No matter
what method of calculation may be applied, however, the principle is that negotiators
should determine their resistance points by comparing the value of agreement at any
stage with the value of no agreement. And the criteria for an acceptable agreement within
this ﬁ'amework should be coming out ahead compared to one’s own alternatives, not
compared to the other party.

The logic of BATNA gives us criteria to evaluate what is often referred to as
status quo and an outcome of no agreement from bargaining. These two points can
undoubtedly be considered at BATNA since they are calculated on the basis of the
amount of losses that each player will suffer from noncooperative outcomes and will
presumably lead each party to pursue its next best available options. A negotiator must
always be conscious of what beneﬁts or harm he will receive in the absence of an
agreement, and should never be pressured or coerced into an agreement that will leave
him worse off than would be the case if he walked out of the negotiation and pursued the
next best alternative to the negotiated agreement (Hopmann, 1996). Failure to agree may
not necessarily be very costly if there is a perfectly satisfactory alternative to rupture.
This means that any failure to reach agreement will not be treated as a defect of the
international political market, nor as breakdown of rationality in a political process in
which states’ interests collide and shift (Blair, 1993).

As long as the resistance points of the parties are not mutually exclusive, there
should exist a range of acceptable agreements between them—the bargaining space. Once

the bargaining space is located, the task is to ﬁnd a precise point of agreement that gives

48

it a larger share of the proﬁt than the others do. Thus, it may be rational for each party to
try to move the agreement along issue dimensions toward positions that will maximize its
interests, while engaged in making concessions until the positions of the parties converge
somewhere within that space. Such a process of locating a point of agreement is
described as concession-convergence process; from the initial positions on issues do the
parties converge to a jointly acceptable outcome based on a process of mutual
concession. Concession-convergence process in negotiation posits that the parties start at
some point of stalemate and approach an outcome through an action-reaction process of
responding to each other’s concessional behavior (Bartos, 1977; Cross, 1969).5 This is
also similar to what Zartman and Berman (1982) call formula, a conscious attempt to
arrive at a joint deﬁnition of problem and solution and to combine both parties’ relevant
positions into a common justiﬁcation for terms of trade as the basis of agreement. Here
the provision of trade-offs forms the essential part of the creation of a formula. For
negotiation is less a matter of making a decision than of putting together a package of
counterbalancing payoffs and establishing the terms of trade among them in order to
make a positive decision possible (Zartman, 1987b). No matter what we call the process,
however, the ﬁnal landing point of negotiation still seems to be dominated by the mixed
motives in negotiations, the simultaneous presence of a common interest in reaching an
agreement within the available bargaining space and a conﬂicting interest in achieving an

unbalanced agreement within that space that favors one’s own position.

 

5 Concession/convergence theory of negotiation even assumes that actors will behave in a predictable way,
and thus, the outcome can be predicted once the bargainer’s concession rates are known. However, this
negotiation theory does not provide a clear explanation about what makes actors concede, what determines
the rate of their concession, and how it affects the process of converging toward outcome. The only way to
deal with this is bringing the concept of power by the back door, which will be seen later.

49

As the negotiation proceeds through concession-conversion, the ﬁnal agreements
rarely leave one party with all of its demands satisﬁed, or the other with no satisfaction,
since there is always some element of compromise involved. Each side avails itself of
some inﬂuence in negotiation to elicit compromise out of the other. Thus, the outcome of
such compromise is not known for certain and even comprises risky components as
lotteries. Negotiations are risky events in that one can expect to obtain something
between one’s own demands and the demands made by the opponent (Bueno de Mesquita
and Lalman, 1992). In negotiation neither participant anticipates ﬁrlly satisfying his or
her initial demand. This is why the utility of the negotiation is expected, not
predetermined; the utility is calculated as the value gained if one’s demands are won,
weighted by the subjective probability of winning, and again summed with the value of
the opponent’s demands, which is weighted by the subjective probability of losing.

Once bargaining is ﬁnished, it is natural to ask whether the negotiation was
successful. What criteria are used to judge success in negotiation after all? A simple sign
of success is the emergence of settlements that two or more sides are willing to sign.
However, this should not be read to indicate that the settlement terms were or were not
the best available alternative to status quo. Although nothing can prove that better
alternative could have been negotiated, neither does agreement on one set of terms prove
that better (or worse) agreements were not possible (Zartman, 1987b). Furthermore, the
emerging set of terms says nothing about mutual satisfaction and equal better-off. Pareto-
optimal criteria do indicate that the parties should be able to improve their positions, with
or without depriving the other parties of their advantages, but insisting that the parties

beneﬁt equally from the outcome may be imposing too stringent a condition for success

50

(Zartman, 1987a). Also, the very nature of game theory does not allow interpersonal
comparison, and success must be evaluated against the problem, not against the
adversary. Players act rationally on the basis of their preferences, which are exogenously
given. It is possible to trace back to the rationality presumed to exist through reasoning,
but not possible to judge the appropriateness of rationality itself through comparison.
This nature of bargaining game makes evaluation of success a matter of subjectivity,
where universal rule of judgements is not available.

A second criterion for success is the degree of compliance by the signatories with
negotiated agreements. Agreements that are essentially one-sided cannot endure
indeﬁnitely. This is due to the fact that the party that has been coerced into accepting an
agreement against its interest will resent that agreement and the other party that imposed
it. The coerced party will have every incentive to avoid, fail to implement, or even cheat
outright on such agreements. However, it is also noted that compliance to agreements
comes down to the domestic level from the international, which is engaged in different
dynamics among concerned players. The causes of noncompliance cannot solely be

ascribed to the process of negotiation.

2.2 FACTORS AFFECTING BARGAINING OUTCOMES

Negotiation outcome is a result of causal process. The analysis of negotiation is
the analysis of outcomes, speciﬁcally, an analysis of what causes a particular outcome
instead of any other (Zartman, 1976). A causal deﬁnition of negotiation focuses attention

on the process of change and the end result of change, namely, an understanding of

51

operating process and investigation of factors that seem to have contributed most to the
attainment of agreement between parties involved. With that regard, notwithstanding that
game theory is a useﬁrl framework for describing how states actually negotiate or for
explaining speciﬁc outcomes of international negotiations, still a good deal of the actual
behavior of negotiators cannot be adequately explained only by axiomatic foundations of
game theory (Rapopport, 1960). Within the bargaining space, many other variables that
are not taken into account by the game theory can also be influential in determining
actual outcomes. In other words, the outcome of a particular negotiation can equally
depend on other factors than mere calculations of the most rational, utility-maximizing
strategy for responding to circumstances. Thus, we need to expand our focus on to those
factors to understand better when and why theoretical expectations of game model will be
realized in any particular international negotiation.

As one of the extended foci, examination of the process of interest change appears
qualiﬁed. Transforming initial conﬂicting positions into compatible ones is the same with
changing players’ interests; the change in interest is brought about by affecting the
negotiation process, which is possible by developing alternative agreements in more
favorable terms to the players. Increasing chances for more favorable terms may be
owing to various conditions exogenous to the bargaining structure that can assist states to
strike agreements in their preferred terms. Therefore, if an agreement is reached, certain
inﬂuence can have been exercised on the players: one or both players have changed their
positions during negotiation, presumably because of certain actions by the other player.

In a similar vein, whereas it is true that opportunities for joint gains often exist, it

is equally true that even mutually beneﬁcial agreements bring greater gains for one party

52

than for the other. The point of agreement is rarely halfway between each party’s
position, but tends to be closer to the preferences of a certain party. Even when
negotiations are aimed at creating a bigger pie for all, the question inevitably arises as to
how much of the bigger pie each side will get, and how much each side should pay to
increase the size of the pie (Blair, 1993). A corollary question to this situation is what
makes one party concede more than the other, what efforts each party deploys to obtain
an outcome favorable to one but at the same time attractive enough for the other to shift
away from its unilateral solution through interest change and to accept the agreement.
Answering these questions necessitates injecting exogenous dynamics into the
negotiation process, namely bargaining strength.

It is important to note that in a search for such bargaining strength, a commonly
cited unit-level variable, bargaining skills, is ruled out. Good diplomacy alone will not
guarantee success in achieving one’s national objectives in a world still characterized by
a great deal of anarchy, although it is a necessary condition for being able to advance
one’s interests (Brady, 1991). The successful negotiation of international agreements
ultimately depends on far more than negotiators with an impressive repertoire of
international skills. Moreover, judging the level of skill in bargaining is a highly
subjective matter, and settling on bargaining skill as an explanatory variable is likely to
derail the research into “analyzing agent rather than the process” (Zartman, 1977).
Bargaining skill is not a factor that can be easily ignored, but a means to measure its level
and inﬂuence should be developed ﬁrst so that it will not be used as an ad hoc
rationalization of outcomes. Even if there were a widely accepted standard of bargaining

skill, assessing the skill manifested in negotiations would require an intimate knowledge

53

 

of the bargaining process, which is neither presently available for this study nor ﬁts the
purpose to generalize necessary conditions for conditionality bargaining. Therefore, this

factor is not considered.

2.2.1 Bargaining Tactics: Commitment, Linkage, and Side Payment

Negotiation tactics are a means of communication; they are a means of
communicating preferences to each other. Tactics are judged as successful when they
alter or modify the interests or alternatives of the other player within the relationship.
However, the elements of mixed motivation in negotiation create a dilemma for
negotiators regarding the tactics that they will employ.

As a starting point, commitment to seek disproportionately large gains in the
distributive phase of negotiations is considered. Commitment, as ‘the power to bind
oneself’(Schelling, 1960), refers to the extent and degree to which a player desires and/or
needs its preferred outcome. Commitment is based on the values that parties attach to
various possible outcomes. This tactic, however, is apt to pave a way to the classical
paradox of Toughness Dilemma: the tougher (more unyielding) a party acts, the greater
its chances for an agreement to its position but the greater the chances for no agreement
at all, whereas the softer (more yielding) a party acts, the greater the chances for an
agreement but the less chances for a favorable one. Thus, the tactic of commitment is
haunted by the dilemma that winning a larger share of good for one party may create a
stalemate and thus detract ﬁom the ability of the parties even to reach an agreement.

Schelling (1960) observes on this tactic that they all run the risk of establishing an

54

immovable position that goes beyond the ability of the other to concede, and thereby
provokes the likelihood of stalemate or breakdown. Unyielding confrontationism can be
unproductive; holding ﬁrm on a one-sided compromise can work to produce an
agreement only when it is done by the stronger side—the side with less to lose in
breakdown—and when the weaker side is not prepared to live with its losses (Zartman,
1987b).

In moving the other player away from its committed position, one can think of a
tactic that a state threatens to take a certain action that would impose signiﬁcant costs on
the other state. In that case, the interest in the issue being negotiated could be overridden
by the interest in avoiding the threatened action (Blair, 1993). Any agreement desired by
the threatening party can be contracted since the cost of non-agreement has gone up due
to the threat; outcomes that were previously only marginally beneﬁcial or perhaps even
negative may now become much more positive relative to the more costly condition of
non-agreement since beneﬁts must always be compared with the value associated with
non-agreement (Hopmann, 1996). The tactic of threat works under the condition of
unequal loss from non-agreement and difference in the ability to cope with it, and
consequently, produces asymmetric outcomes. It is so because the party that stands to
lose less from the failure to agree can threaten the other with an outcome of non-
agreement more credibly and, thus, be in a better position to demand a greater share of
the gain from cooperative outcomes. If both players have equal losses from non-
agreement, then the anticipated compromise outcome of negotiation will be akin to “split

the difference” (Rubinstein, 1982), which is equivalent to the ﬁfty-percent solution. This

55

 

tactic of threat is usually accompanied by a promise of punishment for not consenting
and/or reward afterwards. This is probably better characterized as coercion.

While the strategies of commitment and threat pertain to distributive bargaining,
we can think of two counterparts from problem-solving that deliberately manipulate the
external environment of negotiation. Tactics ﬁ'om the problem-solving perspective are to
resolve distributional disputes by enabling actors to rationalize backing down from prior
commitments (Walton and McKersie, 1991). First, if bargaining space does not exist,
players may try to create one through linkage among nonnegotiable issues if treated
separately but that may produce a negotiable package of issues when combined. An
agreement becomes acceptable to players contingent upon changes in the behavior of the
other with regard to issue having no direct bearing on the negotiation (Friman, 1993).
Linkage is an action of creating trade-offs, in which issues are linked together to create a
mutually beneﬁcial agreement as formula and enhance the prospects for solutions to
disputes. This tactic is more likely to be sought when the problem under negotiation is
complex and involves multiple dimensions, issues or parties to the negotiation, or when
the issue involves very ﬁindamental interests or beliefs that are not readily amenable to
solution by compromise, easily suggesting no presence of bargaining space along a well-
deﬁned dimension (Hopmann, 1996). Also, this is a tactic to make use of instruments or
issues where its bargaining position is relatively strong in order to promote or defend
interests where it is weak (Keohane and Nye, 1989). The effectiveness of linkage tactic in

facilitating agreement depends on the degree of changing the relative costs and beneﬁts

56

associated with agreement versus their BATNA.6 Thus, linkage facilitates bargaining by
offering “more potential quid for the quo” (Keohane, 1984).

Together with the tactic of issue linkage, side-payment refers to a compensatory
measure aimed at facilitating agreement between players by roughly balancing inequities
arising from cooperation. When negotiation outcome is intended to be distributive, side-
payments to make up difference in concessions can offer a means to generate more
egalitarian agreements.7 Hence, side-payment is a tactic that can facilitate a shift toward
a more positive-sum bargaining. This compensatory tactic can take a number of forms
from direct proffer of monetary reward to non-monetary concessions on other issues.8
Furthermore, side-payment is in a complementary relationship with the linkage tactic.
Since issue linkage changes the nature of negotiation from distributive to cooperative by
joining issues for bargaining purposes, offering side-payments is an integral step in order

for issue linkage to take place successfully (Keohane, 1984).

 

‘5 In the text, examination of conditions under which each tactic will succeed and fail is almost skipped
simply to Iirrrit the scope of discussion. As brieﬂy discussed for the tactics of commitment and threat, the
issue linkage tactic also has a possibility of failure and such cases are well compiled by scholars who are
against this tactic. For such topics, see Morgan, T. Clifton. 1990. “Issue Linkages in International Crisis
Bargaining.” American Journal of Political Science 34, no. 2 (May):3l 1-333; and Tollison, Robert D. and
Thomas D. Willett. 1979. “An Economic Theory of Mutually Advantageous Issue Linkages in International
Negotiations.” International Organization 33, no. 4 (Autumn):425-449.

7 The usage of side-payments as a bargaining tactic is not restricted to international bargaining. It can be
used domestically. For domestic use, it is to garner support when faced with other governmental officials or
societal interest groups blocking ratiﬁcation of international agreements. which comes to complement two-
level games by Putnam (1988). As a theoretical discussion, see Mayer, Frederick. 1992. “Managing
Domestic Differences in International Negotiations: the Strategic Use of lntemal Side-payments.”
International Organization 46, no. 4 (Autumn):793-818; and for an application, Friman, H. Richard 1993.
“Side-payments versus Security Cards: Domestic Bargaining Tactics in International Economic
Negotiations.” International Organization 47, no. 34 (Surnmer):379-410.

8 For a detailed typology of side-payments. see Pruitt, Dean G. 1981. Negotiation Behavior. New York:
Acaderrric Press, pp. 143, 157—60.

57

However, the conditions under which policymakers are more or less likely to
attempt bargaining tactics and choose particular tactics remain underspeciﬁed (Friman,
1993). While the nature of tactics does not answer the question of when to shift, it is
easily expected that confrontation is the parties’ initial positions and they will move at
some point toward a more ﬂexible and creative behavior to construct a mutually
satisfactory agreement. It is in this respect that politics of negotiation draws our attention.
So far, the discussion of negotiation has proceeded as if international negotiations took
place in a political vacuum. To the contrary, international negotiations constitute a
subsystem of larger international relations, and the employment of negotiation tactics to
resolve conﬂicts and its outcome depend very much on the state of relationship between
parties within the international context, which makes negotiation a political process.

The context of negotiation is based on the parties’ perception of “relationship,”
that is, “how negotiators visualize their interdependence, the type of commitment the
negotiator has made to the other party and the degree of indebtedness he or she feels”
(Brady, 1991). Most international negotiations are not one-shot affairs but occur in the
context of a preexisting relationship that will be maintained over a long period. Thus, the
history of interaction between the parties and expectations about the nature and value of
continued relationship inﬂuence negotiations. Negotiators make decisions about
strategies and tactics based on previous experience with their negotiating partners. In turn
negotiations not only lead to speciﬁc agreements on issues at hand but also ﬁrnction
progressively to shape attitudes and policies toward ﬁiture relationship.

Thus, political considerations may even override so-called rational decision-

making based on the structure of the negotiation. Viewed from this perspective, the

58

process of bargaining represents not only the convergence of positions but also the
deliberate act of political compromise. Politics inﬂuences decision of when to enter
negotiations; politics shapes negotiating positions; politics enters into the formulation of
interests and objectives; and politics often inﬂuences outcomes after international
agreements are signed (Ibid.). This explains why states under similar conditions have
reaped different outcomes from negotiation with the same counterpart. In short, politics is

the critical dimension of international negotiation.

2.2.2 Resources: National Power in International Negotiation

Incorporating the concluding remarks of the previous section, international
negotiation takes place not in a vacuum but within the international system where
individual states are its components. While denying the existence of higher authority over
them, states are differently placed by their power, and differences in placement explain
their behavior (Waltz, 1979). The difference in power is of special explanatory
importance in such a self-help political system. For the same reason, international
negotiation cannot be separated from power. As noted above, negotiation is a process of
value creation, reﬂecting changes that have occurred to actors’ positions, expectations,
and goals. These changes seem to be caused largely by the tactics mentioned above and
by the existence and use of power for that purpose. The unequal distribution of power
across states in international system results in different abilities to employ the negotiating
tactics successfully. Tactics ﬂow from conscious or unconscious deliberation in which

power is estimated, options assessed, and consequences anticipated (Lawler, 1992). As a

59

conditional statement, states’ behaviors in negotiations and consequent outcomes depend
heavily upon power with which to implement chosen tactics and the credibility perceived
by the other state. Thus, power serves as the foundation of a state’s bargaining strength.
Power lies between basic causal variables and results.

Power is exogenous to the structure of bargaining game. When power is inserted
into it, however, the outcome of the bargaining is no longer predetermined by the
structure of the game; rather it is directly inﬂuenced by the (potential) application of
power to manipulate the positions of the other party. This is why mere emphasis on
negotiator’s bargaining skill to draw preferred outcomes is out of the general context
where international negotiation constitutes a subsystem and why isolated focus on moves
and tactics without looking at their sources or underlying capabilities leaves us with
hollow explanations. Thus, this section examines the conceptualization of power in
international negotiation and the way it affects bargaining outcomes.

To start, noting a persistent theoretical problem with the concept of power is
proper. The concept of power in international relations has been largely illusive.
Sometimes power was portrayed as national interest; at others, it was designated as the
one and only means to achieve national goals.9 Nevertheless, it is inevitable to use the
concept of power in research. No matter how power may be deﬁned, a guideline to
follow is that the concept of power should be not only sufficiently comprehensive but
also purposely relevant to the subject in question, if it is to illuminate what happens in the

causal process that leads to bargaining outcomes. An easy way to meet this requirement

 

9 A representative of example of such an ambiguous use of the power concept is found in Morgenthau.
Hans. 1973. Politics among Nations, 5‘h ed. New York: Alfred A. Knopf.

60

may be to break the concept of power down into its various components and then utilize
the most relevant one in explaining international negotiation.

Power is deﬁned as the way in which actor A uses its resources in a process with
actor B so as to bring about changes that cause preferred outcomes in its relationship with
B (Habeeb, 1988). This conception of power refers to two elements of power, namely,
resources that the states to the negotiation may hold, and their ability to exert inﬂuence
on the other through the process of interaction. For the part of resources, state’s structural
capabilities are usually considered—states’ potential and actual capabilities and thereby
determined relative position vis-a-vis others in the international system. Structural
capabilities are a measure of a state’s total resources and possessions that states may
utilize in various ways in international relations.10 Thus, power is the result of having
resources.

There are of course many elements that make up the potential and actual
capabilities of states. In the past, realist writings on international politics have tended to
stress military capabilities almost exclusively, although more recently a much greater
emphasis has been placed on economic capabilities by so-called neorealists (Waltz,
1979). Regardless of their composition, however, the distribution of capabilities provides
not only a picture of the structure of international system but also that of states’ behaviors
constrained by that structure. Hence, the global hierarchy based on state capabilities was
considered to be an essential factor for the analysis of outcomes in international

negotiation. For often in negotiations, states’ demands and expectations are determined in

 

'0 Refer to Cline, Ray. 1975. World Power Assessment. Washington, DC: Georgetown Center for Strategic
and International Studies; Knorr, Klaus. 1975. The Power of Nations. New York: Basic Books; and
Organski, A.F.K and Jacek Kugler. 1980. The War Ledger. Chicago. The Univ. of Chicago Press for the
assessments of structural power.

61

part by their positions within the hierarchy and corresponding preferences, even before
they are directly communicated by any behavior; states usually determine their opening
positions, in part, on the basis of evaluation of each other’s preferred outcome and
possible response (Hopmann, 1996). Thus, structural capabilities operate as a background
factor that implicitly affects the behavior of states as they negotiate, even without being
explicitly mentioned.

However, possession or existence of capabilities alone does not cause outcomes.
They should be used to create outcomes. The capability—based deﬁnition of power says
nothing about the interaction of actors. A concept of power relevant to negotiation must
contain intentionality and anticipated reactions. States’ capabilities are meaningful only
when states can explicitly utilize them in order to manipulate the negotiation process.“
Power measured in terms of (potential) resources may look different from power
measured in terms of inﬂuence over outcomes. This is why we must also look at the
‘translation’ of power in the bargaining process (Keohane and Nye, 1989). This
translation of power is what is known as behavioral power, which is related to the
‘ability’ of the two elements of power.

As the name implies, behavioral power is concerned with the exercise of power:
the process by which they maneuver and use their resources to achieve preferred
outcomes (Habeeb, 1988). It is through inﬂuencing behavior that capabilities are
converted ﬁom static attributes of states into dynamic instruments that affect interactions
in international negotiations. In negotiation, behavioral power is revealed by states’

tactics to support their interest relative to the others’. The structural capabilities,

 

” The best-known exarrrple of non-performing capabilities is Vietnam War. Although the Unites States
possessed preponderant military capabilities, including nuclear arsenals, over Vietrurm, it could not win.

62

channeled through tactics, can directly serve to modify other states’ interests. Thus,
actual power in negotiations is a combination of the structural power, actor’s position at
the commencement of the process, and the dynamic or behavioral dimension, namely,
changing positions of the counterpart during the course of negotiation.

A problem with the above conception, notwithstanding, is that power is still
conceived not only as structural but also aggregate. This conception is useﬁrl indeed in
picturing an actor’s overall ability in international affairs but of limited use in analyzing
the role of power in speciﬁc international negotiations. It describes power at system level
but not at issue and relationship level; that is, by not specifying the scope and domain of
power, it “implies either highly fungible power resources or a single dominant issue-
area” aldwin, 1979). Yet power resources may not be that ﬁrngible. What ﬁrnctions as
power resources in one policy-contingency framework may be irrelevant in another; the
only way to determine whether something is power resources or not is to place it in the
context of a real or hypothetical policy-contingency ﬁ'amework (Ibid). From this
perspective, the distribution of capabilities for particular issues is likely to better explain
the relationship among states and the outcomes of their negotiations.

Determining the relevance of power in international negotiations is in large part
issue and issue area in question. Issues in international negotiation seem more broadly
deﬁned than in just military and economic terms. Keohane and Nye (1989) say that issues
are not the same as delimited objective problems; issues are problems about which
policymakers are concerned, and which they believe are relevant to public policy. Thus,
an issue is partly circumscribed by a subjective consideration since it needs people, who

especially exert inﬂuence over policy, to perceive a problem as such. An issue area is

63

also formed in a similar manner: when a government sees a set of issues as
interdependent and deals with them collectively, that set of issues establishes itself as an
issue area. Thus, an issue area pertains to actors’ beliefs and behavior rather than the
objective reality of problems. This is why it is difficult to precisely deﬁne the boundaries
of an issue area; this difﬁculty is complicated by the fact that these boundaries can
change over time as issues and their groupings change (Ibid).

Despite the prevailing balance in aggregate power between actors, the outcome of
negotiation is likely to be determined by the balance of power within that issue area. The
premise of issue area power is that power in one issue area loses some or all of its
effectiveness when applied to others. Power resources cannot be easily transferred and,
thus, may not be fungible. In other words, capabilities relevant to one issue area may not
be relevant to another due to the unﬁxed nature of issue area. Each issue area often has
different political structures that are more or less insulated from the overall distribution of
economic and military capabilities among states.12 Within each issue area states will
pursue their relatively coherent self-interests, and stronger states in that issue area will
dominate weaker ones and determine the rule of the game (Ibid.) A state’s net capabilities
will be some combination of its capabilities on all dimensions but its effectiveness may
also vary from issue to issue depending on capabilities most relevant to producing the
negotiated outcomes (Baldwin, 1979). Therefore, it is essential to look at their relative

capabilities at issue dimensions beside overall capabilities of individual states.

 

‘2 Diﬂ’erence in political structures of issue area can be inﬂuenced even by domestic politics, namely their
characteristic patterns of politicization and interest groups that are active in them. For instance,
international trade issue usually brings about broader and keener domestic reaction to international
bargaining than ﬁnancial ones, which makes it more difficult to reach an agreement and get it ratiﬁed by
legislative bodies of the states involved.

64

Whereas aggregate structural power is concerned with an actor’s capabilities and
position vis-a-vis the external environment as a whole, issue-speciﬁc power is concerned
with an actor’s capabilities and position vis-a-vis another actor in a speciﬁc mutual issue
(Habeeb, 1988). Such issue-speciﬁc power comes from the structure of bilateral
relationship than overall constellation of individual states in the system and, thereby,
becomes a particularly relevant component of power to the analysis of international
negotiation. When states enter into a bargaining relationship with one another, it is the
capabilities in that particular issue-area that ultimately count (Jonsson, 1981). The
effectiveness of tactics will also be differentiated issue area by issue area. Their objective
is to alter the issue-power balance, since the issue-power balance determines negotiation
outcomes.

Summing up, power is an important intervening factor between causes and
outcomes. However, power is by no means homogeneous, so the “search for power”
takes many different forms, whose characteristics depend in part on the ultimate goals of
the actor as well as on the particular contexts within which attempts at exerting inﬂuence
take place (Baldwin, 1979). We need to apply it in a more sophisticated manner for a
better explanation of bargaining outcomes. In this sense, power-as-resources
differentiated by issue area appears to be a more appropriate beginning, if not the end.13
The conception of power at an issue-speciﬁc level is also likely to capture the behavioral

aspect of power better than at the structural level.

 

‘3 It will become clear in Chapter 3 why this is only a beginning. but not an end.

65

2.3 ASYMNIETRY IN RESOURCES AND INTERNATIONAL NEGOTIATION

A general model of bargaining assumes equality and symmetry in power between
players so that it is possible to completely interchange them in explaining bargaining
process and reach conclusions that need not be distinguished on the basis of difference in
power. However, not every negotiation in reality takes place under this condition. As far
as our memory stretches out, there have been a lot of negotiations, collectively and
bilaterally, among states of different power in the international arena. Therefore, this
section will be devoted to discussing the feasibility of applying bargaining theory to
asymmetric relations and points to consider in modeling it. This is especially important to
consider since the relationship between the IMF and developing countries, which this
dissertation is to model with game theory, is fundamentally asymmetric.

The relationship between states of different power, mostly between developed and
developing countries dubbed North-South, is characterized not only by conﬂict of
interests but disparity in power. The essence of their relationship is that the two parts of
the world are at different stages of development, so that the South produces raw material
for industry in the North and sells them at ﬂuctuating prices in order to buy more
expensive ﬁnished goods from the North. Although the two parties are locked into
interdependence through the roles they play in world economy, it is hard to miss that
there is neither equality of present status nor equality of opportunity for future, and the
inequality of condition is mirrored and magniﬁed by the inequality of capability of

changing it (Zartman, 1987a).

66

With respect to international negotiation, this prevalent inequality and disparity in
power does trigger a stereotyped conjecture that states with greater capabilities, ceteris
paribus, have an advantage over ones without. Indeed, were North-South negotiations
only a power play, they could be dismissed as not only unimportant but even
uninteresting, given the asymmetrical context (Zartman, 1987b). However, there is a
substantial number of cases that the stronger party did not prevail over the weaker in
negotiations. Often the outcome of negotiations was closer to the preferences of weaker
states, and even when stronger states got their way, it was not always due to their superior
capabilities. What this indicates is that even within the basic structure of international
system that changes slowly, temporary adjustments may occur and basic patterns of
conﬂictual or cooperative interests may also be subject to short-term changes in emphasis
(Hopmann, 1996). In other words, under particular conditions, relatively weak states may
utilize temporarily special capabilities that assist them in negotiating with their stronger
counterparts, hence, achieve a more or less equal footing on bargaining.

These seemingly anomalous results of international negotiations where the
weaker prevail are puzzling in a theoretical venue. They are obviously at odds with
traditional power theory and the causes of this problem seem to lie more or less with the
conceptualization of power in negotiation, which was discussed in the previous section.
The conventional approach to power emphasizes overall distribution of power among
states in international system and does not differentiate it signiﬁcantly among issue areas
in world politics. It predicts a strong tendency toward congruence of outcomes among
issue areas, since power is considered ﬁingible. Power resources are moved around to

secure equal marginal returns in all areas that the stronger are expected to prevail even in

67

their weak issues by using their overall dominance (Baldwin, 1980). In the conventional
conceptualization of power, world politics is treated as a seamless web and power
ﬁingibility ensures congruence between the overall structure of military and economic
power and the pattern of outcomes on any issue area (Keohane and Nye, 1989). Applying
this conception of power, traditionalists have too often jumped directly from the
assessment of states’ capabilities in a particular negotiation to a conclusion about the
outcomes of the negotiation, without examining the negotiation process that intervenes.
Although it is true, indeed, that overall distribution of aggregate structural power
plays a key role in many aspects of international relations, by itself it provides a small
part of the explanation of successﬁrl asymmetric bargaining. Regarding this failure,
Zartman (1971) argues that the aggregate power position of a state cannot be directly
translated into relevant and available power in any particular situation. The mere
existence of superior power resources does not guarantee that they will be effectively
used or that the knowledge of their existence will be sufﬁcient to inﬂuence the behavior
of others. While criticizing the application of traditional conception of power to
asymmetric bargaining, Lockhart (1979) maintained that national power must be utilized
to create useful options within the context of the speciﬁc encounters that states face in
order to have any signiﬁcant inﬂuence on negotiation. Thus, asymmetry in power is not
the determining factor in negotiation. Rather, the situational relevancy of power resources
and the process through which power is channeled into negotiation dominate bargaining

outcomes, even if it does not determine them completely. ’4

 

'4 Negotiation theory based on contextual and procedural elements, through which we can gain insight into
the process of negotiation itself, was recently developed in Duchesne, Erick. 1997. “International Bilateral
Trade and Investment Negotiations: Theory, Formal Model, and Empirical Evidences.” Ph.D. dissertation,
Michigan State University.

68

The situational relevancy of power resources brings the concept of issue-speciﬁc
power of the previous section to any attempt to theorize bargaining between states of
different capabilities. It was concluded that power as an intervening factor in the
bargaining process has a limited ﬁrngibility; power resources in one issue area lose some
or all of their effectiveness when applied to others. Issue-speciﬁc structural power, i.e., an
actor’s capabilities and position vis-a-vis another actor in a speciﬁc issue area of mutual
interest, is a more relevant conception of power to the analysis of international
negotiation. By the same token, therefore, the ability of the weak to escape from their
deﬁnitional inferiority and obtain something from negotiations with the strong, depends
more on their power balance in speciﬁc issue areas than on overall balance.

Hopmann (1978) illuminates the issue-speciﬁc power in depth in the following
way: asymmetry refers to a situation not only in which there is imbalance in power
resources but also in which control of relations is unequal. This means that the
relationship itself between two actors is as important as their overall power resources
since the relationship between them is also a source from which each actor can derive its
power. Thus, negotiation outcomes are based on the dynamics of actor behavior within
the parameters of their relationship. Among other things, unequal costs from failure to
agree, combined with unequal capability to modify the structure of the bargaining
relationship, are two determining parameters of negotiation outcomes; more precisely,
power balance in an issue-speciﬁc relationship is determined by alternatives and control.
Alternatives denote an actor’s ability to substitute relationships with others for the one
with the opposing actor in order to gain similar or better outcomes. Thus, despite an

aggregate power balance favoring the opponent, the availability of alternatives may

69

increase actor’s issue power by decreasing its stakes at bargaining with the opponent and
enable it to achieve its preferred outcomes in a particular issue area. The party that has
many options does have an advantage over the opponent.

Control as a determinant of the issue power balance is deﬁned as the degree to
which one side can unilaterally achieve its preferred outcome despite the costs involved
in doing so. Any interdependent relationship implies that both parties would incur costs if
the relationship broke down. However, it does not imply that each actor would suffer the
same cost if it happens. Actors are not equally dependent upon each other, so the less
dependent actor is in a stronger position than its partner in a particular issue area because
changes in the relationship will be less costly to that actor than to its partner. This often
gives the actor a signiﬁcant political resource rather than economic in that the less
vulnerable actor will try to use such asymmetric interdependence to gain a greater share
of its preferred outcome from negotiation than its opponent.15 Thus, control is a
concomitant effect of asymmetric interdependence that lets the holder obtain more of its
preferred outcomes than its opponent does.

A handy example of alternatives and control in international relations can be
found in trade: the most common indicator of trade power has traditionally been the
degree to which a country relies on exports for its economic well-being, measured in

terms of ratio of exports to GDP, and its share in total world exports. However, a

 

’5 There is a caveat, however, to asymmetrical economic interdependence for political inﬂuence over the
other state. Wagner argues that, by the logic of game theory, the use of economic interdependence for
political inﬂuence still requires that the exchange of economic resources for political concession make both
parties to a relationship better off than they would be if they bargained over the distribution of gains from
economic relationship alone. This means that political inﬂuence from asymmetric economic
interdependence is independent of the degree of asymmetry in the economic relationship or its direction.
For a full explanation, see Wagner, R Harrison. 1988. “Economic Interdependence, Bargaining Power, and
Political Inﬂuence.” International Organization 42, no. 3 (Summer):46l-483.

7O

situation more relevant to asymmetric relationship in trade may be the so-called market
access. That is, if country A is more dependent on access to markets in country B than B
is on A, then B may be in a position to use its relatively lower dependence to exact
concessions from A in trade negotiations or to enforce trade agreements by threatening to
block A’s access to its market, since it would suffer relatively lower costs from cutting
off trade with A (Blair, 1993). It is possible for a country to be relatively weak at trade
talks, while being relatively strong in terms of competitiveness of its industries, if its
economic performance depends on access to foreign markets. Thus, in measuring the
distribution of issue-speciﬁc power in a set of trade negotiations, the competitiveness of
industries under negotiation would not necessarily be considered. Bargaining power in
trade would rather be determined by such factors as relative dependence on the markets
of opponents and the ability to offer subsidies to domestic exporters and suppliers.16

Also, in creating useful options in asymmetric bargaining, as Lockhart said, one
must pay attention not only to the absolute or the relative distribution of capabilities
among states, but also to the speciﬁc processes through which differences in capabilities
are introduced into the negotiations. There are a number of factors that should be taken
into account when evaluating the power balance in international negotiation.
Asymmetries in areas outside the issue in negotiation may be of equal importance;
intangible, or qualitative, factors that operate at state level may play an important role in
the translation of capabilities into control over outcomes, either by facilitating or

hindering the mobilization of capabilities (Blair, 1993). Consideration of these variables

 

‘6 Blair examines the impact of each trade issue structure in export credits, agriculture, steel and ship on
trade agreements within the multilateral framework of OECD and frnds that bargaining positions of OECD
member countries tended to inﬂuence to a great extent by domestic groups.

71

has a legitimate place in explaining the outcomes of asymmetric international
negotiations. Because the direct impact of power resources on negotiation outcomes are
less sure than their advocates would have us believe, it is worth focusing on the process
of negotiation where it is combined with economic, political, and institutional
determinants and translated into actions to produce particular outcomes. 17

A conspicuous example of such process and outcomes in North-South negotiation
was well recorded by John Odell (1980). Using trade negotiations between the United
States and Latin American countries, Odell ﬁnds an explanation for weak states’ victories
in types of strategy. He points out three strategies in particular: (1) the Latin American
countries often took advantage of the pluralist system in the United States to build
domestic allies within the United States to help ﬁght for their cause: (2) in cases in which
the Latin American states possessed a commodity of particular importance to the United
States, or represented a market of signiﬁcant size to U. S. exporters, the threat of
retaliation often resulted in an enhanced negotiation outcome; and (3) the Latin American
states often came to the negotiating table with careﬁrl technical preparation and made
persuasive technical arguments.18 This is a sheer dynamic of negotiation between the
weak and the strong, how behavioral aspects of power affect the negotiation’s outcome. It
indicates that countries lacking traditional capabilities such as military and economic
resources can compensate for this in part by seeking attractive alternative outcomes

through tactics. Weak states are not so inhibited from applying bargaining tactics

 

'7 Again, Duchesne’s dissertation on US-Canada trade and investment negotiations sets a good example of
this.

'8 Odell reaches a similar conclusion about US-South Korean trade negotiations. See Odell, John. 1985.

“The Outcome of Intemational Trade Conﬂicts: The US. and South Korea, 1961-1981.” International
Studies Quarterly 29:263-286.

72

strategies as we have thought, either; for instance, weak states can link unrelated issues
with lower cost compared with their strong counterparts, partly because their domestic
interests are less complex. Such conditions in the weaker states may provide not only a
better option if negotiations fail, but also improve their relative position in inﬂuencing the

outcome of negotiations in a direction favorable to them.

At the closing of this chapter and in bridging it to the next, the following should
be borne in mind. If inequality is the nature of the relationship, negotiation is the more
appropriate means of changing the structure of negotiation on the margins and of
providing new positions that are more satisfactory to both sides (Zartman, 1987a).
Although material resources determine the structural power of an actor and its position
vis-a-vis the external world as a whole, an actor’s power and its position within a
relationship is issue-speciﬁcally established by the degree of dependence. The extent to
which each side can alter its dependence through alternatives and control with
commitment can in part compensate for weakness in aggregate power and accordingly

bring success in various degrees to negotiations in issue areas.

73

CHAPTER 3

MODELING CONDITIONALITY BARGAIN ING GAME

Thanks to the review in the previous chapter, we refreshed ourselves with a
sanguine memory of components both endogenous and exogenous to bargaining models
based on game theory. Among exogenous factors that facilitate bargaining outcomes, the
concept of national power was critically examined and the possibility for bargaining
among actors of asymmetric capabilities in international relations was ﬁrmly conﬁrmed.
Capitalizing on that, we are now well equipped to model bargaining on conditionality
agreements between the IMF and developing countries, whose relationship is allegedly
asymmetric. Thus, this chapter will be devoted to building a game—theoretic bargaining
model for conditionality agreement. A novelty of conditionality bargaining game
developed here, however, lies with a reﬁned concept of power, i.e., leverage. Hence, it

starts with the introduction of leverage theory and its role in conditionality bargaining.

3.1 DEVELOPING A NEW CONCEPT OF LEVERAGE

The general conclusion about the possibility of negotiation among states of
different capabilities becomes a prelude to portraying conditionality agreements between
the IMF and developing countries as negotiated outcomes. Conditionality intrinsically
reﬂects inequality in power due to the fact that one of the two parties involved is the
holder of readily disposable material resources and the other is dependent on those

resources for its economic survival. That inequality surrounding the relationship has been

74

easily taken as the ground to believe that all the developing countries rushing to the IMF
for ﬁnancing would contract much similar conditionality agreements, which are usually
imposed by the IMF, and alter their economic policies as directed. To our surprise,
however, reality has been quite different from this expectation. The IMF, arguably the
stronger party of the two thanks to the abundant resources it controls could hardly be said
to prevail over developing countries, the weaker one, in every encounter either to settle
conditionality agreements or subsequently implement agreed conditions, as mentioned
earlier. When in search for the reason as to why the expectation went awry against the
reality, it was found to be related to the conventional conception of national power—
namely aggregate structural power—projected on the relationship between the IMF and
developing countries.

As a matter of fact, the relationship between the IMF and developing countries is
highly conﬁned to dealing with economic policy change, and the resources conducive to
forming power are exclusively monetary. To rephrase this along theoretical lines, the
relationship between the two is very issue-speciﬁc, even though it does not necessarily
preclude the possibility of linkage with unrelated issues, which makes their power-
balance more or less independent of aggregate capabilities. This is why past attempts to
analyze the relationship between the two with the concept of aggregate capabilities have
been largely unﬁuitful. Therefore, conditionality agreements as bargaining outcomes are
likely to be better illuminated with the issue-speciﬁc concept of power, and theorizing
international bargaining at a general level will be similarly beneﬁted, too.

Insofar as the concept of issue-structural power is invoked, that concept needs

further tailoring in order to explain bargaining outcomes for two reasons. First, as seen in

75

the previous chapter, negotiations proceed to reach an agreement for the realization of a
common interest through joint decision-making, where conﬂicting interests are
simultaneously present (Ikle, 1964; Brady, 1991; Hopmann, 1996). This means that the
core of bargaining situations is reciprocity and the interdependence of decisions, and an
appropriate concept of power should be the one that emphasizes both the interdependent
nature of bargaining and the importance of relative as opposed to absolute inﬂuence
(Young, 1968). Strictly speaking, however, issue-speciﬁc power is still structural power
deprived of fungibility and it remains the same for issue-speciﬁc power that the stronger
party within a concerned issue area will impose change unilaterally on the weaker.
Unilateral achievement of preferred outcomes by issue-speciﬁc power does not meet the
characteristic of negotiation (Habeeb, 1988). Thus, if we are to explain bargaining
outcomes as a causal process with the concept of power, such a concept should capture
not only the ability to impose a change without using brute force (Nagel, 1975), but also
do it reciprocally, satisfying the fundamental nature of negotiation as a chain of mutual
attempts at exercising power.

Second, power is always multidimensional; power varies in degree with respect to
scope, weight, domain, and cost (Baldwin, 1985). Analyzing the outcomes of bargaining
in terms of power at a single dimension can be misleading. In order to make a meaningful
analysis of bargaining outcomes, one must be able to specify who has the capacity to
inﬂuence in what domain with respect to scope. This concomitantly means that the power
relevant to an international negotiation may rest on various bases. Who is affected and on
what basis can be determined only by inquiry into the speciﬁc content of power in a

given situation (Ibid.). There would be situations in which one form of power bases is

76

.r, J
1‘» 5“

“If?

Ing
is

I

U1

‘u.

tn-
'./1

indeed sufﬁcient in international relations. Nonetheless, the relative importance of other
power bases in bargaining cannot and should not be prejudged.

Bound by the above arguments, this dissertation introduces leverage as a concept
more relevant to explaining bargaining outcomes. Although leverage is always deﬁned in
terms of power, it is nonetheless a distinct form of power (Nachmias, 1988).1 In
bargaining literatures as well as everyday language, the word leverage broadly refers to
or is commonly understood as the ability to demand more favorable terms (Werner,
1998). In line with such conceptualization, Barston (1971) noted that a small state’s
ability to resist doing what a larger state wants it to do is a critical source of small state
power, even if it does not have the ability to persuade the large state to do something; and
Bjol (1971) says that a weaker state’s ability to hold out, or to ignore a demand made by
the stronger state, may allow it to achieve its preferred outcome. These two remarks
correctly incorporate mutual inﬂuence in negotiation, which takes place without resorting
aggregate or issue-structural power. The only problem with this conceptualization of
leverage, however, is that it is too vague to empirically test. Its analytical insufﬁciency
becomes noticeable especially when asked what constitutes such ability and how it can be
operationalized.

Inferring from the above and reﬁning ﬁrrther, leverage in this research is deﬁned
as a state’s capabilities or endowments conducive to changing the outcomes of

bargaining within a speciﬁc issue area, which otherwise would not occur. The merits of

 

‘ Leverage is also different from statecraft. Statecraft refers to the selection of means for the pursuit of
foreign policy goals. The main purpose of statecraft is concerned with deciding what kind of policy
alternatives available to a decision-maker is most likely to succeed in pursuing a given set of objectives.
For an in-depth study of statecraﬁ, see Baldwin, David. 1985. Economic Statecraﬂ. Princeton, NJ:
Princeton Univ. Press.

77

this deﬁnition are that it allows for not only meeting the nature of negotiation but also
providing testable analytical distinctions. First, leverage refers to properties and
characteristics of actors as foundation. Of course including capabilities, leverage
comprises other capabilities and endowments, both changeable and unchangeable,
movable and immovable, and convertible and inconvertible to production capability, such
as geopolitical location and historicity.2 For that nature of leverage, Barston (1971)
remarks that, if a small state occupies territory of strategic importance to a larger state, it
may be able to exercise bargaining inﬂuence disproportionate to its “objective”
capabilities. It is this possessional nature of leverage that easily renders it operationalized
without controversy.

Leverage as properties and characteristics needs further explication in two
aspects, which distinguishes it from power. While the traditional concept of power
exclusively focuses on the positive aspect of national capabilities and endowments,
leverage counts not only on positive capabilities and endowments but also on negative or
even dysfunctional ones to the holder but can be effectively used in relation to the other.
Negative capabilities and endowments, like positive ones, can hold potentials to widen
the range of options available to the players in a bargaining situation; the only difference
between them would be the extent, usage, and effectiveness. Second, properties and

characteristics eligible to be leverage consist of both permanent and temporary ones.3

 

2 Some scholars perceive leverage as the outcome of bargaining. In my dissertation, however, leverage is
strictly existing properties or characteristics of states, which affect bargaining outcomes. For comparison.
see Nachnrias (1988).

3 Vital distinguishes further between “permanent and intrinsic” resources and “ephemeral and contingent”
resources. Intrinsic and contingent resources. however, appear to be more or less related to bargaining
strategies discussed in Chapter 2 than the properties and characteristics as leverage. See Vital, David. 1971.
“The Analysis of Small Power Politics.” in Small States in International Relations, ed. August Schou and
Arne O. Brundtland Stockholm: Almqvist & Wiskell.

78

While the former represents ﬁxed material resources of an actor, the latter time-bound
resources, which may be neither available nor useﬁil over time. Agenda for bargaining
are different negotiation by negotiation, and we cannot prejudge the importance of the
former over the latter in determining bargaining outcomes without considering the issues
at hand. Hence, the most relevant and effective resources of states for bargaining cannot
be ﬁxed and will be contingently selected against projected goals.

Due to the above natures, possessing leverage does not necessarily mean a
permanent shift in the power relationship or a change in tangible power bases between
actors. It is frequently asserted that an increase in the power of one actor necessitates a
decrease in the power of another actor.4 Possessing leverage, however, allows a
temporary adjustment within an issue-speciﬁc bargaining relationship that may change
slowly or not at all5 and the pattern of interaction between actors during bargaining may
be subject to this adjustment. In other words, under particular conditions, relatively weak
states may utilize temporarily special capabilities and endowments that assist them in
negotiation and hence achieve their preferred outcomes (Hopmann, 1996). In this sense,

the concept of leverage is even different from issue-structural power, which implicitly

 

4 The entire debate on relative and absolute gains between realists and liberalists was tantamount to a
debate on whether power is zero-sum. Regarding those debates, see Grieco, Joseph M. 1988 “Anarchy and
the Limits of Cooperation: A Realist Critique of the Newest Liberal Irrstitutionalism.” International
Organization 42, no.3: 387-423; Powell, Robert. 1991. “Absolute and Relative Gains in International
Relations Theory.” American Political Science Review 85, no. 4 (December):l303-l320; and Snidal,
Duncan. 1991. “Relative Gains and the Pattern of International Cooperation.” American Political Science
Review 85, no. 3 (September):701-726.

5 There is a consensus that international negotiations are not single-shot affairs but occur in the context of a
preexisting relationship that will be maintained over a long period. Thus, this statement does not preclude a
possibility that current adjustment will lead states to reassess their relationship in future encounters, which
will inﬂuence their dyadic relationship cumulatively. Leverage, when correctly used, may result in more
power.

79

assumes constant power relationship within the concerned issue area, and leverage is
exactly what forms the base of relative bargaining strength between states.

Second, another analytic merit provided by the deﬁnition of leverage is that it
captures the relational or behavioral aspect of power. Negotiation is a process of joint
decision-making through two-directional inﬂuence, and this perspective directs the focus
of attention to the process involved in exercising power rather than to power base. When
leverage is deﬁned as such, not only the common fallacy of equating power with simple
power bases, whether aggregate or issue-speciﬁc, is overcome (Jonsson, 1981), but also
strategic action aimed at controlling the other actor’s policy choices is implied, which is
pertinent to exercising power (Nachmias, 1988). Leverage becomes a means to
manipulate negotiation in order to make the other state conform its policy or actions to
the desires of the state. Leverage intervenes in the bargaining process and gets the other
actor to do something that it would not otherwise do in the ﬁamework of
interdependency. And it is this relational aspect of leverage that evidences the ongoing
inﬂuence between the actors and explains why bargaining outcomes are often not in line
with objective capabilities.

When leverage is strategically cast by the possessor in order to achieve
compliance of the other actor, it is played out through two primary tactics: (I) coercion (a
threat of punishment) and (2) inducement (a promise of reward) (Ibid.). Leverage can be
combined with tactics as examined in Chapter 2; it can be associated either with threat to
inﬂict cost for non-cooperative behaviors or reward for cooperative behaviors through
side-payments and issue linkage in order to manipulate the other actor’s choice. The

success or failure of bargaining, in other words getting preferred outcomes, is likely to

80

-_

 

depend on the skillful application of these two tactics. It is difﬁcult, however, for
observers to distinguish which tactic of the two is applied in a bargaining situation since,
whereas coercive measures may be used without accompanying inducement, rarely are
inducements offered without being preceded or accompanied by coercive measures

(Ibid). Leveraged actors are expected to practice the two tactics simultaneously.

Drawing on the above discussion, I argue that the variance in conditionality
agreements is a ﬁrnction of leverage rather than issue-speciﬁc or aggregate structural
power between the IMF and recipient countries. Although there is an ostensibly huge gap
between the IMF’s and recipients’ capabilities, both sides’ bargaining power should be
measured in terms of leverage and the gap in leverage may not be so big as in
capabilities. The versatile leverage appears to be more responsible for the success or
failure of conditionality bargaining than the ﬁxed capabilities are, and, thus, more
suitable for analyzing bargaining outcomes.

In conditionality bargaining, leverage will enable actors to directly modify a
proposal or induce the other to rearrange its preference orderings, which result in
different conditionality agreements. This is tantamount to stating that without leverage,
conditionality agreements would have been the same across cases. Such a counterfactual
consideration makes leverage a necessary condition to the various outcomes of
conditionality negotiation. In next section I proceed to model conditionality bargaining
framed in game theory in order to show how the two actors reach ﬁnal agreements and
what effect leverage has on conditionality agreements, as exogenous to the game

structure, but as a necessary condition.

81

3.2 A MODEL OF CONDITIONALITY BARGAINING GAME

3.2.1 Strategic Interaction between Donor and Recipient in Conditionality Bargaining

Conditionality bargaining is depicted in Figure 1. Conditionality negotiation
consists of two players, the IMF as the donor, and a developing country as the recipient.
A recipient government, noted as R, initiates a bargaining game. Facing a severe balance-
of-payment gap, a government of developing countries has to decide whom to go for an
emergency ﬁnancial rescue, i.e., multilateral, including regional, ﬁnancial institutions
like International Monetary Fund and World Bank, bilateral aid agencies of developed
countries, or international commercial banks. Such a choice forms its initial move that it
starts a talk with the IMF, noted as D, by making a proposal (PR) for structural
adjustment, or forgoing to make a proposal (~PR). Moving second, the IMF has three
options at node 2: ﬁrst, it can make a proposal (PD) of its own as a response to the
recipient government’s. The IMF ’s proposal could be based on the original proposal by
the recipient country or completely new one but usually expected to be tighter than the
original since the IMF reinforces conditionality by adding more items. Any modiﬁcation
of the proposal can take the form of qualitative as well as quantitative change of
conditions. Second, the IMF can accept the proposal formulated by the recipient
government (AD) as it is, yielding the outcome of Agreement @. And the last option open
to the Fund is reﬁising to proceed with any type of appraisal or negotiation of the
proposal (~AD), resulting in VETOD. When veto is announced by the donor, the

conditionality bargaining game ends there and the recipient has to reforrnulate a proposal

82

 

2 3
~ Pl
P” A' ~ A" PD
SQ
4 AgreementCD VETOD 5
CPR AR ~AR ~AR AR PR

 

 

6 Agreement® VETOR

 

VETOR Agreement©
AD ~AD ~AD AD
(5)
Agreementi NA VETOD Agreement©
~AR
NA

Figure 3. 1 Conditionality Bargaining Game

83

CPD

AR

Agreement®

if it still wants the IMF to bail it out. In the case that the IMF opts for the ﬁrst, the
recipient government at node 4 has to make a choice among three options just as the IIVIF
did at node 2: whether to accept the HVlF’s proposal (AR), producing Agreement @, or to
repudiate it (~AR), which will result in VETOR, or to formulate another proposal for
structural adjustment (CPR). The counterproposal is equivalent to taking the IMF ’s
proposal back to its original one by inserting its interest to a certain degree, which will
eventually pose it at a middle point from each initial position. If the IMF approves the
counterproposal by the recipient country (AD), the conditionality bargaining ends with a
policy package, Agreement @, which has passed through two times of modiﬁcation from
the initial proposal that the recipient country made. If the Fund does not accept the
counterproposal made by the recipient country (~AD), no speciﬁc agreement on structural
adjustment is contracted between the two players and the outcome is called NA (Non-
Agreement). This situation could occur when negotiation breaks down mainly due to both
sides’ intransigence.

The other path leading to conditionality agreements is located on the right hand
side of the tree. If a recipient country does not make a proposal for structural adjustment
(~PR) and the IMF does not take any action on it (~PD), not only nothing is going on
between the players but any effort to revitalize the economic condition of the recipient
country occurs without the IMF’s involvement. Therefore, the outcome is Status Quo
(SQ) at node 3. Nonetheless, the IMF can take an initiative for structural adjustment and
recommend a policy package to that developing country in exchange for a loan at node 5

(PD).6 This might occur due to either the IMF’s ongoing surveillance over member

 

6 A legitimate question at this point is whether bargaining outcomes will be different if the IMF takes the
initiative. On the surface, conditionality bargaining initiated by recipient is based on the IMF Articles of

84

countries or internal lending pressure. At node 5 the options available to the recipient
government are three: simply accepting the proposal proposed by the Fund (AR) in order
to alleviate its economic hardship, accordingly yielding the outcome of Agreement 69, or
making a counterproposal (CPR) that modiﬁes the original proposed by the Fund and
strengthens its position in relation to the Fund, or disapproving it (~AR) so as to yield
VETOR. If the recipient makes a counterproposal and the IMF responds to it with
acceptance (AD) at node 7, the bargaining process ends with a conditionality agreement
modiﬁed from the Fund’s proposal (Agreement {5}). If the IMF repudiates it (~AD), even
though it initiated the bargaining process, the ﬁnal outcome of the bargaining is VETOD.
However, if the IMF proceeds with another proposal that reinstates its position (CPD), the
recipient will have another chance to make a choice at node 8 and wrap up the bargaining
process: if the recipient accepts the second proposal from the IMF (AR), the bargaining
produces an agreement labeled Agreement @; otherwise (~AR), all the bargaining efforts

ﬁom both sides become ﬁrtile, yielding NA.

 

Agreement specifying that a member’s transactions with the IMF in the General Resource Account take
place wholly “on the initiative of such member” and not that of the IMF (Article V, Section 2(a)). Thus.
there is no obligation on the member’s part to approach the IMF at any speciﬁed time with a request for use
of its resources. However, this limitation does not preclude the IMF from entering into discussions with a
member to explore the desirability of a request for possible use of its resources. This procedure was
recognized by the Executive Board in 1992 that the Fund, that is, the Managing Director and staff, acting in
accordance with the general policies adopted by the Executive Board itself, might take the initiative in
discussing with one or more members transactions which it believes suitable for the Fund and helpful to the
members concerned” (Source: Treasurer’s Department. 1998. Financial Organization and Operations of
the IMF. Pamphlet Series, no. 45. Washington, DC: International Monetary Fund). Having all considered, a
proper game-theoretic interpretation of this is that the effect of initiative may be some quantitative on the
outcome but not qualitative, if a game has inﬁnite rounds. That is, the magnitude of the difference in what
the players get may differ slightly, dependent on the modeling approach, but it will not affect which player
gets more than the other depending upon the magnitude of other parameters. For a full explanation, see
Stahl, Ingolf. 1994. “The Rubinstein and Stahl Bargaining Models: A Comparison and an Attempt at a
Synthesis.” EFI Research Paper 6535. Also, the reason why ‘who goes ﬁrst’ does not matter in real
conditionality bargaining will be clariﬁed further at the end of this chapter.

85

In a realistic environment where many developing countries have suffered from
protracted balance-of-payments problems and each of them has contracted more than one
conditionality agreement, it makes more sense to represent the situation as a repeated
game. Under such conditions, the bargaining outcome of a previous encounter may
reshape the context as well as the process of the current negotiation. Integrating such a
nature of conditionality bargaining, however, would add many complexities to a model
that already consisted of many factors. Thus, as a ﬁrst cut of to modeling conditionality
bargaining, this dissertation will be content with a single-shot model but still be attentive
to the phenomenon Of iterated nature of bargaining and will partly capture it in empirical

test later.

3.2.2 Assumptions of Conditionality Bargaining

Modeling conditionality bargaining starts with the assumption of a unitary and
rational actor. The assumption of rationality here is the same as mentioned earlier,
which needs not more explication. It will be enough to give a simple statement that the
two players have goals and attempt to realize those goals through their strategic actions
that they believe maximize their expected utility.

The unitary actor assumption implies that actions and strategies are chosen by a
single, rational leader. This actor is who is actively involved in the bargaining with the
other party and responsible for selecting the strategic actions required to implement its
Objectives to the best of his or her ability, and eventually such decision serves his interest

as well. This assumption, however, is posed with a question of who is the ﬁnal decision

86

 

 

maker and whose interests are served through such decision, if one of the players is an

international organization that is composed of member countries. The IMF as an
international organization has two active groups of actors. As the ﬁrst group, we can
think of top rankers, who are usually representatives of (powerful) member countries,
and, therefore, highly political. As of March 1999, the Fund is composed of 182 member
countries. Each country is represented by a governor (usually either the ﬁnance minister
or a central bank member), whose voting power is determined by the country’s capital
subscription in the Fund. The Board of Governors exerts ultimate control within the Fund
but delegates most of its authority to the twenty-two-member Executive Board headed by
Managing Director. Within the Executive Board, through a system of weighted voting,
ﬁve Directors appointed by the countries with the largest quotas have the most voting
power: the United States, Japan, Germany, the United Kingdom, and France. Quota
subscriptions from these ﬁve countries take around 40% of total fund formation and,
thereby, give them equivalent number of votes to cast. The remaining positions within the
Executive Board are elected by the Board of Governors. Each Director represents a
country or group of countries from the same geographical location and/or with similar
economic interests. In routine practice, formal votes are rarely taken within the Executive
Board; instead, a general consensus among the Directors is ordinarily achieved on each
loan and policy decision.

The second group of actors are “disinterested” and “dispassionate” experts, whose
role is to best interpret the ﬁnancial information at hand in the light of a consistent
economic tradition. The day-to-day operational core of the Fund is a staff of nearly 2,700

economists, ﬁnancial analysts, and functional specialists who work with member country

87

officials on the formulation of macroeconomic policies suitable for Fund ﬁnancing and
who supervise and monitor the progress of those policies. And in this organizational
structure, the Managing Director is especially noteworthy as bridging personnel between
the Executive Board and the rank-and-ﬁle staff. The Managing Director takes part in
conditionality bargaining from the stage of pre-mission preparation through approval of
provisional agreements with member countries, and it is by virtue of his role within the
institution that he has considerable leeway in the application of Fund principles and
brings the Board members and the staff to agree with his interpretation of the problems
and appropriate solutions (Stiles, 1991).

In any event it would be absurd to assume that one group or the other dominates
the whole process of policy making within the Fund. Rather, these groups share inﬂuence
on policy making, depending on the circumstances. Nonetheless, on the basis of the
organizational reality that the ﬁnal decision on conditionality, i.e., approving
conditionality agreements, is conferred onto the Executive Board, the question of
unitarity is solved. This endorsement is supported by the fact that the ﬁve major donor
countries as a collective actor forrn the core of the Executive Board.7 The bulk of the debt
in developing countries is owed to those ﬁve countries, and this is true for oil, non-oil,
low-income, and middle-income developing countries. As the principal sources of
external ﬁnancing, the major creditor countries exercise enormous control over the

overall development ﬁnance (Hardy, 1987). In addition to their common ground as

 

7 The effects of unique institutional arrangement within the IMF, weighted voting system and decision-
making concentrated in a body of limited size, are discussed in Lister, Frederick K. 1984. Decision Making
Strategies for International Organizations: The IMF Model. Monograph Series in World Affairs. Graduate
School of International Studies, Univ. of Denver. Lister views that such mechanism enabled the Fund to
achieve organizational stability and consistency in decision making. This line of assessment broadly
supports the unitary actor model.

88

creditors, there are two more factors that bind them together: high level of consensus on
loan program to members and shared economic ideology. In terms of the scarceness of
the resources available to developing countries, it is agreed upon among those countries
that there must be a way to guarantee the efficient use of resources such as conditionality
(Eckaus, 1986). And in conjunction with the rising tide of opposition to “big
government” in the United States and United Kingdom, neoclassical economics of the
early 1980s in those countries has been conducive to shaping the adjustment policies in
recipient countries later on (Nelson, 1990).

As mentioned in Chapter 1, the IMF projects many objectives of different
priority.8 Which of them will give the highest utility to the IMF? Or is there something
else? As long as conditionality is concerned, the IMF’s utility will be maximized if
structural adjustment can guarantee the repayment of loans by imposing as many
conditions as possible.9 The notion of conditionality is designed to encourage appropriate
economic adjustment and ensure that the member’s use of IMF credit is temporary and
that it will have the capacity to repay the IMF in time. Since the IMF generally makes its
ﬁnancial resources available to countries only in the context of a macroeconomic
adjustment program, the conditionality associated with this support should, in principle,
strengthen a country’s capacity-to-repay prospects (Gold, 1979; Guitian, 1982; Aylward

and Thome, 1998).

 

8 Recall Polak’s classiﬁcation of objectives of conditionality as in Footnote 11, Chapter 1.

9 The IMF’s utility can be set up differently, depending on how we portray it. For an analysis of the IMF
from a narrower public choice model, Vaubel, Roland. I991. “The Political Economy of the International
Monetary Fund: A Public Choice Analysis.” in The Political Economy of International Organizations, eds.
R Vaubel and T. D. Willett Boulder. CO: Westview Press. For an opposite portrayal of the IMF as a
honest broker or catalyst, see Nowzad, B. 1989. “The Debt Problem and the IMF’s perspective.” in Third
World Debt: The Search for a Solution, ed. Graham Bird. London: Edward Elga; and Vasquez, Ian. 1998.
“The IMF’s Dubious Purpose.” The Freeman 48, no. 10 (September):600-601.

89

Such utility for the Fund has practical supports, too. The IMF obtains its usable
funds from three sources: quota subscriptions from member countries, income (periodic
charges on the use of Fund resources and service, etc), and borrowing from member
countries to provide a temporary supplement to usable quota resources. In the past, the
IMF has experienced a new development in its ﬁnancial soundness, the growing problem
of arrears. The failure of some member countries to repay obligations on time, which was
a problem only to other creditors, became a matter of serious concern for the IIVIF toward
the end of the fourth decade of the IMF’s history—around the mid-1980s, a few years
after the debt crisis began. Until 1983, only one member country had experienced
protracted arrears to the IMF, and a handful of others, arrears of short duration. Then,
every year from 1983 until 1990, the amount of outstanding overdue obligation to the
IMF grew substantially. Total arrears rose from SDR 25 million in the ﬁrst quarter of
1981 to a peak of SDR 7 billion in 1992 (5 billion in US dollar), which was equivalent to
almost an eighth of the Fund’s total outstanding credits. Since 1992 the level of arrears
fell sharply, and from 1995 to present, it has changed little. Thus, Over 1989-90 the IMF ’5
Executive Board developed a strengthened cooperative strategy to resolve the problem of
protracted overdue obligations to the IMF"): the IMF affirmed that the Fund’s policy of
non-toleration of arrears to official creditors remained unchanged and emphasized that
countries should build up their reserves in order to ensure that they will be able to meet

their commitments to the Fund.

 

‘0 For a description of the main elements of this strategy and its implementation, see IMF Annual Reports
(1991 and onward). One of IMF measures to enhance recipients’ debt service to the IMF is adjusting the
rate of charge on the use of IMF resources. The rate of charge has been adjusted up- or downward in line
with the size of overdue payments. Although those charges are refunded when overdues are paid, such
adjustments are intended to protect the [M’s ﬁnancial assets by encouraging members to meet their
ﬁnancial obligations to it

90

Mi '
.i'.‘i‘u
K

oi ,
y\i»€r
P'n‘r
I “
in...
‘r: .
3..“
‘ a

if...‘ ‘
' r

T n
IL- “\

‘A‘
is)
Is
I"

The Fund’s utility associated with loan repayment can also be well justiﬁed with
the amount of credits from the Fund to developing countries. Although the Fund may
provide signiﬁcant amounts of ﬁnancial assistance to individual countries expressed as a
percentage of their balance-of-payments current account deﬁcits, it has provided only a
rather small amount of assistance relative to the deﬁcits of developing countries as a
group. While the IMF has been a quantitatively signiﬁcant source of ﬁnance to some
individual developing countries at certain times, it has been quantitatively insigniﬁcant
for developing countries as a whole when assessed against the size of their balance-of—
payments deﬁcits and ﬁnancial ﬂows from other sources (Bird, 1995). At a time when
developing countries have experienced severe payment difﬁculties, the Fund’s ﬁnancial
relationship with developing countries was rather as a net recipient of return ﬂows than as
a large net provider of assistance (Killick, 1995; Sachs, 1989).11

What then makes the IMF so keen about loan repayment? No one can answer this
question more tersely than the IMF’s own staff, Stanley Fischer, First Deputy Managing

Director (1998a):

Let me emphasize that the IMF is not a charitable institution, nor does it
carry out its operations at taxpayers' expense. On the contrary, it operates
much like a credit union. .......They [member quotas] are often described
as an expense to the taxpayer. We are deeply aware in the IMF that our
support derives ultimately ﬁom the legislatures that vote to establish their

 

” This is also graphically shown in Figure 7 in IMFAnnual Report (1998, p. 88). During the period (1985-
1990) when structural adjustment following debt crisis was at its highest peak, the Fund was receiving
repurchases (repayment) more than purchases (loan) by about 100%. This trend was reversed in the early
1990s with a small gap, and repurchases picked up again in 1994. For a better terminology, none of the
above transactions is technically a loan because each is part of a country’s quota. A country purchases
foreign currencies and SDRs with its own currencies and then, over time, has to repurchase its currency
with foreign currencies or SDRs. Consequently, the transactions are not legally or operationally loans.
However, these purchases and repurchases are similar to loans and repayments, and they are often, either
mistakenly or out of convenience, referred to as loans.

91

_:_:-'_ﬁ—v

. .
v. .,
‘ ~‘n

‘h'.’i~
I'
.hbu.r

. ‘r-a
m‘uiik

”7".” r
.l A

“‘"Q:

 

 

2.
u gin;

countries' quotas—their deposits—in the IMF. We must justify that
support.

As long as an international organization depends on members’ donation for its operation,
which requires approval from domestic legislative body, it is so natural for larger donors
to exert pressure on that organization to show productivity and efficiency and for that
international organization to respond to such pressure in one way or another. Under the
circumstances, the IMF has chosen a high rate of loan repayment as a proof of efﬁciency
and worth for continued support.

At a glance, this goal that the IMF projects on conditionality is contradictory to its
imperative to circulate the subscription from member countries, who even established a
new facility in the 1970s in order to stimulate developing countries to use the IMF’s
high-conditionality resources. However, when we look at these from the perspective of
the IMF’s overall institutional purpose, ﬁnancing mandate, repayment as the pursued
goal of conditionality is not odd at all. Simply put, the IMF needs to keep ﬁlnds ﬂowing
in to circulate it. Guaranteeing repayment is nothing but a prerequisite to implement its
ﬁnancing mandate.

The counterpart unitary actor in the recipient country is the national leader who
decides what foreign economic objectives to choose and what strategy to implement in
pursuit of those objectives. Even if we assume the existence of various sources of
inﬂuence on economic policy such as domestic political process, the unitary actor is an
agent charged with implementing actions in pursuit of whatever the Objective may be. In

a word, the unitary actor in the recipient country is who is actively involved in the

92

bargaining with the donor agencies and responsible for selecting strategic actions
required to implement the society’s objectives to the best of his or her ability.

The unitary actor assumption has a special implication to the goal that the
recipient pursues through conditionality bargaining. Each nation’s chief executive,
whether a monarch, a dictator, or a freely elected representative of the people, acts as if
his or her welfare and the preferences of those whose support is needed to retain power
were the same (Bueno de Mesquita and Lalman, 1992). National leaders have strong
personal incentives to tie their own actions to the general welfare of their supporters who
become his or her power base.12 Under prospectively tough circumstances the structure
of the situation procreates to all leaders a signiﬁcant interest in maximizing those interest
Of the state that ensures for themselves the opportunity to survive in a leadership position
and the concomitant opportunity to pursue their more individualistic preferences. During
structural adjustment, various demand-curbing measures hurt particular groups within the
society, although the pain inﬂicted by policies of ﬁnancial restraint is likely to be widely
distributed. The identity and the power of distributional coalitions important to any
particular adjustment effort vary with the structure of the economy (Haggard, 1985).
Against the expected pressure from groups that are hurt by such measures, not all leaders
are able to marshal a sufficiently strong coalition of broad segments of the population that
can be expected to beneﬁt from the structural adjustment measures and ﬁ'om the

avoidance of excessive reliance on demand-management policies, which is critical to

 

'2 The concept of support base is not identical with popularity, which usually ﬂuctuates, and predictably
falls if adjustment is pursued. Rather, the term refers to groups bound to the regime by more durable ties of
ethnic identity, ideological compatibility, ingrained party loyalties, and/or patron-client networks. For
further discussion of political support base and structural adjustment, refer to Nelson, Joan M. 1988. “The
Political Economy of Stabilization: Commitment, Capacity and Public Response.” in Toward a Political
Economy of Development: A Rational Choice Perspective, ed Robert Bates. Berkeley: Univ. of California
Press.

93

their continued political survival. For this reason, the would-be borrowers prefer to use
traditional macroeconomic policies to more intrusive measures of structural adjustment
(Polak, 1991).13 Therefore, aware of possible political turmoil and temporary or enduring
adversary effects on economy, which will have certain impact on his stay in power, the
utility of conditionality agreement will be maximized for the leader when it consists of a
substantially large amount of loan but few conditions. Such a conditionality agreement
would not only cover the country’s projected balance-of-payments gap, but also allow
him to utilize the resources in ways that may prolong his leadership at the same time
(Mosley, 1988).

In any event, however, it should be noted that these utilities are highly simpliﬁed
extreme cases. Needless to say about individual recipient countries, even the IMF may
pursue different goals from each agreement depending on who is the other part of the
negotiation and where a particular agreement ﬁts into the bigger picture of international
political economy. Such particularistic aspects of individual bargaining, however, have to
be dropped in deriving the goals of conditionality bargaining for schematic as well as
generalizable presentation’s sake. These utilities have only the beneﬁt of facilitating
analytical tractability and predictability of the bargaining model. Also, those goals were
deliberately chosen since they appear more or less common and pivotal across individual

conditionality agreements, on top of their substantive issues at stake, and, empirically,

 

’3 Macroeconomic measures to restore equilibrium of economy are traditionally demand control through
monetary instruments such as money growth and credit and ﬁscal restraint. These measures spread the
adjustment burdens throughout the society. On the other hand, structural measures or so-called supply-side
economics are to enhance efﬁciency in resource allocation through strengthened role of market, for
example, privatization, liberalization, and exchange rate changes. Structural measures tend to hurt special
interest groups.

94

comparison of costs and beneﬁts from conditionality agreements becomes easy.
Therefore, they are simplifying, not problematic.

Other than the above ﬁlndamental assumptions of game-theoretic modeling,
which are reconstructed in the context of conditionality bargaining, particular

assumptions relevant to this research are speciﬁed below:

Assumption I: As utility maximizers, players choose strategies to get the outcomes that
they believe maximize the expected utility under the circumstance where their choice of
action is constrained by the other player’s strategies. The players anticipate the
consequences of their action in terms of reaction from the other player to the chosen
course of action.

Assumption 2: The desirability of a conditionality agreement is determined by the
number of conditions and/or the toughness of speciﬁc conditions and the amount of
agreed loan, which can be traded off. However, an agreement that is regarded as desirable
by one actor is not necessarily undesirable to the other.

Assumption 3: Both players prefer to reach an agreement on conditionality rather than to
come to a rupture in which no speciﬁc agreement on conditionality and, accordingly, on
loan is contracted between the two players at the end.

Assumption 4: A package deal of policy reform is an outcome of strategic interaction
between the two players who maximize their expected utilities. Hence, the utility of a
conditionality agreement is not known with certainty and yields an expected value,
assessed according to the subjective probability of reaching conditions preferred by one
player (P') and the subjective probability of reaching conditions preferred by the other
(1 - P'). Such probabilities are restricted between 0 and 1 (0 < P‘ < 1.0).

Assumption 5: The conditionality bargaining process is characterized by efforts to
modify a proposal in a way that accommodates one actor’s needs more. However, the

95

ultimate magnitude or materialization of such modiﬁcation during bargaining, which is
noted as 7 (>0), is not known with certainty. Rather it is directly inﬂuenced by the

probability each player projects on the negotiation. Therefore, it is yiP’.

Assumption 6: Each bargaining outcome can be associated with potential costs. Three
types of costs, domestic political costs (0), reputation costs ((p), and opportunity costs (13)
will selectively affect players’ utilities for bargaining outcomes. All these cost terms are
greater than zero and will be restricted by the subjective probability (P’) in the form of
O,(l - P'), (piP', and [3,-P'.

Assumption 7: Conditionality bargaining takes place over several time periods. The
utility of an outcome after one time period is discounted by 5, which falls between 0 and
1. The sooner actors reach desirable outcomes, the more utility the outcome brings to the
actors.

Assumption 1 stipulates that we are dealing with decisions made in a game-
theoretic manner. Regardless of the internal political or organizational processes
involved, conditionality bargaining carries with them the promise of economic policy
change made according to coherent, well-ordered preferences.

In Assumption 2, unlike many studies of conditionality that assume only conﬂict
of interest and thereby make it convenient to advocate the coercive nature of
conditionality, this model is not predisposed to either a conﬂict or harmony of interests
between the two players. By the partly supplementary nature of goals that the IMF and
recipient countries pursue respectively, structural adjustment could be portrayed as
economic interdependence that involves a possibility of joint gains and joint losses. The
joint gains here mean that recipient countries could improve their economies with loans

contracted with the IMF and the IMF, on the other hand, could get its ﬁlnds circulated

96

and collect accrued interest. The joint losses would be that such gains are not realized and
rather costs are inﬂicted by failing to agree. Mutual awareness of their potential gains and
losses can limit the danger of worsening each actor’s position through overly rigorous
struggles over the distribution of the gains. Their mixed motivation is evidenced by the
fact that around half of the agreed conditions has been adhered to, which indicates that
their interests converge at some points and diverge at others. Therefore, the nature of the
conditionality bargaining game is not ﬁxed in the sense of zero-sum or constant sum.

In relation to the unﬁxed nature of the conditionality bargaining game, the utilities
of winning its own demand on conditions, U'(A,~), and accommodating the other’s, U’I(Aj‘),
should be explicated in comparison with the status quo. The preceding situation to
conditionality bargaining is internally and/or externally perceived economic hardship in
the prospective recipient country, which has latent costs to leaders in power. The value of
leaving economic problems unsolved, which represents status quo, is weighed against
costs involved in structural adjustment and can make it more attractive or less than
accepting the donor’s demand for policy change. The recipient country admits a certain
need for economic policy change but wants to pick palatable ones out of a wide range of
policy options from the consideration of ensuing costs. From this, what determines the
utility of accepting the donor’s demands on structural adjustment is the similarity
between economic policies pursued by both players. The more similar the patterns of
revealed economic policy choices of two players, the smaller the utility of winning any
demand that one player makes on the other and, concomitantly, the smaller the diﬂ’erence

between U’(A,) and U'(Aj-) or vice versa. Conversely, the more dissimilar the revealed

economic policy commitments of two players, the greater the assumed utility for

97

achieving the conditions contained in his demand and the greater the difference between
U’(A,) and U'(A,) or vice versa. A real life parallel is the unorthodox adjustment
implemented by large debtors in Latin America in the 1980s, which indicates a big
difference between U’(A,-) and U’(Aj). This is an essential collateral of conditionality
bargaining that is neither necessarily zero-sum nor nonzero-sum game. Thus, the
relationship of U'(A,~) and U'(Aj) is not automatically opposite, and the utility of accepting
the other’s demand, U'(A,~), is not predetermined to be inferior to status quo.

This logic can also be used to determine the nature of the (initial) proposals made
during bargaining. The intensity of one’s interest inserted in the proposals is not
predetermined simply by the fact that one is making a proposal but under what conditions
the one is making the proposal. In other words, the strength or weakness of the proposer’s
interest stated in the proposal is determined purely by economic urgency for contracting a
conditionality agreement to get a turbulent economic period over, or concoction of
politico-economic calculations based on leverage available, which may overshadow
economic needs and make the difference between U’(A,-) and U’(A,) various. A recipient
country can make a proposal that is immediately acceptable to the IMF from its
judgement that it lacks leverage and needs strict guidance from the Fund in order to
improve its economic conditions. In a similar vein even when the Fund initiates a
negotiation, it can propose a proposal from certain needs that can be accepted by a
recipient country without resistance.

Assumption 4 is concerned with the probability of a recipient or a donor being
ultimately successﬁrl in gaining its goals, given the alternative combinations of strategic

choices by the other player. The probability of succeeding in bargaining is a ﬁmction of

98

the possession of the ﬁnancial capabilities the players expect to have available, although
there could be other factors that inﬂuence the subjective probability of success in
conditionality bargaining. The value of negotiation lottery is, then, the expected
compromise outcome of negotiation. In negotiation, neither player anticipates satisfying
his or her initial demand intact.

Assumption 5 is showing negotiation or bargaining as the process of altering and
modifying values. The modiﬁcation term, y, is based on the concept of leverage and
developed so as to measure value change and creation process. This also incorporates the
concession-convergence process into the model, which was discussed in the previous
chapter.

In Assumption 6, I assume that the conditionality bargaining involves costs. All
types of anticipated costs vary as a ﬁmction of each side’s subjective estimates of
probability of success (P', P") and of the value it attaches to status quo. First, any
resources used to produce certain goods cannot be used to produce others. The values of
the forgone goods measure the cost of that economic activity. The opportunity
materializes ‘alternatives’ of bargaining theory mentioned in the previous chapter.
Alternatives denote an actor’s ability to substitute relationships with others for the one
with the opposing actor in order to gain similar or better outcomes (Hopmann, 1978). The
availability of alternatives may increase actor’s bargaining power by decreasing its stakes
at bargaining with the opponent and enable it to achieve its preferred outcomes. The
opportunity cost is equal to giving up alternatives. To apply this concept to developing
countries engaged in conditionality bargaining, opportunity cost, [3, represents lost

opportunities to bargain with other private and/or bilateral donors, who are known for

99

softer conditionality, while negotiating with the IMF. On the other hand, for the IMF who
is known to be under two kinds of institutional pressures, i.e., disbursing fund and
ensuring loan repayment by recipients, the opportunity cost means losing chances to lend
its ﬁmd to a creditworthier candidate and receive consistent debt service from it. Since
opportunity cost is equal to the value of the required resources in their best alternative
use, it has a positive relationship with the subjective probability of gaining the preferred
conditionality (BF).

Domestic political cost, 9, comes from two sources as follows: Basically, the
introduction of intrusive adjustment to alter the state of economic affairs is likely to
prompt domestic opposition unless the status quo is extremely unbearable. Furthermore,
leaders of leveraged states are likely to face greater cost at the introduction of an
adjustment program compared to those of unleveraged states, all else being equal. The
rationale behind this is that capable states are in a better position to achieve what they
want through other measures and at least to manage a relationship with the donor.
However, if they subject themselves abruptly to foreign economic pressures, not only
failure of their economic policies is publicly declared but so is failure to capitalize on
their leverage in managing their relationship with the international donor. After all, self-
supported economic policy change is generally preferred to one foreign-planted.

The second source of domestic cost is the well-known side effect of structural
adjustment: any policy package for structural adjustment is bound to have various
distributional impacts on different social segments within the society and this sometimes
threatens the stability of the regime in developing countries implementing adjustment

policies (Bienen and Gersovitz, 1985). Leaders of prospective recipient countries are

100

aware of this and try to contract conditionality agreements that minimize domestic
political costs.14 In any event, however, the actual magnitude of political costs will be
adjusted by the subjective belief one has regarding the prospect of success. In particular,
one’s anticipated political cost will decrease as the relative probability of gaining a
preferred conditionality rises. This relationship is captured in 9,(1 - P’). The anticipated
losses from conditionality in the form of political instability and economic adversity will
also vary with leverage. The irony here is that a developing country will suffer from this
cost even when it fails to strike a conditionality agreement with the IMF, because it does
not get a loan it needs for its economic revitalization.

On the other hand, reputation cost, (p, is a different form of political cost, which
only the IMF will pay. There are a few cases when the Fund should be concerned about
its reputation: the Fund can be criticized for being harsh during bargaining; its
inconsistent treatment of recipient countries in conditionality bargaining can also give it a
bad reputation; and defaulted debt service to the Fund and consequential ﬁnancial loss
will raise questions about the soundness of its lending decisions and lessen its
creditworthiness in the international ﬁnancial community that it is expected to lead. In all
these cases, the reputation cost increases with the probability that the Fund will obtain its
preferred conditionality agreement. Thus, (piP'.

Recipient countries are exempt from this cost for two reasons: ﬁrst, repetition of

similar programs in developing countries indicates that present transactions with the IMF

 

'4 The inclusion of domestic political cost is not to argue for a systematic relationship between the
introduction of Fund-supported adjustment program and political instability, as studied by Bienen and
Gersovitz and Sidell (1988). Regardless of its direct causal relationship to political instability, domestic
political cost means to incorporate (temporary and/or lasting) abnormal political reactions that could
coincide with the introduction of an adjustment program and worry politicians in power.

101

are not inﬂuenced much by poor performance, namely not meeting performance criteria,
in past IMF programs. Rather, a case that the IMF hesitates or is inhibited from striking
agreements with a recipient is that it neglected its ﬁnancial obligation as a member to the
IMF. The IMF has a track record of suspending members’ eligibility to use its general
resources and voting rights due to overdue quota payment or loan repayment. Once these
arrears are paid off, however, the IMF resumes the members’ voting rights instantly and
even arranges new ﬁnancial programs for them.15 Second, a recipient’s reputation is
already reﬂected on its opportunity cost. The opportunity cost is in direct relationship
with the size of ﬁnancial windows open to the recipient, and the presence of opportunity
cost to pay altogether connotes its good reputation in international ﬁnancial markets as a
creditworthy borrower. Thus, it is not necessary to add reputation cost term to recipient’s
expected utility ﬁinction.

Assumption 7 is needed to materialize the players’ willingness to continue a
bargaining with an expectation of a better outcome. Both players expect the chances for a
more preferable conditionality agreement to increase, as they continue to negotiate. These
chances, however, are realized with certain costs and players have to decide when to stop
bargaining and accept the result. The volume of discount rate will also be an indicator of
the position that a player holds in conditionality bargaining on the basis of bargaining

leverage.

 

‘5 Announcements of suspension of members’ voting rights due to neglected ﬁnancial obligation and
immediate resumption following fulﬁllment are found in various issues of IMF Survey. To name them.
Guyana (1985), Sudan (1990), Zaire (1994) are just a few among many.

102

3.3 EXPECTED UTILITIES AND PREFERENCES IN CONDITIONALITY BARGAINING

While reviewing bargaining theory, it was conﬁrmed that, as joint decision-
making, negotiations are risky events involving compromise in that one can expect to
obtain something between one’s own demands and the demands made by the other actor.
This makes bargaining outcomes expected utility. The expected utility is calculated as the
sum of value gained if one’s demands are won, weighted by the subjective probability of
winning (P’), and value gained if the opponent’s demands are won, which is weighted by
the subjective probability of losing (1 - P’). The expected utilities of conditionality

bargaining game are formed in this spirit with cost terms added.

3.3.1 Expected Utilities of Outcomes

The recipient and the donor realize gains and losses according to the strategies
available to them. As stated in the assumptions, those gains and losses involve certain
costs that reduce the utility of the demand the players make and are cast as probabilistic
payoffs except for combinations of strategies that lead to immediate acceptance of the
proposal (Agreement C1) and CD), VETO, maintenance of status quo, and Non-Agreement.
The values that each player associates with the event at each of the terminal nodes are

provided below:

103

Expected Utilities for a Recipient Countryi

SQ: U’(SQ)

VETOD: — 0,(1 — P') + [3,-P’

VETOR: B,P’

NA: - [9,(1 - P‘) + [3,-P’]

Agreement 0): U’[A, - 0,(1 — P') - [3,-P']

Agreement 63): 5,{P"[U'(A,- — 9,-(1 — P') - mm + (1 - P’)[U'(Aj— 9,-(1 - P') — (3,1)")1}

Agreement @: 5,2{P’[U'(A, + yr" — 9,9(1 -— P') — am] + (1 — P’)[U’(A,- + yr" — e,(1 - P’) —
51‘1")”

Agreement @1U'[Aj— 0,(l - P') - B,P']

Agreement <5): 5,{P‘[U’(A, + yr" — e,(1 — P’) — 13,135] + (1 - 1>")[U"(Aj + yr" — 9,0 - P’)
-l3rP’)1}

Agreement co; 6,2{P’[U’(A,- — 9,-(1 - P’) — B,P')] + (1 — P')[U'(A,— e,(1 - P’) - B,P')]}

Expected Utilities for a Donor 1

5Q U"(5(2)

VETOD: — (ij’ + 0,11"

VETOR: B,P’

NA: — ((pj-P” + BJ-P’)

Agreement CD: U’I(A,- — (ij’ - Bj-P’)

Agreement @: 5,-{P”[U’I(Aj + yjP’ — (ij’. — [Bl-P’)] + (1 — P’)[U’A,~ + yjP" — (ij’ — B,~P’)]}

Agreement (3): 512{P’IU’(AJ “ (PIP, ‘ [LP/)1 + (1 ‘ PI’)[UI(AI — ‘91P” _ [3111])”

104

Agreement @: U’(Aj — (pl-P” — BJ-P')
Agreement ©: 5,-{P”[U"(Aj — my - (3,9)] + (1 — P’)[U’V(A,- — (1,ij — 13,1351}

Agreement @: 5,2{P,'[U’(Aj + 711)” — (pj-P’ - [31131)] + (1 - 1”.)[U1'Ar + YJ'P’. — (1)11)” ‘ BJP/H}

The basic logic that forms the utilities of VETOD, VETOR, and NA is that these
outcomes do not produce conditionality agreements and neither player’s demands or
preferred conditions (A, and A!) are completely fulﬁlled. Thus, they do not include any
demand term. Facing a huge gap in their preferences, the players may be better off by not
negotiating rather than hammering away incessantly at intractable problems. Neither
would be prepared to accept the other’s offer as it stands, and both might perceive that
nothing would be gained by continuing negotiations in search for better terms; this leaves
them only one option—breaking off negotiations by vetoing. In the case the
conditionality bargaining comes to rupture by the recipient’s veto, that developing
country must be conﬁdent of breaking through its economic hardship even without
bailout by the IMF. Such a recipient may also make domestic political cost caused by its
veto almost obsolete. On top of this, this rupture leaves alive opportunities to contact with
other bi- and multilateral donors. Therefore, the expected utility of veto by the recipient
consists of spared opportunity cost. On the other hand, forced to leave the bargaining
table by veto from the donor, the expected utility for the recipient is composed of
negative domestic political cost in addition to spared opportunity costs. This is related to
two sources of domestic political cost mentioned above: if the recipient is leveraged, its
source of domestic political cost is the ﬁrst type. Its leverage should have properly

managed to have the bargaining done with Agreement CD. This is a case of

105

mismanagement. If the recipient country is unleveraged, the origin of political cost lie
with continued economic hardship with no-loan situation, which it can no longer afford.

On the donor side, veto by the recipient does relieve it from worry about
reputation cost for being tough with the recipient during negotiation. It may be the
recipient country that could be blamed for its uncooperative behavior, not the donor. So
the expected utility for the donor is formed only with opportunity costs put aside. If the
donor vetoes a conditionality agreement, on the other hand, it should expect to incur
consequential reputation cost, even if it spares opportunity cost. So, the combination of
negative reputation and opportunity is the utility for the donor when the bargaining ends
abruptly by its veto.

Non-Agreement has similar disastrous utility for both players. Their desire to win
in the competitive negotiation, claiming more value than they really can achieve, would
have prevented them ﬁ'om reaching a progressive agreement in which they both could
have created a mutually valuable deal. What is worse is they should presumably walk
away even after squandering all other opportunities to sound out other windows of
ﬁnancial contracts while engaged in a long, drawn-out and tough bargaining process that
failed anyway. Therefore, the cost for lost opportunities is added to domestic political and

reputation cost ﬁ'om the failure as the utilities for the recipient and the donor respectively.

3.3.2 Preference Orderings and Restrictions in Conditionality Bargaining

The above assumptions imply that certain restrictions are applied to the ten

outcomes in terms of the order of preferences that leaders are allowed to hold. First, the

106

ranking of SO is very ﬂexible in relation to all other outcomes except for Agreement CD.
A rationale for this is that for the recipient the status quo is equivalent to ongoing
economic hardship regardless of the degree of severity, which may or may not involve
explicitly detrimental costs to the leader of the country. This situation could be
potentially problematic to the regime in power but, fortunately, has not been reiﬁed yet.
In addition to this, opportunities for economic rescue by international donors are not
consumed at all. Therefore, the placement of SQ can be moved up or down according to
the leader’s perception of the situation.

Second, among the non-negotiated outcomes of VETOR, VETOD, and NA,
VETOR is preferred over the others by the recipient since it results in saved opportunity

COStSI

VETOR — VETOD: 13,1)" — [— 9,-(1 — P') + w] = 9,(1 - P’) > 0 ('.'9 > 0,0 < P< 1)
VETOR > VETOD
VETOD —— NA: — 9,-(1 - P’) + 3,1)" — [—e,(1 — P’) - an] > 213,1)" (-.-3 > 0,0 < P <1)

VEHTOD > NA.

Accordingly VETOR > VETOD > NA. VETOR, however, is always inferior to Agreement
CD due to the assumption stating that both players prefer to reach an agreement on
conditionality rather than to come to a rupture.

Third, among the cooperative outcomes, Agreement 0) is superior to all others for
a recipient country. This outcome is so, since negotiations are evaluated as the

expectations from a lottery over the demands made by both parties. Negotiated outcomes

107

are never obtained with certainty, and the probability of gaining its ﬁill demand is never
equal to zero or one: 0 < P < 1.0. As P’ approaches arbitrarily close to 1.0, a recipient

country i prefers Agreement ® to the other negotiated outcomes:

UTA, — 9,-(1 - P’) — B,P'] > P'[U’(A. — e,(1 — P') — B,P')] +
(l - P')[U'(Ar - 9:0 - P') - BrP')1

U'[A,- — 13,-9'] > P’[U’(A, - 13.13)].

Gaining one’s demands with certainty when the other party accepts them is preferable to
negotiating and obtaining them with something less than certainty or obtaining something
less than the initial demand. Thus, the remaining negotiated outcomes are inferior to
Agreement CD.

Fourth, with the same components of utility functions, the preference for
Agreement ® over Agreement ® is determined by the existence of modiﬁcation term, y,

after all other components are canceled out:

91'2{P1[U1(At+ YiPi — 91(1 ‘ Pi) ‘ 91131)] + (1 ‘ P’)[UI(AJ + 1’sz — 91(1 ‘ P') “ 311),)”
> 5:2{PI[U’(A1 - 91(1 ‘ Pi) ‘ 51131)] +
(1 ‘ Pi)[Ui(Aj— 9:“ ‘ PI) ‘ BrP')]}

mR>Ocy>QO<P<l)

108

Through the same process above, Agreement (2) is inferior to Agreement (5):

6.-{P’[U’(Ar — 6,11 - P’) - BrP’)] + (1 - P‘ )[U’(Aj — 9.11 - P') - BtP’H}
—5.{P‘[U’(A. + y,P' - e,(1 - P’) — 13.13)] +
<1 - P')1U"(Ar + W — 9,11 — P') - BrP‘m < 0

—y,P’ <0(-.-y>0,0<P< 1).

Fifth, relationships between Agreement (‘2) and Agreement (:3), Agreement (5.) and
Agreement @, Agreement ® and Agreement (6), and Agreement ® and Agreement ® are
not determined due to the unknown magnitude of y, 0, B and 6 in addition to U'(A,-),
U’(Aj), and P’. For instance, the comparison between Agreement ® and Agreement (3.)

results in:

5r{P’[U’(Ar — 9.0 - P') - B.~P’)1 + (1 - P’)lU'(Ar — 6,-(1 - P’) - BrP')]}
-{8.-2{P"1U’(A.- + n?" — 9.11 — P’) — 13.13)] +
(1- P’)[U’(Aj + YrP’ -— 91(1— P’) - BrP’)1}}
setting x = P’[U'(A, — e,(1 — P') — 13.13)] + (1 — P’ )[U’(A, — 9,(1 — P') — 13.1>’)],
5.x _ 5,2[x + P'(y,P’) + (1 — P’)y,-P']
x - 5,-[x + 7,?)

X(1 — 51) — OI'YI‘PI.

109

As P’ gets arbitrarily closer to 1.0 or O, X becomes U'(A,- — 13,-) or U'(A, — 0,) respectively.
Although (1 — 5,) is always greater than zero (0 < 5) <1), the sign of the equation X(l — 5,)
- 5mP’ will take either of positive or negative according to the real magnitude of each
ingredients, since any relationship among cost terms is not presumed.

Sixth, Agreement G) is another outcome with certainty. However, it consists only
of utility of the demand made by the donor and related costs for the recipient. In any case,
U'(A,) will be greater than U'(A,-) and what determines the ultimate utilities of U’(Aj-) is
necessity of policy reorientation, as mentioned before. Therefore, the recipient country’s
preference for this outcome is purely up to the perceived similarity of policy packages
that both sides bring to the bargaining table, which is situational in proportion to
leverage.

On the donor side, ﬁrst, similarly, SQ does not involve clearly visualized costs.
Keeping things as they are saves the organization from the risk of lending ﬁands to less
creditworthy borrowers and international criticism for being a tough bargainer. What
might hold the IMF back from going for status qua is the so-called internal lending
pressure. With the dramatically reduced demand for the IMF ’5 fund from the developed
states, staying out of development ﬁnancing will give rise to a problem of generating
revenue by circulating it. However, the utility of SQ could be very ﬂexible below
Agreement 69.

Second, among the non-negotiated outcomes, preferences go in the order of

VETOR, VETOD, and NA. This is proved as shown below:

110

VETOR —VETOD: 13,11" — (—tp,P’ + 131)")
(pJP’> 0 (-.-tp,> 0,0 < P<1).
VETOD - NA: -(ij’ + 13,.1)’ —( up)!” - 13,1N‘)
213,11" > o (-.-13,> 0,0 < P<1.0).

Therefore, VETOR > VETO o > NA.

Third, in terms of cooperative outcomes, the donor’s preference over them is
reversed of the recipient’s. Agreement CD is preferred by the donor for the same reason
that Agreement CD is preferred by the recipient. Despite the probable costs, this outcome
is better than any other negotiated result since it is a sure thing. The placement of
Agreement 0) is variable as a reﬂection of the donor’s perception of its leverage and
dissimilarity of the adjustment proposal the recipient tries to advance to its own. Its
ranking among negotiated outcomes cannot be predetermined but are situational.

Fourth, the comparison between Agreement ® and (9, is the only difference of

which is the modiﬁcation term, and results in yj-P’ that is invariably greater than zero:

51{V[U’(Ar + 101” - <1)/1’ - BM] + (1 - 1011M + Yr-P’ — <1)/1’ - 131*P")]}
‘51{P"[UA(A1 - (DJ-P]. — 1311’" )l + (1 ‘ P’)[U"(A, - (PJ'P’ ‘ [311” )1}
setting Y = P’[U’(A, — my - 13,P’)] + (1 - P’)[U’A,- - qt)?" - B,P’)],
6,(Y+ yJP’) - 5er

Y+ y,1>1 —Y = ”it > o (.11 > 0,0 < P < 1).

111

Through the same process of subtracting one from the other, Agreement (is) is left with
yjP”, which is proved above to be greater than zero when subtracted by Agreement (3).
Therefore, Agreement C6) is preferred to Agreement C3).

Fifth, as with the recipient’s case, the relationship between Agreement (22> and
Agreement (3), Agreement (5) and Agreement ©, Agreement ® and Agreement (.6),
Agreement ® and Agreement (3) cannot have ﬁxed signs due to the unknown terms of
utility components. For instance, the relationship between Agreement ® and Agreement

(6) turns out:

61{P"[U’(Ar — (pr-P" -BJ-P)] + (1 - P)[U’(Ar — (DJPI - [ii-PH}

- 63113-111113,- + n?" — «nit — 13.13)] + (1 — P)[U’Ar + W — tn?" — m1}
setting Y = P"[UJ(A, — (1)113” — 13,9")1 + (1 — Mum, — (pJ-P" — 13,116],

51(Y) — 5,2[Y +(y,1y)1>'+(1 — P’)y,-P’]

Y — 51(Y + yr!)

Y(l - 5,) - 5jyj-P’.

1f the subjective probability of winning its demands on conditionality, P”, approaches 1.0,
Y is equal to U’(Aj — (p, - [3,), and if the probability goes to 0.0, Y remains to be U’I(A,).
All three cost terms of (1)), B), and 5) are positive magnitudes and there is no presumed
relationship such that Y(l —5j) — 5jyj-P’ can take either a plus or minus sign. Therefore,

nothing is told as to which one is the greater of the two.

112

The following are all the ﬁxed restrictions on preference orderings put together:

 

 

Restrictions to Recipient’s Preferences Restrictions to Donor’s Preferences
Agreement 0) > all others Agreement @ > all others
Agreement 69 > Agreement (.6) Agreement C6) > Agreement 63)
Agreement ’6) > Agreement (2.) Agreement CD > Agreement (5)
VETOR > VETOD > NA VETOR > VETOD > NA

With the above restrictions, we are getting one step closer to the entire picture of
the conditionality bargaining process and its outcomes. Utility functions simply represent
an individual’s preference over gambles as willingness to accept risk, not his “true”
intensity of preference (Morrow, 1994). Preference intensity is difﬁcult to identify, let
alone measure, hence we are allowed only to arrange preferences ordinally. In
conditionality bargaining, the way the players put the expected utilities in ordinal ranking
is the indicator of how willing each player is to impose its demands on the other or
accommodate the other player’s in order to facilitate the negotiation. Thus, the preference
ordering is the key factor that determines the so-called types of players. The order in
which the players arrange possible outcomes tells us whether they are a hard bargainer or
a soft one. The types of players are closely related to the concept of BATNA. The types
of players reﬂect how seriously they will consider non-agreement or status qua as their
best alternative to negotiated agreements and where they will place them among the
possible outcomes. By all means this underscores a propensity for reaching an agreement

as the ﬁnal outcome of bargaining.

113

In addition to the types of players, the intensity of self-interest inserted in the
proposal determined by the perceived situation should be considered in order to properly
capture the conditionality bargaining. Driven by the urgency of contracting an agreement
in one way or another, the players tune the tone of self-interest and its correspondence
with its counterpart’s, using leverage. The tone of likeness between the two self-interests
stated in the proposals is revealed by the difference between U'(A,») and U’(A,~). Therefore,
a proposal of strong self-interest is likely to have a sufﬁciently big difference between
them, whereas a weak proposal is expected to show a small difference.

By combining the types of players shown as a propensity for agreement under the
restriction of assumptions with the types of proposal determined by the intensity of self-
interest, I can set up three cases of conditionality bargaining: bargaining of strongly self-
interested proposal with high propensity for reaching an agreement, bargaining of
strongly self-interested proposal with low propensity for an agreement, and bargaining
with weakly self-interested proposal with high propensity for agreement.16 Therefore, it is
possible to form nine different games in total by multiplying three cases for each player
and map the nature of conditionality with the outcomes generated by those nine games.
One last note to the preference orderings of outcomes for each case is that it is only a set
of many possibles since the undetermined relationship among utilities allows more

arrangements. Note that ten is the highest ranking, and one is the lowest.

 

'6 I purposely omit the combination of accommodative proposal with low orientation toward agreement on
purpose because a proposer who makes an accommodative proposal and risks rupture at the same time does
not make much sense theoretically and realistically.

114

mm

Strong proposal Strong proposal Weak proposal

with High propensity with Low propensity with High propensity

for agreement for agreement for agreement
SQ 4 7 4
VETOD 2 2 2
VETOR 3 3 3
NA 1 1 1
Agreement ® 10 10 10
Agreement €23 7 5 7
Agreement 69 9 9 6
Agreement @ 5 4 9
Agreement © 8 8 8
Agreement (6) 6 6 5
130—1le

Strong proposal Strong proposal Weak proposal

with High propensity with Low propensity with High propensity

for agreement for agreement for agreement
SO 4 7 4
VETOD 2 2 2
VETOR 3 3 3
NA 1 1 1
Agreement (D 5 4 9
Agreement ® 8 8 8
Agreement C3) 6 5 5
Agreement 60 10 10 10
Agreement 6) 7 6 7

Agreement (6) 9 9 6

115

3.4 EQUILIBRIUM OF CONDITIONALITY BARGAINING WITH PERFECT INFORMATION

This conditionality bargaining game is built on perfect and complete information.
Each player knows the payoffs of the outcomes and preference over them. This
information is common knowledge in that all players have that information and that the
other player knows that information, and so on. Equilibrium for nine cases of
conditionality bargaining games will be sought in a manner of subgame perfect. A
subgame perfect equilibrium for a game is an equilibrium for every part of the game from
that node forward and requires players’ forward-looking behavior rather than short—
sighted. Players select strategies from alternative paths available to them, anticipating
their subsequent effects and reactions of the other to the chosen course of action.

In a game where both players make proposals reﬂecting their respective interest
strongly with high hopes that they will be able to induce the other to agree on their
proposal, their interaction proceeds PR, PD, CPR, AD -> Agreement @. In a game that the
recipient remains unchanged but the donor is determined to conclude a conditionality
bargaining only with an agreement that reﬂects its interest strongly, and, accordingly, not
so enthusiastic about hammering out an agreement with the recipient country, their
interaction generates Agreement (33 through PR, PD, CPR, AD, too. In the case in which the
recipient of this type faces a donor who is very willing to accommodate the developing
country’s demand on conditions, the moves of the game are PR, AD —* Agreement 0).

In the game where the recipient wants to conclude a negotiation only with an
agreement that carries its interest fully, accordingly, coming to the bargaining table with

low inclination to concede to the donor, and the donor, on the other hand, expresses a

116

high hope not only for reconciling but ﬁilﬁlling its demands to the maximum, the
conditionality bargaining follows the process of PR, PD, CPR, AD —> Agreement @. If both
players are less enthusiastic about reaching an agreement, adhering to their interests, the
game ends up with Agreement ® through PR, PD, CPR, AD. However, when the recipient
who holds ﬁrm onto its interests with little willingness to concede confronts a donor who
is by contrast very willing to accommodate the developing country’s demand in
conditionality agreement, their interaction produces Agreement CD through PR, AD.

However, in a different setting in which the recipient does not advance its
interests strongly and take any chance to rupture the bargaining process with a donor who
is passionate about reaching an agreement and realizing its interest to the maximum at the
same time, the conditionality bargaining goes through ~PR, PD, AR to yield Agreement 69
on the right hand side of the game tree. If such a recipient confi'onts a donor who wants to
swing the bargaining to its interest, accordingly, coming to the bargaining table with low
inclination to concede to the recipient, the conditionality bargaining is likely to proceed
~PR, PD, AR to obtain Agreement @. On the other hand, if both players are very
accommodative in dealing with each other’s demand, the conditionality bargaining
moves back to the left hand side of the tree and produces PR, AD -+ Agreement (0.

The table below summarizes the above outcomes. The information contained in
the table supports the validity of the theory developed, which states that leverage is a
necessary condition for different conditionality agreements. A closer look at the table
enables us to observe the following: ﬁrst, both players’ propensity toward producing
negotiated agreements, in other words the types of players, is not the determinant of

different conditionality agreements, if both players possess some sort of leverage to

117

Table 3. 1 Outcomes of Conditionality Bargaining with Perfect Information

 

 

 

 

Recipient SH SL WH
Donor
SH Agreement ® Agreement ® Agreement @
SL Agreement 6) Agreement ® Agreement @
WH Agreement 0) Agreement CD Agreement CD

 

Note: SH: Strong proposal with High propensity for agreement
SL: Strong proposal with Low propensity for agreement
WH: Weak proposal with High propensity for agreement

mobilize to the bargaining. Similar outcomes cluster by the nature of the proposal based
on leverage, not by the propensity for agreement. It appears that the outcomes of
conditionality bargaining are determined by whether or not the players can mobilize
leverage and how effective it is rather than how willing they are to negotiate despite the
lack of leverage. Second, under the circumstance that both players mobilize leverage for
bargaining, the recipient seems to use it with more efﬁciency than the donor, since at the
conclusion of a game, the outcomes bring higher utility to the recipient than to the donor.
Third, in a bargaining situation where both players possess leverage, the recipient can
endure a lengthy bargaining process and eventually attain a relatively desirable outcome,
Agreement ©, which is ranked second highest. Fourth, when conditionality bargaining is
asymmetric, that is, only when one party of the bargaining possesses leverage and
capitalizes on it, this player can end the bargaining with its most preferred outcome,
namely, Agreement CD or Agreement @. This is similar to taking advantage of the other

player’s constrained situation. Finally, if players lack leverage, there is a tendency for

118

them to take a course that can quickly ﬁnish the bargaining with outcomes that do not
involve discounting, which is in both players’ interest.

In interpreting the outcomes, three Agreement (D’s on the bottom row need
special attention. Both cases where a strong recipient meets a weak IMF and a weak
recipient meets a weak IMF produce the same outcome, Agreement CD. However, the
content of the agreement or its meaning of three of them will differ. That is, in the case
that both players are weak, their priority is contracting an agreement more than anything
else and, thus, their policy proposals will show high similarity to each other, which will
make the difference between U’(A,) and U’(Aj-) {U’(A,) -U’(Aj)} small as explained earlier.
On the other hand, in a bargaining where a strong recipient deals with a weak IMF, their
proposals have more dissimilarity than similarity so that the difference between U’(A,)
and U’(A,~) is substantial and most of U'(A,-) will be satisﬁed. Therefore, Agreement CD
resulting from negotiations between players with low leverage means joint gains for both
players, while negotiations involving asymmetric leverage mean that players will beneﬁt
differentially.

In conclusion, the results of conditionality bargaining games show that the
imposition of strict conditions by the IMF is neither inevitable nor universally applied.
The conditionality bargaining games, with only three sets of preference orderings, show
enough diversity in outcomes and identify leverage to be an important factor that
determines such differences. The presumed dominance of the donor in conditionality
agreements, rather, seems to be a matter of frequency: that is, what we count as leverage
has been relatively simple and limited and, thus naturally, the number of leveraged

developing countries is smaller than unleveraged ones, which make cases of imposition

119

more noticeable. For instance, if we count oil endowment as leverage, the number of
developing countries leveraged in that sense is only 14 out of 169 (8.3%) in the low— and
middle-income brackets.l7 Often, conditionality bargaining involves a strong IMF and a
weak recipient, whereby both parties exhibit a high propensity for agreement. In such a
situation, a high ﬁequency of Agreement @ is expected. If that is the case, this
automatically resolves the issue of whether the results will be different depending on the
bargaining initiator. Not many developing countries might initiate negotiations with the
IMF unlike the IMF’s ofﬁcial statement on the initiation of conditionality dialogue. ’8 To
the contrary, they would wait for the IMF to approach them ﬁrst and instantly accept the
solutions that IMF comes up with.19 On the other hand, leveraged developing countries in
any case would capitalize on their ability in order to reap preferred outcomes from the
IMF even in a crisis, although it is not that observable to outsiders. The point to be made
here is that, as the outcomes of the game model developed in this chapter show, the
possibility of negotiation and subsequent outcomes exists in every bargaining situation,
no matter what developing country may be involved. The differences in results come

from whether or not and how they take advantage of such possibilities.

 

'7 The classiﬁcation of countries and the number are from World Bank. 1994. World Development Report.
Washington, DC: World Bank

‘8 Recall Footnote 6 of this chapter.

‘9 Evidence supporting this is implied in the IMF ’5 response to the criticism of its intrusive and same old
measures for Southeast Asian countries that had experiences in indigenous economic development The
IMF said that intrusiveness, if any. was inevitable due to the urgency of the crisis; many developing
countries, including Southeast Asians, did not heed warnings from the IMF nor come to the IMF until their
economic problems became a total disaster. Thus, while coping with crises with little time and prioritizing
stabilization, IMF-supported programs had been perceived as such. For the full text, see Fischer, Stanley.
“The IMF and the Asian Crisis.” Forum Funds Lecture at UCLA, Los Angeles. [Online] Available
http://www.imforg/external/np/speeches/l998/032098.htm, March 20, 1998b.

120

3.5 THEORETICAL PROPOSITIONS OF CONDITIONALITY BARGAINING AND HYPOTHESI S

The equilibrium outcomes serve as the basis for generalizing conditionality
bargaining. The propositions that follow specify under what conditions tough or soft
conditionality agreements are contracted. The underlying theory is that toughness and
softness are a function of leverage. The propositions of conditionality bargaining are
categorized into two groups: the ﬁrst is concerned with the impact of utility components
on the equilibrium outcomes; and the second pertains to the effects of exogenous factor,
i.e., leverage, on conditionality agreement. Each proposition is accompanied by proof

and, at the end, testable hypotheses are derived from them for future empirical analysis.

Proposition 1: In a conditionality bargaining, as a recipient country applies a higher
subjective probability of leading the bargaining in its preferred way, the toughness of
conditionality will decrease. For the IMF, as its subjective probability increases, the

toughness of conditionality will increase.

The subjective probability is a comprehensive indictor of players’ conﬁdence in
bargaining, which is based on self-evaluation of related economic, political, social, and
international conditions. Thus, for a recipient whose utility is maximized with soft
conditionality, a high subjective probability should be in an inversed relationship with the
toughness of conditionality. On the other hand, for the IMF, of which the utility is
maximized with loan repayment by the recipient, tough conditionality as a means to
achieve its goal is in a direct relationship with its subjective probability of concluding a
bargaining. Mathematically speaking, as the players’ subjective probability increases up

to 1.0, (l - P’) and (l - P’) get closer to zero, which are combined with the utility of

121

accepting the opponent’s demand. This eventually makes the second part of the utility
function obsolete and the negotiated outcome becomes similar to their most preferable
outcomes that do not involve negotiation lottery such as Agreement CD and Agreement @.
Therefore, the toughness of conditionality as utility will be adjusted in their preferred

directions, as the players’ subjective probabilities increase.

Iivpothesis 1a: If a recipient has capital inﬂows adequate enough to meet its
ﬁnancial needs for balance-of-payments gap, its subjective probability goes up
and the toughness of conditionality will decline.

Hypothesis 1 b: If the IMF has adequate capacity to meet recipients’ loan requests,

 

it can apply a high subjective probability to conditionality bargaining and thereby
impose tougher conditions on recipients.

Proposition 2: In bargaining over conditionality, as the expected opportunity cost
incurred by the negotiation process rises, the IMF will increase the toughness of
conditionality. On the other hand, a recipient country, which has a mounting opportunity
cost for contracting a conditionality agreement with the IMF, will not be forced to
contract an agreement with tough conditions.

The opportunity cost represents loss in exchange for beneﬁt from a conditionality
agreement and ﬁinctions to reduce the size of beneﬁt, assuming that ﬁxed preference over
the magnitude of (A1) and (A). If the opportunity cost comes into the utility function,
reducing the total utility of a conditionality agreement, adjusting the ﬁxed magnitude of
conditionality will be a way to compensate for the loss and maintain the total utility of
conditionality, holding other things constant. Under that circumstance, the IMF will

increase its demand for the toughness of conditionality as its opportunity cost increases.

122

On the other hand, the recipient’s incentive to continue a negotiation with the IMF will be
maintained if conditions get less severe as its opportunity cost increases. Otherwise, it
will quit bargaining with the IMF and talk to other donors available that attach soft
conditions to loans. Thus, the opportunity cost has a direct relationship to the toughness

of conditionality for the donor but an inversed relationship to the recipient.

Hypothesis 2a: If the IMF expects an increasing part of its funds to be
consequently apportioned for a recipient in bargaining, its opportunity cost goes
up, whereby it will intensify the toughness of conditionality.

Hypothesis 2b: Ifa recipient has more capital ﬂowing in from private lenders than
the IMF, its opportunity cost for negotiation with the IMF will increase, whereby
the toughness of conditionality will decline.

Proposition 3: In a conditionality bargaining, as the domestic constraints on the contents
of agreement rise in the recipient country, the toughness of conditionality will decrease.

As with opportunity cost, the existence of domestic political cost itself functions
negatively to the total size of utility for the recipient, and thereby, adjustment in the
magnitude of demand for toughness will follow to compensate for the loss relative to the
size of the domestic political cost. Even if the domestic political cost is combined with
the recipient’s subjective probability of losing, the inﬂuence pattern of domestic political
cost remains the same. Holding the recipient’s subjective probability constant, larger
domestic political cost should be offset by increase in the demand for soft conditionality.
Thus, either combined with the subjective probability or not, the toughness of

conditionality will decrease as the domestic political cost increases.

123

Iivgothesis 3: In a recipient country, if its regime is more democratized, the

 

domestic political cost that conditionality bargaining with the IMF causes to it
will also be on the rise, whereby the toughness of the conditionality agreement it
contracts will decrease.

Proposition 4: In bargaining over conditionality, if the reputation cost for the IMF has
already increased or is on the rise, the toughness of conditionality will decrease.

The IMF will acquire a bad reputation if it treats recipients harshly or Fund
supported adjustment programs do not produce expected results in recipient countries.
The existing reputation cost, which is carried over from its past lending, lessen the size of
utility that the IMF pursues as the other cost terms do. The Fund, which believes that
tough conditionality increases the probability of loan repayment by the recipient, will
strengthen the conditionality if the existing reputation cost looms large for the future.
Thus, the IMF will intentionally adjust its demand for tough conditions upward in order
to level the reputation cost in the current bargaining. Thus, as the reputation cost for the

Fund is on the rise, the toughness in conditionality agreement will increase.

Hypothesis 4: If it suffers from increasing arrears in loan repayment from
previous conditionality agreements, the IMF will tighten ﬁirther conditionality to
increase debt service from the recipient.

Proposition 5: In a conditionality bargaining, as the cost for delaying agreement
decreases for the recipient, the bargaining will be repeated over rounds, which will
decrease the toughness of conditionality. On the other hand, the low delaying cost means
increase in toughness of conditionality to the donor.

124

The value of goods is reduced by time factor. In conditionality bargaining the
value of negotiated agreement will be discounted as much as it comes late. Thus,
assuming that the ﬁnal outcomes have the same value, obtaining it now as undiscounted
brings more utility to the recipient than discounted. Nonetheless, if the delaying cost or
discounting is sufﬁciently low for the recipient, the recipient would rather prolong
bargaining to next rounds and wait for a proposal with softer conditions to be presented
than accept a proposal against its interest. On the other hand, if the donor discounts the
future value a lot, concluding the negotiation “quickly” becomes more important for it
than imposing high conditionality on the recipient and, thus, the toughness of
conditionality will be lessened as the discounting rises. A donor under internal lending
pressure would behave like this. Thus, the less the recipient discounts ﬁiture agreement,
the more likely it is to contract agreements in their favor for soft conditionality. For the

donor, the opposite case is applied.

Hypothesis 5a: If a recipient has a declining discount rate, it values the future
bargaining outcome as much as the present one, and can lessen the toughness of
conditionality agreement which it will be contracting with the IMF.

Hypothesis 5b: If the IMF suffers an increasing cost as it delays conditionality
agreement, it will rush to conclude the negotiation with fewer tough conditions.

Proposition 6: In a conditionality bargaining, as the cost involved in rupturing the
negotiation for both players or at least for one is on the rise, the likelihood of concluding
the bargaining will increase.

125

The utility from rupturing conditionality bargaining, whether by vetoing in the
middle of bargaining process or at its last round, does not include any demand term but
only cost terms. Non-agreement after consuming all of the opportunities for loan
arrangement with other lenders and borrowers is the worst among them because, while
vetoes save at least one cost term, opportunity cost for both players, NAs have the players
relentlessly pay all the costs involved in bargaining, which even makes it more attractive
to accept a proposal made by the opponent. As showed in the above propositions, the
presence of cost terms should be compensated for by changes in demand for toughness of
conditionality in the players’ preferred deﬁnition, namely increasing or decreasing
toughness. Thus, unless the players can afford these bare costs, they have a strong

incentive to ﬁnish the bargaining with some sort of agreement rather than no result.

Proposition 7: Holding other conditions constant, the more leverage players have, the
more likely they will contact a conditionality agreement that is closer to their preference.

All negotiated outcomes have similar utility ingredients. What distinguishes them
one from another is discount factor and the ability to modify proposals, y, presented by
the other player. The ability to modify proposal combined with subjective probability has
a positive inﬂuence on the total utility, which can be identiﬁed as leverage. Under that
circumstance, unless the subjective probability is zero, such ability compensates for the
loss incurred by other costs in addition to upgrading the demand for toughness of
conditionality. Moreover, leverage is the factor that enables the players to extend
bargaining into subsequent rounds. Considering that the player’s total utility gets smaller

by time factor as the bargaining is extended, the players will be better off by settling for

126

an agreement in earlier rounds that do not involve discounting than continuing
negotiation unproductively if it goes without inputs from leverage. Thus, if the players
possess the ability to modify presented proposals, the conditionality agreement that they
contract will get closer to their preference because they have leeway to modify the terms

of agreement to a certain degree.

Hypothesis 7a: If the ﬁve Executive Board member countries increase their

 

ﬁnancial contribution to the IMF, they are more likely to request a tough
conditionality agreement.

Hypothesis 7b: If the IMF has disbursed loans to a particular developing country
disproportionately out of its total lending budget, the likelihood that the Fund will
impose tough conditions on that country will decline.

Hypothesis 7c: If the ﬁve major donors within the Fund have a special bilateral
relationship with a recipient, they are more likely to lessen the toughness of
conditionality for the recipient.

Hypothesis 7d: If the ﬁve Executive Board member countries have contributed to
a recipient’s economy with official development aid, commercial investments and
export market access, they can increase the toughness of conditionality by linking
bilateral economic relationship to multilateral conditionality.

Hypothesis 7e: When a developing country has been experiencing a deteriorating

Vi

 

economic condition before a negotiation with the IMF, such a economic condition
will have a direct impact on intensifying the toughness of conditionality
agreement contracted with the IMF.

127

Hypothesis 7f: As the recipient country accumulates more external debt, the
possibility that it lessens the toughness of conditionality agreement by threatening
debt moratorium will increase.

Hypothesis 7g: If a recipient has depended more on the IMF as a source of foreign

 

capital than on other international lenders, that recipient is more likely to
conclude a bargaining with the IMF with an agreement of tough conditionality.

Hypothesis 7h: If a recipient has an economic system that is characterized by
endowment of natural resources valuable to the ﬁve major donors, the recipient is
less likely to contract a tough conditionality agreement with the IMF. On the other
hand, if its economic system depends much on trade with other countries or is
under systemic transition, the recipient is likely to receive tougher conditions.

Hypothesis 7i: If a recipient has provided proﬁtable investment opportunities for
the ﬁve major donors within the IMF, that recipient country will be able to reduce
the toughness of conditionality by using its economic potentials.

Hypothesis 71: If a developing country negotiating with the IMF is dependent on
domestic market more than on foreign markets for its economic growth and has
been pursuing import-substitution strategy with a heavy public sector, the
likelihood of high conditionality will decrease.

Hypothesis 7k: If a conditionality negotiation is not the ﬁrst encounter between a
developing country and the IMF, such a record of previous transactions will
increase the likelihood of high conditionality for the IMF’s sake.

Proposition 8: In a conditionality bargaining, the more a prospective recipient lacks
leverage, the less likely it is to initiate bargaining with the IMF and a greater chance it
will have to conclude the bargaining with a tough agreement preferred by the IMF, once
it begins.

128

As seen in Proposition 7, leverage compensates for the loss from costs and
discounting. Without leverage, the recipient can avoid loss from discounting by
concluding the bargaining in the rounds that do not involve discounting. The outcomes
that come without discounting are Agreement (I) and VETOD on the left-hand side of the
tree and Agreement 69, VETOR, and SO on the right-hand side. Among these, VETOD and
VET OR, are not affordable outcomes due to their costliness as shown in Proposition 6.
Thus, the choice is narrowed down to Agreement CD, Agreement © and status quo. If a
weak recipient is to initiate the negotiation and conclude it with Agreement (1) without
help from leverage, it has to make a proposal in which its demand for toughness of
conditionality, namely (A,), is equal to (A!) in order to assure that the IMF will accept the
proposal immediately, knowing that the IMF is also weak. However, if the recipient is
not sure of the IMF’s strength and accidentally makes a proposal where (A,) and (A!) are
not of the same magnitude, initiating the bargaining, the outcome could be Agreement ®
at the node 4. In this case, Agreement ® is not attractive to the recipient lacking leverage
since its accordingly low subjective probability makes the ﬁrst part of utility function
close to zero and its actual utility will become the same with Agreement @ but
discounting. If Agreement ® and 60 have the same value but the former is discounted, the
recipient will be better off by choosing Agreement G), and thus, it will not initiate the

bargaining with the IMF unless it is leveraged.

Proposition 9: In bargaining over conditionality, as the recipient’s leverage increases, the
credibility of a threat from it to rupture the bargaining increases, which will decrease the
likelihood of tough conditionality.

129

Facing a proposal made by the IMF at node 2, a leveraged recipient will not
accept the proposal but make a counterproposal to the IMF at node 4 with the anticipation
that the IMF will accept its proposal and thereby conclude the bargaining with its second
best alternative, Agreement ©. The leveraged recipient can threaten to rupture the
bargaining at the node 4 in order to strengthen the IMF’s incentive to accept its
counterproposal at node 6, which will otherwise inﬂict more damage to the IMF than
ﬁnishing the bargaining with Agreement © does. Regardless of its strength and
propensity for agreement, the IMF will be better off with Agreement ® than NA by the
given preference orderings. For a strong recipient, VETOR is always inferior to
Agreement ©. Nonetheless, the possession of leverage can make the utility of VETOR
equivalent to saved opportunity cost enlarged while relatively reducing the utility of
Agreement ®. Thus, threat from a recipient with leverage is more credible than that
lacking and the IMF will be induced to accept a counterproposal from it, which will

contain fewer tough conditions than the previous one.

Proposition 6 is not given hypotheses derived from it. The reason is that
Proposition 6 itself is related to the endogeneity of agreement as bargaining outcome, that
is, whether or not the players contract a conditionality agreement. An assumption of the
conditionality bargaining model says that the players prefer contracting an agreement at
the end of the bargaining to rupturing it and, as the model is constructed so, it is related
mainly to what kind of agreement they contract rather than to whether or not the IMF and
recipients strike a bargain. In addition to this theoretical reason, not forming a testable

hypothesis about Proposition 6 can be justiﬁed with the existing numerous conditionality

130

agreements. The long-term dependency between the IMF and developing countries was
mentioned earlier, and such recidivism in conditionality agreement evidences the size of
costs that the opposite case, namely, not-contracting-agreement, inﬂicts on both players
and their strategy to avoid them. Thus, a hypothesis for Proposition 6 is already tested.
Propositions 8 and 9 also lack pertinent hypotheses. They partly share qualities of
Proposition 6 in deriving testable hypotheses. However, a more severe obstacle to
verifying those Propositions does not lie with deriving testable hypotheses from them but
with testing them with proper empirical records. As for now, there are no compiled or
publicly available data on who initiated negotiation with the IMF and whether any
bargaining ruptured at the end. Thus, Propositions 8 and 9 should be content with their

theoretically proven validity for the time being.

This research represents a modest attempt to move beyond the conceptualization
of conditionality based on unilateral imposition by the IMF in order to explain various
outcomes of conditionality bargaining. In this regard, this chapter has successfully proved
that conditionality agreements are an outcome of circumstantial creation depending on
players’ strategy and exogenous factor called leverage, but not of predeterrnination. The
subsequent propositions generalized conditions under which the nature of conditionality
agreement is differentiated in terms of toughness. Although the entire ediﬁce is
constructed on theoretically driven assumptions of the world, we should remain assured
of its robustness due to the exhaustive list of possible outcomes that it provides. Even the
conventional argument of imposition by the IMF, which has been forwarded without

scientiﬁc delves, is explained by the model as part of the diverse outcomes rather than

131

discarded for being incorrect. Our remaining interest from now on is to see its structural
stability and ﬁrmness: that is, whether the ediﬁce can hold up even when it is thrown into

the real world full of random noises. Its next challenge is to weather empirical tests.

132

CHAPTER 4

INTERMEZZO: BRIDGING ABSTRACT TO CONCRETE

The previous chapter theoretically proved that conditionality agreements are a
complex product of factors that players consider as important. Among those factors,
leverage was presented as an essential component in shaping the product in their
preferred ways with regard to toughness or softness of conditionality. Now the validity of
the conditionality bargaining model should be tested by its correspondence with the
record of conditionality agreements contracted in the past. Deriving testable hypotheses
from generalizing propositions at the end of the previous chapter was to serve this
purpose. In this chapter a statistical model for those hypothesis tests will be discussed.
Operationalization of various concepts included in the hypotheses—such as utility
ingredients and leverage—as well as their measurements in variables will be laid out in
detail. Later in the chapter, the issue of selecting a statistical method that properly reﬂects

the purpose of the analysis and collected data will be discussed.

4.1 DATA FOR THE ANALYSIS OF CONDITIONALITY BARGAINING

No matter what statistical method may be chosen to test the hypotheses with,
leverage must be placed on the right-hand side of equation as an independent variable
that likely affects the outcomes of conditionality bargaining. Selecting independent
variables needs to remain relevant to the speciﬁc issue, conditionality, and operationalize

leverage, not power. The following is operationalization and measurement of dependent

133

and independent variables and expected utility components for each player. All of the
data for ﬁnancial transactions between the IMF and recipient countries came from IMF
Annual Report of various years, and 1998 Wold Development Indicator CD-Rom was the
key source for other economic data of recipients. Other sources of data are listed at the
end of the variable description. Abbreviated capitals of variables will be used in each

speciﬁcation. Boxes contain descriptive statistic of variables grouped by their nature.

4.1.1 Dependent Variable: TITNESS

The most important factor to consider in statistical modeling of conditionality
bargaining is choosing a proper dependent variable that can capture the operation of
leverage. The operation of leverage is not directly observable and we can only offer
conjecture that leverage was mobilized and had a certain effect on the bargaining process
by looking at outcomes comparatively. Thus, it is critical to such a conjecture to choose a
correct dependent variable that likely reveals the operation of leverage and thereby
variance in conditionality agreements in terms of toughness.

Although measuring variance in conditionality agreements can be addressed in
many ways, the number of conditions each recipient country agreed upon with the IMF,
TITNESS (tightness) is proposed as a possible outcome of leverage. Supposing that the
conditions are generated through a process of unobservable factors, it is also plausible
that those numbers are positively correlated with the operation of leverage. The

underlying effect of leverage is estimated backward, using those numbers. Even with this

134

 

simple arithmetic number of conditions per program, it should sufﬁce to indicate the
toughness of a conditionality agreement.1

The most reliable source for counting conditions is by all means the Letters of
Intent (LOIs), ﬁnal documentation of negotiation between the two parties. Nonetheless,
LOIs have never been systematically available in the public domain and are not likely to
be so in the near ﬁxture, either.2 Although there exist alternative sources for extracting
conditions such as country reports and working papers prepared by IMF staff, the number
of agreed conditions for structural adjustment will be counted from press releases
contained in IMF Survey, the semimonthly newsletter of the IMF.

Press releases in IMF Survey were deliberately chosen for several merits it
provides. First, press releases are not independent of original LOIs; they are based on
them. Modeling bargaining process with game theory is unavoidably limited to the part of
process publicly observable, and LOIs convey the outcomes from. such process. Press
releases of conditionality agreements constitute a concentrated public face of LOIs.
Therefore, the information and data extracted from press releases are good enough,
although not perfect, for the analysis. Then, a consequential question is how close are

press releases to original LOIs? To answer this, content comparison of two documents,

 

’ Measuring the tightness of adjustment programs in terms of the number of conditions was developed ﬁrst
for the World Bank’s conditionality in Mosley, Paul, Jane Harrigan, and John Toye. 1995. Aid and Power:
The World Bank and Policy-based Lending, Volume 1 Analysis and Policy Proposals. 2"d ed. London:
Routledge.

2 As part of an ongoing initiative by the IMF toward greater openness with respect to its operations and
activities, the IMF is expanding public access to documentary materials in the IMF ’s archives, including
Executive Board documents that are more than ﬁve years old, and other materials that are more than 20
years old. The IMF moved ﬁrst in 1996 to declassify all materials in its archives that were more than 30
years old, but LOIs were excluded because they contained highly conﬁdential information. LOls started to
be publicized from 1997 only with agreement of the involved governments. For the currently available
ones, visit http://www.imforg/extemal/np/loi/mempubhtm.

135

press releases and their publicly available original LOIs, was attempted. The conclusion
from this is that, with some risk of extrapolation, there exists a satisfactorily high
association between them: in the basic area of agreement—such as type of arrangement,
amount, and program length—the sample press releases showed the same content with
the original LOIs.3 More importantly, press releases are judged as faithful to LOIs in
delivering core conditions or themes of adjustment programs. The major difference of
press releases from LOIs lies with the degree of detail: that is, while the sample LOIs
described conditions to be implemented in detail—mentioning policy measures, target or
in-charge government agencies, and sequencing—the press releases dropped many
speciﬁc instruments and time-lines for implementation. The other merit of the IMF
Survey is its coverage. The press releases deliver all agreements in a disinterested
manner, no matter what country may be the other party of contract, and no matter what
type of arrangements it may be. Hence, they enable us to see a complete universe of
conditionality agreements, whilst country reports and staff working papers do not.
Finally, press releases were written to an implicitly ﬁxed format so that information is
delivered in a systematic and consistent manner rather than desultory, which makes data
coding less arbitrary. On the other hand, country reports are more research-oriented,
focusing on issues of interest. Therefore, medium to long-term effects of policy
implementation preceded by conditionality agreements are viewed from fairly theoretical
perspectives. This type of text makes it more difﬁcult for readers to discern conditions

under one arrangement from another. Therefore, for these merits, despite the availability

 

3 This conclusion was drawn from the comparisons of 1997 LOI of South Korea; 1998 LOIs of the
Philippines, Brazil, and Bulgaria; and 1997 Policy Framework Paper of Uganda

136

of original LOIs in limited number and other sources, conditions were solely counted
ﬁom the press releases in IMF Survey.

Now turning to type of arrangements and data collecting point, this research used
398 cases of Stand-By Agreement (SBA), Extended Fund Facility (EFF), Structural
Adjustment Facility (SAP), and Enhanced Structural Adjustment Facility (ESAF). SBA
has been in operation since 1953, EFF since mid-1970, and SAF and ESAF since mid to
late 1980s, accounting for 916 in total. Including other ﬁnancial operations of the IMF—
such as Buffer Stock Financing Facility (BSFF), Compensatory and Contingency
Financing Facility (CCFF) and Systemic Transformation Facility (STF)——the total will
easily go over 1000. However, these facilities were excluded from the counting targets
due to their unconditional or temporary nature. In the case of BSFF and CCFF, member
countries can borrow money from the IMF without the requirement of changing their
economic policies if their economic difﬁculties were caused by detrimental developments
in international economy that were largely beyond the control of the member. Loan
repayment is the only condition attached to these facilities. And STF, while requiring
policy change, was established in 1992 to help economies in transition in Eastern Europe
and Russia and terminated in 1995.

The 398 cases are almost a complete universe of conditionality arrangements in
relatively recent history over the past ﬁfteen-year period, 1983-97.4 The rationale behind

the selection of this time period is that the year of 1983 marked the watershed of IMF’s

 

" The actual number of SBA, EFF, SAF and ESAF since 1983 is slightly more than 400. Missing cases.
fewer than ten, occurred simply because their press releases could not be found, although they were
mentioned in the list of arrangements.

137

involvement in development ﬁnance.5 Although the IMF had been supporting member
countries in restoring their macroeconomic balances since its founding, it was not until
debt crisis broke out in August 1982 when it contracted a series of agreements with Latin
American members, starting with Mexico in January 1983, that the IMF got
systematically involved in policy-based lending in the name of structural adjustment.
Moreover, although IMF Stand-by Agreements had long linked speciﬁc policy measures
to ﬁnancial support, the policy-based lending since the 19803 has been directed to a much
broader range of policies than traditional IMF Stand-bys, been considerably more
directive, and affected more countries than before (Nelson, 1990). Aside from changes in
theoretical underpinnings of programs, simple statistics also evidence changes in Fund
ﬁnancial activities: the number of four types of arrangements for 30 years up to 1982 was
518 out of 925 and total amount of commitments was 55,389 million SDRs. On the other
hand, for the past 15 years, the IMF contracted 407 agreements worth 110,758 million
SDRs. Compared to the ﬁrst 30 years, the average number of arrangements per year
increased by more than 50% and the total amount of commitments by almost 200%.

Once the source of data was identiﬁed, discussion of the range and scope for
condition counting seems to be in order.6 Brieﬂy speaking, the conditions to be counted
from the press releases are policy instruments at a medium level of speciﬁcity.

Adjustment decisions vary along three dimensions: timing, scope and content. The focus

 

5 Margaret G. de Vries, IMF historian, also conﬁrmed that since August 1982, the IMF’s main
preoccupation had been with the external debt problems of the developing countries. Her full article is
“Selected Topics: Historian Traces Origins and Development of Fund Involvement in World Debt
Problem.” IMF Survey (January 7 and 21, 1985).

6 There were some cases that the IMF calls “augmentation” under which initial programs are extended with
increased amount and additional conditions in the middle of or near the end of the program. Such changes
were excluded from counting with a view to maintaining consistency in collecting data and capturing
bargaining strength intact.

138

of variance in this dissertation is content. Each episode of conditionality agreement has
two main components, macroeconomic and structural, under which eight to ten common
policy areas are identiﬁed.7 What is meant by a policy instrument at medium speciﬁcity
is a policy instrument designed to achieve projected policy goals. For instance, if a policy
goal set in the agreement is reducing overall government budget deﬁcit, the recipient
country can attain it through either cutting government expenditures or increasing
revenues or through a combination of both. These two measures can be even further
speciﬁed to limiting increases in government payrolls and increasing income taxes by a
certain rate, respectively. From such a cascade of speciﬁcity, policy instruments to be
counted at medium speciﬁcity in the ﬁrst place are limited to reducing government
payrolls and increasing income tax rates. If conditions at the targeted level are not
available in the press releases, however, counting goes one level up. Although policy
instruments of one level down (i.e., staged implementation of a policy with timeline)
were released in some cases, those speciﬁc policies were excluded from counting in order
to maintain consistency in counting.8

Another counting item from the press releases concerns various numeric

macroeconomic targets. Strictly speaking, numeric targets are not conditions. However,

 

'7 The components of adjustment programs can be categorized into three: macroeconomic, structural, and
mesoeconomic. While macroeconomic and structural components relate to the generation of wealth,
mesoeconomic components mainly relate to transfers and distribution of resources through expenditure
policies in a given macro-context. However, some meso-policies can affect generation of wealth directly
and, thus, two-categorization of policies by inserting meso-policies into either macro or structural as
appropriate was adopted for simplicity’s sake. For the role of meso-policies in generating wealth, see
Stewart, Frances. 1987. “Altemative Macro Policies, Meso Policies, and Vulnerable Group.” in Adjustment
with a Human Face, Volume I Protecting the Vulnerable and Promoting Growth, eds. Giovanni A. Cornia,
Richard Jolly, and Frances Stewart Oxford: Clarendon Press; and Stewart, Frances. 1995. Adjustment and
Poverty: Options and Choices. London: Routledge.

8 Condition-counting scheme, which is broken down into policy components, is attached at the end of the
dissertation as Appendix 1.

139

conditions for structural adjustment are policy measures to achieve those targets and,
accordingly, they delineate policy options as conditions. In essence, such targets are the
ones that tell us the degree and extent to which recipient’s economy should be contracted
in order to undertake structural adjustment; thus, they are a true indicator of qualitative
toughness of conditionality agreements. Delicate adjustment of macroeconomic targets is
as much an item for serious bargaining as is including and excluding certain policy
instruments (Such importance of numeric targets will be well shown later in the chapter
concerning the Korean case). Therefore, a rationale for counting them as conditions is
that they conﬁrm whether or not those targets were explicitly agreed, although their
qualitative nature is completely lost while being counted. The following is descriptive
information on the frequency of arrangement types (ARRGTTYP) and the dependent

variable, TITNESS.

Table 4. 1 Descriptive Statistic of Type of Arrangement

 

 

ARRGTTYP Freq. % Valid % Cumul. %
SBA 258 64.82 64.82 64.82

EFF 34 8.54 8.54 73.37

SAF 37 9.30 9.30 82.67

ESAF 69 17.34 17.34 100

 

Table 4. 2 Descriptive Static of Tightness of Conditionality

 

N Min. Max. Mean Std.Dev.

 

TITNESS 395 3 33 12.89 5.03

 

140

Retrieving the press releases of conditionality agreements had an additional merit
that it provided a clear-cut answer to the question of where the variance in conditionality
agreements is located. All the conditionality agreements are similar in that they are
structured to cover up to ten macroeconomic and structural areas of economy.9 The
conventional argument for the inﬂexibility of conditionality agreements seems to be the
result of looking only at the general structure of conditionality agreements or monetarist
assumptions.10 The real ﬂexibility of conditionality agreements, however, seems to lie
with policy instruments at the medium and lower level of speciﬁcity, which could not be
properly observed without reviewing the original LOIs or similar documents, such as
press releases. The only problem with this simple counting of conditions is that it is likely
to lose sensitivity of recipient countries to particular conditions. Even if all the recipient
countries agree upon the supposedly same set of conditions, it will have different impact
on individual members due to the difference in their economic context where those
conditions will be implemented. It is not impossible to measure the sensitivity of speciﬁc

conditions with recipient countries, but it does not appear appropriate to use a relatively

 

9 A similar content-analysis was conducted by Stewart (1995). He suggests that the content of IMF
adjustment programs have three elements: demand restraint, switching policies emphasizing exchange rate
reform, and long-term supply policies. Demand restraint policies have dominated IMF programs, being a
universal feature, while over half the programs include exchange policies and somewhat lower proportion
long-term reforms. However, this conclusion appears relatively indecisive and less informative in that he
used only 93 cases between 1980-84 (Table 1.5, p. 9).

'0 The variance in conditionality cannot be identiﬁed with fundamental changes in Fund programs. This
variance is more or less related to relative ﬂexibility, given the ﬁxed ﬁame for program design Seen from
the perspective of program design, or basic philosophy of economics, ﬂexibility in the content of IMF
conditionality has not incorporated any fundamental movements on the provisions that have traditionally
formed the bedrock of Fund programs (Killick. 1995). Thus, the inﬂexibility of Fund-supported programs
seems to have to do with the level of perception, fundamental assumptions vs. speciﬁc content interactive
with individual economies. As long as Fund-supported programs are based on monetarist assumptions and
structured that way, they would not look changing. However, if the focus is on speciﬁc conditions and their
impact in a particular economic context, the IMF has been differentiating its approach to developing
countries.

141

large number of cases, since this requires a match between each condition and each
characteristic of an economy, or the cause of economic crisis in each recipient country.

Such in-depth examination better ﬁts small-N studies.

4.1.2 Expected Utility Components

The expected utility functions for the conditionality bargaining game were
explained in Chapter 3. I, therefore, move directly into how to measure components of

these equations.

0 Subjective Probability: DSUBPROB, RSUBPROB

The most daunting job in testing the conditionality bargaining model is to form
formulae for subjective probabilities for each player. Subjective probability is
ﬁmdamentally private information known only to individual players who supposedly
calculate it under bargaining situation and can be based on anything, as the adjective
‘subjective’ indicates. It is still more for researchers who work with conjecture.

As the lender of last resort, the IMF’s subjective probability with respect to
conditionality bargaining is formulated here to represent its possible liability or liquidity
position in relation to member countries from the developing world. The IMF has
institutional arrangements that allow it to borrow from a particular group of member
countries in order to meet extraordinary needs for fund, which indicates that not every

unit of SDR in the pool of ﬁrnd from member quotas can be used to support members or

142

are immediately available.11 Thus, the IMF’s ﬁnancial capacity in relation to members
from the developing world during any ﬁscal year is likely to be determined by two
considerations: legitimate borrowing requests from all of them within pre-set access
limits12 and undisbursed loan carried over from the previous ﬁscal year. The undisbursed
loan cannot be ignored in the consideration of liability since disbursement, as a due
course of periodic installments or sudden request for resumption of installments, could
take place any time during the ﬁscal year. Against this need for ﬁrnds, the size of
immediately usable ﬁrnds without external borrowing will determine the IMF’s
subjective probability. Even though the IMF is a basket of currencies, not all of the
participating currencies are internationally recognized and used as an effective medium
of international transaction. Most countries maintain around 80% of their foreign reserves
with gold and major ﬁve currencies, which are those of ﬁve Executive Board member
countries (see IMF Annual Report, 1998). Therefore, considering the worldwide

acceptance of their currencies, the size of immediately disposable fund will be

 

” Under Article VII of the Articles of Agreement, the IMF is authorized to borrow currencies for the GRA
when the IMF’s liquidity (its stock of SDRs and usable currencies) is low relative to the demand for its
resources. Under General Arrangements to Borrow (GAB) since 1962, the IMF may borrow, in certain
circumstances, speciﬁed amounts of currencies, for four- or ﬁve-year each, from 11 industrial countries or
their central banks: the United States, Germany, Japan, the United Kingdom, France, Italy, Canada, the
Netherlands, Belgium, Sweden, and Switzerland. Following the Mexican ﬁnancial crisis in 1994-95,
concern that substantially more resources might be needed to respond to future ﬁnancial crises prompted a
call on the Group of Ten and other ﬁnancially strong countries to develop ﬁnancing arrangements that
would double the amount available under the GAB, namely New Arrangements to Borrow (NAB).
Representatives of 25 countries reached agreement in principle in 1996 on the features of NAB, which was
subsequently adopted by the Executive Board in 1997. The amount of resources potentially available to the
IMF under the GAB and the NAB combined will be up to 34 billion SDRs. Source: Treasurer’s
Department. 1998. Financial Organization and Operations of the IMF. Pamphlet Series, no. 45,
Washington, DC: International Monetary Fund.

‘2 The pre-set access limits do not include enlarged access to IMF resources to a multiple of quota or
without limits in relation to quota under the Supplemental Reserve Facility (SRF) approved in late 1997.

143

determined by the portion of quotas that the ﬁve Executive Board member pay in their
currencies and securities denoted in their currencies and retained SDRs. ’3 Thus, a formula

for the IMF ’s subjective probability is as follows:

(Currencies and Securities from 5 Major Donors + SDRs) within GRA
(Access Limits x Total of Member Quotas except for 5 Major Donors) + Undisburs ed Loans '

 

The above formula is used only for SBA and EFF, which are run with resources
from General Resource Account (GRA). SAF and ESAF, on the other hand, are disbursed
from an account appropriated to them (Special Loan Account within GRA) and different
lending principles from those of SBA and EFF are applied. Another reason that SAF and
ESAF have to have a separate subjective probability formula is that while all the
members—regardless of their economic size and level of development—can access SBA
and EFF, SAF and ESAF are available only to member countries within lowest income
group designated by the IMF. '4 Therefore, applying one formula to all types of agreement

is misleading. The IMF’s subjective probability for SAF and ESAF is computed from

Special Loan Account within GRA + Accrued Interest from Investment
Access Limits x Total of Eligible Member Quotas + Undisburs ed Loan of SAF/ESAF '

 

The structure of the above formula is fairly the same with that for SBA and EFF

except for Accrued Interest from Investment. The reason for its inclusion in the formula

 

’3 Each member can pay up to 75% of its quota in its own currencies and 25% in SDRs.
‘4 SAF started with 61 lowest income countries in 1986, although China and India indicated that they did

not intend to draw on the resources, and the number of eligible countries ﬂuctuates with inclusion of new
members and graduation of some of them.

144

is because the unused portion of SAP and ESAF ﬁmd is invested in global ﬁnancial
markets until it is needed and both the amount invested and interest gains from it are
substantial. While GRA resources not in use are also invested, the amount invested and
accrued interest gains are almost negligible, compared with those of SAF/EASF fund.
One thing to keep in mind about these subjective probabilities for the IMF is that
they are annual subjective probabilities, not one-on-one bargaining partner-speciﬁc.
Under the circumstances, formulating such a formula is not feasible, although it is more
desirable. To understand it better, imagine that the Fund’s annual subjective probability

imitates the notion of reserve requirements (capital adequacy ratio) in the banking sector.

For the recipient, its subjective probability is determined by its ability to mobilize
fund internationally against its domestic capital needs, indicated by balance-of—payments
current account deﬁcit. Turning to the concept of balance of payments, deﬁcit in current
account should be offset by surplus in capital account, and vice versa, which keeps the
country’s economy in balance. Faced with severe balance-of-payments gap, however,
developing countries have two ﬁnancing options: running down reserves or borrowing
from foreign sources, either commercial banks or ofﬁcial ﬁnancial agencies. However,
since reserve inadequacy has already become a serious problem to many developing
countries, depleting reserves to correct maladjustment in their economies can hardly be
considered as an option for them, and mobilizing ﬁrnds abroad looks more practical
(Bird, 1995). Mobilizing funds internationally includes borrowing from multilateral
lender—such as the IMF, the World Bank, and regional development banks, bilateral aid

agencies in the developed countries and some oil-producing Arab countries, and

145

syndicated banks by issuing national bonds. Therefore, the recipient’s subjective
probability of contracting a preferred conditionality agreement with the IMF is a function
of ﬁrnd externally raised, especially net hind-raising from private lenders, which is
enough to enable it to bypass the IMF virtually leaving the IMF as the lender of last
resort.15 Besides excluding credits ﬁ'om the IMF, borrowing from other ofﬁcial lenders
should also be left out since the ﬁve Executive Board members are taking part in other
multilateral lenders in a similar manner.16 Thus, from the relationship between debt-
creating capital inﬂows from private creditors—usually bonds issued to syndicated banks
(DEBTINFL), and the size of capital need to be replenished, a formula for the subjective

probability a recipient country could apply to its bargaining with the IMF is derived:

Net Debt Creating Capital Inﬂow
Balance of Payments Current Account Deﬁcit ’

 

0 Discount Rate: DDISCOUT, RDISCOUT
In this research the IMF is portrayed to discount negotiation outcomes with the
interest rate it pays to members who hold SDRs more than required from member

obligations. This SDR interest rate is set weekly in accordance with ﬂuctuation in the

 

‘5 Even in case that developing countries have no ﬁnancial windows but the IMF, they can have differing
abilities to withstand lack of access to international ﬁnancial markets without resorting to the IMF, and this
ability is private knowledge. Although this ability is eligible to be subjective probability, this research
sticks to international fund-raising capability as the subjective probability.

'6 This does not eliminate a possibility of borrowing from other oﬂ'rcial lenders. Intricate cross-currents of
support and rivalry among various international institutions involved in development ﬁnance exist (Nelson,
1990), and this could leave ﬁnancial windows open for the recipient However, at least formal or informal
cross-conditionality among creditors and donors was pronounced, and commercial banks have been less
spontaneous in lending to developing countries for a considerable period since the debt crisis in 1982.
Thus, capital ﬂows from private lenders could be a sheer indicator of ﬁnancial ability of the recipient.

146

exchange rates of the ﬁve major donors’ currencies and annual average of them was used
as the Fund’s discount factor.

The recipient country, which is in economic crisis, might discount bargaining
outcomes with the rate at which its foreign reserve assets are depleted. Since the level of
foreign reserves affect imports, the ratio of non-gold reserves to imports expressed in
terms of the number of weeks covered by the stock of non-gold reserves is a proxy rate of
depletion. This data was collected from International Financial Statistics Yearbook

published by the IMF.

0 Recipient’s Domestic Political Cost: RPOLCOST

The political structure of decision making constrains policy choices, and such
institutional constraints are different across institutional setting—namely, regime type.
Usually a democratic regime is known as imposing more institutional constraints than
authoritarian does; a democratic regime has a separate body retaining the right to ratify
conditionality agreements, and citizens of that regime have freedom to express their
dissatisfaction through protesting and vote casting at next election. Whether it is
demonstrations or losing elections, it is a form of domestic political cost resulting from
conditionality agreements, and the magnitude of such cost will be differentiated by
regime type. Therefore, recipient’s regime type or degree of democratization are used as
a proxy measure of domestic political cost. This variable is based on the score of
[democracy — autocracy] within POLITY IIId data set modiﬁed from POLITY III by
McLaughlin et al (1998), which ranges ﬁ'om negative 10 to positive 10. The more a

recipient country is democratized, the less likely it is to agree upon tough conditions.

147

o Opportunity Cost: DOPPOCST, ROPPOCST

The opportunity cost that the IMF has to pay if it concludes an agreement with the
recipient country is lost opportunity to contract agreements with other recipients, which
equally means that certain amount of fund and proceeds from it are withheld. Thus, the
opportunity cost for the IMF is operationalized by ﬁrnd apportioned for an agreement and
interest accruing from it. The sum of agreed amount of loan and periodic ﬁnancial charge
for that loan was taken to measure the magnitude of opportunity cost for the IMF. For
convenience’s sake, ﬁnancial charge was calculated only for the ﬁrst year. For the IMF,
the opportunity cost has a direct relationship to the number of conditions.

On the other hand, opportunity cost for a recipient means forgone chances for
loans with easier terms from other lenders if it has alternative ﬁnancial windows. Capital
ﬂows from bilateral lenders and private ﬁnancial institutions are indicators of the
existence of such opportunity. Therefore, the sum of net loans from bilateral agencies in
the developed countries and private banks will measure this variable, and this variable is

likely to have a negative relationship to the number of conditions.

0 Donor’s Reputation Costs: DREPUCST

Conditionality is a means for the IMF to ensure debt service by the recipient. Any
arrears in repayment can hurt the IMF’s reputation in international ﬁnancial matter, since
arrears suggest that Fund-supported programs could not deal with economic problems in
developing countries properly. Whether it is real cause or not, such situation would raise
a question as to the IMF’s performance. Actually, arrears have become a serious problem

to the IMF during the 19803, particularly towards the end of the ’805, and the IMF

148

emphasized internally and externally timely debt service by recipients. Thus, using
arrears in repayments would be a nice operationalization of reputation cost that the IMF
has to pay. Data of arrears started to be methodically reported in Annual Report from
1986 and so does this variable. In addition, although only handﬁrl of developing countries
had arrears to the IMF, annual total amount of arrears to the IMF rather than one-on-one
country speciﬁc was put in the data. This variable is likely to have a positive relationship
to the toughness of conditionality by having the IMF’s tighten conditions so as to ensure

repayments.

0 Previous Engagement: PREENGAG

Structural adjustment is characterized by recidivism. Many developing countries
contracted conditionality agreements with the IMF over and over and program after
program. This long-term dependency between the IMF and recipient countries is
expected to reveal information about the other player’s type and strategies, and it is likely
to inﬂuence the outcomes of future bargaining. Hence, the outcomes of their interaction
may be on the learning curve, based on the evaluation of past transactions and
performance rather than independent of the past. Although this aspect of interaction
between the IMF and developing countries is not properly captured in the bargaining
game model by iteration, the effect of repeated encounter on the conditionality agreement
can be statistically measured. In other words, transaction records, regardless of the types
of arrangement prior to current negotiation, can measure the iteration. As it will be coded

binary, not the number of previous encounters, this variable does not measure the degree

149

of iteration, however. If the two players have been engaged in multiple ﬁnancial

transactions, the toughness of conditionality will increase.

Table 4. 3 Descriptive Statistic of Expected Utility Components

 

 

 

 

 

EU Comp. N Min. Max. Mean Std.Dev.
DSUBPROB 398 0.14 1 0.43 0.19
DDISCOUT 398 3.85 9.78 6.38 1.84
DOPPOCST 398 1.50 16182 354.24 1255.15
DREPUCST 331 182.7 3496 2324.16 1034.57
RSUBPROB 220 0 1 0.27 0.33
RDISCOUT 348 0 71.7 12.36 11.83
ROPPOCST 388 0 10461.53 563.55 1109.28
RPOLCOST 296 -9 10 -0.83 7.16
PREENGAG Freq. % Valid % Cumul. %
No Previous ENGAG 28 7.04 7.31 7.31

Previous ENGAG 355 89.20 92.69 100

Missing 15 3.77

 

4.1.3 Independent Variable: Leverage

Operationalization of leverage is done only with economic indicators that are

considered to be as relevant to conditionality bargaining, and this is partly a reﬂection of

characteristics of the contemporary world economy, which is highly interdependent.

150

4.1.3.1 Operationalization and Measurement of Leverage for the IMF

0 Size of Lendable Funds: FUNDSIZE

This variable measures ﬁnancial capacity of the IMF, namely how well the IMF is
prepared to lend. If the IIVIF is concerned about its liquidity position17 vis-a-vis its
members, as mentioned earlier, the size of ﬁrnd the IMF can dispose of will be a factor
that determines the tightness of conditionality. Having said this, a problem with this
operationalization is determining how much money the IMF can apportion for lending,
given the ﬁxed pool of ﬁrnd based on members’ quota subscriptions. The IMF might
have principles for allocating its budget, although not shared with the rest of the world,
and such principles could determine the whole size of fund lendable to member countries
during ﬁscal year.18 What is more critical for operationalization, however, is how to
divide the big pie into pieces for separate lendings. Simple amount of annual budget for
lending does not sufﬁce to tell us about how much the IMF can afford individual
members’ requests. Thus, FUNDSIZE enables us to capture a more bargainer-speciﬁc
picture of IMF lending practice than the subjective probability of the IMF does.

Assuming that the annual lending budget is equals to the size of immediately
usable ﬁrnd without resorting to external borrowing, the IMF ’s preparedness for each

lending can be measured as the ratio of the size of disposable fund (the sum of currencies

 

’7 The Fund’s liquidity position vis-a-vis developing countries is different from its overall liquidity
position, which is calculated by the relationship between its immediately usable assets, comprising SDRs
and usable currencies in the GRA, and its liquid liabilities consisting of members’ liquid claims on the
Fund (reserve positions, undrawn balances of ordinary resources and lenders’ claims on the Fund). The
Fund’s holdings of gold are not included in the category of immediately usable resources.

‘8 According to Williamson (1983b), about one-half of the sum total of Fund quota, as a nile of thumb, is

available for lending at any one time. Though in a slightly different context, namely raising quotas,
Williamson’s calculation of supply of and demand for IMF funds is worth referring to mp. 657—659).

151

and securities from the ﬁve major donors and SDRs in GRA) less undisbursed loans, to
the maximum amount of loan that a recipient can request within the pre—set access limits.
As with the subjective probability for the IMF, the ratio is obtained from two formulae in

accordance with the type of arrangement. The following is for SBA and EFF:

(Currencies and Securities from 5 Major Donors + SDRs) within GRA - Undisburs ed Loans
Access Limit by Arrangement Type x Member' 5 Quota '

 

For SAF and ESAF, it follows as:

(SLA + Accrued Interest from Investment) - Undisburs ed Loan
Access Limit by Arrangement Type x Member' 5 Quota °

 

As the size of ﬁrnd increases within which the IMF can comfortably approve loan
requests, the tightness of conditionality will decrease. In other words, as the ratio gets

larger, the IMF is likely to be more liberal and the number of conditions will decline.

0 IMF’s Exposure to Recipients: IMFEXPOS

This variable measures the size of transactions that the HVIF has had with
recipients thus far. It is measured by the cumulative amount of credits in SDR from the
IMF to the recipient up to the time of current arrangement. This is the sum of all credits
regardless of the type of arrangements through which credits were transferred, even
including BSF, CFF, and STF. And it is likely to be in a positive relationship with the

dependent variable.

152

0 Degree of Exposure: DEGEXPOS

While the IMFEXPOS measures the total amount of transactions between the two
players in absolute terms, it needs to be re-viewed from a relative perspective, namely, as
a portion out of total. This variable will eventually be a test statistic of weight the IMF
has put on individual recipients. The conversion of IMFEXPOS into ratio will be done by
dividing it by total outstanding credits to all the developing countries arranged by the

IMF. As the ratio goes up, the number of conditions is likely to decline.

0 Five Major Donors’ Quotas: SMQUOTAS

The ﬁrnd that the IMF manages depends on members’ quota subscriptions, which
is determined by their size of economy at their admission. Quota is a direct indicator of
members’ economic capacity and ﬁnancial contribution to the IMF. Thus, the degree to
which the interest of the ﬁve Executive Board member countries, as a collective actor, is
represented in conditionality or the inﬂuence that they exert on conditionality decision
can be partly measured by their quotas. This variable is the arithmetic sum of quotas in
SDR ﬁ'om the ﬁve Executive Board members. The greater the contribution they make to

the IMF is, the more conditions are likely to be issued.

0 Votes Cast by Five Major Donors: VOTECAST

The IMF elected for a weighted voting system as its decision making rule, which
is based on member’s quota subscription to the IMF. This means that their economic
power within the IMF is directly converted into political one, with which members can

maneuver lending policies toward developing countries in their favor. The percentage of

153

votes cast by the ﬁve Executive Board members out of total votes of the year, was taken
to measure their political power within the IMF. Traditionally the votes cast by the ﬁve
major donor countries took up over 40% of the total, but this is now on the gradual
decline. The more political power they have, the more conditions they are likely to

impose.

0 Amount of Official Development Aid: ODAAMNT

Although it has been traditionally project-oriented, official development
assistance took substantial part of GNP, around 30% at minimum and 80% at maximum,
in many developing countries. Aware of that, the major donors may capitalize on it,
linking tomorrow’s bilateral aid to today’s conditionality bargaining as an attempt to
achieve their preferred conditions. Thus, official development assistance from the ﬁve
Executive Board members to LDCs claims a legitimate entry into the data set. In
measuring the ofﬁcial aid ﬂow to LDCs, it should be noted that total official aid from 20
Development Aid Committee (DAC) countries within Organization for Economic
Cooperation and Development (OECD) and certain Arab countries was used altogether,
not only from the ﬁve major donors. Statistics of official aid published by the DAC
breaks down the ﬂow of aid by origins and destinations. Nonetheless, DAC data based on
creditor reporting system (CRS) concentrate on major recipients and many small
countries were omitted from the list. On the other hand, World Bank data based on debtor
reporting system (DRS) cover all the recipients, but ignores the origin of aid. This
research used World Bank data in order to increase the number of available cases. It is

obvious that using such data will inﬂate the amount of ODA from the ﬁve Executive

154

Board members. However, the official aid from the ﬁve Executive Board members still
makes up more than 70% of total ODA (IMF, 1998a) and, thus, the amount may not be
outrageously inﬂated. The ofﬁcial development assistance is likely to have a positive

relationship with the number of conditions.

0 Amount of Commercial Investment ﬁ'om Five Major Donors: COMINVET

Capital ﬂows from the developed world to the developing do not consist only of
aid. Commercial investments, a broader term for pursuit of proﬁts from commodity
production and ﬁnancial activities, are the other form of capital ﬂow to developing
countries. They can choose one among foreign direct investments, portfolio investment in
equity and bond, and syndicated bank lending. As with the official aid, for capital-
stricken developing countries, hosting commercial investments from the capital-abundant
donor countries could be an important survival strategy. Thus, the donor countries may
take advantage of this as bargaining leverage. The sum of net private capital inﬂows to
LDCs was used to measure this variable. Again, this data was not broken down by the
origin of the capital, but the numbers do not appear to be unduly exaggerated. The
number of conditions is likely to increase as the size of commercial capital inﬂux to a

recipient gets larger.

0 Allowing Access to Five Major Donors’ Market: EXPOACCS
A characteristic of the trade system of developing countries is lack of diversity in
trading partners. Their imports and exports are directed to small number of countries,

mostly former colonizer countries, and this could be a potential problem to them in

155

bargaining process. That is, if country “A” is relatively more dependent on access to
markets in country “B” than “B” is on “A,” then “B” can be in a position to use its
relatively lower dependence to exact concessions from “A” in trade negotiations or to
link it to other negotiations, since “B” would suffer relatively lower cost from reducing or
cutting off trade with “A.”19 This can happen during conditionality bargaining, if the
major donors attempt to increase the toughness of conditionality by changing the degree
of market openness to exports from the recipient. Thus, the degree of market openness to
foreign trade, especially to exports, constitutes one form of leverage. It is measured by
the percentage of exports directed to the ﬁve Board member countries out of the total
exports from the recipient, which was extracted from Direction of Trade Statistics
Yearbook published by the IMF. The more the ﬁve Board members open their market to

the recipient, the more conditions will be agreed upon.

 

'9 This argument should be balanced with the opposite situation that trade dependence helps the debtor. In a
debt rescheduling agreement, a conﬂict of interest among different groups in the creditor country, pitting
investors who want to enforce maximum repayments against consumers and exporters who want to
maintain normal trade relations, can increase the relative bargaining power of the debtor. In punishing a
debtor who repudiates, lenders may be inﬂicting damage on creditor-country citizens who beneﬁt from
trade with the debtor. Debtor countries have some leverage because foreign lenders do not directly beneﬁt
by cutting them off from world markets. See Bulow, Jeremy and Kenneth Rogoff. 1989. “Multilateral
Negotiations for Rescheduling Developing Country Debt.” in Analytical Issues in Debt, eds. Jacob A.
Frankel, Michael P. Dooley, and Peter Wickham. Washington, DC: International Monetary Fund.

156

Table 4. 4 Descriptive Statistic of Donor's Leverage

 

 

DLeverage N Min. Max. Mean Std.Dev.
SMQUOTAS 398 25596 57839 44511.88 11228.50
VOTECAST 398 38.8 41.28 40.06 1.01
FUNDSIZE 398 -17.88 3640.41 177.37 376.13
IMFEXPOS 392 0 6430.67 329.65 630.62
DEGEXPOS 392 0 17.46 1.09 2.04
ODAAMNT 392 —243.95 5438.5 336.17 435.58
COMINVET 387 —467.8 20719 479.98 1899.28
EXPOACCS 328 3.70 93.45 41.95 18.45

 

o Adjacency: ADJCNCY

Geographical proximity has been regarded as a factor that provokes conﬂicts or
facilitates cooperation in international relations. Economic disaster in one country easily
spreads over others in the region and a leader country of the region would be willing to
minimize contagion, containing the crisis at the epicenter, in order to maintain regional
stability. Thus, the ﬁve major donor countries are likely to consider developing countries
in the same region more favorably than those which are not. Following the World Bank’s
classiﬁcations of countries by geographical location, recipients are coded binary if they

are located in the same region with any of the ﬁve major donor countries.

0 Colonial Ties: COLOTIE
It is a well-known fact that many developing countries peg the exchange rate of
their currencies to those of their former colonizers and their trade is highly concentrated

with them. Their former colonizers are now dominant powers within the Executive Board

157

and this history-laden bilateral relationship between them could affect the interaction
within the boundary of the IMF and work favorably toward the recipients.20 Thus,
colonial relationship in the past, which is coded binary, is included in the data set as

political context variable.

0 Military Alliance: ALLIANCE

An indicator of the strength of political relationship between two countries is
whether they are bound by a defense pact. If a recipient county is integrated into an
alliance system led by one of the major donor countries, economic policies are likely to
be inﬂuenced by political considerations. Thus, whether a recipient country is allied to
any of the major donors, especially to the United States, during the year of agreement
will be coded binary and its effect on the number of conditions will be observed. The

source of data is COW alliance data updated by D. Scott Bennett.

0 Military Aid from Five Major Donors to Recipient: MILITAID

Another way of operationalizing political ties between the ﬁve major donors and
recipients is military aid. Military aid can be offered to developing countries within or
outside alliance systems in the form of arms trade, technology licensing and transfer, and
military training. The only problem with operationalizing this variable is the lack of hard
data on military aid. Data on military expenditure as a whole have been plagued with
secrecy and accordingly inaccuracy, and data on military aid as part of them share the

same problem. As a way of getting around this problem, a dummy variable of arms trade

 

2° For examples of the inﬂuence of colonial ties combined with that of geographical proximity, recall
Footnote 20 of Chapter 1.

158

between developing countries and the ﬁve Executive Board members was created. This
assumes that transfer of conventional arms from the ﬁve major donors to LDCs is partly
subsidized by the suppliers, and that subsidy is taken as military aid. Lists of arms trade
during the period of 1983 through 1997 were obtained from SIPRI Yearbook on
Disarmament and Arms Control, and arms trade between the ﬁve Executive Board
members and recipients in the year previous to conditionality agreements was coded

binary.

0 International Context: INTCONXT

Other than bilateral political relationship, the general international context within
which agreements for structural adjustment take place is also important.21 The most
dramatic change in the contemporary international relations may be the end of Cold War.
The demise of communism in Eastern Europe and Russia not only has changed the
landscape of international relations; it has also resulted in world leaders altering their
national interest and the scheme for calculating it from those during the Cold War period.
This effect could be extended to conditionality bargaining. Thus, to trace back up the
impact of the changes in international context on conditionality, a dummy variable of
Cold War distinguished from the year of 1990 was created.22 It is likely that the major
donors were less harsh with recipients during the Cold War period than Post-Cold War

period.

 

2' The effect of context on international relations is well examined in Goertz, Gary. 1994. Context of
International Politics. Cambridge, MA: Cambridge Univ. Press. He classiﬁes contextual effect in three
types, i.e., as barrier, cause, and changing meaning. Applying his theory to conditionality bargaining,
context functions as a cause since it would change states’ interest in Fund lending and conditionality.

22 The breakdown of the Berlin Wall in 1989 is considered as the beginning of post-Cold War period in this
dissertation. However, this breakdown might have effected in international relations from 1990.

159

Table 4. 5 Frequencies of Donor's Leverage in Binary Form

 

 

Freq. % Valid % Cumul. %
ADJCNCY
Not Same Region 285 71.61 71.61 71.61
Same Region 113 28.34 28.34 100
COLOTIE
No Colonial Ties 184 46.23 46.23 46.23
Colonial Ties 214 53.77 53.77 100
ALLIANCE
Not Allied 314 78.90 78.90 78.90
Allied 84 21.11 21.11 100
MILITAID
No Military Aid 305 76.63 76.63 76.63
Military Aid 93 23.37 23.37 100
INTCONXT
Cold War 188 47.24 47.24 100
Post Cold War 210 52.76 52.76 52.76

 

4.1.3.2 Operationalization and Measurement of Leverage for the Recipient

0 Total External Debt: TOTDEBT

One of the ways that the recipient countries can handle the situation in their favor
is capitalizing on their ﬁnancial and economic relations with the major donors by
threatening total moratorium. This is a situation that can be succinctly illustrated by

Keynes’s oft-quoted dictum “if you owe the bank a thousand dollars, you have a

160

problem; if you owe the bank a million dollars, the bank has a problem.” With
sufﬁciently large debt cumulated, developing countries can threaten default of debt
service, and this can drag creditors into economic mishaps. Any breakdown in bank-
debtor country relations would necessarily have repercussions elsewhere and debt crisis
becomes a public affair with creditor governments and international organizations
(Aggarwal, 1987; Chan, 1993). Thus, the total external debt the recipient has
accumulated is eligible to be bargaining leverage. Total external debt, regardless of its
origins and even including credits from the M, will be used. This variable is likely to be

in an inverse relationship with the number of dependent variable.

0 Severity of Economic Crisis

- Balance-of-Payments Current Account Deﬁcit: BOPCUADF;
- Inﬂation Rate: INFLAT;

- Unemployment Rate: UNEMPLO;

- Overall Government Budget Deﬁcit to GDP: BUDGDEF

Many, but not all, adjustment talks with the IMF are preceded by economic crisis
in a prospective recipient country.23 Thus, many aspects of bargaining with the IMF, such
as outcomes and length of negotiation, are likely to be affected by the severity of its
economic crisis. Although economic crisis can be diagnosed with many other symptoms,

this research uses balance-of—payment current account deﬁcit, inﬂation rate of consumer

 

2’ Some recipient countries go into adjustment agreements with the IMF as a precautionary step, even
though their economies do not show obvious signs of economic crisis. Stand-by Agreement for the
Philippines in 1998 was such case, where the Filipino government initiated a talk with the IMF, since it
expected an abrupt depreciation of its currency in the wake of currency crisis in the region summer of 1997.
This was also to cope with balance-of-payments problems in the foreseeable future, which would be caused
by foreign exchange crisis.

161

price index, unemployment rate, and ratio of consolidated budget deﬁcit to GDP as
indicators of economic crisis. Actually these are the problems that the IMF wants to
tackle through structural adjustment in developing countries and, thus, it will be of our
interest to observe their relationship to the toughness of conditionality. The more severe

the economic crisis is, the more conditions a recipient is likely to agree upon.

0 Use of IMF Credit: IMFCRDTS

This variable uses the same data with the variable of IMFEXPOS for the IMF.
However, the meaning it has to the recipient will not be the same as to the IMF. Who will
be beneﬁted from this shared part by both sides in conditionality bargaining will be

determined by other interacting variables.

0 Foreign Direct Investment Inﬂows: FDIINFLO

Moving production facilities to foreign countries of low cost or investing in them
to acquire a lasting management interest (10% or more of voting stock) became common
moves for developed countries, and these multilateral economic activities are the main
driving force behind globalization. While foreign direct investment must be a lucrative
opportunity for both investors and host countries, it is also true that there are associated
risks—the risk of nationalization, as seen in the movement of New International
Economic Order (NIEO) in the late 19703, or when hosting countries become entangled
in domestic political upheavals, or when the relationship between home and host
countries turns sour. Therefore, being a popular target for foreign direct investments or

hosting many foreign direct investments is an asset to that country, which can be taken

162

advantage of through nationalization in case. Thus, foreign direct investment as leverage
in conditionality bargaining will be measured by the size of net foreign direct investment
inﬂows from the major donor countries since they are home countries of the MNCs most
active in foreign direct investment. The more a recipient hosts foreign direct investment,

the less likely it will agree upon tough conditions.

0 Portfolio Equity Investment: EQIINVET

Unlike foreign direct investment, portfolio equity investment does not involve
participation in management. Equity ﬂows are the sum of country funds, depository
receipts, and direct purchases of shares by foreign investors. They reﬂect the
internationally perceived value of the stock market of the country and accessibility. The
global accessibility looms more important than ever as international investors turn their
eyes to the so-called emerging markets with expectation of fast and high returns. Thus, a
recipient country which has provided good investment opportunities may use its
accessibility and proﬁtability as leverage in conditionality bargaining. The amount of the
net portfolio equity investment will be used for this variable, which is likely to be in an

inverse relationship with the number of conditions.

0 Ratio of Trade to Gross Domestic Production: TRADEGDP

A country’s economic structure is also likely to affect its bargaining position vis-
a-vis the IMF, especially if the economy is structured to be self-sufﬁcient or dependent
on other states for food and material resources. Autarky, as state of economy or a national

policy of restricting economic exchanges across borders, is often deliberately pursued

163

 

with a view to maximizing the state’s independence of inﬂuence from other countries and
to making a country more easily govemable. While autarkic economies are likely to
insulate them from vicissitudes in international economy, interdependent economies are
more likely to be subject to external changes and lose their leeway in managing economy.
Thus, the import of economic structure will stand out at juncture rather than in times of
peace, and conditionality bargaining preceded by economic crisis will be one of those
cases. In examining the inﬂuence of structure of economy on conditionality bargaining,
the weight of trade in gross domestic production will be a proxy measure of the degree of
self-sufﬁciency. Recipients that have a high ratio of trade to GDP are more likely to

receive more conditions than the ones who have a low ratio.

0 State-Owned Enterprise to GDP: SOEGDP

One common policy for structural adjustment is reducing the range and scope of
public sector in economic activities and expanding those of private sector. What is
usually done to achieve this goal is restructuring and privatizing state-owned enterprises
(SOEs), accompanied by massive lay-off. This sudden increase in unemployment puts
political pressure on the government under adjustment and can even lead to political
instability. These expected political consequences of structural adjustment may be used to
excuse recipients from tough conditions. Thus, possible domestic rent-seeking behavior
or politics of special interests will be surrogated by the economic activity of SOEs in the
economy (GDP). Here economic activity of SOEs is value added as the sum of their
operating surplus (balance) and wage payments. The magnitude of SOEs’ economic

activity is likely to be in an inverse relationship with the number of conditions.

164

0 Ratio of Imports to Gross Domestic Production: IMPOGDP

Theoretical reason for this variable is the same with TRADEGDP but from the
opposite direction—that is, how much an economy is domestic oriented and how large
domestic market is. However, looking at the structure of economy from the angle of
imports is intended to catch not only the economic, but also the political meaning in
relation to politics of special interest. In many developing countries import-substitution-
industrialization has political rationality to protect particular social segments and
constituencies through economic policy, which will affect conditionality bargaining in
one way or another. Thus, the ratio of imports to GDP is another proxy measure of rent-
seeking behavior in developing economy. Recipients that have a higher ratio of imports
to GDP are less likely to receive a higher number of conditions than others who have a

lower ratio.

0 Size of Urban Population: URBANPOP

Implementing structural adjustment is known to have a different inﬂuence on
social segments. Especially if adjustment accompanies drastic government spending cuts
in terms of subsidies and payrolls, and price increase of imported foodstuff through
devaluation, the social segment that is hit hardest is urban labor class, who was once
beneﬁciary of urban bias. Overurbanization in developing countries sets the principal
conditions for the occurrence and severity of protests against structural adjustment
involving international agencies such as the IMF (Walton and Ragin, 1990), and the
urban labor class is easily mobilized to represent its interest through mass movements.

This urban resistance to adjustment affecting their interest is one of the obstacles to

165

overcome in order for the recipient government to implement externally required
economic reform. Thus, the urban population is used as an indicator of the presence of a
politically charged urban labor class in the recipient country.24 This variable is likely to

reduce the tightness of conditionality as its size increases.

Table 4. 6 Descriptive Statistic of Recipient's Leverage

 

 

RLeverage N Min. Max. Mean Std.Dev.
BOPCUADF 367 -29660 10723 —731.00 2751.08
INFLAT 304 -100 11750 140.49 1051.35
UNEMPLO 154 0.1 56.3 10.55 10.68
BUDGDEF 208 —35.56 6.38 -4.27 4.93
TOTDEBT 392 3.7 140010 11468.89 21769.28
IMFCRDTS 392 0 6430.67 329.65 630.62
FDIINFLO 391 —217.7 10972 223.03 839.20
EQIINVET 391 0 5528.8 83.36 485.95
DEBTINFL 388 -2484.2 9629.5 201.12 1024.55
TRADEGDP 387 6.32 227.93 67.33 36.11
IMPOGDP 387 2.982 131.74 37.69 21.35
SOEGDP 120 0.4 50.1 9.32 8.24
URBANPOP 398 5.2 90.5 42.24 20.26

 

0 Signiﬁcance to International Finance: INFINSIG
While the INTCONXT measures the impact of a broader international context,

more narrowly deﬁned international context, which is relevant to issues at band, should

 

2" Another way to test the impact of urban population on structural adjustment is using the degree of
urbanization measured by residuals from the regression of level of urbanization on GNP per capita, as
Walton and Ragin did in their work.

166

also be considered. The Fund’s activities reﬂect its founding purpose of stabilizing
international economy through monitoring balance of payments and foreign exchange
adjustment in member countries. With the advent of a highly mobile and integrated
international ﬁnancial system in the early 19903, however, the IMF has paid keen
attention to developments in international capital markets, which is also of interest to the
major Executive Board member countries. Furthermore, in conjunction with their
overseas investment activities, the ﬁve Board members are likely to give special attention
to conditionality agreements with the so-called emerging market economy countries. The
signiﬁcance in international ﬁnance of a developing country seems to be hinged on two
things: their investment potentials and the seriousness of its economic problems to draw
attention from international investors. Once a developing country is designated as an
emerging market economy, its investment potentials are quasi officially recognized.
Thus, the ﬁnancial signiﬁcance of a developing country is coded binary if it has been
listed as an emerging market economy in The Economist since 1993 and arranged loans
with the IMF, which were equivalent to 100% or more of it quota subscription as the

indicator of seriousness of its economic problems.

0 Resource Endowment: RESOENDW

Two times’ oil embargo in the 19703 had crippled the world economy and made
Western industrialized countries, which were dependent on imported oil, vulnerable to
external forces. Since then, endowment with natural resources, especially oil, has been
given a special position in international political economy. Even though the power of oil

and the inﬂuence that oil exporting countries exert on the world economy has faded, their

167

status, even if they are in an economic crisis, will have a bearing on lenders’ perception
of the situation, and this will be the case as long as oil remains the vital fuel source for
industrial activities. Thus, incorporating the power of oil into the model will illustrate the
effectiveness of natural endowments as leverage in external ﬁnancial relations. Coding
fuel-exporting countries follows the World Bank’s country classiﬁcation: a recipient is a
fuel-exporting country if oil exports take at least 50% of the total export of goods and

services. If a recipient is endowed with oil, it is less likely to receive tough conditions.

0 Economic System: ECONSYS

The demise of communism in Eastern Europe and Russia meant not only political
democratization, but also transformation of economy into capitalist system. With the
admission of former communist counties and spin-off countries from the Russian
Federation into the IMF, the IMF undertook a mission to march market economy to those
countries, and the transition process in these countries is likely to affect their position in
conditionality bargaining vis-a-vis the IMF. Thus, the economic system that recipient
countries have, whether they are economies in transition or long-standing capitalists, will
be treated as a dummy variable. And being an economy in transition is likely to increase

the number of conditions, rather than reduce it.

168

Table 4. 7 Frequencies of Recipient's Leverage in Binary Form

 

 

Freq. % Valid % Cumul. %
INFINSIG
Not Significant 392 98.49 98.49 98.49
Significant 6 1.51 1.51 100
RESOENDW
Non Oil Endowed 363 91.21 91.21 91.21
Oil Endowed 35 8.79 8.79 100
ECONSYS
Capitalist 316 79.40 81.65 81.65
Former Communist 71 17.84 18.35 100
Missing 11 2.76

 

There are two more independent variables, which will be used for statistical
analysis: member quota subscriptions (QUOTA) and the ratio of amount of agreed loan
to quota (AMNTQOTA). The descriptions of these variables are omitted from this
section since they do not require any operationalization and measurement. They use
information as given by the IMF in Annual Reports and the press releases in IMF Survey.

For reference purposes, their descriptive statistics are provided below.

Table 4. 8 Descriptive Statistic of IMF-Recipient Financial Transactions

 

 

TRANSACT N Min. Max. Mean Std.Dev.
QUOTA 398 4 4313.1 290.42 501.51
.AMNTQOTA 398 15 1939 89.96 117.50

 

169

4.2 STATISTICAL METHOD: POISSON MODEL FOR COUNT DATA

The number of agreed conditions was chosen to be the dependent variable for
statistical analysis. Those numbers are discrete, not continuous, and can be said to be
counts of events that occur in a particular way: events that an underlying social system
produces randomly during a ﬁxed time period and, at the end of the period, only the total
of these events is observed. Although it remains unobserved, the underlying process that
drives the observed counts is of considerable substantive importance, which statistical
analysis ultimately seeks for, and an event count is the aggregation of this underlying
process (King, 1989). Event counts that results from a process are distributed according

to the Poisson distribution. The Poisson distribution is

[apt-wt)"
ytl ’

 

r. ~ myth) =

where Y. is the dependent variable, t refers to time, y, is a particular realization of Y,, and
A, is the rate of event occurrence. This distribution is appropriate for event count analysis
because it can take only on the values of non-negative integers and does not have upper
bound, which theoretically enables us to count an inﬁnite number of events in each
episode. Applying the event count approach to conditionality, component conditions of
an adjustment package are events and the total number of conditions is the count of them.
It could be assumed that there are certain unobservable underlying continuous effects on
the number of conditions. This assumption can also be expanded into that the rate at

which recipient countries agree upon conditions with the IMF can take a probability

170

distribution. That is, the rate at which countries agree upon conditions is the probability
of leverage that each player mobilizes to the bargaining. Thus, the observed number of
conditions is used to estimate the underlying variable. Plugging components of
conditionality data into the distribution formula, Y, is the number of agreed conditions, 1
refers to particular episodes of agreement, y, is a particular item of condition of Y,, and A,
is the rate of condition occurrence or probability of leverage. Since count of conditions
takes only positive values and no speciﬁc restriction to the number of negotiable
conditions has been identiﬁed, the Poisson distribution ﬁts the nature of data.

There are two things noteworthy in applying the Poisson distribution to the data
of conditionality. First, the Poisson regression model in this analysis uses data implicitly
truncated at zero. That is, since the dependent variable is the number of conditions agreed
upon, zero might represent qualitatively a situation that the recipient country did not
contact the IMF at all, let alone reach an agreement on conditionality. On the other hand,
values greater than zero represent the number of conditions that a recipient country
consented to implement in addition to indication of preceding contact with the IMF. In
such a setting, it might make sense to conﬁne our attention to the non-zero observations,
thereby the distribution of dependent variable is truncated at zero. And, deriving the
Poisson distribution requires two key substantive assumptions that events accumulating
during observation period i are independent and they have a constant rate of occurrence,
A. This directs to the principle of complete randomness, where ideally the variance is

equal to the mean, but it is not very natural or practical for all situations (Consul, 1989).

171

In practice this means that the rate of agreeing upon conditions is the same across
individual agreements, which might not be the case.25

The Poisson distribution is one of special probability distributions formed through
experimental situations and used as models of observed phenomena. That is, by listing
expected probability of a random variable, those distributions only allow us to conﬁrm
whether the underlying process approximates a speciﬁc distribution. Those distributions
in themselves are not to measure the effects of particular independent variables, which is
of interest to the majority of cases in econometrics. To obtain the results we intend to
with probability distributions, the Poisson model should be transformed into one that
combines regression concepts so that we can estimate the unobserved rate from
independent variables (Martin, 1992). Thus, from the Poisson distribution, a Poisson
regression model to ﬁnd the unobserved continuous process is speciﬁed as a function of

these explanatory variables,

E( Y t) = A: = CXPOCA B),

where x, is a vector of explanatory variables and B is a parameter vector, indicating the

effect of each explanatory variable as the underlying rate of leverage, A.

 

2’ If either of the assumptions is not met, a different distribution for total count, Y ,, will be produced. That
is, by assuming that A varies within an observation—or that a particular form of “contagion” which affect
one another negatively or positively—occurs among individual events making up Y,, negative binomial
distribution is formed In the negative binomial distribution, A is still the expected rate of events. Whereas
Var(Y,) = A, in the Poisson distribution, then negative binonrial distribution sets it as Var(Y,) = A,exp(y),
where y registers contagion. The more events within observation i that either have heterogeneous A or are
positively correlated, the larger the parameter (I2 will be. Although 02 cannot equal one in this distribution,
the smaller 02 is, the closer the negative binomial distribution is to the Poisson. For a full explanation, see
King, Gary. 1989a. “Event Count Models for International Relations: Generalizations and Applications.”
International Studies Quarterly 33123-147; and 1989b. Unifying Political Methodology: The Likelihood
Theory of Statistical Inference. New York: Cambridge Univ. Press. Additional analysis results using
negative binomial distribution are given in Appendix 2.

172

The Poisson model will be estimated using maximum-likelihood techniques with
both systematic and stochastic components, rather than least squares. Conventional least
squares estimation is characterized, among others, by the assumption of linearity of a
parameter or parameters and normality of disturbances and, thereby, dependent variable
is also normally distributed. In addition, least squares method treats negative and positive
observations symmetrically, and involves sums of observations that weight all
observations equally. While only continuous variable can be normally distributed, the
Poisson model is based on discrete dependent variable and, in principle, a nonlinear
regression. The discrete nature of the number of conditions and the preponderance of
small ﬁnite or countable inﬁnite values in the Poisson model suggest that the
conventional least squares technique is not appropriate. On the other hand, maximum
likelihood estimation is a statistical method that makes more explicit use of information
about the presumed shape of the disturbance distribution and the logic of which is best
illustrated in the setting of a discrete distribution (Greene, 1993). Thus, it will be easier to
estimate the parameters of nonlinear equations of Poisson model with maximum
likelihood techniques than with ordinary least squares. The likelihood function of the

Poisson regression is:

L(ﬂly) = Z [vi/3'2. — exp( ex.) — lnytlj.

For the last, it is important to note that the numerical value of maximum
likelihood estimate is interpreted as the probability at maximum of observing a particular

sample, assuming that a hypothetical model with unknown parameter, [3, generated the

173

data. Therefore, the likelihood parameters of the model are not necessarily the marginal

effects as the coefﬁcients in the linear regression model are read.

The interpretation of the theoretical ﬁndings in the previous chapter will be
limited if they are not statistically tested. Thus, this chapter represented a preparation en
route to empirical tests of them in the following chapter. Dependent and independent
variables and a matching statistical method were speciﬁed, which would be used to test
the underlying process of leverage in the conditionality agreement. The key criterion for
selecting these variables was whether or not they were relevant as leverage in the issue
area of conditionality bargaining, and it also took into account the broader contextuality
of conditionality bargaining so that characteristics of international system could be listed
as eligible leverage. Now we move onto testing the speciﬁed hypotheses by grouping

and/or combining the listed variables.

174

CHAPTER 5

EMPIRICAL TESTS OF CONDITIONALITY BARGAINING MODEL

Having the hypotheses and the statistical method speciﬁed, 1 shall now proceed to
empirical tests of the conditionality bargaining model, using the variables described in
Section 4.1.1 through 4.1.3. The focus of analysis is on identifying the unobserved rate of
recipient’s agreeing upon conditions with the IMF (i.e. the impact of leverage on the
bargaining outcomes). In other words, how does the existence or nonexistence of leverage
affect the toughness of conditionality agreement indicated by its number? Since both
parties of the bargaining can inﬂuence the number of conditions, TITNESS, statistical
analyses will be conducted in separation for each player with groups of independent
variables of similar nature. First, it levels ground by testing conﬂicting views on IMF

conditional lending.

5.1 CONTROVERSIES ON CONDITIONAL LENDING PRACTICE

As a reply to criticisms of its inﬂexibility in conditionality, the IMF claims that it
treats member countries fair by strictly applying rules based on quota and credit tranche.
The Fund’s external relations documents (IMF, 1998a) say that conditionality is basically
dependent on whether or not their borrowing is exceeding their ﬁrst credit tranche (25%
of quota), and that high conditionality is automatically followed if a recipient requests loan

of upper tranche or exceeding its quota subscription. Thus, a model speciﬁcation to verify

175

IMF’s claim is formed below, using the members’ quota subscriptions (QUOTA) and the
proportion of agreed amount to their quotas (AMNTQOTA).

Remarking on the tables of statistical results hereafter, robust standard errors are
added to regular estimates of coefficients, normal standard errors, and p-values. The
robust standard errors are White’s heteroscedasticity-corrected variances,‘ which test
heteroscedasticity common in cross-sectional data. Although tests designed to detect
heteroscedasticity are usually applied to ordinary least squares residuals, the White test of
heteroscedasticity is general and does not rely on the normality assumption of classical
linear regression model (Gujarati, 1995), and thus it is applicable to this dissertation.
White’s heteroscedasticity-corrected standard errors can be larger or smaller than
uncorrected standard errors, and the statistical signiﬁcance of regressors can accordingly
change. The statistical signiﬁcance in the following analyses, however, are judged by
normal standard errors, and p-values that lost or gained statistical signiﬁcance with the

White standard errors at 0.05 level are marked with an asterisk (*).

Table 5. 1 Quota Proportionality and Conditionality

 

 

Number of obs = 395

LR chi2(2) = 101.46

Prob > chi2 = 0.0000

Log likelihood = -1 191.7903 Pseudo R2 = 0.0408
TITNESS | Coef. Std. Err. R.Std.Err. z P>|z|
QUOTA | .0001871 .0000223 .0000408 8.382 0.000
AMTQOTA I .0004954 .0000816 .000086 6.070 0.000
_CONS I 2.449968 .0176228 .0221846 139.023 0.000

 

 

' For White’s standard errors, refer to White, Halbert. 1980. “A Heteroscedasticity-Consistent Covariance
Matrix Estimator and a Direct Test for Heteroscedasticity.” Econometrica 48, no. 4:817-838.

176

The above result indicates that the IMF abides by its proclaimed policy. The
number of conditions increases as the percentage of the agreed loan to member’s quota
increases. What appears to be rather perplexing in this result, however, is the sign of
coefﬁcient and the signiﬁcance of QUOTA. Theoretically, the quota per se in absolute
terms is not likely to have a direct bearing on high conditionality until it is transformed
into relative terms as the ratio to the amount of arranged loan, which means there should
be no systematic relationship to conditionality. Or, alternatively, it will sound more logical
if it shows a negative coefﬁcient sign, since the quota subscription is a measure of
members’ economic capacity in the world economy and ﬁnancial contribution to the IMF.
Thus, a fair expectation is that the more quota a member country subscribes, the less
tough conditions it will receive. However, my research did not support this contention.

The IMF’s defense would be strengthened if it can repute an opposite conventional
allegation against conditionality. Structural adjustment is known to have produced more
or less disappointing results and this record was worse in the poorest countries. A putative
blame mounted against the Fund is that it is tougher with the poorest members and usually
puts more conditions on arrangements of concessional nature, such as SAF and ESAF
(Killick, 1995). This is an irony to many, since those arrangements were established to
support special needs of heavily indebted poorest countries. Thus, it is put on a test
whether facilities of concessional nature added extra toughness on conditionality besides
the rule of proportionality above. The Fund’s unequal treatment for poor countries can be
easily operationalized with types of arrangement. If types of arrangement have a different
effect on the number of conditions, the allegation will be veriﬁed. A model for this test has

two dummy variables for arrangement types, EFF and SAF/ESAF, of which SBA is the

177

base or omitted category. By doing so, arbitrarily ranking those types is avoided and
statistical results will reveal their relative inequality. During data collection, it was
observed that some recipient countries contracted more than one agreement with the IMF
at the same time, and this is interpreted as an effort on the recipient side to get around
access limits in order to ﬁll up the balance-of-payments gap. Hence, another dummy
variable for multiple arrangements (JNTARRGT)2 is added to the model.

Table 5.2 indicates that the conventional allegation was not groundless. Since the
variables are dummies, the difference in coefficients does not mean difference in slope, but
difference in intercepts. The facilities with concessional elements turned out to start at a
higher intercept than SBA’s, which means the base number of conditions is higher. And
EFF, although not concessional, requires additional efforts on the IMF side to arrange
external borrowing on behalf of recipient countries, and is also associated with a higher
conditionality than SBA. On the other hand, contracting multiple arrangement at one time
did not show a theoretically expected coefficient sign, nor did it have signiﬁcant inﬂuence
on the toughness of conditionality.

Interestingly enough, the results from the two models support both sides of the
debate on conditionality. The IMF has advocated for its fair practice in conditionality,
invoking the rule of proportionality of quota and conditions, and this defense is upheld. On
the opposite side, a criticism in the ﬁeld that the IMF was not only arrogantly inﬂexible in
designing adjustment programs but also biased against poor countries (which are more

likely to be unleveraged than not), was sustained, too. What is inferred from this result is

 

2 In my data set 48 (12.06 %) cases out of 398 were identiﬁed to be joint arrangement. They were usually
combination of SBA and one of SAP and ESAF.

178

Table 5. 2 Type of Arrangement and Conditionality

 

 

Number of obs = 395

LRch12(3) = 62.33

Prob >ch12 = 0.0000

Loglikelihood=-1211.3519 PseudoRZ = 0.0251
TITNESS | Coef. Std. Err. R.Std.Err. z P>|z|
131le .2817696 .0472401 .0634486 5.965 0.000
SAFESAFDI .2103832 .0320927 .0385329 6.555 0.000
JNTARRGTI -.075636 .0439919 .0555368 -1.719 0.086
_CONSI 2.479571 .0183646 .0265961 135.019 0.000

 

that each side of the debate has chosen one aspect of the phenomenon for its convenience
and unduly emphasized it. Viewed in a more objective manner, this statistical result, while
conﬁrming both sides’ grounds, indicates that IMF lending practice has been basically
standing in the middle ground and that it is the applied method in formulating
conditionality that distinguished results. Conditionality agreements have not been
contracted invariably through one process; in some cases, the IMF fairly abided by the rule
of proportionality, and in others, it did not. This is nothing but a statistical evidence for the
presence of bargaining process in conditionality agreements. That is, whether a recipient
will contract a conditionality agreement following the rule of proportionality or imprudent
treatment is not predetermined but dependent on its bargaining with the IMF. Now the
question is what factor determines the ﬁnal choice of conditionality formula between the

rule of proportionality and harsh treatment? The answer seems to lie with leverage.

179

5.2 THE IMF AND LEVERAGE

If the maintained proportionality is the only factor that affects the toughness of
conditionality, difference in agreements should have a ﬁxed number of conditions in
proportion to the ratio of agreed loan to quota. Apparently, that is not the case, and the
following is an analysis of the IMF’s leveraged behavior to purposely increase or decrease

the number of conditions beyond the proportionality:

0 Expected Utility Components and Conditionality

The components of expected utilities of conditionality bargaining are neither
directly related to leverage, nor do they ﬁrnction like leverage. However, those ingredients
still affect the total value of bargaining outcomes as seen in the propositions, and
accordingly, players’ preferences over them.’ Thus, the relationship between the changes
in the magnitude of components and the number of conditions is examined with the
following speciﬁcation. The double entry of opportunity and reputation costs (simple and
quadratic terms) is due to their curvilinear relationship to the dependent variable. This is a
combined speciﬁcation that directly tests the four hypotheses regarding utility ingredients
(Hypotheses 1b, 2a, 4, 5b and 7k derived from Propositions 1, 2, 4, 5 and 7) at the end of

Chapter 3.

 

3 In a sense, the outcomes of the bargaining model in Chapter 3, especially Agreement 0) through 0, can
be differentiated by the number of agreed conditions, if qualitative comparison of contents is not feasible.

180

Table 5. 3 Donor's Expected Utility Components and Conditionality

 

 

Number of obs = 316

LRchi2(7) = 250.10

Prob > ch12 = 0.0000

Log likelihood=-855.93359 PseudoR2 = 0.1275
TITNESS I Coef. Std. Err. R.Std.Err. z P>|z|
DSUBPROBI .0303856 .0858001 .0851678 0.354 0.723
DDISCOUT I -.0828879 .010829 .0117315 -7.654 0.000
DOPPOCSTI .0001744 .0000239 .0000241 7.300 0.000
DOPPOCSZ I —9.41e-09 1.85e-09 2.39e-09 -5.074 0.000
DREPUCSTI .0003124 .0000785 .0000872 3.979 0.000
DREPUCSZ I —5.15e-08 1.96e-08 2.16e-08 —2.624 0.009
PREENGAGI .0149339 .056665 .0595303 0.264 0.792
_CONSI 2.601685 .1322186 .1363316 19.677 0.000

 

In Table 5.3, the subjective probability is an overall measure of conﬁdence level of
the IMF. Its coefﬁcient sign shows that the number of conditions will go up as the IMF
feels conﬁdent about the bargaining situation based on its ﬁnancial capacity as
theoretically expected, but unfortunately, its statistical signiﬁcance is not there. On the
other hand, other utility ingredients gave satisfactory results. The coefficient of discount
rate and its signiﬁcance are supportive to the model. The more the IMF discounts the
bargaining outcomes, the less tight conditionality agreements are. Also, the coeﬂicient
signs of opportunity cost are in line with theoretical expectations of the model and
comfortably signiﬁcant. As the opportunity cost increases, the Fund compensates for it by
intensifying the toughness of conditionality. However, its quadratic term indicates that

such increase in toughness will be halted at a certain point. The reverse of trend in

181

opportunity cost and tightness of conditionality may be related to fear for default in the
future. That is, the opportunity cost is in exact proportion to the size of presumed loan. If
the size of loan is large enough, the IMF may not reﬂect the face value of opportunity cost
for fear of default by the recipient in the future and, thus, the number of conditions will not
increase continuously.

The different coefﬁcient signs of reputation cost indicates that the number of
conditions will increase as the reputation cost increase, but it will start to decline after the
reputation cost goes beyond a certain point. In this analysis, the reputation cost was
measured by arrears to the IMF, and this trend reveals an interesting behavior of the IMF
in conditional lending. That is, there could be many reasons why recipients fall behind its
debt service to the IMF. A possible cause that some scholars suggest is that structural
adjustment was too defected to generate resources to recipients for debt service, but the
IMF has not realized it (Sachs, 1989; Bird, 1995). The fact that the number of conditions
increases as the reputation costs increases reﬂects this argument. The Fund remains
desensitized to this problem until the arrears become a burden on its ﬁnancial capacity, and
starts to correct program defect from that point as indicated by its quadratic term. The
variable of PREENGAG was included to measure the effect of the repeated encounter
between the IMF and recipients, and the learning effect as an alternative to modeling an
iterated game. The result says that the Fund’s previous ﬁnancial relationship with the
recipient inﬂuences positively the number of conditions but not signiﬁcantly. This means
that the ﬁve major donor’s interest and decision in conditionality is more likely to be based
on other issues and circumstances at the moment than the evaluation of the past

performance of the recipient. This explains the recidivism in conditionality agreements.

182

0 Institutional Arrangement within the IMF and Conditionality

The institutional arrangements within the IMF that allow the ﬁve Executive Board
members to exert inﬂuence on the decision making process in general are likely to affect
the number of conditions, too. In this respect, the most important institutional
arrangement of the IMF appears to be the scheme for ﬁrnd formation: member quota
subscription. About 40% of annually required funds are provided by the ﬁve Board
member countries, and the other 60% comes from the other 177 members. Another
institutional setting of the IMF to be considered is the decision making process using a
weighted voting system, which is in turn based on members’ quota subscription. Within
the IMF, economic power and voting power are two sides of one coin. The overwhelming
ﬁnancial contribution to the IMF enables the ﬁve major donors to collectively enjoy up to
40% of total votes cast. The ﬁve Board members can inﬂuence the number of conditions
in proportion to their ﬁnancial contribution to the IMF through their voting power at
program approval stage. With this speciﬁcation, Hypothesis 7a of Proposition 7 will be

tested.

Table 5. 4 IMF Institutional Setting and Conditionality

 

 

Number of obs = 395

LR chi2(2) = 260.88

Prob>chi2 = 0.0000

Log likelihood=-1109.7953 PseudoR2 = 0.1052
TITNESS I Coef. Std. Err. R.Std.Err. z P>|z|
SMQUOTAS I 4.45e-06 2.85e-06 3.58e-06 1.559 0.119
VOTECAST l -.1781124 .0321259 .0397333 -5.544 0.000
_CONS I 9.467126 1.401448 1.738459 6.755 0.000

 

183

According to the above table, the number of agreed conditions would be on the
rise, as expected, when the ﬁve major donors put more money into the pool of fund, but it
does not have a signiﬁcant impact on the toughness of conditionality. On the other hand,
their voting power converted from their ﬁnancial contribution to the IMF has a signiﬁcant
inverse relationship with the intensity of conditionality. That is, as the portion of votes cast
by the ﬁve donors goes up, the number of conditions to be issued to recipient countries
will decline. The insigniﬁcance of SMQUOTAS and crossing curves of SMQUOTAS and
VOTECAST is seemingly odd, but is a correct reﬂection of trend in ﬁrnd formation and
voting power by the major donors. That is, since 1992, the ﬁve Executive Board members
have donated money in absolute amount more than ever, but still their portion in total fund
formation has declined below 40%, and so has their voting power. Compared with the
situation before 1992, the ﬁve major donors used to enjoy more voting power while
providing less resource. The plummeted conversion rate of l SDR of quota into votes
indicates that their privilege has been edged off for the past ﬁve years. Nonetheless, the
inverse relationship of voting power to the toughness of conditionality in the 19903
demonstrates that changes in internal and external environments surrounding the NF,
which will be seen later, ﬁlled in the major donors’ weakened institutional capabilities and

provided them with opportunities to tighten conditionality"

 

" This discrepancy is similar to what Leech calls “voting paradoxes” at collective level—where a
member’s share of the votes has increased but its share of power fallen or vice versa. For the detailed
description of weighted voting and actual voting power in the context of the IMF, see Leech, Dennis.
1998. Power Relations in the International Monetary Fund: A Study of the Political Economy of a priori
Voting Power using the Theory of Simple Games. University of Warwick CSGR Working Paper 06/98.

184

0 Financial Relations and Conditionality

As the lender of last resort, two types of ﬁnancial relations will affect the number
of conditions that the IMF likes to issue: ﬁrst, the IMF ’3 ﬁnancial capacity (i.e., the degree
of preparedness to accommodate recipient’ request at the onset of bargaining, which will
be a private information); and its ﬁnancial involvement with recipients (i.e., the degree of
exposure to the recipient). The degree of exposure receives special attention since, in case
that negotiation should rupture, fear for default will be a ﬁrnction of outstanding credits
that the IMF has provided for the recipient so far. Thus, modeling with variables for
present and past ﬁnancial relations of the IMF to the recipient will be appropriate in
examining the operation of leverage. This speciﬁcation includes the IMF’s subjective
probability, since it is formulated with the notion of the IMF’s liquidity position vis-a-vis

recipient countries. This speciﬁcation tests Hypothesis 7b of Proposition 7.

Table 5. 5 Financial Relations and Conditionality

 

 

Number of obs = 388

LR chi2(6) = 56.97

Prob > chi2 = 0.0000

Log likelihood = -1184.4156 Pseudo R2 = 0.0235
TITNESS I Coef. Std. Err. R.Std.Err. z P>|z|
DSUBPROBI .289684 .0719904 .0924084 4.024 0.000
FUNDSIZEI -.000083 .0000414 .0000635 -2.006 0.045*
IMFEXPOSI .0006141 .0002035 .0002684 3.018 0.003
IMFEXPOZ I -5.92e-08 2.90e-08 3.85e-08 —2.040 0.041*
DEGEXPOS I -.2057179 .0658543 .0857881 -3.124 0.002
DEGEXPOZI .0096807 .0035157 .0048862 2.754 0.006
_CONSI 2.434571 .0391532 .0547982 62.181 0.000

 

185

In adjusting the degree of the tightness of conditionality with ﬁnancial leverage, the
IMF’s subjective probability plays a signiﬁcant role, unlike in the speciﬁcation only with
utility components. As an indicator of conﬁdence level in liquidity position vis-a-vis
developing countries, the higher subjective probability the Fund has, the more conditions
was requested. Its signiﬁcance says that it has impact on trimming the number of
conditions. The IMF also appears to put more conditions to its loans in order to increase
the probability of debt service by recipient countries when its liquidity position is
squeezing, which is measured by the ratio of lendable fund to possible maximum request
from the recipient (FUNDSIZE). On the other hand, the IMF becomes liberal and
generous with conditionality, if it manages relatively abundant ﬁrnds.5 Especially the
inverse relationship between the size of lendable funds and the number of conditions
agreed should be regarded as a realization of the assumption about the IMF’s utility in
ensuring loan repayment. That is, the IMF, as the lender of last resort, is concerned about
maintaining usable ﬁrnd at a certain level, and thereby, tries to increases the probability of
debt service by recipient countries through high conditionality (Bird, 1995).

The direct relationship between the size of outstanding credits to the recipient and
the number of conditions is apparently correct, taking into account Fund utility in loan

repayment. Its quadratic term, IMFEXPOZ, on the other hand, indicates that the

 

5 This result is in line with a general cyclicity in IMF conditionality noted by Williamson (1983), Cooper
(1983), and Cornelius (1988): IMF conditionality is stricter when the world is in a recession than when
there is a boom. Large fund, which is made possible from a worldwide economic boom, enables members
to pay their quotas on time and clear arrears and/or enough private lending to bypass the IMF. Less
demand for IMF credits, or the IMF’s reaction to the declining use of its credit potential, leads to easier
conditionality. The IMF is likely to make itself attractive as a lender by relaxing conditionality. On the
other hand, when the world economy is in recession, the IMF tightens conditionality to reduce the demand
for IMF credits.

186

toughness of conditionality will be molliﬁed if the IMF has lent beyond a certain level, and
this is interpreted as related to potential threat for default. The only problem with
examining the relationship between the accumulated credits in absolute amount and the
number of conditions is that it is likely to fail to capture a bigger picture of the dynamics
between the IMF and the recipient. Rather the dynamics between the IMF and the
recipient is likely to be better portrayed with the relative size of credits against the total
credits. Thus, the variable measuring this relativity, DEGEXPOS, turns out to have an
inverse relationship with the number of conditions with signiﬁcance. The more portions of
total credits were lent to a particular recipient, the fewer conditions were imposed on that
country for fear of default. Given the curve-linear relationship of DEGEXPOS to the
dependent variable, the quadratic term of DEGEXPOS indicates that the number of
conditions will decrease only until it reaches a certain point of degree of exposure. After

passing that point, the number of conditions will eventually go up again.

0 Bilateral Political Relations and Conditionality

Special bilateral relationship between the donor and the recipient, which is
exogenous to the IMF’s ﬁnancial activities, could be transmitted to the interaction within
the boundary of the IMF and work favorably toward the recipient. Gathering variables that
measure bilateral political relationship based on history, geography and military, a model is
speciﬁed, as shown below. This speciﬁcation is a collection of dummy variables for
whether or not they had such relationship with the major donors in order to test

Hypothesis 7c of Proposition 7.

187

Table 5. 6 Bilateral Political Relations and Conditionality

 

 

Number of obs = 395

LRchiZ(5) = 229.58

Prob>ch12 = 0.0000

Log likelihood = -1125.4455 Pseudo R2 = 0.0926
TITNESS l Coef. Std. Err. R.Std.Err. z P>|z|
ADJCNCY | —.0386721 .0554701 .083053 -0.697 0.486
COLOTIEI -.063427 .0308537 .0372842 -2.056 0.040*
ALLIANCE | -.0113362 .0641274 .0955718 -0.177 0.860
MILITAIDI .0257644 .0353666 .0480312 0.728 0.466
INTCONXT I -.4140473 .0308306 .0370782 -13.430 0.000
_CONSI 2.770656 .0247148 .0294001 112.105 0.000

 

The above results show that the number of conditions is also affected by factors
exogenous to what the IMF is legitimately expected to deal with. Except for military aid,
having geographical proximity, colonial afﬁliation, and alliance bond with one or more of
the ﬁve major donors is better than not, although only colonial relationship among them
has statistical signiﬁcance. And the international political context characterized by the
Cold War up to 1990 ﬁrnctioned to signiﬁcantly reduce the number of conditions, as
expected. Confrontation with communism seemed to procreate pro-stability bias within
the IMF due to the major donors’ interest in maintaining pro-West regimes in recipient
countries and hence help recipients receive fewer conditions, compared with the post-Cold

War period when the base number for agreed conditions moved upward.6 In any event,

 

6 Public choicests such as Vaubel (1991) ascribe the increase in the number of agreed conditions to
organizational politics within the IMF. Stiles (1991) also emphasizes bureaucratic politics among
departments, while building a bargaining model between the IMF and developing countries. Seen from
this point of view, the inﬂuence of international political context on the toughness of conditionality can be

188

the ﬁve major donors seemed to play favoritism toward recipients connected to them in
one way or another. This indicates that, although conditionality is a highly economic topic,
political consideration is always working as the background condition.

These results, however, should be read carefully. Although overall political
context does matter to conditionality bargaining, neither concept of issue speciﬁc power
nor leverage, which this research supports, is disrupted. The two military related variables,
ALLIANCE and MILITAID, are usually regarded as important assets of national power.
However, their statistical insigniﬁcance connotes that they are not effective leverage to
inﬂuence the other, at least in conditionality bargaining, which means that even issue
linkage between economic and military hardly takes place. This is especially true when the
ﬁve major donors organizationally support programs contradictory to their business in
military, such as cutting excessive military budget in developing countries. In these cases,

military related leverage is likely to work negatively rather than positively.7

o Bilateral Economic Relations and Conditionality
Just as bilateral political relationship affects the outcome of conditionality
bargaining, bilateral economic relationship can do the same thing. This is especially the

case with asymmetric dependence of one actor on the other for trade and ﬁnance, whereby

 

spurious, and a learning effect weighs more. However, the inﬂuence of the Cold War on IMF
conditionality appears to be real, when considering the different slopes for increase in the number of
agreed conditions before and after 1990. The chart in Appendix 3 shows this graphically. Thus, a fair
eclectic solution from these two different evaluations of the Cold War effect on IMF conditionality may be
that the increased toughness of conditionality is an outcome of synergy between the learning effect from
bureaucratic politics and the changes in international context. The internal activism of IMF technocrats
would be spurred on by permissible contextual changes.

7 Recall that cutting military budget was newly added to structural adjustment as a secondary objective
(Polak, 1991).

189

each party of the relationship will suffer a different cost of loss when their economic
relationship is disrupted. Naturally, the party who will lose less is likely to take advantage
of it to draw yield from the party who will lose more. The asymmetric dependence
between the ﬁve Executive Board member countries and recipients is easily noticed in the
unidirectional ﬂow of aid and commercial investment from the former to the latter, and the
needed access to the ﬁve Board member countries’ markets for exports. Thus, a model of
leverage for conditionality bargaining manifested by asymmetric economic dependence is

explored below. This speciﬁcation corresponds to testing Hypothesis 7d of Proposition 7.

Table 5. 7 Bilateral Economic Relations and Conditionality

 

 

Number of obs = 317

LR chi2(4) = 121.72

Prob > chi2 = 0.0000

Log likelihood = -945.08062 Pseudo R2 = 0.0605
TITNESS I Coef. Std. Err. R.Std.Err. z P>|z|
ODAAMNT I .000283 .0000389 .0000613 7.285 0.000
COMINVET | .0000341 6.10e-06 5.52e-06 5.602 0.000
EXPOACCS l -.0057731 .0033715 .0040674 -1.712 0.087
EXPOACCZ I .000014 .0000376 .0000431 0.371 0.710
_CONS | 2.622162 .0710782 .0866744 36.891 0.000

 

According to Table 5.7, the ﬁve Board member countries tended to link their
bilateral aids to conditionality bargaining so as to get the recipients to agree on more
conditions. As the ﬁve countries increase the amount of bilateral aid to developing

countries, the number of conditions increase in the same manner. At a profound level, this

190

is also a manifestation of the so-called cross-conditionality. Cross—conditionality is deﬁned
as where acceptance by the borrowing country of the conditionality of one ﬁnancial
agency is made a precondition for ﬁnancial support by the others (Griﬂith-Jones, 1992).
In this case the ﬁve major Board member countries make IMF lending an almost
obligatory part of bilateral debt-rescheduling/new money packages, and loans granted by
the IMF are stringed with tighter conditions in proportion to the size of bilateral loans.
Thus, a multilateral relationship of the recipient with the IMF is in fact transformed into a
bilateral one with bilateral donors through cross-conditionality.

Like ofﬁcial assistance, commercial investments in developing countries from the
major donors also work favorably for the donor to increase the number of conditions
signiﬁcantly. This will be an effective leverage, particularly against recipient countries if
foreign capital inﬂux takes a substantial portion in their balance of payments or GNP.
Furthermore, this leverage can be used as a tool for the donors with which to punish
stubborn recipients when the bargaining comes to rupture or to sanction them in other
cases.

Allowing recipients to access to their markets is also an important leverage. In
simple terms, EXPOACCS says that it has an inverse relationship with the number of
conditions, but comes slightly short of statistical signiﬁcance. However, the positive
direction of the coefﬁcient sign for the quadratic term, despite no statistical signiﬁcance,
indicates that more conditions will be requested as the ﬁve Board member countries grant
more market access to recipient countries. As in trade argued by Blair (1991), asymmetric

dependence becomes leverage in conditionality bargaining as well.

191

5.3 THE RECIPIENT AND LEVERAGE

On the recipient side, leverage that recipients can mobilize in order to elevate their
bargaining position vis-a-vis the IMF is somewhat different than that for the donors.
Basically, recipients countries do not possess institutional tools to effectively deal with the
donors within or outside the IMF, in part because economic crisis hits different countries
at different times with varying intensity, and more importantly, because large developing
countries have more to gain by negotiating singly than jointly (Aggarwal, 1987: Chan,
1993). Despite creditors’ early fears, debtors’ cartel has never been formed, and their fate
in conditionality bargaining seems to be dependent on their individual ability, rather than

as a collective actor. This section is laid out in the similar manner to that for the IMF.

0 Expected Utility Components and Conditionality

As with the IMF, the recipient will see conditionality bargaining from an economic
perspective—that is, whether or not the negotiation will bring it more beneﬁt than cost.
To ﬁnd a way that the recipient chooses the ﬁnal landing point in conditionality bargaining
with cost efficiency, a speciﬁcation only with utility components is formed below, which
intends to test Hypotheses 1a, 2b, 3, 5a and 7k of Propositions 1, 2, 3, 5 and 7.

The subjective probability, which is based on the recipient’s ability to replenish its
ﬁnancial needs abroad, has not only an inverse relationship with the number of conditions,
but also has signiﬁcance. While the subjective probability for the IMF has a direct

relationship with the toughness of conditionality, the recipient’s subjective probability has

192

Table 5. 8 Recipient's Expected Utility Components and Conditionality

 

 

Number of obs = 142

LR ch12(7) = 51.63

Prob>ch12 = 0.0000

Log likelihood = -390.32497 Pseudo R2 = 0.0620
TITNESS l Coef. Std. Err. R.Std.Err. z P>Iz|
RSUBPROB I -.3032604 .0877815 .103158 -3.455 0.001
RDISCOUT] .0039871 .002429 .0029911 1.641 0.101
RPOLCOSTI .0109071 .0037613 .0042106 2.900 0.004
RPOLCOSZ | —.0009615 .0010107 .0011028 -0.951 0.341
ROPPOCSTI .0003112 .0000576 .0000585 5.400 0.000
ROPPOCSZ I -3.16e-08 8.67e-09 8.223-09 —3.646 0.000
PREENGAG! -.106602 .1244539 .1804403 —0.857 0.392
CONSI 2.456332 .13847 .1970486 17.739 0.000

 

a negative relationship, the crossing curves of which seems to be a theoretically and
practically inevitable aspect of conditionality bargaining. In any event, this result indicates
that a recipient country which can raise enough ﬁrnds abroad will receive a smaller number
of conditions than one which cannot, and even is less likely to rush to the IMF for ﬁnancial
help. In the previous chapters, recipients’ desire to bypass the IMF and its actual threat to
the IMF’s raison d’etre was mentioned many places. With this result at hand, we can tell
under what conditions it will be feasible: abundance in world liquidity at system level and
value of individual countries for investment and credibility to draw capital in.

Domestic political cost (measured by regime type) is in a curve-linear relationship
with the toughness of conditionality, hence its quadratic term was added. The ﬁrst term of

regime type shows signiﬁcance, but retains a wrong sign of coefficient. Theoretically, the

193

number of conditions should go down as the domestic political cost rises. On the other
hand, the second order of regime type has a correct coefficient sign but not signiﬁcant. As
long as it is insigniﬁcant, it is not worth discussing the variable further. However, the
parabolic curve of regime type indicates that recipient countries on the brink of
democratization or in the process of consolidation right after democratization (which are
located around zero on the scale of autocracy and democracy), receive the highest number
of conditions. Both absolute authoritarian and absolute democratic regimes enjoy low
conditionality, but the former occurs because it has zero domestic political cost, while the
latter has a narrow win-set for bargaining—as constrained by domestic political process
(Haggard and Kaufman, 1989; Putnam, 1988). The revealed curve-linear relationship
between conditionality and regime type could be a case that rekindles the debates on the
causal direction between structural adjustment and political instability in developing
countries (see Bienen and Gersovitz, 1985; Sidell, 1988).

A recipient’s opportunity cost with quadratic term, in addition to its ﬁrst order, has
a direct relationship with the number of conditions upon initial inspection. However, the
theoretical reasoning contained in opportunity cost is very similar to that of subjective
probability, and thus, it should be in a negative relationship with the number of conditions.
This expectation is well captured by the quadratic term of the variable; that is, the increase
in number of conditions will be reversed at a certain point and eventually decrease as the
recipient’s opportunity cost keeps increasing. This appears to indicate that there is a
minimum opportunity cost, or that there is a minimum amount of foreign ﬁrnd raising in

order for a recipient to reduce the number of conditions.

194

Although it does not have statistical signiﬁcance, the coefficient sign of the
recipient’s discount rate is odd. Since it was operationalized with the number of weeks
that the recipient’s foreign reserves can cover imports, the more weeks of imports the
recipient’s foreign reserves cover, the better it can endure lengthy bargaining with the
IMF, which will eventually reduce the toughness of conditionality. Considering this, rather
than pure theoretical concept of discount rate, the coefficient sign of RDISCOUT should
be directed to negative, not positive, as the recipient’s discount rate increase. Previous
loan agreements indicate long-term dependency of the recipient on the IMF for external
ﬁnancing and, thereby, reveal to the IMF the recipient’s performance record and strategic
preferences, which can be taken advantage of by the IMF. However, contrary to the
hypothesis, the history of previous transactions with the IMF is associated with decreasing
the toughness of conditionality (i.e., lower intercept) though insigniﬁcant. Therefore, this
variable needs a new measurement in the ﬁrture with the actual number of previous

engagement rather than a binary coding.

0 Economic Crisis and Conditionality

Conditionality bargaining takes place ﬁrst with internally and externally sensed
economic crises in a recipient country. Thus, it is appropriate to check out the impact of
economic crisis in individual countries on the bargaining process that drove them to the
IMF for ﬁnancial bailout. The severity of economic crisis will have certain ﬁrnction to
conditionality bargaining in the respect that it inﬂuences the utility ingredients, such as

subjective probability and discounting, and eventually determines the contents of economic

195

reform package. What follows is a model with indicators of economic crisis. This

speciﬁcation tests Hypothesis 7e derived from Proposition 7.

Table 5. 9 Economic Crisis and Conditionality

 

 

Number of obs = 76

LR chi2(4) = 17.72

Prob>chi2 = 0.0014

Log likelihood = -214.84502 Pseudo R2 = 0.0396
TITNESS I Coef. Std. Err. R.Std.Err. z P>|z|
BOPCUADFI .0000473 .0000121 8.75e—06 -3.911 0.000
INFLATI .0000415 .0000458 .0000356 0.905 0.365
UNEMPLOI .0059683 .0060896 .0068607 —0.980 0.327
BUDGDEFI .0074728 .0074973 .0082706 0.997 0.319
_CONSI 2.571146 .0686321 .0769482 37.463 0.000

 

The founding purpose of the IMF is to provide temporary ﬁnancial help for
member countries experiencing imbalances in their balance of payments. As its
organizational purpose directs, the size of the balance-of-payment current account deﬁcit
was the single most signiﬁcant factor, among other indicators of economic crisis, that
weakens the bargaining position of recipient countries. The wider the gap of balance of
payments is (which is to be ﬁlled with ﬁnancing from the IMF), the more conditions
recipient countries have to endure to receive loans from the IMF. Surprisingly,
hyperinﬂation, unemployment, and chronic central government budget deﬁcit in
developing countries have neither expected coefﬁcient signs nor statistical signiﬁcance.

This result could be ascribed in part to the insufﬁcient number of cases for analysis. The

196

effective number of cases used in this speciﬁcation is only 76 out of 398, which is the
lowest of the whole speciﬁcations in this research. This dramatic reduction of observations
might come mainly from the unemployment data, which has only 154 cases available, and

eliminating missing cases by list-wise would decrease usable cases ﬁrrther.

0 Politics of Debt and Conditionality

Foreign debt itself is part of economic crisis in developing countries. Nonetheless,
developing countries can convert their problem into a useﬁrl leverage by threatening
default, through which the ball is hit into the creditors’ court from debtors. Facing the
possibility of moratoria and the ripple effect on their ﬁnancial system, creditors are likely
to roll over the debt or initiate talks for debt rescheduling. A model using variables of
accumulated debt can show how recipient countries can draw out yields from the ﬁve
Board members who strive to prevent possible economic damages. Even in using debt as
leverage, recipients’ strategy can be two-tiered by distinguishing debts by origin (i.e., debt
owed solely to the IMF and total debt stock ﬁom all types of creditor, multilateral,
bilateral, and private sources). This speciﬁcation also utilizes quadratic terms of the
variables, since they are in a curve-linear relationship with the dependent variable. And
INFIN SIG was included to examine the impact of recipients’ status in the world economy
and recognition of their debt problems in conditionality bargaining. This speciﬁcation tests
Hypotheses 7f, 7g, and partly 7i of Proposition 7 together.

The following table reveals an intriguing fact: the effect of leverage played out
with external debt on the number of conditions is circumstantially determined by the origin

of debt. As a recipient accumulates debt, regardless of types of creditors, it coincidentally

197

increases the number of conditions. However, if that country’s debt is large enough, this
effect will be reversed so as to reduce the number of conditions—as the coefficient sign of

quadratic term of total debt, TOTDEBTZ, indicates with signiﬁcance.

Table 5. 10 External Debt and Conditionality

 

 

Number of obs = 388

LR ch12(5) = 99.09

Prob>ch12 = 0.0000

Log likelihood = -1 163.352 Pseudo R2 = 0.0408
TITNESS l Coef. Std. Err. R.Std.Err. z P>IzI
TOTDEBT I 8.88e-06 2.00e-06 2.54e-06 4.447 0.000
TOTDEBTZ I -5.52e-11 2.03e-11 2.18e-11 -2.724 0.006
IMFCRDTS I -.0001054 .0000509 .0000663 —2.069 0.039*
IMFCRDSZ I 1.74e-08 9.98e—09 1.07e-08 1.748 0.081
INFINSIG I .6790757 .1534382 .162366 4.426 0.000
_CONS | 2.493964 .0194144 .0247107 128.459 0.000

 

On the other hand, debt owed to the IMF have a completely opposite effect: a
heavy use of IMF credits, which indicates limited overseas ﬁrnd-raising ability of the
recipient, does not help strengthen its bargaining position vis-a-vis the IMF. Departing
from the initial pattem, IMF credits appear to increase the number of conditions after
passing a point that can be regarded as a threshold. This result is especially fascinating,
recalling that, as a shared variable with the IMF, which measures its vulnerability as a
creditor (IMFEXPOS, IMFEXPO2), their coefficient showed opposite signs with
signiﬁcance (Table 5.5). Taking into account the opposite direction of the curves of

IMFEXP02 and IMFCRDSZ, usage of IMF credits or exposure beyond a certain point

198

appears to make both sides concerned about their positions in the ﬁnancial relationship
with each other and, thus, more compromising than before the threshold point.

In passing, more results are reported about the relationship between the toughness
of conditionality and total external debt. The variable of TOTDEBT was not normally
distributed8 and, thereby, the unbiasedness and efficiency of the estimate was suspected.
To correct this problem, additional functional forms for TOTDEBT such as logging the
variable were attempted in order to ﬁt the normal curve better. Nonetheless, the ﬁrnctional
form did not give satisfactory results in terms of the direction of coefficients and statistical
signiﬁcance. This disappointing result from the transformed variables is interpreted as that
as long as total external debt is concerned with conditionality, its sheer size matters, and
not anything else. This interpretation is agreeable admitting that total external debt is not
only debtor’s problem, but creditor’s as well. The amount of debt accumulated by the
debtor is the exact amount that the creditor should collect and, hence, as the size of debt
gets bigger, the creditor becomes vulnerable. This simple number tells the creditor the
severity of ﬁnancial damage it should swallow if the recipient defaults, which is why a
threat of default from a recipient country can be an effective weapon. Therefore, total
external debt as leverage should be seen from the creditor’s viewpoint rather than the
debtor’s.

The direction of coefﬁcient and the statistical signiﬁcance of INFINSIG is rather

surprising. Theoretically, the emerging market economies are likely to receive a milder

 

8 Among 392 valid cases out of 398, the smallest total external debt is $ 3.70 million and the largest is
$ 140,010.00 million. The frequencies of external debt below the mean of $ 11468.90 million counts 78%
of the total, which means that the mean was boosted by a few cases of extremely large debt making up the
remaining 22%. Graphically the distribution is highly skewed to the right.

199

treat from the IMF than those who are not. Apparently, however, the recipients’ rising
economic status in the world economy did not help them with the toughness of
conditionality: those countries agreed on more conditions for economic policy change. A
possible interpretation of this result may be that those emerging market economies,
instead, received large loans from the IMF, which far exceeded pre-set access limits. That
is, the amount of agreed loan could be tremendously large for their quotas and hence
sufﬁciently compensate for the toughness of conditionality. If such was the case, in reality
the conditionality agreements for emerging market economies could still be softer than
those for developing countries that have not yet reached that status. In a sense,
(temporarily) troubled emerging market economies are baptized with a mixed blessing of
tough conditions and enormous ﬁrnds poured into them, which seems to come from the
idea that emerging market economies, not only quickly normalized but complying with the
IMF’s rule of game, are important to prosperity of world economy. However, it remains
to be seen whether this implicitly discriminatory lending practice of the IMF can be
upheld, especially depending on how rapidly Southeast Asian countries will bounce back.
The recipient’s debt leverage means, in one sense, that conditionality was bent
over for the major donor countries whose private interest weighs more against official. For
recipients who mainly depend on the IMF for their external ﬁnancing, the size of
accumulated debt may not be an effective tool to reduce the number of conditions, and the
ﬁve Executive Board members’ interest in conditionality bargaining corresponds to
protecting the IMF’s ﬁnancial turf. However, if recipients have accumulated commercial
credits substantially more than credits from the IIVIF in their total debt stock, the ﬁve

Board member countries’ priority is switched to avoiding wrinkles on their domestic

200

economy, and recipient countries can take advantage of this to reduce the toughness of
conditionality. This outcome is also a byproduct of complex interdependence where direct
interstate relations are affected by the presence of important transnational actors (Keohane
and Nye, 1989). More speciﬁcally, in face of debt crisis, banks’ priorities are simply to
avoid default while limiting the extent of any new loan exposure, and their presence
looming large not only affects the general foreign policy environment but also substantially
alters the issues of salience for policy or the nature and scope of policy options available
to the governments (Cohen, 1985). As a result, the goals and contents of Fund programs
are compromised to some extent.

These results substantiate two arguments regarding debtors’ behavior introduced
earlier. First, what Sachs (1989) called double-edged behavior of debtor countries seems
to be well evidenced with these results: debtor countries are aware that the IMF is an
ongoing institution and will not default to the IMF for fear that it will rupture its
relationship with this institution, which leads to tougher conditions. On the other hand,
recipients will dare to default debt service to private syndicate lenders, since it might not
forestall ﬁrrther borrowing from new lenders elsewhere due to their ﬁrndamental
competition for proﬁt-generation from lending, and this is-why recipients can mollify the
toughness of conditionality.

The second argument this result substantiates is moral hazard problems in recipient
countries. The necessity for strong structural adjustment has been coupled with rooting up
brinkmanship in those countries, but it was not easy to pinpoint the conditions under
which it becomes plausible and feasible. As the Fund argues (Fischer, 1998c; Kapur,

1998), the notion that the availability of IMF programs encourages reckless behavior by

201

debtor countries may be far-fetched, considering the fact that the economic, ﬁnancial,
social, and political pain is simply too great—and that no country shows any great desire
to enter IMF programs unless they absolutely have to. Thus, it does not make much sense
to argue that countries would deliberately court such a crisis, knowing that international
assistance would be forthcoming. However, given the reaction from the creditors in terms
of the number of conditions, recipient countries whose debt approaches the highest point
of parabolic debt curve (approximately 80 billion in US. dollars), may have incentives at
least to neglect their economic problems in order to threaten a debt crisis with the hope

for debt roll-over.9

0 Economic System and Conditionality

Haggard and Kaufman (1989) and Chan (1993) argued that economic
characteristics of a recipient country are regarded as important to resisting to pressure
from the IMF to accept particular conditions. While staying stubborn and persistent,
recipients can prolong the bargaining process to next rounds, and it will eventually work
to reduce the number of conditions. Thus, a model composed of variables that capture
characteristics of an economy will reveal to us what effect they have on conditionality
bargaining. Unlike the others, this model includes a recipient’s subjective probability. Its
inclusion is more or less decided by the nature of the other variables; they are a ﬁxed

nature of economy rather than ﬂexible and, as such, are hard to change pending

 

9 Kapur (1998) says that moral hazard is in both borrowing countries and creditors and that perhaps the
greater moral hazard is among creditors who believes that the IMF offers implicit insurance of bailouts.
His argument is not groundless considering that the ﬂow of capital was reversed from the developing
world to the developed due to increased repayments during 19803.

202

negotiation with the IMF .10 The subjective probability can deteriorate prior to negotiation
with the IMF, but is not likely to improve dramatically. Thus, the subjective probability
from ﬁnancing capability can also be treated as a ﬁxed characteristic of an economy. This

speciﬁcation will be successﬁrlly testing Hypothesis 7h of Proposition 7.

Table 5. 11 Economic Characteristics and Conditionality

 

 

Number of obs = 215

LR chi2(4) = 37.27

Prob > ch12 = 0.0000
Log likelihood=-635.71567 PseudoR2 = 0.0285
TITNESS I Coef. Std. Err. R.Std.Err. z P>|z|
RSUBPROB I -.0438199 .0601914 .0719006 -0.728 0.467
RESOENDW l -.1368356 .0799052 .0927954 -1.712 0.087
TRADEGDP I -.0013185 .0005337 .0006279 —2.470 0.013
ECONSYSI .2515626 .0465242 .0484485 5.407 0.000

_CONS I 2.590707 .0422086 .0517985 61.379 0.000

 

Above, the subjective probability keeps the theoretically expected coefﬁcient sign,
but it is not signiﬁcant. RESOENDW, which mainly deals with oil endowments, also has
the expected sign, but comes short of conventionally accepted signiﬁcance level (P=0.05).

TRADEGDP, which indicates whether or not a recipient country is self-sufficient, is

 

’0 Low elasticity in factor allocation and mobility of resources characterizes developing economies, which
would hinder them from appropriately responding to short-term unexpected external changes. Adjustment
policies in face of balance-of—payments difficulties require the relative transfer of resources out of
nontradables into tradables, to boost exports and reduce import requirements. The ability to do this with
reasonable ease is therefore an important attribute. This difference in economic characteristics would
affect the recipient’s bargaining ability to avoid conditions ensuing adverse effects. See for more
discussion, Killick, Tony. 1993. The Adaptive Economy: Adjustment Policies in Small, Low-Income
Countries. EDI Development Studies. Washington, DC: World Bank.

203

signiﬁcant, but the coefficient sign is opposite to the theoretical expectation. That is, a
high ratio of trade to GDP indicates low self-sufficiency and less ability to resist to the
Fund’s pressure to change economic policy, and thus, the number of conditions will
increase as the ratio does. A possible explanation of the opposite direction of the
coefficient sign with signiﬁcance is that theoretical beneﬁts that make autarky an attractive
policy for a state are nothing but an illusion, damaging nations more than it helps them
(McCarthy, 1996). In our own century, there were many examples of semi- and complete
autarkic practices being adopted by nations, only to be undermined by information ﬂows
and the mobility of people——and they are not sustained for long. Furthermore, as indicated
by the fact that a state’s economic signiﬁcance in the world economy is often measured by
its share in the world trade—which is quite contrary to the concept of autarky—an open
economic system that contributes to generating global wealth through a large volume of
trade in specialized and value added products has more leverage than a closed system
today. With the variable of ECONOSYS, Eastern European countries, under systemic
transformation from a centrally planned economy to a market-oriented capitalist, are
shown to have received more conditions than other long-standing capitalist economies.
This is interpreted as necessary to create market economies from scratch in these countries
and concomitant desire of the IMF to integrate them into the mainstream of global

economy as fast as possible through institutionalization of market economy.

0 Global Economic Opportunities and Conditionality
The international economy in the late twentieth century is characterized by

intensiﬁcation of globalization. Multinational corporations in the developed countries have

204

looked for economic opportunities in promising emerging market economies, with or
without the intention to participate in management. Developing countries which can
provide such opportunities for the major donor countries may use that asset as leverage in
conditionality bargaining. The ﬁve Board member countries which do not want their
economic privileges in those countries forfeited will properly respond to recipients’
requests. Thus, a model with variables measuring economic or investment opportunities is
built below. All the variables in this speciﬁcation have a curve-linear relationship with the
dependent variable, and hence, their quadratic terms were used together. With these
variables, Hypothesis 7i, and thereby, the validity of Proposition 7, will be empirically

verified.

Table 5. 12 Global Economic Opportunities and Conditionality

 

 

Number of obs = 384

LR chi2(7) = 92.46

Prob > ch12 = 0.0000

Log likelihood = -1155.8875 Pseudo R2 = 0.0385
TITNESS | Coef. Std. Err. R.Std.Err. z P>|z|
FDIINFLOI .0000821 .0000449 .0000447 1.831 0.067
FDIINFLZ I -1.37e-08 4.86e-09 4.18e-09 -2.825 0.005
EQIINVETI .0002807 .0001046 .0001177 2.683 0.007
EQIINVEZ | -4.91e-08 2.23e-08 2.32e-08 —2.197 0.028
DEBTINFLI .0000257 .0000355 .0000365 0.725 0.468
DEBTINFZ I —6.71e-09 4.87e-09 4.43e-09 -1.377 0.169
INFINSIGI .8356056 .1283103 .0624643 6.512 0.000
_CONSI 2.518689 .0156671 .0203565 160.763 0.000

 

205

The above results reﬂect new characteristics of international economy in which
ﬁnance is emphasized more than trade. Thanks to telecommunication technology and
capital account liberalization in the developed countries, the total amount of ﬁnancial
transaction outnumbers that of commodity trade these days. Seen from this perspective, if
a recipient hosts large foreign direct investments, it can be used to reduce the number of
conditions, but with no efﬁciency. On the other hand, if the recipient is a host of large
equity investments, it can lessen the harshness of conditionality agreement with efficiency.
Any artiﬁcial barrier intended to block access to lucrative emerging markets must hurt
international investors’ interests (Chan, 1993), especially from countries where domestic
markets are saturated, and this is well reﬂected on the parabolic curve of EQIINVE2 with
signiﬁcance. DEBTINFL, the pure measure of net private bank and trade-related lending
and portfolio bond investments, has a coefficient sign in line with theoretical expectation,
but is not signiﬁcant. Its quadratic term says that the number of conditions will eventually
be on the rise, but with no signiﬁcance. As a matter of fact, debt-creating capital inﬂows
are part of total debt stock, and hence, the same theoretical expectation is projected on
this variable as on total debt stock. Let alone the insigniﬁcance, however, the signs of
coefficients of two DEBTINFL’s are opposite to those of total debt stock. It needs ﬁrrther
study of developing countries’ debt structure and attitude towards creditors by type.
Again, the variable INF IN SIG ﬁrnctions to directly increase the base number of conditions

with signiﬁcance, as it did in the other speciﬁcation.

206

0 Politics of Special Interest and Conditionality

A new emphasis in international bargaining theory is the concept of a two-level
game developed by Putnam (1988). National bargainers are supposed to build consensus
on two fronts to have negotiation succeed: ﬁrst, he or she has to reach an agreement with
the foreign negotiator on issues at hand; second, he or she has to achieve domestic
consensus on the agreement. In so doing, domestic institutional setting determines the size
of win-set within which the national negotiator can maneuver the bargaining outcomes and
even draw a preferred outcome from bargaining using (private) information of the given
win-set.

The focus of domestic political process in the context of structural adjustment is
politics of special interests, or rent seeking behavior, which is blamed for a low degree of
implementation in conjunction with overambiguous and overambitious targets. To test this
argument, a model is built with variables that could be regarded as a proxy of special
interests and other political and economic system, which are intertwined with one another
to determine the size of win-set. In developing countries, public sector is one of the most
conspicuous sources of special interest, which usually becomes a target for restructuring
and streamlining in the wake of adjustment. The contribution made by state-owned
enterprises to GDP will give us a clue as to how much opposition will be entailed to
implementing conditions. And as related to this, the size of urban population—which
consists mostly of public sector employees and labor workers in manufacturing sector who
are also potential political protesters—will give us a broader picture of the political

dynamics in developing countries.

207

Regime type was used in other speciﬁcation as the indicator of domestic political
cost in case a recipient government agrees with the IMF upon a set of unfavorable
conditions. Domestic political cost can be a proxy of the size of win-set, too. In many
developing countries, import-substitution-industrialization (181) was deliberately chosen
for political reasons, rather than economic and helped the scope and range of state-owned
enterprises in economic activities expand. The ratio of imports to GDP is a dummy proxy
for whether a recipient is delivering 181 as economic development strategy. This
speciﬁcation will test Hypothesis 7j of Proposition 7, and any signiﬁcant result of
RPOLCOST in this speciﬁcation will reconﬁrm the validity of Proposition 3, since that
variable was already used in testing the impact of domestic political cost as utility

ingredient.

Table 5. 13 Politics of Special Interest and Conditionality

 

 

Number of obs = 111

LR chi2(5) 16.54

Prob > ch12 = 0.0055

Log likelihood = -306.64576 Pseudo R2 = 0.0263
TITNESS | Coef. Std. Err. R.Std.Err. z P>|zI

RPOLCOST I .0113294 .0052987 .0070783 2.138 0.033*

RPOLCOSZ I -.0033591 .0011981 .0014303 —2.804 0.005
IMPOGDP I -.0053501 .00216 .0025553 -2.477 0.013
SOEGDP I .0092185 .0038928 .0042665 2.368 0.018
URBANPOP I -.0002814 .0016901 .0021793 -0.167 0.868
_CONS I 2.650135 .1291183 .1563398 20.525 0.000

 

208

In the above model, the impact of regime type on the number of conditions
remains the same with in the model of utility ingredients, but the impact of quadratic term
on reducing the toughness of conditionality has a clear signiﬁcance. Thus, democratic
regime helps reduce the number of conditions by narrowing the size of win-set. Also, if a
recipient’s economy is based on ISI, it effectively reduces the number of conditions to be
consented upon, since the size of special interest that hinders structural adjustment appears
large. The size of public sector in economy turns out to be signiﬁcant, but has a coefficient
sign opposite to the theoretical expectation. On the other hand, URBANPOP shows a
theoretically expected coefficient sign, but not signiﬁcant. If a recipient country is less
urbanized and most of its population dwell in rural area, the government is limited in
setting the win-set for negotiation in its favor and seems to sign an adjustment agreement
with more conditions than not. However, the reliability of this speciﬁcation cannot be ﬁrlly
veriﬁed due to the small number of observations. This is mainly from the data of

SOEGDP, which was one of the poorest in the data set, together with unemployment.

5.4 COMPREHENSIVE SPECIFICATION FOR CONDITIONALITY BARGAINING

So far, the statistical models have been focused on exploring possible forms of
national and organizational leverage from as many different angles as possible. Those tests
suffice to open eyes to a world of variety in leverage to be deployed for conditionality
bargaining, depending on issues in question, but not enough to give us a comprehensive
picture of the relationship between leverage and conditionality bargaining. Thus, a model

that can illuminate conditionality bargaining with a bird’s-eye-view is necessary.

209

Exploring a comprehensive model for conditionality bargaining should be based on
the theoretical considerations discussed above. However, it should not be an anthology
from each speciﬁcation, which consists of variables that showed correct directions of
coefficients and statistical signiﬁcance. A major concern of the comprehensive
speciﬁcation is parsimony, which can still capture most of the characteristics of
conditionality bargaining with as few variables as possible. And while individual
speciﬁcations examine the effectiveness of leverage in one direction—as if not checked
and balanced by that of the other player—this hardly takes place during negotiation, and
any outcome of bargaining is produced through clash of leverage. By putting leverage
(mobilized by each player) together in one model, the effectiveness of its leverage will be
weighed one against another.

In a sense, the comprehensive speciﬁcation should be a condensed narrative of
conditionality bargaining of how it starts, what is at stake, and how it ends—and choosing
variables for the speciﬁcation follows suit. Hence, the story goes that the entire process of
conditionality bargaining unfolds with economic crisis in developing countries; a basic rule
of the game is that a prospective recipient can draw rescue ﬁrnd from the IMF within the
pre-set limits and the number of conditions will increase as it approached the limits.
Agreed conditions for structural adjustment are also a means to build consensus between
debtors and creditors on handling snow-balling debt problems in developing countries with
a view to preventing mischievous consequences at the system level. The ﬁve Executive
Board member countries try to inﬂuence conditionality in proportion to their ﬁnancial
contribution to the IMF; the IMF’s lending policy is inﬂuenced by its ﬁrndamental

imperative as the lender of last resort, or provider of global liquidity. And the leading

210

powers within the IMF adjust the level of conditionality in accordance with their vision for
world management, constrained by the context within which conditionality bargaining
takes place. Thus, variables corresponding to each story line are balance-of—payments
current account deﬁcit, the ratio of agreed loan to quota, ﬁve major donors’ quota
subscription, the size of lendable fund, international context, and accumulated external

debt

Table 5. 14 Comprehensive Speciﬁcation for Leverage and Conditionality

 

 

Number of obs = 361

LR chi2(7) = 290.86

Prob > ch12 = 0.0000

Log likelihood = -975.5674l Pseudo R2 = 0.1297
TITNESS | Coef. Std. Err. R.Std.Err. z P>Iz|
BOPCUADF I .0000149 6.00e—06 8.73e-06 2.480 0.013*
AMNTQOTA I .0010166 .0002605 .0003093 3.903 0.000
TOTDEBT I 6.80e-06 1.83e-06 2.15e-06 3.710 0.000
TOTDEBT2 I —3.46e-11 1.71e—11 1.94e-11 —2.031 0.042*
SMQUOTAS | .0000137 2.08e-06 2.27e-06 6.561 0.000
FUNDSIZE I -.0000441 .0000433 .0000444 —1.020 0.308
INTCONXT I -.1752262 .0474759 .0524387 —3.691 0.000
_CONS I 1.875059 .1193949 .1298989 15.705 0.000

 

The above results are roughly in line with those of individual speciﬁcations:
AMNTQOTA, INTCONXT, TOTDEBT, and TOTDEBT2 retained the theoretically
expected direction of coefficients and more or less the statistical signiﬁcance as they did in

individual models. On the other hand, SMQUOTAS not only kept the correct sign of

211

coefficient, but also its signiﬁcance jumped to signiﬁcant level. FUNDSIZE kept the
correct sign, but lost signiﬁcance; and BOPCUADF obtained a wrong coefficient sign, but
remains signiﬁcant, which is atheoretical.

To test the robustness of this comprehensive speciﬁcation, another speciﬁcation is
built with more variables. This mission is also to overcome a potential weakness of the
comprehensive speciﬁcation—disconnection from the components of expected utilities—
by including appropriate ones. Among the various utility components, the most vital part
is players’ subjective probabilities. Player’s subjective probabilities are a succinct
expression of players’ tangible and intangible capabilities to be transformed into leverage.
Although the formulae for each player’s subjective probability were formed only with
tangible assets, due to problems with operationalization and measurement, the idea at the
bottom is that the effect of the intangible on the subjective probability cannot be ignored in
reality, and thus, it pretends to have incorporated them into the formulae.

The results from the speciﬁcation (Table 5.15) with players’ subjective
probabilities included are that the presence of subjective probabilities does not change the
results much from the one without subjective probabilities. The only difference is that the
quadratic term of accumulated external debt gained a clear-cut signiﬁcance, and the
indicator of economic crisis exchanged the correct direction of coefﬁcient for statistical
insigniﬁcance. Above all, the subjective probabilities themselves failed to show statistical
signiﬁcance, despite their theoretically correct coefficient signs. Thus, the robustness of
the comprehensive model without subjective probabilities appears to be sustained with

these results.

212

Table 5. 15 Comprehensive Speciﬁcation with Subjective Probabilities

 

 

Number of obs = 214

LRch12(9) = 156.14

Prob > ch12 = 0.0000

Log likelihood = -571.21333 Pseudo R2 = 0.1202
TITNESS I Coef. Std. Err. R.Std.Err. z P>Iz|
DSUBPROBI .115779 .108549 .1043283 1.067 0.286
RSUBPROB I -.0659206 .0653784 .070381 -1.008 0.313
BOPCUADF I -8.20e-06 .0000121 .0000182 —0.675 0.500
AMNTQOTAI .0011201 .0003472 .000409 3.227 0.001
TOTDEBTI 9.86e—06 2.40e-06 2.54e-06 4.100 0.000
TOTDEBT2 I -9.82e-11 2.68e-11 3.54e-11 -3.664 0.000
SMQUOTASI .0000114 2.79e-06 2.92e-06 4.081 0.000
FUNDSIZE I -.0000379 .0000581 .0000747 —0.652 0.514
INTCONXT I -.1868513 .0639664 .0666755 -2.921 0.003
_CONS I 1.910639 .155501 .1683266 12.287 0.000

 

The main theme of this dissertation is to identify bargaining leverage as the source

of variance in conditionality agreements (i.e., the degree of tightness in conditionality), and

thus, all the statistical analyses used the number of agreed conditions as the dependent

variable. One of the assumptions of the game model says that the desirability of a

conditionality agreement is determined by the toughness of conditionality and the amount

of agreed loan, which can be traded off. If the model is established with such an

assumption, the amount of agreed loan must be another measure of bargaining leverage,

and conducting an additional analysis with that variable seems to be worth while in that it

will demonstrate the explanatory power and versatility of the model, overcoming the

restrictive nature of the original dependent variable. Moreover, if the statistical model

213

were well speciﬁed, switching the dependent variable from the number of conditions to the
amount of agreed loan, while keeping the independent variables, will have the coefﬁcient
signs switched as necessary, since the two dependent variables are theoretically in a trade-
off relationship.

Using the amount of agreed loan as the dependent variable, however, is poised
with a problem. The amount of agreed loan is already reﬂected on the variable of
AMNTQOTA in the comprehensive model, which causes endogenity to the model. In
looking for a substitute, the role of Fund lending in relation to the size of balance-of-
payments gap is good enough, but this has the problem of endogenity, too, since the
balance-of-payments current account deﬁcit is used as an independent variable. To solve
this problem, the ratio of the amount of agreed loan to GDP (AMNTGDP, %) will be
used. This new variable transforms the absolute amount of agreed loan into relative term,
namely the IMF’s contribution to recipients’ economies. At a superﬁcial level, there
should be normality in the size and distribution of Fund lending in relation to the level of
development of borrowing countries, which is measured by GDP. If there exists any
different relationship between Fund lending and the level of economic development,
however, we can also suspect that bargaining leverage affected the extraordinary
distribution of IMF ﬁrnds. Unlike the number of agreed conditions, AMNTGDP, the
alternative dependent variable, is continuous, and thus, the statistical method is switched

to ordinary least squares regression from Poisson.

214

Table 5. 16 Descriptive Statistic of the Alternative Dependent Variable

 

N

Min.

Max .

Mean

Std.Dev

 

AMNTGDP

387

.16 33.13

2.88

3.

40

 

Table 5. 17 Comprehensive Speciﬁcation with the Alternative Dependent Variable

 

 

Number of obs = 359

F(7, 351) = 22.51

Prob>F — 0.0000

R-squared = 0.3098

Adj R-squared = 0.2960

RootMSE = 2.2236
AMNTGDP I Coef. Std. Err. R.Std.Err. t P>|t|
BOPCUADF I .0002251 .0000559 .0000505 4.029 0.000
AMNTQOTA | .0204927 .0020843 .0029429 9.832 0.000
TOTDEBT I -.0000701 .0000159 .0000129 -4.400 0.000
TOTDEBT2 I 3.05e-10 1.54e-10 1.17e-10 1.986 0.048
SMQUOTAS I .0000407 .0000165 .0000141 2.470 0.014
FUNDSIZE I -.0010097 .0003292 .0003474 -3.067 0.002
INTCONXT I .2241728 .3636633 .2928286 0.616 0.538
_CONS I .0333385 .9312187 .7339856 0.036 0.971

 

In Table 5.17 above, the variable AMNTQOTA is the ratio of agreed loan to

recipient member’s quota subscription; thus, the higher the ratio, the more the actual

amount of arranged loan is and the more contribution is made to the recipient’s economy.

The coefficient sign of the AMNTQOTA is so naturally positive and signiﬁcant, as

indicated by high t-ratio (9.83 2). Quotas from the ﬁve Executive Board member countries

215

are a signiﬁcant base for ﬁrnd formation within the IMF. Theoretically, increased quota
allocations to them means that overall liquidity position of the IMF is improved and the
usable resources are affluent, which has more or less a positive impact on recipient
economies. The variable of SMQUOTAS testiﬁes this reasoning behind with a positive
coefficient sign and statistical signiﬁcance.

The most interesting part of the results above is the coefficient signs of
TOTDEBT. When TITNESS was used as the dependent variable, the coefﬁcient signs of
ﬁrst and second order of TOTDEBT were positive and negative, respectively. Now, the
dependent variable is switched to the ratio of arranged loan to GDP and the coefﬁcient
signs of two TOTDEBT’s are negative and positive, respectively. This means that as the
size of total external debt gets larger, the IMF’s contribution to recipients’ economies gets
smaller. However, the positive coefﬁcient of the quadratic term indicates that the IMF’s
contribution to the recipients’ economies will be on the rise again, after passing a certain
point in the size of total external debt. The logic behind this is the same with that for the
number of conditions; that is, as the size of accumulated external debt increases, the IMF
fears for moratorium by the debtor country and makes more contribution to the debtor
country’s economy to help it continue debt service to the creditors.

With the number of conditions on the left-hand side of the equation, the severity of
economic crisis indicated by the balance-of-payments current account deﬁcits
(BOPCUADF) did not give us the theoretically expected result. However, with the ratio
of arranged loan to GDP, the result is satisfactory in terms of the direction of the
coeﬂicient sign and its signiﬁcance. This indicates that a slight imbalance in economy is

easy to have a loan arranged with the IMF and the size of arranged loan is likely to be

216

larger than necessary. However, when the imbalance is severe, the recipient has to go
through an uphill battle with the IMF for a meager contribution from the IMF. The IMF
has been criticized for discriminating against heavily indebted poor countries whose
balance of payments problem is chronic and structural (Bird, 1995), and this criticism is
empirically supported 3 by this result.

Although it turned out to be insigniﬁcant, the meaning of the coeﬁicient sign of
INTCONXT should be noted. The variable of INTCONXT indicates that IMF lending has
made a greater contribution to recipients’ economies during the Cold War than the post-
Cold War. This result is in contrast with its coefficient sign, when the dependent variable
was the number of conditions. In the previous analysis, the IMF was read to request fewer
conditions (negative sign) during the Cold War period than during the post—Cold War.
This contrast means that the curves of the two regression lines cross each other, rather
than go in parallel, and thereby substantiates the trade-off relationship between the number
of conditions and the amount of agreed loan. The only utterly odd result is the coefficient
sign of FUNDSIZE. The variable FUNDSIZE measures the preparedness of the IMF to
lend. Theoretically, this means that the more the IMF is ready to lend, the more
contribution it makes to recipients’ economies. The expected coefficient sign is positive,
but it turned out to be negative with a high signiﬁcance.

As with the speciﬁcation with the number of agreed conditions as the dependent
variable, the comprehensive speciﬁcation using the ratio of agreement loan to GDP is also
slightly modiﬁed with the addition of subjective probabilities. The result (Table 5.18) is
quite similar to the original speciﬁcation without the subjective probabilities. As expected,

no ﬁrrther interpretation is necessary; simply, interpreting what the coeﬂicients of the two

217

Table 5. 18 Comprehensive Speciﬁcation with the Alternative Dependent Variable

and Subjective Probabilities

 

 

Number of obs = 213

F( 9, 203) = 10.86

Prob>F = 0.0000

R-squared = 0.3249

Adj R-squared = 0.2950

RootMSE = 2.5506
AMNTGDP | Coef. Std.Err. R.Std.Err. t P>It|
DSUBPROB I -.8507406 .9971009 .9261574 —0.853 0.395
RSUBPROB I -2.092324 .5881145 .4249873 —3.558 0.000
BOPCUADF I .0003741 .0001108 .0000952 3.377 0.001
AMNTQOTA I .0207657 .0031413 .0039302 6.611 0.000
TOTDEBT I —.0000547 .0000238 .0000164 —2.294 0.023
TOTDEBT2 I 3.56e-10 2.50e-10 1.63e-10 1.424 0.156*
FUNDSIZE I —.00127 .0004952 .0005226 —2.564 0.011
SMQUOTAS I .0000413 .0000249 .0000219 1.660 0.098
INTCONXT I .1408132 .5574756 .4664202 0.253 0.801
_CONS I 1.056218 1.370929 1.031344 0.770 0.442

 

subjective probabilities mean is enough. The recipient’s subjective probability has a

negative coefficient in this speciﬁcation and this is a correct reﬂection of the nature of the

recipient’s subjective probability. That is, the recipient’s subjective probability is equal to

its ability to raise ﬁinds abroad so as to replenish balance-of-payment current account

deﬁcit; a higher subjective probability means that it would borrow most ﬁrnd from

international capital markets and loan arrangement with the IMF would be only a back-up

plan in case of shortage. Thus, as the recipient’s subjective probability increases, the

economic contribution from the IMF will decrease (i.e., the negative sign of the coefficient

218

is theoretically correct). On the other hand, the donor’s subjective probability makes
another odd case along with the size of lendable ﬁrnd (FUNDSIZE) in the speciﬁcation.
The donor’s subjective probability is formulated to be the Fund’s annual liquidity position
vis-r‘r—vis member countries from the developing world, which means that as the IMF’s
fund position gets more liquid, the more economic contribution to developing countries is
feasible. Considering this, a theoretically correct coefficient sign of the donor’s subjective
probability is positive, but it came out negative though insigniﬁcant. Overall, the addition
of the subjective probabilities did not affect the model enough to reveal speciﬁcation error.
With or without the subjective probabilities, the two different comprehensive
speciﬁcations are judged to illuminate well the relationship between the IMF and

developing countries, which is determined by mobilized leverage.

5.5 THEORETICAL IMPLICATIONS OF FINDINGS

The main purpose of the statistical analyses was to verify the external validity of
the theoretic model of conditionality bargaining. The beginning of this research was a
question of whether the existing conditionality agreements are arguably all alike, and what
could explain such difference if they are not. To answer the question, a conditionality
bargaining model was built using the concept of leverage. The subsequent quest was
empirically testing the relationship between leverage and variance in agreements
represented by the number of agreed conditions with all the risk of oversimpliﬁcation of
the contents. On average, the correspondence between the outcomes predicted by the

model and real world experience was satisfactory. There were a total of 39 independent

219

variables as operationalization and measurement of leverage in 11 small and 2
comprehensive speciﬁcations,ll 23 (58.97%) of which showed theoretically expected
coefﬁcient signs and statistical signiﬁcance. And the number of speciﬁcations within which
more than half of the used independent variables showed signiﬁcant explanatory power
was 11 out of 13. This indicates that the models were properly speciﬁed to test the
hypotheses. Through a variety of operationalization, it was revealed that the difference in
conditionality agreements was a direct outcome of leverage that each party of the
bargaining brings to the negotiation table.

In passing, there were 8 cases of independent variables that changed statistical
signiﬁcance with robust standard errors obtained through the White test. This
phenomenon was not ubiquitous but sparsely scattered over the speciﬁcations that it is
safe to say that the asymptotic unbiasedness and consistency properties of maximum
likelihood estimators in large samples was not destroyed. Nonetheless, they are no longer
efficient and the lack of efficiency makes the usual hypothesis-testing procedure of
dubious value (Gujarati, 1995). Therefore, correcting those errors would be desired in the
ﬁrture.

A possible remedial measure to correct the inefﬁciency of some estimates identiﬁed
by the White test is to transform those variables in functional forms such as square rooting
or logging to reﬂect suspected types of heteroscedasticity and thereby make their
variances homoscedastic. The target variables for ﬁrnctional transformation will even

include the dependent variable, especially the number of agreed conditions. This variable

x

1’ The statistical analyses that conﬁrm partial grounds of both sides of the debate (Table 5.1 and 5.2) and
the comprehensive speciﬁcations with the subjective probabilities (Table 5.15 and 5.18) are not counted.

220

was obtained through content analysis and was ﬁrst-cut in the ﬁeld. Therefore, re—counting
the number of agreed conditions, ﬁrst, is suggested in order to enhance its objectivity and
accuracy, and then ﬁrnctional transformations should be attempted as a remedial measure.
Another remedy to this problem is increasing the size of samples so as to cover
conditionality agreements during the entire presence of the IMF. If this is done, this will
not only correct the efﬁciency problem of estimates haunted by heteroscedasticiy and
enhance their asymptotic unbiasedness as they approach their true values, but also expect
to end the controversy especially over the variable of INTCONXT since the analysis will
include more conditionality agreements during the Cold war period.

This research was built against a conventional skepticism of the relationship
between developed and developing countries in general, or possibilities of negotiation
between them from the perspective of aggregate power structure in international relations.
Rather, the model was narrowly focused on issue-speciﬁc bargaining structure, with the
concept of leverage and the statistical results supported this notion given issue area of
conditionality. Most of the variables used for the analyses were monetary and ﬁnancial,
such as the total external debt and bond/equity portfolio investment, which are more
volatile and vulnerable to domestic and international disruptions than commodity
economy; and derivatives from an economic “relationship” between the two parties, such
as market accessibility and ﬁnancial dependency, rather than hard data on usable
resources, human and natural, or production capacity. Such second-tier economic data
neither have direct bearings on aggregate power in international relations, such as GNP
and population, nor can be transformed into military capabilities if needed. Nonetheless,

analyses based on those variables, without resorting to power parity, showed that both

221

sides not only considered the other as qualiﬁed negotiation partner, but also drew out
conclusions on interaction through bargaining, given the issues to be settled.

Another fact—conditionality bargaining was better explained with the concept of
leverage within a speciﬁc issue area—is evidenced in that the variables of alliance and
military aid did not have signiﬁcant explanatory power. These variables are highly
regarded components of aggregate-power-based theory of international relations, and
within that theoretical tradition, the outcomes of interaction among states are pretty much
predictable due to power ﬁrngibility, regardless of the nature of the issue at hand. Hence,
if the aggregate structural power were the correct conceptualization underlying the
conditionality bargaining, those variables should show signiﬁcance with theoretically
correct coefficient signs, which did not happen. In the similar vein, this unsubstantiated
conceptualization of aggregate structural power may indicate that conditionality
bargaining is even selective in issue-linkage, or has a limited scope of issues to be linked
to. The other economic variables, although originally presented as operationalization of
leverage, are in a sense a manifestation of issue linkage taking place successﬁrlly. This
selectivity in issue linkage may be related to the fact that one party of the bargaining is an
international organization. That is, although the ﬁve major Executive Board members are
portrayed as having tremendous sway in decision making within the IMF and insert their
interests into its operation, their action is, in any event, under the umbrella of an
international organization, which will delineate the scope and range of feasible issue
linkage for them more than in bilateral relations.

Third, at a substantive level, IMF conditionality has been entangled in debates over

many aspects. The bottom line of the debates was that IMF conditionality could not

222

achieve intended goals; its feet were only dragged into muddy global debt problems deeper
and deeper. Many attempts to chase this problem to the origins closed their research with
a hasty conclusion that the IMF dominated the entire process of structural adjustment with
one design. Against this single-minded conclusion of coercive conditionality, a game-
theoretic bargaining model showed theoretically that there could be many different
outcomes on the continuum of absolute coercion to absolute compromise, and statistical
analyses in this chapter showed empirically that this could not take place, unless it was
intervened with a necessary condition of leverage. Furthermore, the statistical results
supported or reputed conventional arguments in conditionality literature published by the
IMF and those critical scholars who have been accepted without systematic delve. For
example, both arguments for fair application of conditionality in line with credit tranche
(Gold 1979; Guitian, 1981) and discriminative treatment by the Fund of the poorest
countries (Killick, 1995), especially eligible for SAF/ESAF, which is conﬂicting with the
former, were given a ﬁrm ground by the analysis. External debt (Sachs, 1989) and moral
hazard problems in developing countries (Cornelius, 1988 and Vaubel, 1991 vs. Bird,
1995 and Fischer, 1998), along with bittersweet backlash of colonialism (Haggard and
Kaufman, 1989), were also evidenced to be real by the statistical analysis. Those statistical
analyses were based on almost complete universe of conditionality agreements, which
immunizes them from selection bias. Therefore, their results vindicated the veracity of
conventional wisdom and enhanced the level of generalization. In addition, they could also
capture the overlooked ﬁrnctional mechanism of that wisdom.

Fourth, the statistical results of this chapter testify the close connection between

the game model and empirical expectation. Although driven by abstract in the ﬁrst place,

223

various ingredients of the model, such as cost terms and discount rate, showed
surprisingly high explanatory power to the research question. This does not necessarily
mean that the Fund and recipient countries calculate their utilities in the exactly same way
in reality. However, at least the results assure us of two things: players in conditionality
bargaining are highly likely to consider such components seriously; and the difference in
those components plays a certain role in conditionality outcomes.

Lastly, the statistical results vindicate the external validity of the model in terms of
expected bargaining outcomes. As proﬁled in Table 3.1, conditionality bargaining game
produced three different results and leverage was identiﬁed as the underlying necessary
ﬁmction for those results. On the other hand, it was shown at the beginning of this chapter
that contentions from both the IMF and critical scholars have credible empirical records.
The veriﬁcation of those contentions is not contradictory to the outcomes from the model,
but are securely nested in the model. That is, the two contentions can be regarded as two
extremes on the continuum of coercion to compromise, and it is leverage that moves
conditionality agreements from middle ground to either direction. Compared with game
theoretic outcomes, they are equivalent to maximum imposition by the IMF, Agreement
(9, and maximum technical application of conditionality, Agreement 0), respectively. And
Agreement 6) comes to exist between these two extremes by the ﬁrnction of leverage. The
subsequent statistical analyses aimed at materializing the existence of leverage, playing in
the middle ground and its effectiveness. Many independent variables that are congruent
with theoretical expectations plainly say that there exist more sources of leverage to give
players ﬂexibility in conditionality bargaining than we have thought, from institutional

settings of the IMF and recipient countries to various opportunities in international politics

224

and economy. A windfall from the established validity of the conditionality bargaining
model based on leverage is that conditionality is prospective rather than retrospective.
IMF conditionality will change its face, continuously interacting with newly emerging
trends in international political economy, and both sides’ bargaining style and appropriate

leverage will adapt themselves to them.

225

CHAPTER 6

LEVERAGE, CAPITAL MOBILITY AND CONDITIONALITY:
INTERNATIONAL BAILOUT OF KOREA, 1997

On December 3, 1997, Michel Camdessus, Managing Director of the IMF,
announced that Korea and the IMF concluded a Stand-by Agreement. Through this
agreement, Korea would receive a loan worth around $55 billion. The IMF would provide
$21 million in Stand-by credits over three years, the World Bank would provide up to $10
billion, and the Asian Development Bank would provide up to $4 billion. Some countries,
including the United States, Britain, Japan, Germany, France, Canada, Australia and Italy
would pitch in a total of up to $20 billion. This aid package could be the largest one in
IMF loan history. In the summer of 1997, Indonesia and Thailand received assistance from
the IMF which totaled about $60 billion combined, and Mexico negotiated a package of
$50 billion to ease out its ﬁnancial crisis in 1995, which had been the biggest IMF loan up
to that date. Agreed policy measures for structural adjustment comprised disciplined ﬁscal
and monetary operation, far-reaching ﬁnancial sector reforms, and radical liberalization of
trade and capital accounts, as well as improvements in the structure and governance of
Korean corporations.

At a glance, this recent international bailout of Korea is one of the ordinary Stand-
by Agreements that have been contracted between the IMF and developing countries since
the debt crisis broke out in 1982. However, the Korean case is distinct from previous ones
in many respects: the international economic context of the bargaining was considerably

changed from that of the debt crisis, the origins of the problem were located within the

226

private sector rather than the public sector, and the ﬁnal terms of agreement contained
unusual items. The entire process of bargaining took only two weeks from the request for
bailout to approval by the Executive Board of the IMF. What is more important is that
such distinctiveness does not work to segregate it from other conditionality agreements.
To the contrary, it appears to be the culmination of factors shared across the conditionality
agreements, which form the very core of the conditionality bargaining with the IMF.
Therefore, this chapter delves into Korea’s recent experience with the IMF.

A case study at this point, after all the efforts to generalize conditionality
bargaining, is intended to put a concrete look on the theoretical model. So far, discussions
of conditionality bargaining have been abstract and general. While theorizing and
generalizing conditionality bargaining was attempted, many details of individual
conditionality bargaining—which might be sometimes more relevant to explaining its
outcomes and, thereby, differentiate it from others—were inevitably sacriﬁced. The case
study in this chapter calls those details back into the picture and brings down the
discussion from general to particular level. Through a detailed study of a speciﬁc case, we
will be able to get practical sense of many components of the conditionality bargaining
game. That is, what form each ingredient of the game can take in reality, and how well
each part of the model ﬁts a speciﬁc case. This comparison between theory and reality is
indispensable for establishing validity of the theory, and a good generalization should
remain valid even after idiosyncrasy of each case is taken into account. Also, the
distinctiveness of individual cases will be better understood when they are viewed against
the common pattern piercing through them all, rather than staying aloof. Now, we start

with a chronicle of what unfolded in Korea late 1997.

227

6.1 SUMMARY OF EVENTS

January through October 199 7: Korea’s macroeconomic performance in 1997 was
broadly favorable, notwithstanding a widening current account deﬁcit, mainly as a result of
deteriorating terms of trade. However, since the beginning of the year, an unprecedented
number of large business conglomerates had moved into bankruptcy. This reﬂected a
number of factors, including excessive equipment investment in certain sectors—such as
steel and autos, which was ﬁrrnished by domestic banks borrowing capital from
international ﬁnancial markets for short-term (due in a year or less)—and a weakening in
the proﬁtability associated with the cyclical downturn. Those bankruptcies undermined the
ﬁnancial system, and non-performing loans rose sharply to the equivalent to 7.5% of GDP.
Preceded by crisis in Thailand and Indonesia, foreign banks began to reﬁrse to roll over the
country's short-term debt. The won, Korean currency, plunged precipitously as banks and

corporations scrambled to buy dollars to repay loans suddenly called in.

November 18: The Korean government, and more speciﬁcally, the Ministry of Finance and
Economy (MOFE), adamantly denied a spreading rumor that a request for help from the
IMF was under review. The MOFE assured that any bailout by the IMF was not being
considered, saying that Korea's foreign currency reserves were not at a worrying level yet,

which required emergent IMF loans.

November 19: The Korean government announced ﬁnancial market stabilization measures

as a self-help attempt, which comprised restructuring of the ﬁnancial system through inter-

228

bank mergers and acquisitions (M&As), as well as disposing of bad debts at banks and

merchant banks specializing in short-term lending.

November 20: Stanley Fischer, First Deputy Managing Director of the IMF, arrived in
Seoul and delivered the IMF’s concern about the Korean economy. This assured the
Korean authorities of the IMF’s ﬁrll support to Korea’s request for rescue if it occurred.
Also, Lawrence Summers, Deputy Secretary of Treasury of the United States, said at a
meeting with Chang-Yeul Lim, Minister of MOFE, that the United States believed that
financial support to countries with problems should only be provided in the context of an

IMF program.

November 21: In an overnight reversal of policy stance, the Korean government
announced that it decided to seek a rescue loan from the IMF to lift itself out of an
ongoing ﬁnancial crisis, and that Korea and the IMF were already engaged in unofﬁcial
negotiations over the amount and terms of the emergency loan. Following the Korean
authorities’ announcement, the IMF management welcomed Korea’s decision and pledged
a ﬁtll support. Anxiety about Korea spread to other markets in Asia and other regions to

contribute to a 129.80-point drop in Dow Jones Industrial Average in the United States.

November 26: Hubert Neiss, Director of Asia-Paciﬁc Region of the IMF, arrived in Korea
with other seventeen IMF delegates and started conducting research on the Korean

economy.

229

December 1: The Minister of MOFE announced that the Korean government and the IMF
delegation team reached a provisional agreement on a bailout program, including the

general outlines of amount and terms.

December 2: The Managing Director of the IMF denied the Korean assertion on a
package settlement in Malaysia, where he was attending a meeting of Asia-Paciﬁc ﬁnance
ministers and central bank officials. He said that Seoul’s announcement of reaching an
agreement was considered to be part of Seoul’s negotiating tactics. In the following
telephone talk with the Minister of MOFE, Camdessus expressed negative views about
parts of the tentative accord worked out by the working-group officials of the two sides,
and he demanded that the Korean government delay the announcement until every last
detail was settled. The same day, the Korean authorities suddenly suspended operations
of nine badly ailing merchant banks until the end of the year. Those banks would be closed

for good unless they devised a plan to right themselves by March 31, 1998.

December 3: Arriving in Seoul, Michel Camdessus had a direct talk with the Minister of
MOFE. In the last-minute negotiation, Minister Lim accepted demands by the IMF for
immediate liquidation of seriously ailing banks and merchant banks. He also agreed to
allow foreign banks and security companies to establish subsidiaries in Korea by mid-

1998, and to extend upper limits for foreigners acquiring per item shares.

December 4: The Executive Board of the IMF approved Korea’s request for a three-year

Stand-by credit equivalent to 15.5 billion SDRs (1,939% of Korea’s quota of 799.6

230

million SDRs), using the accelerated procedures established under the Emergency
Financing Mechanism (EFM).l At the announcement of the package deal, major stock

markets around the world buoyed as it soothed investors’ fear.

In the Letter of Intent,2 the Korean government pledged to implement the
following conditions: macroeconomic objectives were to be set to narrow the external
current account deﬁcit to below 1% of GDP in 1998 and 1999, inﬂation was to be
contained at or below 5%, and the real GDP grth was to be decelerated to about 3% in
1998 and following recovery in 1999. Monetary policy was to be tightened immediately
with call rate raised up to 21%, and ﬁscal policy was also to be maintained tightly in line
with monetary policy in order to achieve budget balance through both revenue and

expenditure measures.

The structural components of the program included a wide-range of ﬁnancial
sector restructuring: the Bank of Korea Act was to be revised to provide for central bank
independence, and commercial banks failing to meet the Bank for International
Settlements’ (BIS) net worth ratio of 8% were to be immediately liquidated through third-

party takeovers or M&As. Also, capital accounts were to be liberalized to accelerate

 

' The EFM was adopted in September 1995. It strengthens the IMF’s ability to respond swiftly in support
of a member country facing an external ﬁnancial crisis and seeking ﬁnancial assistance from the IMF in
support of a strong economic adjustment program. The policy favors larger loans from the IMF in a
shorter time during periods of market crisis, but these loans would carry higher interest rates and shorter
repayment periods than traditional IMF assistance. Source: IMF. 1998. “Financial Facilities and Policies:
IMF Financing Helps Members Pursue Sound Policies.” IMF Survey Supplement on the Fund (September)
or visit http://www.imf.org/external/pubs/ﬁ/survey/sup0998/08.htrn.

2 Its ﬁrll text is available on the IMF’s website: http://www.imforg/extemal/np/loi/ 1203 97 .htrn.

231

foreign investment in the Korean equity market by allowing the establishment of
subsidiaries of foreign banks and security companies by mid-1998, and by extending the
upper limit for foreigners acquiring per item shares from 26% to 50% and to 56% in 1998.
Timetables for trade liberalization were also to be set in compliance with the World Trade
Organization (WTO) commitments. In addition, reforming corporate structure and its
governance was required: the Korean government was not to intervene in bank
management and lending decisions except policy lending (agriculture, small business, etc);
and a timetable was set to improve the transparency of corporate balance sheets. Finally,
labor market was to be reformed to facilitate the redeployment of labor, in parallel with

ﬁrrther steps to improve labor market ﬂexibility.

6.2 RECONSTRUCTING THE CASE TO THE CONDITIONALITY BARGAINING

To demonstrate that the Korean case conforms to the conditionality bargaining
model, the contents of the above chronology should be reorganized by the game
component. Blurring timelines, components of the events will be organically reassembled
by their shared theoretical aspects, which will make it easier to link causes to the
outcomes.

As already seen, despite the imminent crisis, the MOFE representing Korean
government dismissed speculations that Korea would seek assistance from the IMF. While
the MOFE admitted that the domestic ﬁnancial market was unstable, particularly in the
aftermath of serial corporate bankruptcies throughout 1997, it did not hesitate to show

conﬁdence in the nation's economic capability, saying “it is strong enough to overcome the

232

ongoing ﬁnancial sector turmoil” (Yoo, 1997a). The ﬁnancial stabilization package
announced in November was a last-ditch measure in order to avoid the IMF, but it failed
to prevent the country from sliding ﬁrrther into ﬁnancial turmoil. Such reluctance to open
a talk with the IMF was a sheer reﬂection of the cost that the Korean government
expected to pay when it occurred. The Korean government was concerned that an IMF
rescue ﬁrnd would create many short-term economic problems, such as a rise in
unemployment, bankruptcies of ﬁnancial institutions, mounting tax burdens, and negative
GDP growth. Furthermore, “[t]he inﬁision of IMF emergency bailout money into Korea, if
implemented as planned, would surely result in a loss of economic sovereignty and seismic
upheavals of the national economy” (Yoo, 1997b). However, a bigger concern behind it
was that it would not stay isolated as economic cost, but turn into political cost easily.

The bailout by the IMF and ensuing economic cost came at a critical time for the
country's political leadership. In two weeks, Korea was supposed to hold a presidential
election. Politicization of economic cost entailed to an IMF bailout would undoubtedly
affect the election. Knowing that various side-effects would follow a rescue plan by the
IMF, it was not a readily acceptable option for the incumbent government and party, in the
midst of presidential campaign, to pursue such a controversial aid package publicly. This
would unquestionably divert votes from their candidate to competitors. If bypassing the
IMF had not been unavailable and the political costs linked to the bailout had not loomed
so large, possibly the best course for the Korean government to follow would have been
to ask for a $55 billion loan from the IMF after December's presidential polls, as the
chairman of the US. Institute of International Economics suggested (Yoo, 1997a).

Unfortunately, however, the situation was too urgent to wait until the presidential election

233

was over in order to reduce the political cost. It is hard to tell whether the ﬁnancial crisis
was an offspring of a concealed political-business cycle in Korea. However, there was, at
least, evidence that the authorities in charge neglected warnings from the society and
procrastinated in taking appropriate actions, although the government and party in power
did not blatantly manipulate macroeconomic policy to maximize their electoral chances.
In any event, the outgoing regime’s behavior before the crisis was broadly in line with an
argument that when a country already had economic problems, the government was less
likely to initiate a stabilization program in the year before or during an election (Haggard,
Kaufman and Webb, 1992). And it appears to be mainly due to the domestic political cost
that it identiﬁed with the expected result of the presidential election.

The above concern was well grounded, considering the presence of highly
mobilized labor organizations in Korean politics and economy. Korea had been rocked by
a month of worker unrest earlier in 1997 when the congress passed a controversial labor
law that would have made layoffs easier. Aggressive and persistent labor movements
forced the Korean government to postpone its taking effect until 1999. When the bailout
program was announced, as expected, the militant Korea Confederation of Trade Unions
promised to wage “all-out strikes” if companies pushed for layoffs as a result of bailout by
the IMF. And, imaginably, opposition political parties were eager to exploit voters'
feelings of anger and shame ahead of presidential election. President-elect Tae-Jung Kim,
who was a long-time opponent of the regime and linked his policy platform to the interest

of labor class, was one of the candidates who advocated re-negotiation with the IMF for

234

his presidential bid.3 His victory may have been, in one sense, beneﬁted from political
damage inﬂicted on the candidate ﬁ'om the incumbent party.

Aware of collateral side-effects of an IMF bailout, the Korean government secretly
approached its allies for a bilateral aid package so that it could bypass the IMF. The
Minister of MOFE suggested that Korea would prefer an aid package negotiated with
major trading partners, such as Japan and the United States, rather than submission to a
multilateral IMF bailout. However, Korea’s efforts to secure loans from the central banks
of the United States and Japan failed because they said that it was politically impossible for
them to loan tens of billions of dollars required to patch up the world's 11th largest
economy, and that they would extend loans only if Korea ﬁrst accepted an IIVIF aid
package (Chen, 1997).

The United States was known to have ﬁrmly rejected Korea’s bid for bilateral aid.
From the start, the Clinton administration said that it would not do in Asia what it did in
Mexico: taking the lead with a bailout plan that puts American taxpayer dollars at risk.
Instead, it insisted that Korea implement reforms that were needed to get it back on the
right path (Sanger, 1997). At the APEC summit, where Korea’s ﬁnancial crisis became a
hot potato, Japanese Prime Minister Hashimoto expressed the same view of the measures
the Korean government had to take to overcome its ﬁnancial troubles. He said, “our

government will provide the best possible support to Korea within the framework of the

 

3 The IMF was so keenly aware of the risk in bailing out the outgoing Young-Sam Kim’s regime that it
asked three leading presidential candidates to officially pledge a ﬁrll respect for the agreement with the
IMF when one of them took office. Nonetheless, President Kim promised re-negotiation with the IMF
during his campaign, even for a short time, which caused the Korean stock market to plummet further and
added difficulty to subsequent debt rescheduling talks in December. Later, he was known to change his
policy, following advice from Jong-Keun You, his top economic adviser.

235

IMF assistance” (Chon, 1997). This was not a good sign to Korea in that it made the IMF
bailout not a matter of choice from necessity, but inevitability. This meant that all other
possible opportunities for ﬁnancing abroad were already consumed. Even before the
negotiation with IMF started, the Korean government was freed from opportunity cost
that might be incurred if it accepted the bailout program by the IMF without consulting
other donors. It might be this total loss of opportunities that drove Korea to the IMF,
which would subsequently reduced Korea’s conﬁdence.

Initially, many expected that $20 billion would be enough to help Korea out of its
ﬁnancial crisis, which was mainly caused by a slew of corporate failures and their impact
on its banking system. However, when the government announced its plan for requesting
a bailout by the IMF, it suggested that the amount of bailout would be more than
expected. By the time the IMF delegation team ﬁnished research on the Korean economy,
it was revealed that the debt accumulated by the banks reached up to $120 billion, which
far exceeded expectations; thus, more assistance became necessary. Aside from burdening
the economy, such a sizable debt was poised as an immediate problem to the bargaining
with the IMF in terms of conﬁdence or subjective probability of winning the negotiation.
With other ﬁnancial windows closed, the IMF was the only source that the Korean
government should depend on for the needed capital. Furthermore, considering the
amount of loan, the conditions which would be traded off with the loan would not be easy
ones. As statistically tested in the previous chapter, the subjective probability of winning
negotiation with a preferred policy package is boosted by capital inﬂows from abroad that
meet ﬁnancial needs of the recipient. If such is the case, Korea’s subjective probability of

winning its preferred deal was very low. On the other hand, expectations that Korea

236

would receive tough conditions for the loan would be naturally high.

To make things worse, Korea also had to act quickly. While its foreign reserves
had dwindled below $10 billion during crisis, which the IMF judged as a perilous level,
Korea had to ﬁnd a ﬁnancial source to defray $20 billion short-term debt that was due at
the end of December. Korea was desperate for assistance because of this ticking time
bomb; therefore, once a bailout by the IMF became official, the Korean government
wanted an early conclusion of the talk with the IMF by accepting most of IMF’s demands
in return for rescue money. The Minister of MOFE pledged to exercise maximum
ﬂexibility in accepting the IMF's possible calls for adjustments of the nation's major
economic indices, except for the IMF's direct intervention in domestic industrial
restructuring (Yoo, 1997c). Thus, it is hardly imaginable that the Korean government
would have valued ﬁrture bargaining outcome as much as the present one with a hope for
a better deal. For the Korean government, prompt loan disbursement, followed by a quick
deal, must have looked more attractive than prolonged bargaining over minor changes in
conditions. As if vindicating this situation, the negotiation was over in a week. “Sensing
the seriousness of the Korean currency crisis, including its shallower-than-expected
foreign exchange reserves, the two sides reached the agreement on the early introduction
of IMF ﬁrnds,” commented Hubert Neiss (Kang, 1997). And the Korean government
stated that it had decided to accept most of the conditions presented by the IMF, based on
the judgement that the IMF-set terms would not give much burden to Korea's economic
policy operation (Ibid).

As easily inferred from its attempt to bypass the IMF, the Korean government’s

ﬁrst priority was getting loans from allies without strings attached. However, when it

237

turned out to be unavailable, the Korean government, like many others, tried to maximize
its utility in negotiation with the Fund by securing a loan matching its ﬁnancial needs,
while minimizing harmﬁrl conditions, which would partly compensate for the poised costs.
Unfortunately, however, this never happened, and the bargaining results tell us that the
Korean government was forced to make a choice between two equally undesirable trade-
offs, “larger loans but at harsher terms and small loans but at more favorable terms”
(Chan, 1993), and its choice went for the former.

On the IMF side, the actor who played the most conspicuous role throughout the
process was unarguably the Executive Board represented by the United States and the
Managing Director. As the largest shareholder within the IMF, the United States holds
tremendous power over Fund decisions" While rejecting Korea's call for direct ﬁnancial
help, the United States still directed how to quell a ﬁnancial crisis with the toughest
elements of the economic prescription, especially ﬁnancial reform (Passell, 1997).
Furthermore, the Managing Director who has considerable latitude in the application of
Fund principles could, and did, intervene in the bargaining process to deliver the intention
of the Executive Board as witnessed in the chronology. While the negotiation in Korea
was proceeded by the mission delegates who interpreted the ﬁnancial information in a
disinterested manner, their self-proclaimed technical decisions were blocked and
permeated by political factors fiom the Executive Board, and the Managing Director
supportive to it, as Assetto (1988) analyzed ten years ago.

In discussing the role of the Executive Board in conditionality bargaining

 

4 The United States’ voting rights as the largest shareholder is over 18% of total shares. Many key issues
require an 85% majority, so that the United States effectively has a veto over major Fund decisions.

238

particularly in East Asian countries’ cases, it is noteworthy that Japan had been, until then,
the biggest creditor to Asia, and it would be posed under the direst damage by any default
in the region. According to Moody's, about half of all external debt in Southeast Asia was
held by the Japanese banks, with a much smaller portion, between 10% and 20%, held by
lenders in the United States (O’Brien, 1997). Unlike the Latin American debt crisis of the
19803, big American banks, with the exception of Citicorp, had a limited lending presence
in Asia. The Asian debt crisis involved lenders from a number of countries, with European
and Japanese banks facing much larger potential losses than American banks, although it
would still be troublesome for some American banks if those loans were defaulted.
Nevertheless, Japanese banks exposed to the troubled region were not properly
protected by any substantial leadership from the Japanese authorities to present solutions
to the problem.5 Japan's role in the Asian crisis was largely as banker and follower, taking
the roles assigned by the United States, despite its economic size and location. As ﬁnancial
crises were sweeping through Asia, provoking fears of a global recession, Japan was too
mired in its own problems to be of much help. Rather, Japan had prolonged its own
economic torment and contributed to the Asian ﬁnancial crisis by denying its problems
(WuDunn, 1997). Japan was unwilling to help Korea independently, nor would it push
ahead with its own plan, if any, without Washington's approval. Japan put up $10 billion
for the Korean rescue package, which was twice as much as the United States’

contribution, but the rescue plan was orchestrated from Washington. In a sense, this may

 

5 The relationship between the degree of debt concentration and bargaining power in debt negotiation was
empirically tested. Taking this into account, Japan should have taken the leadership to assemble a bailout
plan for Korea in its favor. For the empirical test, see Fernandez, Raquel and Sule Ozler. 1999. “Debt
Concentration and Bargaining Power: Large Banks, Small Banks and Secondary Market Prices.”
International Economic Review 40, no. 2 (May):333-355.

239

indicate the degree of unity achieved among the ﬁve Executive Board member countries.
Although there were potentials for conﬂict among them due to different degree of
exposure to economic damage caused by the crisis in the region, a rift among them was
hardly disclosed, at least to outside world.6

Meanwhile, the United States was making its second-line contribution more
politically palatable to itself What the United States, or the IMF in general, was pursuing
in bailing out the sinking boat of Korea was solving a two-way moral hazard problem,
which is broadly in line with guaranteeing loan repayment. Rubin, Secretary of Treasury
of the Unites States stated the following:

the availability of IMF resources made debtor countries less willing to

pursue adjustment programs needed to enable them to meet their

outstanding debt-service obligations. This would effectively reduce the

risks faced by investors in poorly run economies and then investors would

not demand that national leaders enforce economic discipline, if they

become convinced that the IMF always stands ready to forestall default.
(quoted ﬁom Passell, 1997)

From this, the IMF appeared to discipline prospective borrowers through tough conditions
on behalf of unruly investors in advance, on one hand, and to secure debt service to the
IMF by them on the other.

Before the ﬁeld research was launched, the actual size of short-term debt that

Korean banks had accumulated was kept almost in secret. When its number was ﬁnally

 

" This contrasts with inter-country differences during debt crisis in the 19803 argued by Chan (1993). He
says that the US with a comparatively large exposure to Latin American countries generally took a harder
line than European counterparts in negotiating with the debtor government, insisting on drastic
adjustment efforts by the debtor government as a condition for extending new loans. Instead, the ﬁrst
official disagreement among the ﬁve Board member countries came out at the annual meetings of the
World Bank and International Monetary Fund (IMF) in October, 1998 over world economic plan
aftermath of Asian crisis. Online articles on their conﬂicting plans are available at http://cnn.com.

240

ﬁgured out, the IMF also learned the size of loan it would have to provide. Financing
Korea’s imminent $20 billion short-term debt meant that a huge portion of the IMF’s
usable ﬁrnds would have to be appropriated for one loan; thereby, not only would the
Fund’s overall liquidity position be signiﬁcantly inﬂuenced, but also the level of lending to
other members would have to be adjusted downward.7 Facing such enormous opportunity
cost from bailing out Korea, the IMF automatically might want certain compensation for
the deprived opportunities through tough conditions.

In addition to the opportunity cost, the IMF should be concerned about criticism
of it if Korea fails in economic restoration only to cause additional debt arrears to the
IMF. However, the IMF might not be unduly concerned about it, compared with the
opportunity cost, since Korea had a track record of success in economic adjustment with
Fund support in the 1980’s”, and general debt arrears to the IMF have been declining since
1992. Rather, the reputation cost the IMF suffered was post hoc. As soon as the
conditionality agreement was announced, the IMF came under severe attack for the wrong
treatment for the country that had enjoyed high economic growth until then; ﬁrrthermore,

it became entangled into debates on abolishing the IMF. From then on, the IMF had to

 

7 The impact of ﬁnancial crisis in Southeast Asia on the Fund’s ﬁnancial position is shown in graphs in
1998 Annual Report. The Fund’s annual lending reached the highest in 1997 (Figure 7, p. 88) and
liquidity position fell from the previous year by more than 100%. Related to this, during its 11th quota
review, the Board of Governors requested quota increase by 45% in January 1998 and, as of March of
1999, 153 members consented to it.

8 Korea implemented adjustment programs in 1981, 1983 and 1885. It should be noted, however, that
putting Fund-supported adjustment programs forefront as the factor for success is controversial in the
sense that the economic development model that Korea was following in the 1980’s depended more on
government intervention in economy than efficient market function, which is opposite to Fund’s structural
adjustment.

241

devote substantial energy to justifying its existence and ﬁinction without a systemic role.9

Taking its prompt reaction to Korean situation in account, quick conclusion of the
bargaining might be important to the IMF, too. As an isolated case, the IMF might not
have serious reasons to hurry up in order to bail out Korea. However, the Fund was vastly
concerned about the systemic impact of the Korean crisis, namely, contagion of crisis to
other emerging markets in Latin America and Russia, and a global recession aftermath.
The Asian economy was buoyant for the past decade and led world economic growth. In
the wake of crisis, the IMF adjusted its forecast for world output rates from 4.4% (both
for 1998 and 1999, which were published before the crisis) to a negative 1.1% for 1998
and negative 1.2% for 1999 (IMF World Economic Outlook, 1997 and 1998).10 Thus,
prioritizing containment of crisis and stability in the world ﬁnancial market, the IMF
shared interest with Korea in the fast resolution of the negotiation.

With those costs in mind, the two parties negotiated hard on a few issues, which
would be vital to maximizing utility, if they were won as preferred. And each party’s
standing place appears to have had a signiﬁcant impact on the ﬁnal outcome of the
negotiation. The ﬁrst issue for haggling was annual grth target. The IMF called for
Korea to adjust the nation's GDP growth target for 1998 drastically from the initially set
6% to less than 2.5%, whereas Korea offered 4.8% or a minimum expansion of 3%. The

Korean authorities feared that a GDP growth of 2.5%, and the resulting belt-tightening,

 

9 This is well indicated by the number of ﬁles on IMF website that sells the IMF. Since the end of 1997,
the IMF’s website has been inundated by documents that explain the Fund’s role in Asian crisis and its
new mission for capital market stability.

1° This was a forecast in May report. It was upgraded to 2.0% and 2.5% respectively in the October report.

242

would sharply raise the nation's jobless rate leading to overall social unrest. However, it
was settled for 2.5%, as the IMF wanted.

The other issue that the two sides failed to narrow their differing perceptions on
was in relation to liquidating troubled ﬁnancial institutions. It was inevitably expected that
ﬁnancial sector restructuring would be a troublesome point of the negotiation, taking into
account the IMF’s great interest in the ﬁnancial sector during its ﬁeld research, as well as
the Korean government’s circumventive announcement earlier. Thus, negotiations focused
on policy options to shake off Korea's ﬁnancial system through M&As, capital increases,
and third-party takeover. As concrete measures for restructuring ﬁnancial industry, the
IMF proposed that Korea sharply expand its ﬁscal spending and mobilize more taxpayers
money, which it said would help reduce non-performing loans at banks and merchant
banks. This automatically meant the revision of the domestic tax system, too. The IMF
also strongly demanded that some insolvent banks, as well as several ailing merchant
banks, be liquidated or merged with foreign banks. Other banks must tidy their portfolios,
exchanging short-term liabilities for long-term debt that is less subject to speculative
meltdown. Most of all, the IMF presented that more rapid capital account liberalization
would be required to provide momentum for the overall reform process by weakening
entrenched vested interests and the bureaucratic control on private sector activities.

The IMF also challenged the economic methods that ﬁreled the nation’s grth by
asking for the restructuring of business groups, so-called chaebol. As a prerequisite for
assistance, the IMF insisted that the Korean government discard policies favoring chaebal
in ﬁnancing and taxation, while ending the policy of forcing unsound businesses to merge

into solid ones. The conglomerates needed restructuring to leave them with less debt, and

243

to change to a management style which measures success in proﬁts rather than market
share (Passell, 1997). Linked to corporate governance, the IMF demanded from the
ﬁnancial institutions greater accountability, more responsible lending and investment
practices, and a move away from the back-room deals that had brought great wealth to
politicians and favored businessmen.

The Korean government was reluctant to accept the IMF’s demand for an
immediate shutdown of heavily-indebted banks and merchant banks. Korea wanted a
softer measure on those issues, citing enormous negative impact of banks' liquidation on
the overall Korean economy. Regarding the banking sector restructuring, Korea asked the
IMF to give three to six months to resolve the ailing banks' problems in accordance with
its own program, unveiled in November. The Korean authorities worried about rising
unemployment, labor unrest and the possibility of runs on banks by frightened depositors;
therefore, they argued for allowing the ailing institutions to be revamped or merged with
healthier ones. This request was understandable considering Korean banks’ long-standing
role in the economy. The Korea government’s perception on the potential closing of the
merchant banks was well pointed out by a researcher at a securities company:

It is being taken very seriously in Korea because, for their scale, they have

provided very sizable credits to Korean business and industry. In particular,

the merchant banks are important sources of loans for smaller and midsize

businesses. The effects of any merchant bank closings on employment

would be small because the average merchant bank employs 100 to 300

people. But the government is more worried about setting a precedent for

the liquidation of commercial banks, which are far larger than merchant
banks and employ thousands of people. (quoted from Pollack, 1997)

244

The effort to ﬁnd a formula for liquidating troubled banks was ﬁustrated because the two
parties had quite different conceptions of what would constitute such a formula. And it
was in this matter that Camdessus perceived Korea as “coming round to the IMF’s stand
on the most sensitive point, what should be done with the country's merchant banks”
(Yoo, 1997d). Nonetheless, the MOFE had no other choice but to suddenly halt the
operation of nine merchant banks in the middle of bargaining in the hope of facilitating
negotiations, as well as easing the terms. However, even such a desperate action could
not reverse the course of event. The Korean government accepted the most risky
conditions set by the IMF, of which implementation is certain to expose all local industrial
and ﬁnancial enterprises to indiscriminate foreign mergers and acquisitions, simultaneously
opening capital markets to foreign speculators (Yoo, 1997c).

A legitimate question at this point is how the theory of leverage can explain this
negotiation result, and what would have been done differently given the underlying power
relationship between the IMF and Korea. To begin with, the Korean crisis attracted
immediate attention ﬁ'om the IMF, and the procedures gained momentum quickly once it
was set in motion. This seemed mainly due to Korea’s status in the world economy. There
was a presumption that immediately following the debt crisis in late 1982, the IMF would
be anxious to avoid a general international banking and ﬁnancial collapse (Bird, 1995). By
the same token, when Korea’s crisis was divulged to the rest of the world, both the IMF
and the United States underscored the risks of much broader trouble for the world
economy if they failed to act. The United States, which views relative stability in
international ﬁnancial markets as an important foreign economic policy goal because the

alternative is still seen as a source of potential political instability at both domestic and the

245

international level, feared that Korea’s failure could resonate in Japan and around the
globe (Altbach, 1998). An economically strong Asia that imports as well as exports, and
thereby supports global growth, was also good for global interest. Thus, the economic
cost was expected to be enormous if the IMF failed to rescue Korea. This was leverage
that Korea had at a systemic level.

At a particular level, on the other hand, Korea has a signiﬁcant factor to concern
the lenders, which might be called leverage. “We are convinced that economic stability is
important to our national security concerns especially in Korea, one of the last surviving
arenas of cold war conﬂict,” Rubin argued (Sanger, 1997). At that time, four-party talks
(North and South Koreas, China, and the United States) were coincidentally scheduled in
Geneva over a peace treaty that would ﬁnally end Korean War. The United States had a
lot riding on a stable and prosperous Korea on the comer of northeast Asia. If Korea’s
ﬁnancial health continued to deteriorate, the diplomatic pressure on Washington to step in
could become excruciating.

Other than the security concerns, there were some other palpable economic
interests at stake in Korean ﬁnancial crisis, too. Rubin confessed in an interview with the
New York Times (12/18/97) that “you have to remember that the crux of our concern is
America's interests. Our economy can certainly be affected if a sufficient number of
emerging nations head into a prolonged crisis.” This statement pertains to economic
(inter)dependence as leverage. Asian products account for 11% of all imports to the
United States, and Japan alone accounts for an additional 17%. The Asian ﬁnancial crisis
helped drive down the price of imports to the United States——with goods from Korea,

Taiwan, Hong Kong and Singapore dropping an unprecedented 1.2% in November of

246

1997, reﬂecting the steep decline in the values of their currencies relative to the dollar
(Stevenson and Sanger, 1997). Tumbling price of imports from Asian countries was a
boon to consumers and a big help in holding down inﬂation. However, the proﬁts of
multinational corporations were already under pressure by the downturn in Asia. In
Korea, US goods got nearly twice as expensive with the country's suddenly devalued
currency, which was squeezing many US. exporters—from auto parts manufacturers to
farmers—and could imperil the jobs those exports support. With inﬂation already down
to about 2%, the Federal Reserve Board avoided raising interest rates despite the tight
labor market because the turmoil in Asia would lower US. exports to Asia, and thereby
would cut economic grth in 1998 by half a percentage point or more. Higher interest
rates would not only aggravate the Asian distress by ﬁrrther weakening their currencies,
but would risk slowing the US. economy too much. Under the circumstances, the near-
certainty was that the US. trade deﬁcit would be driven sharply higher by the economic
crisis sweeping through Asia and have profound economic and political implications
(Ibid.). This would have an effect to drive the trade deﬁcit back to, or above, the records
set in the late 1980’s, although it would still be a smaller share of economic output than in
the '803. After years in which the strong economy has dampened confrontations over
international trade, the issue would be poised to ﬂare again against a backdrop of turmoil
in Asia as imports surge and exporters ﬁnd it more difficult to sell their goods abroad
(Ibid.). Just as Korean needed the Untied States as an export market, the United States
also needed Korea and other Southeast Asian countries as an export market. Both sides
appeared to have equal footing in terms of leverage indicated by economic

interdependence.

247

This was the leverage that both sides of the bargaining table commonly recognized.
Before and during the bargaining, the Korean government wanted to maximize its utility in
the loan it needed and easy terms, and such expectation was reasonably coming from the
self-perceived signiﬁcance in the world economy and security signiﬁcance in the Northeast
Asia. It was fairly contextual leverage. Unfortunately, however, that was all. Korea’s
attempt to chase two rabbits at the same time did not bear fruit. It turned out to be so,
since Korea failed to transform its positional signiﬁcance into bargaining power to bring
substantial change to outcomes (i.e., softening conditions while maintaining the level of
loan due to the lack of relevant leverage to back it up). The circumstantial weakness—the
need for immediate rescue with a huge loan—was so overwhelming that it limited the
range of options it could choose and let its economic status be preyed upon through
drastic ﬁnancial market opening.

The lack of effective leverage on the Korean side cannot be better shown than in
abandoning its plan for ﬁnancial market opening, which was formed in accordance with
the United States’ request a few years ago. Moreover, the conditions that the Korean
government agreed upon appear to be more than corrective measures for restoring
macroeconomic balance (The Economist, 1997). It is even dubious whether such a ﬁrll
opening of ﬁnancial market was indispensable to reestablishing macroeconomic
equilibrium.ll For fairness’ sake, the causes of the crisis, as will be seen later, were

pertinent to governance and transparency of the ﬁnancial system, which was recently

 

” The inclusion of structural components in the conditions caused debates later on between the IMF and
critical scholars. For instance, Feldstein (1998) argued that such measures endangered the IMF ’s
effectiveness in dealing with the problems. As a response to this, Fischer (1998d) argued that such
measures were absolutely necessary, since the causes of the ﬁnancial crisis were structural ones rather
than balance-of-payments problems.

248

added to the menu for Fund-supported structural adjustment. Although may of the
objectives underlying the program were laudable, capital account liberalization might be
neither a necessary nor a sufficient condition to establish such transparency, which reﬂects
a troubling lack of institutional self-restraint under pressures from major shareholders
(Kapur, 1998). Let alone the weak theoretical grounds that it was holding toward
ﬁnancial market opening, it was a quasi tradition within the Korean government to protect
its ﬁnancial market, and it could not be overcautious about opening its markets. The lack
of leverage, however, not only brought an end to what the Korean government had
protected in the most defensive manner, but also deprived it of opportunities to take its
pace with autonomy. Throughout the negotiation between the IIVIF and the Korean
government, the ﬁnancial sector restructuring was the most hostile issue, where the power
relationship between the two players was ultimately tested.

The overall process of the negotiation between the IMF and Korea and its
outcome ﬁt Agreement ® of Figure 1. At ﬁrst, it appeared to be an encounter of two
strong proposer with high propensity for agreement, which might yield Agreement C33 as in
Table 3.1. Various attempts by the Korean government to circumvent the IMF’s demands
at the beginning hardly make it a weak bargainer. It is also unreasonable to call Korea a
strong proposer with low propensity for agreement, since the imminent short-term debt
did not allow the Korean government to be intransigent at all, although it wanted to avoid
particular conditions such as ﬁnancial market opening. If it could be a hard bargainer,
Korea might have walked away from the negotiation by declaring veto at node 4. Unlike
the beginning of the talk, the circumstances were rather unfavorable to Korea, and were

even getting worse as time went by. Having low subjective probability, high domestic

249

political cost, and no luxury to discount ﬁiture bargaining outcome already counted, the
intervention by the Managing Director in the middle had an effect of almost vetoing the
proposal made by the Korean government. Conﬁonting the donor who was much stronger
than it expected at the ﬁrst round, the Korean government settled for Agreement ® at
node 4 by accepting a proposal by the IMF. Such turnaround in the ﬁnal outcome of the

negotiation signiﬁes the role of leverage in conditionality bargaining.

6.3 IMPLICATIONS TO DEBATES ON CONDITIONALITY

Theoretical reconstruction in the previous section conﬁrmed that there exists
conformity between the conditionality bargaining model and the Korean case. The
effectiveness of leverage that each player mobilized to the negotiation was ruthlessly
contested with the utilities, which were maximized in terms of toughness of conditionality.
Speaking of the toughness of conditionality, radical capital account liberalization appears
to be its nucleus, and the saliency of issues in ﬁnancial market opening (a policy extrinsic
to returning an economy to equilibrium) leaves us perplexed about the nature of IMF
conditionality. Searching for the meaning of such characteristics in the Korean bailout to
the general debates regarding conditionality requires examination of the origins of the
crisis. As will be seen, they are not merely causes to a particular crisis, but they reﬂect the
context of the bargaining and set the course for what would be pursued through
conditionality.

Needless to say, the current international economic system is characterized by high

interdependence and capital mobility. International barriers to the ﬂow of capital have

250

been disappearing rapidly, and investors and ﬁnancial institutions move money around
wherever they wish. This increased ﬁnancial integration holds government hostage to
foreign exchange and capital markets, forcing greater ﬁscal and monetary discipline than
they might otherwise choose. National systems intended to supervise banks in their home
countries have proven unable to keep pace with the rapid development of a global
ﬁnancial marketplace that pays little attention to borders. The United States and other big
countries also attempted to bring changes to governments and banks outside their borders
by encouraging greater disclosure of ﬁnancial data.12 Unfortunately, however, those
efforts to bolster the soundness of ﬁnancial systems around the world have so far yielded
little, if any, success.

In addition to this, there is no international body able to play the role of global
regulator. While the IMF, other multinational bodies, and national governments are willing

to come to the rescue at crises,l3 they lack the crucial power that usually goes hand-in-

 

” Most standards, guidelines, principles and practices have been developed by Basel Committee on
Banking Supervision within Bank for International Settlements (BIS) since 1974. The Basel Committee,
which consists of 12 industrialized countries, continuously revises and adapts them to market
developments. Although focusing on banking system, its jurisdiction extends to non-bank ﬁnancial
intermediaries, such as securities, insurance, and pension funds, in order to harmonize rules, regulations,
and practices in individual countries, as well as internationally. The BIS held a conference on how to
measure aggregate market risk and regulate excessive market exposures in the spring of 1997, just when
the East Asian crisis began to unfold, and concluded that not many regulatory schemes were feasible. For
more information on international capital rules, visit http://www.bis.org.

'3 The IMF’s regulatory role in international capital markets is represented by the establishment of the
Interim Committee of the Board of Governors on the International Monetary System in 1974, the Fund’s
ministerial oversight body on the management and adaptation of the international monetary system, as
well as dealing with disturbances that might threaten the system. It decided in April 1997 to seek to
amend the Articles of Agreement of the Fund so as to bring capital movements formally under its
jurisdiction, and make the liberalization of international capital movements 3 central purpose of the Fund.
For details, visit: http://www.imforg/extemal/np/cm/ 1997/CM970428A.htm. Related to that, the IMF’s
efforts to enhance the soundness of the ﬁnancial system has focused on the banking system, the
framework of which is well discussed in IMF. 1998b. Toward a Framework for Financial Stability.
Washington, DC: International Monetary Fund. The IMF also established Special Data Dissemination

251

hand with status as the lender of last resort (i.e., the ability to demand changes in a bank's
management before crisis hits).14 A result from this is that, to a remarkable extent,
individual nations are suddenly more at risk because of the ineffectiveness of obscure
banking regulators in far-off countries. Thus, one nation’s trouble in international
payments system might have serious implications everywhere, too. What is different from
the past is that the stakes are now too high that collapse of an important country's banking
system could produce a domino effect (Norris, 1997).

The recent crises in East Asia were nothing but the manifestation of such a risky,
but irreversible, trend in the international ﬁnancial system. Korea, Thailand and Indonesia
were among the countries paying a steep price for the problems with their banks. These
countries failed in common to dampen overheating pressures that were manifested in large
external deﬁcits and property and stock market bubbles, and their lax prudential rules and
ﬁnancial oversight led to a sharp deterioration in the quality of banks' loan portfolios
(Fischer, 1998b). As the crises unfolded, political uncertainties and doubts about the
authorities' commitment and ability to implement the necessary adjustment and reform
exacerbated downward pressures on currencies and stock markets. However, it should be
admitted that large private capital ﬂows to these markets—in an imprudent search for high

yields by international investors without due regard to potential risks—contributed, to an

 

Standard (SDDS) in 1996, which guides data disclosure on economic, banking, and other ﬁnancial
statistics to the public.

14 One of IMF’s ﬁrnctions is surveillance on developments and problems in the international economy
and in individual economies. Once a year, the Fund’s staff prepares an Article IV report for each country,
an in-depth analysis of the country’s economic policies and performance. The IMF also issues warnings of
impending problems. However, heeding such warnings is member counties’ discretion. Fischer (1998b)
suggested later that the Fund warned Thailand of an impending crisis, but that actions were not taken.

252

important degree, to the buildup of the imbalances that eventually led to the crises.15
Market participants were aware that macroeconomic conditions and export performance
in the region were not so positive as in the past, and that speculative activities and short-
terrn capital ﬂows into the region were becoming more pronounced. Nonetheless, they
chose to ignore all of these considerations and to focus instead on the region’s high
growth, comparatively low inﬂation rates, lack of signiﬁcant budget deﬁcits, openness of
the region’s markets, and high rates of return that their past investments in the region had
yielded (UN, 1998).

Korea’s capital market was also vulnerable to the currency turmoil that swept
other Asian countries in the second half of 1997. After the crash in the Hong Kong stock
market in late October of that year, Korea's economy was attacked by the same forces that
raided other South Asian countries, primarily unmanageable overseas debt loads. Short-
terrn debt grew fast in 1995.16 Partly reﬂecting a jump in trade credit associated with the
expansion of trade and the liberalizing steps. In 1997, about $66 billion of Korea's
estimated $120 billion in overseas debt was short-term, of which, about $20 billion was
due at the end of the year. Such debt loads had dragged down the value of the won and

forced many companies into bankruptcy. Seeing its currency plummeting unchecked,

 

15 Up to the current crisis, Asia attracted almost half of total capital inﬂows to developing countries,
nearly $100 billion in 1996. Capital movements in both matured and emerging markets before and after
the crisis are extensively analyzed in Adams, Charles, Donald J. Mathieson, Garry Schinasi, and Bankim
Chadha. 1998. International Capital Markets: Developments, Prospects, and Key Policy Issues
(September). Washington, DC: International Monetary Fund.

16 Looking at debt to GDP ratio, Korea’s external debt increased from 14.7% of GDP at the end of 1994
to 17.2% at the end of 1995, and by a ﬁuther $ 6.5 billion (1.3% of GDP) in the ﬁrst quarter of 1996.
Accordingly, interest on short-term debt increased sharply in 1995, but at 5'/2% of exports, although the
overall debt-service ratio remained moderate. Source: Adams, C., J. G. M. Schulze-Ghattas, V. Arora, J.
Zhou, and A. Bakastov. 1996. Korea, Selected Issues. IMF Staﬂ Country Report 1996/136.

253

foreign banks began reﬁrsing to roll over Korea's burgeoning short-term debt after back-
to-back corporate bankruptcies, which shook up the fragile banking system.

One thing noteworthy in that debt problem is that almost all of the debt was owed
by the private sector. Unlike elsewhere, the Korean government itself was relatively free
of debt, which made the situation dramatically different from the crisis that afflicted
Mexico in 1994-95, or the Latin American countries in the 1980’s. The IMF evaluated in
its country report that the ﬁscal situation in Korea was sound as a result of a long tradition
of ﬁscal conservatism: over the years, ﬁscal expenditure, which was relatively low by
international standards, has been broadly in line with revenue, resulting in relatively low
levels of government debt and high government saving (Adams et al, 1996).

Korea’s immediate problem lay in the crushing debts of its banks and chaebol.
Nonetheless, the origin of the crisis was rather deep-seated: state-manipulated ﬁnancial
systems, long a key formula the Korean government prescribed for a decade of
extraordinary growth (Alexander, 1998). In Korea, backed by a seemingly limitless supply
of cheap loans from govemment-controlled banks, a few dozen politically connected
conglomerates had long grew at an incredible pace along with the economy and dominated
the industrial landscape (Choi, 1993).17 Corporate bonds, as a way of mobilizing capital,
was hardly used until a system of bank guarantees for corporate bond issues was
introduced. In 1988, about 97 per cent of corporate bonds were guaranteed by banks,
which meant effectively that they were converted into a means of indirect bank borrowing

(Mahler, 1990). As a result, the conglomerates' equity stretched paper-thin, and the banks

 

'7 His article methodically reviews economic development through ﬁnancial support for business groups
from the Korean goverrunent.

254

went deeply into debt. The risk in holding corporate bonds was not the bankruptcy of the
company, but the bankruptcy of the guaranteeing bank.

Under the circumstances, any default of debt service obviously didn’t mean that
the Korean government failed to honor the nation's sovereign debt, but that some of the
nation's banks and corporations were unable to pay back loans. However, even when such
private default takes place, it signiﬁes a far more serious disaster for the Korean
government. Until then, banks in Korea had close ties to the government, and default by
them would, by any means, would do harm to the nation's reputation and complicate the
prospects for a recovery (Kristof, 1997). For this reason, the Bank of Korea was forced to
pay $10 billion in short-term foreign debt on behalf of Korean commercial banks just
before the outbreak of crisis, which turned out to be beyond its ability with far low foreign
reserves.

It is undeniable that an excessive unhedged foreign borrowing by the domestic
private sector—and a lack of transparency about the ties between government, business,
and banks—both contributed to the Korean crisis and complicated efforts to deﬁrse it.
The resulting ﬁnancial sector weakness accelerated the currency crisis to a certain extent,
since such weakness was seen as limiting the scope or willingness of authorities to use
interest rates to defend the currency, or cast doubts on the prospects of the economy more
generally, or cast doubts on the political autonomy and integrity/governance of ﬁnancial
institutions (Fischer, 1998c). And this formed a natural environment under which capital
ﬂight was encouraged, regardless of the capital control framework. The Korean
government was once praised by scholars as the prototype of developmental state for its

ability to penetrate into and transform the society for development purposes (Evans,

255

1992). Now it has become the nation’s nightmare.

Seen from this perspective, the nature of the Korean crisis was a temporary
illiquidity, not ﬁrndamental insolvency (F eldstein, 1998). And the crisis broke out because
of bad governance, not because of insuﬂicient ﬁnancial market opening. If that was the
case, the matching conditions to cure this illness do not intuitively need wide ﬁnancial
market Opening. Nonetheless, the IMF insisted on addressing domestic issues with
external treatments, and the interface between the two foreign domains should be found in
international capital mobility and opportunity for investment: the remarkable performance
of foreign portfolio investment in Korea that had yielded high returns with every type of
investment.

Considering this performance, it is not surprising that considerable excess demand
by foreigners for participation in the Korean capital market has emerged. The magnitude
of this demand is clearly reﬂected on the extent to which the stock in the Korea Fund has
traded at high premiums over its net asset value. When they were initially issued in 1984,
Korea Fund shares traded at a premium in the 10-20% range, which increased gradually to
reach a remarkable 157% at the end of August 1987, before the stock market crash in
New York in October 1987. Even after that incident, such a strong foreign interest in
investing in the Korean stock market continued, combined with the desire to ﬁrrther
diversify their portfolios in emerging market economies. From a long-term perspective on
performance of the Korean economy, it was reasonable to expect the Korean stock market
to have a superior performance in the ﬁtture, too. Although the prices of Korean stocks
and Korean corporations’ price-earnings ratios increased sharply in the past seven years,

they did not appear to be unreasonably high in light of the superior performance of most

256

Korean corporations (Mahler, 1990). The hefty premium in the price of Korean Fund
shares indicated such international viewpoint, and suggested that Korean stocks were still
signiﬁcantly underpriced.

Although the balance of payments was switched ﬁom surplus to deﬁcit in the early
1990’s, it was comfortably ﬁnanced by the positive response of capital inﬂows to a slight
relaxation of controls, which could be interpreted as materialization of the above
expectation. The net private capital account strengthened sharply in 1991, reﬂecting an
increase in foreign bond issues by Korean corporations, and in 1992, net portfolio inﬂows
nearly doubled, reﬂecting the partial opening of the stock market to foreign investors. Net
portfolio investment inﬂows continued to strengthen over the years, with net private
capital inﬂows reaching $16 billion in 1995. Korea continued to be a net exporter of direct
investment, but increasingly became a major net recipient of portfolio investment inﬂows.

In contrast to such seemingly abundant potential for investment, ﬁnancial policy of
Korea that frames investment environment had exhibited a surprisingly low level of
openness, particularly given the country’s strong export orientation. Throughout its rapid
economic development, the Korean economy was characterized by signiﬁcant government
intervention and a ﬁnancial sector which lagged the industrialization of the economy.
Given the proﬁtability of the Korean market, therefore, it appears natural that the pent-up
demand for ﬁnancial services in protected markets increased the interest of foreign
ﬁnancial institutions in lowering the barriers to entry. As the so-called emerging markets
have grown, the opportunity costs of being closed out of them have grown
correspondingly, and foreign ﬁrms have become active lobbyists for liberalization

(Haggard and Maxﬁeld, 1996). US. based ﬁnancial ﬁrms have been particularly

257

aggressive in securing diplomatic support for their interests. These political pressures
have played out at a number of different levels, from the formulation of the services at
GATT (now WTO) to regional negotiations (such as the North American Free Trade
Agreement) to bilateral consultations.

A common cause of ﬁnancial market opening across a large number of developing
countries was more frequently balance-of-payments crises, which was more typically made
manifest in speculative attacks of exchange rate, rapid capital ﬂight, and a sudden collapse
in the availability of external lending. From the government’s perspective, even under
acute international pressures, capital account liberalization helps resolve both short and
long-term foreign exchange problems by increasing the credibility of the government’s
economic policy stance in the eyes of creditors and earners of foreign exchange (Haggard
and Maxﬁeld, 1993). However, compared with other middle-income developing countries,
Korea had been relatively immune to balance-of-payments trouble. In fact, from 1985
through 1988, Korea experienced large current account surpluses, peaking at over 8% of
GNP in 1987. These surpluses generated pressures that were exactly the opposite of those
experienced in the crisis cases. In the absence of pressing balance-of-payments constraints,
the Korean government was under little pressure to liberalize. Thus, its policy shift did not
come as a response to balance-of-payments difficulties.

Instead, the main sources of pressure on Korea came from the US government.
Under the 1988 trade bill, the US Treasury Department was authorized to determine
whether countries manipulated their exchange rates to prevent effective adjustment, or to
gain competitive advantage. The department found that Korea was manipulating its

exchange rate, and in February 1990, Financial Policy Talks between the two countries

258

were launched. These talks coincided with an array of complaints from American banks
that they experienced discrimination in their Korean operations, and from American
securities ﬁrms that began to develop an interest in gaining access to the lucrative Korean
market (Haggard and Maxﬁeld, 1996). In that talk, the MOFE presented a blueprint for
the “comprehensive” liberalization of the ﬁnancial sector, which was tailored to American
requests. The United States, acting at the behest of American banks, pressed Korea to
open its ﬁnancial markets, which was how the capital account liberalization started in
Korea.

Historically, ﬁnancial-sector reform in Korea was an ongoing effort, with the ﬁrll
effects of some of the already-implemented measures still to take hold, and with further
measures on the agenda. Since it started in the early 1980’s, it accelerated in the late
1980’s with pressures from the United States. Over the years, a range of reforms aimed at
the development of domestic money, security, and foreign exchange markets developed.
These measures paved the way for a broader opening of capital account for portfolio
capital ﬂows in the 1990’s. Moreover, as long as Korea was to join the Organization for
Economic Cooperation and Development (OECD) in the near ﬁrture, the liberalization of
the domestic capital market was considered to be a necessary condition.18 Eventually,
complete liberalization measures that place limits on access to trade, credit, and
quantitative ceilings on the amount of foreign borrowing were going to be abolished by
2000, provided that stable macroeconomic conditions can be maintained throughout and

the interest differential has narrowed to 2% points. Following this schedule, even in the

 

'8 Originally Korea was expected to join OECD in the late 1980’s. However. its actual join was delayed
until 1996.

259

ﬁrst half of 1997, several measures were taken to ease investment inﬂows. ’9

The progress in capital market internationalization shows two major concerns for
the Korean authorities: to ﬁrst ensure that domestic ﬁrms not be taken over by foreigners,
and then to ensure that the scope of potential inward or outward capital ﬂows not be so
large as to seriously disrupt the stability of the domestic capital market or the won
exchange rate. In light of these concerns, it was considered desirable to proceed gradually
with internationalization of the capital market, which was also conﬁrmed by the IMF in a
research paper (see Mahler, 1990). The reasons that the Korean government was likely to
open the stock market only gradually are: ﬁrst, there was a strong demand by foreigners to
invest in Korea, and this, if granted, would put pressure on the won to appreciate.
Second, there is a desire to avoid foreign purchase and control of Korean ﬁrms. Third,
there was a concern that foreign investors would have a speculative, short-term focus,
which might lead to destabilizing movements in and out of the Korean market, particularly
during and any periods of serious domestic unrest. Lastly, it was believed that Korean
securities companies need more time to develop before they were in a position to compete
with foreign securities ﬁrms. It is obvious that if liberalization of ﬂows was accompanied
by liberalization of entry, these gains might accrue to foreign ﬁnancial intermediaries,
which have access to large pools of foreign ﬁrnds, superior technology, and a
sophisticated knowledge of foreign market opportunities. Although they were the

rationale for gradual liberalization of capital account, such policy brought the opposite

 

‘9 For example, a capital increase of about $200 million was authorized for the Korea Fund; the limit on
foreign ownership of Korean equities was raised from 20% to 23%; foreign investors were allowed to
purchase non-guaranteed bonds of small and medium-sized companies with maturities of 3 years and
above and up to 50% of the issue; and restriction of the usage of long-term loans with maturities of over 5
years brought into the country by foreign manufacturers were abolished.

260

result, losing transparency in the relationship between the state and the market.

Also, the approach to the liberalization of capital inﬂows and outﬂows depended
on the priorities adopted by the country. In the case of Korea, management of balance-of-
payments current account appeared initially to be the overriding consideration, and Korea
sought to manage current account surpluses and deﬁcits through regulatory and other
changes which inﬂuenced capital inﬂows and outﬂows (Johnston et al, 1997). The
authorities’ gradual approach to capital account liberalization aimed to balance the long-
run beneﬁts of a more open capital account with the potential adjustment costs associated
with a more rapid opening.

Even before the actual negotiations for bailout began, pressure for market opening
was quite expected in Korea. Major trading partners of Korea, particularly the United
States, may raise their calls for ﬁrrther opening of domestic markets, especially
liberalization of those sectors in which they enjoy advantages by using the forthcoming
rescue package of the IMF as a leverage, which turned out to be the ﬁnancial market
(Chun, 1997). Targeting ﬁnancial market was not groundless, taking into account that just
recently the WTO, after long negotiations, concluded a ﬁnancial services pact that ﬁrrther
opens up the banking, securities, and insurance industries in scores of nations. Therefore,
such radical ﬁnancial market opening was introduced at the expense of the previously
established plan under the pressure from the IMF, more correctly, the Executive Board,
and it was not surprising that Rubin expressed his content, saying that “the United States
welcomes the program, which includes signiﬁcant ﬁnancial restructuring measures to open
Korea's ﬁnancial markets to foreign participation and improve corporate governance as

well as important macroeconomic policy changes” (CNN, 1997).

261

The fact that the condition of ﬁnancial market opening was driven by the
Executive Board can be indirectly proved by the different views held by IMF experts on
capital account liberalization. As well known, capital account liberalization can have
signiﬁcant beneﬁts for economic growth and welfare. At the same time, however, the
opening of capital account entails certain risks if not accompanied by the necessary
structural reforms and macroeconomic polices. Therefore, there exist three different
views with regard to the liberalization of capital account: one view says that capital
account liberalization should follow liberalization of current account and the domestic
ﬁnancial system (Edwards, 1984); another view argues for simultaneous liberalization of
current and capital accounts (Krueger, 1984); and Hanson (1994) suggests that a stable
macroeconomy and domestic ﬁnancial liberalization to a signiﬁcant degree are
preconditions to international ﬁnancial liberalization. Against these theoretical backdrops,
IMF experts (Johnston et al, 1997) hold a stance that attention needs to be paid to
sequencing capital account liberalization with structural measures, especially in the
monetary and ﬁnancial sector, and to pacing liberalization in conjunction with the
development of appropriate macroeconomic policies. The liberalization of capital account
and other aspects of economic and ﬁnancial sector reform are a good deal more complex
and not readily open to stylized presentation on sequencing. Different components and
aspects of capital account should be liberalized at various stages, along with aspects of
current account and domestic ﬁnancial sector in line with countries’ overall
macroeconomic objectives (Bhattacharya, 1997). Lack of coordination between the

domestic ﬁnancial sector and capital account reforms can create distortions and regulatory

262

incentives for capital movements that are unrelated to the underlying economic conditions,
thus risking greater instability in capital movements. 20

IMF economists (Johnston et al, 1997) predicted that capital account liberalization
would continue in a number of developing countries. It follows that they assume that the
Korean authorities will continue to implement their program to gradually liberalize capital
account and domestic ﬁnancial sectors. Moreover, given Korea’s strong ﬁrndamentals and
proﬁtable investment opportunities, the deﬁcits in overall balance of payments will be
comfortably ﬁnanced by a mix of debt and non-debt-creating capital ﬂows. However,
somewhat faster liberalization would imply somewhat larger current account deﬁcits,
together with higher grth and investment in the near term. Larger current account
deﬁcits would be comfortably ﬁnanced, but could reduce the ability of the economy to
respond to unfavorable external developments, such as unexpected deterioration of terms
of trade (Adams et al, 1996). Most importantly, however, a much faster pace of external
liberalization would risk complicating short-run macroeconomic management and could
place signiﬁcant upward pressure on the exchange rate.

Through the conditions attached to the Korean bailout, the politicized Executive
Board revealed its deep commitment to mobilizing international capital, reducing risk for

investors, and pressing countries to open markets at virtually any cost. A growing number

 

2" Stanley Fischer also said at an IMF seminar that capital account liberalization had to be phased
appropriately so as to retain some capital controls in transition. He favored market based controls, such as
were maintained by Chile, over quantitative limitations on ﬂows. This stance was reconﬁrmed in an IMF
seminar this year. Distinguished economists had one voice in advocating orderly liberalization of capital
account. See Fischer, Stanley. “Capital Account Liberalization and the Role of the IMF.” speech delivered
at IMF seminar “Asia and the IMF,” Hong Kong. [Online] Available
http://www.imf.org/extemal/np/apd/asia/ﬁscher.htrn, September 19, 1997; and IMF Economic Forum.
“Capital Account Liberalization: What’s the Best Stance.” IMF, Washington, DC. [Online] Available
http://www.imf.org/extemal/np/tr/ 1998/TR981002A.htm, October 2, 1998.

263

of critics in the United States and across Asia argue that the IMF is not only acting at
Washington's bidding, but that it is protecting foreign investors, including many American
banks and businesses that made foolish investments in Asia (Sanger, 1997). There was an
evidence that large Japanese, European and American banks institutions missed warnings
signs in Korea, but became, in some ways, the beneﬁciaries of Korea's $55 billion bailout
package. While the United States would only put American taxpayer dollars at risk as part
of a “second line of defense,” backing up the IMF and other international institutions, it
was exerting considerable effort to get Korea to open up their ﬁnancial markets to the
Merrill Lynches of the world. Although the events of 1997 have highlighted the risk of
investing in emerging markets, the process of global portfolio diversiﬁcation is likely to
resume over the longer term (Knight, 1998). And this means that the IMF's conditionality
will continue to reﬂect pressure from world ﬁnancial centers on developing countries to
liberalize their capital accounts prematurely.

The Korean bailout brought in the old criticisms revolving around IMF
conditionality. The IMF called for high interest rates and ﬁscal austerity to bring back
foreign capital and reverse the depreciation of the Korean won. A related concern was
that the prescriptions the IMF is enforcing in Korea and other Asian countries will squeeze
the growth out of those countries, touch off unemployment, and risk social unrest.21 Thus,
Jeffrey Sachs has critically proclaimed that the IMF's seal of approval is a seal of doom;

with such deep uneasiness, austerity will push Asia into recession and political instability

 

2' Again, this was defended by the IMF on the ground of stemming the outﬂow of capital, while making a
decisive start on the longer-term tasks of restructuring the ﬁnancial sector, bringing ﬁnancial sector
regulation and supervision up to international standards, and increasing domestic competition and

transparency (Fischer, 1998c).

264

(Passell, 1997). It was also folly to call for policies that reduce consumption and
investment in a country like Korea with high savings rate and a high-technology industrial
sector. Alternatively, Sachs suggested that the United States and Japan should have
moved quietly months ago to stem panic in the Korean currency market, in essence acting
as the lender of last resort to an illiquid, but basically solvent, economy (Ibid).22 The
heavily leveraged IMF laid out detailed conditions on Korea that it would have a rather
self constraining effect and may not leave room for ﬂexible and prompt adjustment to
ﬁrture contingencies in the international economy.

Supportive to Sachs, Asia's economic crisis deepened as Moody's and Standard &
Poor's, the world's largest credit-rating agencies, downgraded the sovereign debt of Korea
with Indonesia and Thailand together to "junk" status, and sent its ﬁnancial market into
another violent plunge. The declaration of a bailout by the IMF was aimed at persuading
private investors around the world to keep investing in Korea, and that they could
continue to loan Korea the cash it needs to avoid default. However, the country's credit
rating dropped precipitously, and the "junk bond" status of the nation's credit made it far
more difficult and expensive for the country to obtain loans just at the moment it needs
them the most. The IMF involvement is not only an indicator of past economic
difficulties, but also a lead indicator of future problems, particularly given the recidivist
tendencies of some of the Fund’s clients. Here, the involvement of the Funds stands as a
proxy for declining creditworthiness, and in this case, a negative catalytic effect is the

expected and rational outcome (Bird, 1995).

 

22 Feldstein (1998) shares many of Sachs’ views on the nature of the Korean crisis, wrong adjustment
program, and concerted efforts by Japan and the United States rather than publicized bailout by the IMF
for ﬁnancial provision.

265

Overall, the Korean bailout is a recent case showing that the IMF conditionality is
not a simple technical ﬁx. Rather, it subtly adapts itself not only to a rapidly changing
context in international economy, but also to the power relationship between the two
players, which is revealed through mobilized leverage. Although some of the conditions
Korea agreed to implement were typical technical application, the negotiation was
disproportionately focused on ﬁnancial market opening that may have a considerable
domestic and international inﬂuence on the economy. Even the Fund perceives that the
programs for East Asian countries mark departure from the kind of programs the IMF has
traditionally supported in the past. This is because the centerpiece of each program is not
a set of austerity measures to restore macroeconomic balance, but rather a set of forceﬁil,
far-reaching structural reforms to strengthen ﬁnancial systems, increase transparency, open
markets, and restore market conﬁdence (Camdessus, 1998). The saliency of ﬁnancial
issues in the Korean bailout undeniably had to do with the high tide of international capital
mobility and played a key role in setting the direction of conditions and toughness.23
However, consenting or not, on that condition was the outcome of battle of leverage;

thus, this drives the nature of IMF conditionality to “political.”

 

2’ The East Asian crisis revived debates on capital account liberalization, and strengthening the
architecture of international monetary system has become a new mission for the IMF. This is easily
understood if we visit the IMF ’3 website, which is inundated with speeches and reports of reforming
international ﬁnancial system.

266

CHAPTER 7

FINALE: CONCLUDING REMARKS AND FUTURE RESEARCH

FINDINGS AND INIPLICATION S

This dissertation aimed at obtaining a better understanding of structural adjustment
by tracking down the process of conditionality agreement between the IMF and
developing countries. Departing from conventional wisdom in IMF conditionality, the
process of agreement was portrayed as a dynamic bargaining where both parties involved
rely on their visible and invisible assets, so-called leverage, in order to produce their
preferred outcomes. While the possibility of bargaining between the IMF and developing
countries as recipients of loans has been grossly ignored in the ﬁeld or even a burgeoning
curiosity of it has not been sufficiently supported by reﬁned analytical tools, this
dissertation succeeded in developing a game-theoretic model and testing it empirically at
both general and particular levels.

The most important ﬁnding from this research was that leverage is a necessary
condition that sets the course of the relationship between the IMF and developing
countries and it plays the key role in producing difference in conditionality agreements.
Corresponding to the three different theoretical results of Table 3.1, there exists variance
in conditionality agreements in a practical sense, and such variance is quite related to the
dynamics in interaction between the two actors, largely determined by leverage. Leverage

as a direct cause or intervening variable functions like a prism refracting invisible solar ray

267

ﬁnite-m

into seven different colors. The ray of conditionality invariably hits the prism of leverage,
and the speciﬁc color of a conditionality agreement that results depends on the size and
strength of the leverage. The enumerated propositions generalize such conditions across
conditionality agreements.

The statistical analysis of the conditionality bargaining model was taken as a
necessary step to build a palpable body that enables us to feel it empirically. Testing the
hypotheses, directly derived from the propositions, served as a link between the theory and
the reality. With the number of agreed conditions, the empirical analyses supported the
theory that leverage is an important aspect of the conditionality bargaining process, which
effects the difference in conditionality agreements. More speciﬁcally, ﬁnancial capacity;
structure of ﬁnancial (inter)dependence; domestic and organizational institutional
arrangements; the use, or threat, of ﬁnancial retaliation; severity of economic crisis;
domestic economic structure; and historical and international political context were all
signiﬁcant and important variables to determine the degree of toughness in conditionality.
In addition to the individual speciﬁcations of leverage for each player, the two
comprehensive speciﬁcations explored a big picture of leverage and conditionality, and
they were in culmination of presenting the explanatory power of the model, retaining the
theoretically expected coefficient signs and statistical signiﬁcance. With respect to the
parameters of the bargaining model, it appeared that the utility ingredients also play a
signiﬁcant role in determining the bargaining outcomes as much as leverage does. Thus, it
will be safe to say that it is the combination of leverage and utility parameters that better
represents the nature of a conditionality agreement. Indeed, such powerﬁrl results of the

statistical analyses not only indicate that the statistical model itself and the variables

268

 

 

specify the bargaining model properly, but also vindicate, among other things, the external
validity of the propositions. Thus, conditionality agreements are a complex product of
factors that players consider as important, and, among those factors, leverage is the
essential component in molding the product in their preferred way in terms of toughness
or softness.

The above ﬁndings have considerable academic and practical implications. First,
while the whole research design of this dissertation was developed to mainly serve a
narrow and limited topic of IMF conditionality, it is located in the broad theoretical
context of international relations. Although power has been a concept that is most widely
used in international relations, it was reduced and reﬁned to the concept of leverage in this
research in order to account for the bargaining intrinsic to IMF conditionality, which is
highly interactive and has a circumscribed agenda. The concept of leverage is certainly
similar to the traditional concept of power, but it is distinguished from power for its
relevance to and effectiveness in the speciﬁc context of bargaining. This characteristic of
leverage is deemed to be the main factor that enhances the explanatory power of
bargaining models and thereby improves their predictive ability.

Second, the results of this research provide an answer to the controversies
regarding structural adjustment, for which there has been no clear consensus in regard to
the ﬁrndamental causes of failure and their remedy. By unlocking its bargaining process,
which had been put aside as a black box thus far, it disclosed that the seed of failure might
be planted at the very beginning of the program. Conditionality agreements are produced
under the inﬂuence of various factors, both domestic and international, and they can be

plagued by political compromise more than what is necessary. Having said this, it is

269

indispensable to think ahead and consider what factors are likely to have what inﬂuence on
conditionality bargaining. This means that studies of structural adjustment, especially
those of implementation, should be reoriented so as to incorporate this prior intervening
process of leverage. That is, when the economic problems in developing countries are
double tallied with actual solutions obtained through compromise, and theoretical
solutions found without compromise, the issue of the effectiveness of Fund-supported
adjustment will come into our grips, and search for appropriate solutions to the identiﬁed
problems will start from there.

Third, if conditionality agreements are the compromised outcome ﬁ'om clash of
political rationality, it is possible that the agreed upon conditions may not properly ﬁt the
problems. While bargaining is proceeding, conditions vital to restoring balances in the
economy could be dropped for political reasons and replaced with irrelevant ones. What
is worse in this situation is that even if the parties involved in conditionality bargaining
admit that political rationality does more harm than good to solving problems in hand, it is
hardly plausible to remove the digressing effect of political rationality from conditionality
agreements. Therefore, a realistic and achievable policy alternative is a better usage of
political rationality in order to reduce overambitious and overambiguous conditions in
agreements, and to attain targeted results slowly, but surely. This does not mean to
overlook political rationality played through leverage, but to change the tactics to play it
with. As this research indicates, the nature of conditionality bargaining is not
predetennined to be zero-sum or non-zero-sum, and players’ ﬁrndamental interests
converge at circulating ﬁrnds without leaks. With this interest shared, political rationality

in conditionality bargaining should be used to have their reform proposals gradually closer

270

to shared preference under the shadow of the ﬁrture so that it can grind a way to
successﬁrl adjustment through iterated interactions.

For a wiser use of political rationality, conditionality agreements need to include a
condition of building a permanent communication channel between the IMF and recipient
countries, which is staffed by indigenous experts rather than expatriates transplanted by
the IMF. This concomitantly requires that conditionality agreement put more emphasis on
building state capacity to generate informed and indigenous analysis of economic problems
and options, which obviously affect the speed and coherence of adjustment responses. To
some extent, this capacity is determined by a very basic featuree—the number of well-
trained and experienced staff in economic agencies within recipients’ governments.
Maintaining balance in economy is as important as restoring it through adjustment
programs, which requires a high level of management skill. Of course, management is at a
human dimension, not at a favorable external environment. Thus, enhancing intellectual
capacity in recipient countries should be a part of the reform program beyond emphasizing
elementary education within social safety net. This policy alternative comes down to the
ownership of program, as the IMF likes to call. Ownership of a reform program is created,
not given, through involvement.

Fourth, statistical analysis for this research had difficulty in obtaining reliable data
on agreed conditions. Although this problem was resolved to count conditions from the
press releases contained in IMF Survey, it did not fail to turn our attention to related
practical issues in conditionality—transparency and implementation of conditionality.
Recently, the IMF has started to tackle transparency of its own. Low transparency in

Fund lending can be partly indicated by the secrecy of Letters of Intent. Disclosure of

271

LOIs does not trace back farther than late 1997. Nonetheless, the IMF alone should not
be blamed for the secrecy of LOIs, or low transparency, considering that the counterpart
of conditionality agreement shares the same interest. An easy justiﬁcation for sealing
LOIs off is to minimize negative reactions in recipient countries before introducing
adjustment programs, and to facilitate implementation. To the contrary, however, this
practice appears to have exacerbated the problem rather than helped it. Undisclosed LOIs
are more likely to nurture an environment favorable to slippage, succumbing to domestic
political rationality. Even if it does not become an international public good, as Vaubel
(1991) thinks, policy conditions are more likely to be implemented if they are made public.
It will also make it easier for the IMF to monitor the implementation of conditionality and
eventually enhance the degree of compliance. In addition, sharing information among
many different groups outside the IMF is a way to make evaluation of program design and
its effectiveness more reliable.

Furthermore, disclosing LOIs is important in that it will ﬁll in the missing link
between successﬁrl structural adjustment and politics. Nowadays, the IMF emphasizes
transparency, governance, and democracy side-by-side with structural adjustment in
recipient countries. What is overlooked in this discourse is that democracy, as a necessary
condition to putting derailed economies back on track, requires forming a broad consensus
on reform process, and without disclosing the content of LOIs, such a democratic process
will not take place. Transparency about the details of Fund-supported adjustment
programs, and dialogues among concerned groups, are integral parts of both enhancing
public support and developing informed programs in the ﬁrture. This will also be a way to

establish ownership of a reform program.

272

FUTURE RESEARCH AGENDA

This dissertation has opened the door for other related research. First, correlation
between the types of conditionality agreements and the degree of implementation should
be examined. Bargaining theory argues that a criterion of successﬁrl negotiation is the
degree of compliance by the signatories with negotiated agreements. Agreements that are
essentially one-sided cannot endure indeﬁnitely, due to the fact that the party that has been
coerced into accepting an agreement against its interest will resent that agreement and the
party who imposed it. This is also applied to conditionality agreement. Thus, the
relationship between imposition or compromise and the degree of compliance should be
examined, which is possible by classifying the existing conditionality agreements to the six
bargaining outcomes within the model developed in this dissertation. This research will be
expedited if the meaning of Agreement @, ®, @, G), @, and 6) is explored; that is,
practically what constitutes them and how different they are from one another. A method
to classify conditionality agreements in realistic terms would use the duration of
bargaining, correspondence between the causes of economic crisis in the involved
developing country and prescribed policies, usage of leverage, agreed amount of loan, and
so on. This information can be extracted from other detailed case studies. With this, an
array of studies of structural adjustment becomes complete, from conditionality agreement
to implementation and its effectiveness.

Second, so far the bargaining game is solely based on complete information.
However, the state of the world is hardly known to us with certainty, and this gives us a

reason to ﬁrrther reﬁne the model with incomplete information to reﬂect reality better. In

273

the conditionality bargaining game, a key factor that makes players’ knowledge of their
payoffs incomplete will be uncertainty about the type of players determined by their
leverage. In other words, a key factor is whether players are leveraged or not. Thus, an
expected modiﬁcation of the model to capture this feature is having two sets of the same
game trees side-by-side, and beginning the game with a chance move that determines
whether or not the recipient is leveraged. By distributing probabilities to the types of
recipient on each side, equilibrium of the game will be sought in a Perfect Bayesian
manner. This Bayesian update is expected to provide valuable information about the role
of leverage in repeated relationship between the IMF and recipients. The Bayesian update
will enable us to explain why remarkably different or consistent outcomes are produced
with particular recipient countries over time. This will check on learning effect in
conditionality, as well.

Third, as another way of examining the impact of leverage, it is possible to look at
the toughness of conditionality agreements from a perspective of their internal and
external orientation. Comparing the number of conditions that covers each sector, we
could classify conditionality agreements to their orientation and run regression with
maximum likelihood technique. The classiﬁcation of agreements could be simple binary or
ordered scale on the continuum of most internal to most external oriented, and statistical
methods will be selectively applied from probit and ordered probit in accordance with the
type of dependent variable. This research is expected to reveal two aspects of

conditionality agreement. First, the orientation of conditionality agreements reﬂects the

 

* This research idea is based on Prof. Beth Simmons’s comments on my paper presented at Midwest
Political Science Conference (April, 1999). I appreciate her inputs.

274

degree of sudden exposure of an economy under adjustment to external economic forces
or integration into global system. Killick (1993) suggests that choice between relatively
open or closed economic policies is a “big” one that would have many ramiﬁcations for
the economy’s structure. Thus, internally oriented conditionality is likely to give a
recipient country more time to concentrate national energy on rebuilding their economic
system and laying inroads to gradual opening later, minimizing external disturbances.
Considering the low adaptability and resiliency of developing economies, internally
oriented agreement is likely to be preferred to externally oriented, and again, the degree of
agreed orientation would be a litmus test of bargaining strength. Second, in terms of
frequency, the 1990’s is likely to witness more externally oriented conditionality
agreements than internally oriented due to more globalized economic activities. Thus, this
will reveal the interaction of conditionality with the changing context of international
political economy, in which the major Executive Board member countries’ interest is
sitting tight. Signiﬁcant results of this test will reconﬁrm the validity of the theory that
this dissertation advocates.

Fourth, with a view to corroborating ﬁrrther the generality of the developed
bargaining model, it should be applied to other bargaining relationship with international
ﬁnancial institutes. Since the 1980’s the World Bank has been involved in policy lending,
in addition to its traditional project lending. Debt rescheduling with syndicated banks and
the Paris Club is also a good candidate for application. The Paris Club represents oﬂicial
lenders in the industrialized countries; thus, the elements of collective political rationality
will not be diluted in forming a uniﬁed front among them toward recipient countries.

Examining bargaining with individual bilateral aid agencies in the ﬁve major donor

275

countries is worth pursuit for their extensive involvement in aid-giving and international
relations as a whole. Subsequent statistical analysis just requires adapting the dependent

variable to key issues within each bargaining relationship.

In the wake of economic crisis in Southeast Asian countries, voice for the abolition
of the IMF has become louder than ever. Even if it is true that the existence of an
organization to bail them out only exacerbates moral hazardous behavior in ﬁnancial
matters by developing countries, it is dubious as to whether the global economic shape,
especially since 1982, would have been dramatically different without the IMF. This is
easily understood when seen against the ﬁndings of this dissertation so far. The process
through which a new role of ﬁnancing development was assigned to the IMF indicates that
by and large it serves a high interest of developed countries as a collective actor in the
changing water of international monetary affairs, and that conditionality is a safeguard
against delinquency by developing countries in ﬁnancial transactions with them. The
application and implementation of conditionality, however, has been equally tampered and
hampered by political calculations from developing countries using leverage. It is obvious
that such compromise would have occurred at a bilateral level between individual
developed countries and developing countries anyway, reﬂecting a complicated nature of
international relations, if it had not been for the IMF. The only difference is that the
presence of the IMF has provided a multilateral forum where political bargaining over
conditionality takes place in a systematic and collective manner. Thus, bringing order to

international development ﬁnance by abolishing the IMF, or replacing it with something

276

 

else, is not only a wishful thinking, but also a passing vogue that can hardly be incarnated
into actions.

Having said this, what makes discussions about the IMF more productive may be
admitting the existence of the IMF here now, and trying to provide a viable mechanism for
sharing responsibility for a better management of the international monetary system, but
not such counterfactual discussions. The IMF ’s effectiveness derives from the fact that it,
as an international institution, can carry on a policy dialogue with member countries and
make policy recommendations. Fund programs should be effective in improving economic
performance and reducing the demand for its lending to ﬁnance balance-of—payments gaps
in developing countries. The effectiveness of Fund-supported programs has a strategic
connection to its organizational longevity. The long-living IMF is of interest to developing
countries as well; therefore, developing countries are equally responsible for the
effectiveness of the IMF. Their equal responsibility in the effectiveness of the IMF means
that they should capitalize on conditionality as a means to achieving long-term
development, and that they should bargain hard to lay out achievable goals for structural
adjustment, rather than create another troubled cycle of austere structural adjustment from
recklessness. All in all, sharing responsibility for the effective IMF and stable international
development ﬁnance requires both parties involved in conditionality bargaining to use their
leveraged political rationality prudently from a long term perspective, not from anything

else.

277

 

APPENDICES

APPENDIX 1

METHOD OF DATA COLLECTING:
COUNTING THE NUMBER OF AGREED CONDITIONS

COUNTING RULES

General objectives of the program are not counted.

Principal economic and ﬁnancial objectives (targets) of 8 policy areas are counted as
conditions.

Speciﬁc policy instruments of medium range under each policy area receive priority
as conditions. If this is not possible, count the number of 8 policy areas instead.

Narrow steps or timetables under each policy instrument are not counted.

5. For arrangements with annual installment and performance review such as SAF and

ESAF, conditions are counted only from the initial arrangement. Also, augmentation

and/or extension of the arrangement are excluded from counting.

o COUNTING FORMAT
Country Name
1. Type of Agreement, Arrangement Date - Expiration Date (Duration by Months)
11. Total Amount Agreed (SDR Millions), Quota Percentage
111. Total Number of Agreed Conditions
IV. Macroeconomic Components
1. Fiscal Policy
2. Monetary Policy
3. Exchange Rate Policy
V. Structural Components

4. Trade Account

5. Capital Account

278

6. Privatization of PSE

7. Labor Market/Income/W age Policy
8 Sectoral Reform/Liberalization

9 Governance

10. Data Dissemination

TWO EXAMPLES OF CONDITION COUNTING

Mexico

I. EFF 010183-123185 (36 months)
11. SDR 3,410.63 M 450 % onuota

III. 19

VI. Macroeconomic Components

1. Fiscal Policy
0) reduction of the PS deﬁcit from 16.5 % of GDP in 82 to 8.5% in 83, 5.5%
in 84, and 3.5 % in 85
® increasing productivity of PS
(33 holding down current expenditure
- eliminating subsidies to public and private sectors
- cutting back the grth of wage bill
- review of public investment projects
@ increasing public sector revenues
- adjustments in PSE pricing policies and administrative fees
(9 tax reform
- widening tax base, elimination of special exemption, adjusting tax
rates
(6) current account deﬁcit reduction to US $ 4.25 billion in 83
2. Monetary Policy
® Bank of Mexico subject to limints in line with the anti-inﬂation and BOP

aims

279

® private sector receives an appropriate share of domestic ﬁnancial savings.
(3) inﬂation reduction from 100% to 87% in 55% in 83
@ ﬂexible management of interest rate to allocate credit

3. Exchange Rate Policy
0) ﬂexible exchange rate policy

V. Structural Component

4. Trade: Not Identiﬁed

5. Capital Account
CD reducing the scope and weight of the restriction on external transactions
® lowering reliance on external ﬁnancing
C33 debt rescheduling

6. Privatization of PSE
Cl) rationalizing PS by consolidating agencies

7. Labor Market/Income/W age Policy
0) forward looking wage policy consistent with declining inﬂation and

protection of employment

8. Sectoral Reform/Liberalization:
C1) decontrol of consumer goods price

9. Governance: Not Identiﬁed

10. Information Dissemination: Not Identiﬁed

Korea

I. SBA 120497-120300 (36 months)
11. SDR 15,500 M 1939% of Quota

III. 30

V1. Macroeconomic Component
1. Fiscal Policy
C1) limiting the slowdown of GDP growth in 98 and recovering in 99
® orderly reduction in the external current account deﬁcit

C3) ﬁscal stance shifting to deﬁcit

280

6) additional ﬁscal measures covering 1-1.5% of GDP for balance or a small

surplus

- increasing mineral oil taxes

- broadening VAT base and selective increase in income and corporate
taxes

- cuts in current expenditures in infrastructure and other capital

expenditures

2. Monetary Policy

(D containing inﬂation at 5% or below

® building international reserves to more than 2 months of imports by end-

L3)

98

tight monetary stance

- reversing large liquidity injection

- raising money market rates sharply and maintained at a high level

- day-to-day conduct of monetary policy guided by movements in the

exchange rate and short-term interest rates

3. Exchange rate Policy

CD maintaining ﬂexible exchange rate regime, with limited intervention to

smoothing operations

V. Structural Component
4. Trade: Not Identiﬁed

0) time table for trade liberalization in line with WTO commitments

® streamlining import certiﬁcation procedures

5. Capital Account

CD increasing the ceiling on aggregate foreign ownership of listed Korean

®
®

('9

shares from 26% tp 50% by end-97 and to 55% by end-98

increasing the ceiling on foreign ownership from 7% to 50% by end-97
liberalizing foreigners' access to Korean money markets

ﬁrrther reduction in restriction on foreign direct investment by simplifying
approval procedures

setting a time table to eliminate restrictions on foreign borrowing by

281

corporations

6. Privatization of PSE: Not Identiﬁed

7. Labor Market/Income/W age Policy

0) enhancing the ﬂexibility of labor market by easing dismissal restrictions

under mergers and acquisitions and corporate restructuring

® strengthening employment insurance system and private job placement

operations

8. Sectoral Reform/Liberalization

G) strengthening ﬁnancial system

rapid resolution of troubled ﬁnancial institutions in compliance with
Basel Committee capital standards

mergers and acquisition by domestic or foreign institutions

deposit guarantee by 2000 and replacement with regular deposit
insurance system for small depositors

auditing ﬁnancial statements by international standards

legislation to set up an agency to consolidate supervisory ﬁrnctions and
independence of the Bank of Korea

promoting competition through allowing foreigners to establish bank

subsidiaries and brokerage houses by mid-98

9. Governance

0) measures to reduce reliance of corporations and ﬁnancial institutions on

short-term debt and allow a better diversiﬁcation of risk in the economy

10. Information Dissemination

CD ﬁnancial data disclosure

282

APPENDIX 2

STATISTICAL RESULTS OF NEGATIVE BINOMIAL REGRESSION

1. CONTROVERers ON CONDITIONALITY LENDING PRACTICE

0 Table A2. 1 Quota Proportionality and Conditionality

 

 

 

 

 

 

Number of obs = 395
LR ch12(2) = 52.08
Prob > ch12 = 0.0000
Log likelihood = -1159.2646 Pseudo R2 = 0.0220
TITNESS I Coef. Std. Err. z P>Iz|
QUOTA I .0001905 .0000334 5.697 0.000
AMNTQOTA I .0005249 .0001367 3.839 0.000
_CONS I 2.445897 .0239314 102.204 0.000
/LNALPHA I -2.934642 .180416 -16.266 0.000
ALPHA I .0531498 .0095891
Likelihood ratio test ofalpha=0: ch12(l) = 60.72 Prob >ch12 = 0.0000
0 Table A2. 2 Type of Arrangement and Conditionality
Number of obs = 395
LR ch12(3) = 32.17
Prob > chi2 = 0.0000
Log likelihood = -1 169.2183 Pseudo R2 = 0.0136
TITNESS I Coef. Std. Err. Z P>|zl
EFFD I .2762633 .0657496 4.202 0.000
SAFESAFD | .2093691 .0434371 4.820 0.000
JNTARRGT I -.0707466 .0588498 -1.202 0.229
_CONS l 2.479655 .0242032 102.452 0.000
/LNALPHA | ~2.78741 .1634643 -17.052 0.000
ALPHA I .0615805 .0100662
Likelihood ratio test ofalpha=0: ch12(l)= 81.09 Prob>chi2 =0.0000

283

2. THE IMF AND LEVERAGE

0 Table A2. 3 Donor’s Expected Utility Components and Conditionality

 

 

 

Number of obs == 316
LRch12(7) = 181.53
Prob>chi2 = 0.0000
Log likelihood=-855.9336l PseudoR2 = 0.0959
TITNESS I Coef. Std. Err. z P>|z|
DSUBPROBI .030367 .0858011 0.354 0.723
DDISCOUTI -.0828879 .0108291 —7.654 0.000
DOPPOCSTI .0001744 .0000239 7.299 0.000
DOPPOCSZ I -9.41e-09 1.85e-09 -5.074 0.000
DREPUCSTI .0003125 .0000785 3.980 0.000
DREPUCSZ I -5.15e—08 1.96e-08 —2.624 0.009
PREENGAGI .0149354 .0566655 0.264 0.792
_CONS | 2.601688 .1322197 19.677 0.000
/LNALPHAI -13.58643 241.3616 -0.056 0.955
ALPHA I 1;26ef06 _ .0003035
Likelihood ratio test ofalpha=0: ch12(l)= 0.00 Prob >chi2 = 1.0000

0 Table A2. 4 IMF Institutional Setting and Conditionality

 

 

 

Number of obs = 395
LR ch12(2) = 164.34
Prob > chi2 = 0.0000
Loglikelihood=-1103.l367 PseudoR2 = 0.0693
TITNESS I Coef. Std. Err. z P>Iz|
SMQUOTAS I 4.49e-06 3.22e-06 1.394 0.163
VOTECAST I -.1783848 .0361717 -4.932 0.000
_CONS I 9.475981 1.578376 6.004 0.000
/LNALPHA I ’—3.83732 .3276949 -11.710 0.000
_ ALPHA I .0215513 .0070622
Likelihood ratio test ofalpha=0: ch12(l) = 13.32 Prob >ch12 = 0.0003

284

0 Table A2. 5 Financial Relations and Conditionality

 

 

 

Number of Obs = 388
LRchi2(6) = 29.62
Prob > ch12 = 0.0000
Log likelihood=-1146.5659 PseudoR2 = 0.0128
TITNESS I Coef. Std. Err. z P>Iz|
DSUBPROB I .2967343 .0991421 2.993 0.003
FUNDSIZE I -.0000829 .0000528 -1.569 0.117
IMFEXPOS I .0006417 .0002719 2.360 0.018
IMFEXP02 | ~6.70e-08 4.20e-08 -1.594 0.111
DEGEXPOS I —.2155992 .0877231 -2.458 0.014
DEGEXPOZ I .0106466 .0051093 2.084 0.037
_CONS 1 2.431847 .0525519 46.275 0.000
/LNALPHA I -2.802075 .1674655 ~16.732 0.000
ALPHA I .060684 .0101625
Likelihood ratio test ofalpha=0: ch12(l) = 75.70 Prob > chi2 = 0.0000

0 Table A2. 6 Bilateral Political Relations and Conditionality

 

 

 

Number of obs = 395
LRch12(5) = 141.05
Prob>chi2 = 0.0000
Log likelihood=-lll4.778l PseudoR2 = 0.0595
TITNESS I Coef. Std. Err. z P>IzI
ADJCNCYI -.0418189 .0645883 -0.647 0.517
COLOTIE | -.0641433 .0361835 -1.773 0.076
ALLIANCE I -.0034904 .0746122 -0.047 0.963
MILITAIDI .0231609 .0412816 0.561 0.575
INTCONXTI -.413717 .0356464 -11.606 0.000
_CONS | 2.77074 .0294293 94.149 0.000
/LNALPHAI -3.577371 .2698067 -13.259 0.000
ALPHAI .0279491 .0075408
Likelihood ratio test ofalpha=0: chi2(1) = 21.33 Prob >ch12 = 0.0000

285

0 Table A2. 7 Bilateral Economic Relations and Conditionality

 

 

 

Number of obs = 317
LR ch12(4) 64.14
Prob >chi2 = 0.0000
Log likelihood=-922.57856 PseudoR2 = 0.0336
TITNESS I Coef. Std. Err. z P>|zI
ODAAMNT I .0002857 .0000528 5.411 0.000
COMINVET | .000034 8.77e-06 3.884 0.000
EXPOACCS I -.0064279 .0043993 —1.461 0.144
EXPOACCZ I .0000211 .0000487 0.434 0.664
_CONS I 2.633649 .0930376 28.307 0.000
/LNALPHAI -2.979918 .2077722 -14.342 0.000
ALPHAI .050797 .0105542
Likelihood ratio test ofalpha=0: ch12(l) = 45.00 Prob > ch12 = 0.0000

286

3. THE RECIPIENT AND LEVERAGE

0 Table A2. 8 Recipient’s Expected Utility Components and Conditionality

 

 

Number of obs = 142
LR ch12(7) = 34.86
Prob > ch12 = 0.0000
Log likelihood=-388.15214 PseudoR2 = 0.0430
TITNESS I Coef. Std. Err. z P>|z|
RSUBPROBI -.301033 .098572 -3.054 0.002
RDISCOUTI .0039904 .0027373 1.458 0.145
RPOLCOSTI .010692 .0042479 2.517 0.012
RPOLCOSZ I -.0009522 .0011443 —0.832 0.405
ROPPOCSTI .000312 .000067 4.658 0.000
ROPPOCSZ I —3.18e-08 1.01e-08 -3.142 0.002
PREENGAGI —.1132061 .1409647 -0.803 0.422
_CONS I 2.460957 .1560336 15.772 0.000
/LNALPHA 1 ‘—3.737273 .5651836 -6.612 0.000
ALPHAI .023819 .0134621

 

Likelihood ratio test of alpha=0:

ch12(l) = 4.35

287

Prob > ch12 = 0.0371

0 Table A2. 9 Economic Crisis and Conditionality

 

 

 

 

 

 

Number of obs = 76
LR chi2(4) = 1 1.88
Prob>chi2 = 0.0183
Log likelihood=~213.62777 PseudoR2 = 0.0271
TITNESS I Coef. Std. Err. z P>|zI
BOPCUADFI -.0000466 .0000145 -3.217 0.001
INFLAT | .000043 .0000521 0.826 0.409
UNEMPLOI -.0061386 .0068856 -0.892 0.373
BUDGDEFI .0078921 .0085113 0.927 0.354
_CONS | 2.57476 .078008 33.006 0.000
/LNALPHAI -3.817794 .7545288 -5.060 0.000
ALPHAI .0219762 .0165817
Likelihood ratio test of alpha=0: chi2(l)= 2.43 Prob >ch12 =0.1187
0 Table A2. 10 External Debt and Conditionality
Number of obs = 388
LR ch12(5) = 50.32
Prob > c1112 = 0.0000
Log likelihood=-ll36.2183 PseudoR2 = 0.0217
TITNESS I Coef. Std. Err. z P>Iz|
TOTDEBTI 9.51e-06 2.73e—06 3.488 0.000
TOTDEBT2 I —6.07e-11 2.79e-11 -2.174 0.030
IMFCRDTS I -.0001195 .0000687 -1.739 0.082
IMFCRDSZ I 1.92e-08 1.45e-08 1.328 0.184
INFINSIGI .7027564 .2292394 3.066 0.002
_CONS I 2.493314 .0249551 99.912 0.000
/LNALPHAI -2.966685 .1877219 -15.804 0.000
ALPHAI .0514737 .0096627
Likelihood ratio test ofalpha=0: ch12(l)= 54.27 Prob >chi2 = 0.0000

288

a Table A2. 11 Economic Characteristics and Conditionality

 

 

 

Number of obs = 215
LRchi2(4) 22.19
Prob>ch12 = 0.0002
Log likelihood=—623.43099 PseudoR2 = 0.0175
TITNESS I Coef. Std. Err. z P>Iz|
RSUBPROB 1 —.0419593 .0756342 -0.555 0.579
RESOENDWI -.1365599 .0975041 -1.401 0.161
TRADEGDP I -.001316 .0006605 —1.992 0.046
ECONSYS I .2524292 .0597833 4.222 0.000
_CONS | 2.589819 .0526914 49.151 0.000
/LNALPHAI -3.089796 .271673 -11.373 0.000
ALPHAI .0455112 .0123642
Likelihood ratio test of alpha=0: chi2(l)= 24.57 Prob >ch12 =0.0000

0 Table A2. 12 Global Economic Opportunities and Conditionality

 

 

 

Number of obs = 384
LR chi2(7) = 45.63
Prob > chi2 = 0.0000
Log likelihood=-1127.8826 PseudoR2 = 0.0198
TITNESS Coef. Std. Err. z P>|zI
FDIINFLO I .0000832 .000061 1.362 0.173
FDIINFLZ I -1.39e-08 6.67e-09 -2.088 0.037
EQIINVET | .0002939 .000143 2.055 0.040
EQIINVE2 I —5.16e-08 3.02e-08 -1.707 0.088
DEBTINFL I .0000233 .0000465 0.501 0.617
DEBTINF2 I -5.90e-09 6.82e-09 -0.866 0.387
INFINSIG l .8154012 .1993105 4.091 0.000
_CONS I 2.517888 .0203053 124.001 0.000
/LNALPHA I -2.936915 .1856367 -15.821 0.000
ALPHA I .0530291 .0098441
Likelihood ratio test of alpha=0: ch12(1)= 56.01 Prob>chi2=0.0000

289

0 Table A2. 13 Politics of Special Interest and Conditionality

 

 

 

Number of obs = 111
LR chi2(5) = l 1.59
Prob > chi2 = 0.0409
Log likelihood=-303.39154 PseudoR2 = 0.0187
TITNESS | Coef. Std. Err. z P>|z|
RPOLCOSTI .0115667 .0062371 1.854 0.064
RPOLCOSZ I —.0034072 .0014194 —2.400 0.016
IMPOGDPI -.0053107 .0025 -2.124 0.034
SOEGDPI .0093389 .0045748 2.041 0.041
URBANPOPI -.0002679 .0019719 -0.136 0.892
_CONS | 2.649875 .1511199 17.535 0.000
/LNALPHAI -3.359159 .4883161 -6.879 0.000
ALPHAI .0347645 .0169761
Likelihood ratio test ofalpha=0: chi2(l)= 6.51 Prob>ch12=0.0107

290

4. COMPREHENSIVE SPECIFICATION FOR CONDITIONALITY BARGAINING

0 Table A2. 14 Comprehensive Speciﬁcation for Leverage and Conditionality

 

 

 

Number of obs = 361
LR ch12(7) = 201.93
Prob > ch12 = 0.0000
Log likelihood=~975.37494 PseudoR2 = 0.0938
TITNESS I Coef. Std. Err. z P>Iz|
BOPCUADFI .0000144 6.25e-06 2.303 0.021
AMNTQOTAI .0010103 .0002676 3.775 0.000
TOTDEBTI 6.82e-06 1.89e-06 3.610 0.000
TOTDEBT2 I -3.50e-11 1.76e-11 -1.984 0.047
SMQUOTASI .0000137 2.13e-06 6.412 0.000
FUNDSIZE I —.0000446 .0000442 -1.009 0.313
INTCONXTI -.1752791 .0485099 -3.613 0.000
_CONS I 1.874869 .1221796 15.345 0.000
/LNALPHA I -5.594907 1.666612 -3.357 0.001
ALPHAI .0037167 .0061944
Likelihood ratio test ofalpha=0: ch12(1)= 0.38 Prob>ch12=0.5350

291

0 Table A2. 15 Comprehensive Speciﬁcation with Subjective Probability

 

 

 

Number of obs = 214
LR ch12(9) l 17.80
Prob > ch12 = 0.0000
Log likelihood=-57l.21334 PseudoR2 = 0.0935
TITNESS I Coef. Std. Err. z P>|z|
DSUBPROB | .1158501 .1085484 1.067 0.286
RSUBPROB | -.0658933 .0653782 -1.008 0.314
BOPCUADF | -8.21e-06 .0000121 -0.676 0.499
AMNTQOTA l .0011201 .0003471 3.226 0.001
TOTDEBT I 9.86e-06 2.40e-06 4.101 0.000
TOTDEBT2 I -9.83e-11 2.68e-11 -3.664 0.000
SMQUOTAS I .0000114 2.79e—06 4.082 0.000
FUNDSIZE I -.0000379 .0000581 -0.653 0.514
INTCONXT I —.18683 .0639676 -2.921 0.003
_CONS I 1.910441 .1555035 12.286 0.000
/LNALPHA | -14.88023 306.4768 —0.049 0.961
ALPHA I 3.45e-07 .0001057
Likelihood ratio test of alpha=0: chi2(l)= 0.00 Prob> ch12: 1.0000

292

APPENDIX 3

TREND IN TIGHTNESS OF CONDITIONALITY

20157

 

16523-

 

12389 e Tightness or Conditi
onality
9.254 S UCL

 

Average = 12.89

5620 LCL
1983 1985 1987 1989 1991 1993 1995 1997
1984 1986 1988 1990 1992 1994 1996

htearr

 

 

 

SgnmleveIS

Figure A3. 1 International Context and Trend in Tightness of Conditionality

293

BIBLIOGRAPHY

REFERENCES

Adams, C., J. G. M. Schulze-Ghattas, V. Arora, J. Zhou, and A. Bakastov. 1996. Korea,
Selected Issues. NF Staff Country Report 1996/136.

Adams, Charles, Donald J. Mathieson, Garry Schinasi, and Bankim Chadha. 1998.
International Capital Markets: Developments, Prospects, and Key Policy Issues
(September). Washington, DC; International Monetary Fund.

Aggarwal, Vinod K. 1987. International Debt Threat: Bargaining among Creditors and
Debtors in the I980s. Berkeley, CA: Institute of International Studies.

 

. 1996. Debt Games: Strategic Interaction in International Debt
Rescheduling. Cambridge: Cambridge Univ. Press.

Alexander, Arthur J. 1998. “Asia’s Financial Crisis: Linked to Its Economic Miracle?”
Japan Economic Institute Report, no. 19A (May 15): 1—13.

Altbach, Eric. 1998. “The Asian Crisis and the IMF: After the Deluge, the Debate.”
Japan Economic Institute Report, no. 17A (May 1): 1-13.

Anunobi, Fredoline O. 1992. The Implications of Conditionality: The International
Monetary Fund and Aﬁica. Lanham: University Press of America, Inc.

Assetto, Valerie J. 1988. The Soviet Bloc in the IMF and the IBRD. Boulder, CO:
Westview Press.

Aylward, Lynn and Rupert Thome. 1998. “Countries’ Repayment Performance vis-a-vis
the IMF.” III/IF Staﬂ Paper 45, no. 41595-617.

Baldwin, David. 1979. “Power Analysis and World Politics: New Trends versus Old
Tendencies.” World Politics 31, no. 2 (January): 161-194.

. 1980. “Interdependence and Power: A Conceptual Analysis.”
International Organization 34, no. 4 (Autumn):471-506.

 

. 1985. Economic Statecraft. Princeton, NJ: Princeton Univ. Press.

 

Barston, Roland P. 1971. “The External Relations of Small States.” in Small States in
International Relations, eds. August Schou and Arne O. Brundtland. Stockholm:
Almqvist & Wiskell.

Bartos, Otomar J. 1977. “Simple Model of Negotiation: A Sociological Point of View.”

294

Journal of Conﬂict Resolution 21 (December):565-5 79.

Bennett, D. Scott. 1997. “Testing Alternative Models of Alliance Duration, 1816-1984.”
American Journal of Political Science 41, no. 3. (July):846-878.

Bergsten, C. Fred. 1990. “From Cold War to Trade War.” International Economic
Insights (July/August).

Bhattacharya, Rina. 1997. “Pace, Sequencing and Credibility of Structural Reforms.”
World Development 25, no. 7: 1045-1061.

Bienen, Henry S. and Mark Gersovitz. 1985. “Economic Stabilization, Conditionality,
and Political Stability.” International Organization 39, no. 4 (Autumn):729-754.

Biersteker, Thomas J. 1993. “International Financial Negotiations and Adjustment
Bargaining: An Overview.” in Dealing With Debt: International Financial
Negotiations and Adjustment Bargaining, ed. Thomas Biersteker. Boulder, CO:
Westview Press.

Birch, Melissa H. 1993. “The International Monetary Fund.” in Dealing with Debt:
International Financial Negotiations and Adjustment Bargaining, ed. Thomas
Biersteker. Boulder, CO: Westview Press.

Bird, Graham. 1995. IMF Lending to Developing Countries: Issues and Evidence.
Londdon: Routledge.

Bjol, Erlin. 1971. “The Small State in International Relations.” in Small States in
International Relations, eds. August Schou and Arne O. Brundtland. Stockholm:
Almqvist & Wiskell.

Blair, David. 1993. Trade, Negotiations in the OECD: Structures, Institutions, and States.
London: Kegan Paul International.

Brady, Linda P. 1991. The Politics of Negotiation: America ’s Dealings with Allies,
Adversaries, and Friends. Chapel Hill, NC: The Univ. of North Carolina Press.

Bueno de Mesquita, Bruce and David Lalman. 1992. War and Reason: Domestic and
International Imperatives. New Haven: Yale Univ. Press.

Buira, Ariel. 1983. “IMF Financial Programs and Conditionality.” Journal of
Development Economics 12: l 1 1-136.

Bulow, Jeremy and Kenneth Rogoff. 1989. “Multilateral Negotiations for Rescheduling
Developing Country Debt.” in Analytical Issues in Debt, eds. Jacob A. Frankel,
Michael P. Dooley, and Peter Wickham. Washington, DC: International Monetary
Fund.

295

Camdessus, Michael. “The IMF and its Programs in Asia.” speech delivered at the
Council on Foreign Relations, New York. [Online] Available
http://www.imforg/extemal/np/speeches/ 1998/020698.htm, February 6, 1998.

Chan, Steve. 1993. “Relative Bargaining Power in International Debt Negotiation:
Collective Action, Sovereignty En Garde, or Mutual Partisan Adjustment?”
Conﬂict Management and Peace Science 13, no. 1:29-60.

Cheru, Fantu. 1992. “Structural Adjustment, Primary Resource Trade and Sustainable
Development in Sub-Saharan Africa.” World Development 20, no. 4: 497-512.

Choi, Byung-Sun. 1993. “Financial Policy and Big Business in Korea: The Perils of
Financial Regulation.” in The Politics of Finance in Developing Countries, eds.
Stephan Haggard, Sylvia Maxﬁeld, and Chung H. Lee. Ithaca, NY: Cornell
Univ. Press.

Chon, Shi-Yong. “Hashimoto Pledges Support for IMF Bailout at Summit with Kim.”
[Online] Available http://www.koreaherald.co.kr, November 25, 1997.

Chun, Sung-Woo. “Market Opening Pressure Likely to Increase after IMF Loan.”
[Online] Available http://www.koreaherald.co.kr, November 26, 1997.

Cline, Ray. 1975. World Power Assessment. Washington, DC: Georgetown Center for
Strategic and International Studies.

CNN. “US, Allies Haggle over World Economic Plan.” [Online] Available
http://cnn.com, October 4, 1998.

CNan “IMF bails out South Korea: Seoul to Receive a Record $55 Billion Financial
Rescue Loan and Other Aid.” [Online] Available http://cnn.com, December 3,
1997.

Cohen, Benjamin J. 1982. “Balance-of-Payments Financing: Evolution of a Regime.”
International Organization 36, no. 2 (Spring):457-478.

. 1985. “International Debt and Linkage Strategies: Some Foreign-
Policy Implications for the United States.” International Organization 39, no. 4
(Autumn):699-727.

 

Collier, Paul. 1997. “The Failure of Conditionality.” unpublished manuscript.

Consul, P. C. 1989. Generalized Poisson Distributions: Properties and Applications.
New York: Marcel Dekker, Inc.

Cooper, Richard N. 1983. “Panel Discussion.” in IMF Conditionality, ed. John
Williamson. Washington, DC: Institute for International Economics.

296

Comia, Giovanni A., Richard Jolly and Frances Stewart, eds. 1987. Adjustment with a
Human Face, Volume I Protecting the Vulnerable and Promoting Growth.
Oxford: Clarendon Press.

Cornelius, Peter. 1988. “Some Arguments against a Global Stabilization Policy by the
IMF.” Economia Intemazionale 41 :3-18.

Cross, John G. 1969. The Economics of Bargaining. New York: Basic Books.

Dell, Sidney. 1983. “Stabilization: The Political Economy of Overkill.” in IMF
Conditionality, ed. John Williamson. Washington, DC: Institute for International
Economics.

de Vries, Margaret G. 1985. “Selected Topics: Historian Traces Origins and
Development of Fund Involvement in World Debt Problem.” III/IF Survey
(January 7 and 21).

Dooley, Michael P., Donald J. Mathieson, and Liliana Rojas-Suarez. 1996. Capital
Mobility and Exchange Market Intervention in Developing Countries. IMF
Working Paper 1996/31.

Duchesne, Erick. 1997. “International Bilateral Trade and Investment Negotiations:
Theory, Formal Model, and Empirical Evidences.” Ph.D. dissertation, Michigan
State University.

Eckaus, R. S. 1986. “How the IMF lives with its conditionality.” Policy Science 19:
23 7-252.

Edwards, Sebastian. 1984. The Order of Liberalization of the External Sector in
Developing Countries. Princeton Essays in International Finance, no. 156.

Faruqee, Hamid. 1992. “Dynamic Capital Mobility in Paciﬁc Basin Developing
Countries.” IMF Staff Paper 39, no. 3:706-717.

Feldstein, Martin. 1998. “Refocusing the IMF.” Foreign Aﬂairs 77, no. 2 (March/April):
20-33.

Fernandez, Raquel and Sule Ozler. 1999. “Debt Concentration and Bargaining Power:
Large Banks, Small Banks and Secondary Market Prices.” International
Economic Review 40, no. 2 (May):333-355.

Ferguson, Tyrone. 1988. The Third World and Decision Making in the International
Monetary Fund: The Quest for Full and Eﬂective Participation. New York: Pinter
Publishers.

Finch, David. 1988. “Let the IMF be the IMF.” International Economy

297

(J anuary/F ebruary).

 

. 1989. The WF: The Record and the Prospect. Princeton Essays in
International Finance, no. 175.

Fischer, Stanley. “Capital Account Liberalization and the Role of the IMF.” speech
delivered at IMF seminar “Asia and the IMF,” Hong Kong. [Online] Available
http://www.imf.org/extemal/np/apd/asia/ﬁscher.htm, September 19, 1997.

. “The Asian Crisis: A View from the IMF.” speech delivered at the
Midwinter Conference of the Bankers' Association for Foreign Trade,
Washington, DC. [Online] Available
httpz/lwww.imf.org/extemal/np/speeches/I 998/012298.htm, January 22, 1998a.

 

. “The IMF and the Asian Crisis.” Forum Funds Lecture at UCLA,
Los Angeles. [Online] Available
http ://www.imf. org/extemal/np/speeches/ 1 998/03 2098.htm,_March 20, 1998b.

 

 

. 1998c. “The Asian Crisis and the Changing Role of the IMF.” Finance
and Development 35, no. 2 (June):2-4.

 

. 1998d. “In Defense of the IMF: Specialized Tools for a Specialized
Task.” Foreign Aﬂairs 77, no. 4 (July/August):103-106.

 

. “On the Need for an International Lender of Last Resort.” paper
delivered at the joint luncheon of the American Economic Association and the
American Finance Association, New York. [Online] Available

http ://www. imf. org/extemal/np/speeches/ 1999/0103 99.htm, January 3 , 1999.

Fisher, Roger 1987. “What is a ‘Good’ U.S.-Soviet Relationship — And How Do We
Build One?” Negotiation Journal 3, no. 4 (October):3 19-327.

Fisher, Roger and William Ury. 1991. Getting to Yes: Negotiating Agreement without
Giving in. New York: Penguin.

Friman, H. Richard. 1993. “Side-payments versus Security Cards: Domestic Bargaining
Tactics in International Economic Negotiations.” International Organization 47,
no.34 (Summer):379-410.

Gates, Scott and Brian D. Hume. 1997. Games, Information, and Politics: Applying
Game Theoretic Models to Political Science. Ann Arbor: The Univ. of Michigan

Press.

Goertz, Gary. 1994. Context of International Politics. Cambridge: Cambridge Univ.
Press.

298

Gold, Joseph. 1979. Conditionality. Pamphlet Series, no. 31. Washington, DC:
International Monetary Fund.

Greene, William H. 1993. Econometric Analysis, 2nd ed. Englewood Cliffs, NJ: Prentice
Hall.

Grieco, Joseph M. 1988. “Anarchy and the Limits of Cooperation: A Realist Critique of
the Newest Liberal Institutionalism.” International Organization 42, no. 3
(Summer):485-507.

. 1993. “Understanding the Problem of International Cooperation: The
Limits of Neoliberal Institutionalism and the Future of Realist Theory.” in
Neorealism and Neoliberalism, ed. David Baldwin. New York: Columbia Univ.

Press.

 

Grifﬁth-Jones, Stephany. 1992. “Cross-conditionality or the Spread of Obligatory
Adjustment: A Review of the Issues and Questions for Research.” in Cross-
C onditionality Banking Regulation and Third World Debt, eds. Ennio
Rodriguez and Stephany Grifﬁth-Jones. London: MacMillan Press.

Guitian, Manuel. 1981. Fund C onditionality: Evolution of Principles and Practices.
Pamphlet Series, no. 38. Washington, DC: International Monetary Fund.

. 1987. “Adjustment and Economic Growth: Their Fundamental
Complementarity.” in Growth-Oriented Adjustment Programs, eds. Vittorio
Corbo, Morris Goldstein, and Mohsin Khan. Washington, DC: International
Monetary Fund and World Bank.

 

Gujarati, Damodar N. 1995. Basic Econometrics, 3rd ed. New York: McGraw Hill, Inc.

Habeeb, William M. 1988. Power and Tactics in International Negotiation: How Weak
Nations Bargain with Strong Nations. Baltimore: The Johns Hopkins Univ. Press.

Haggard, Stephan. 1985. “The Politics of Adjustment: Lessons from the IMF’s Extended
Fund Facility.” International Organization 39, no. 3 (Summer):505-534.

Haggard, Stephan and Chung H. Lee. 1993. “The Political Dimension of Finance in
Economic Development.” in The Politics of Finance in Developing Countries,
eds. Stephan Haggard, Sylvia Maxﬁeld, and Chung H. Lee. Ithaca, NY: Cornell
Univ. Press.

Haggard, Stephan and Robert Kaufman. 1989. “The Politics of Stabilization and
Structural Adjustment.” in Developing Country Debt and Economic Performance:
The International Financial System, ed. Jeffrey D. Sachs. Chicago: Univ. of
Chicago Press.

299

. 1995. The Political Economy of Democratic
Transitions. Princeton, NJ: Princeton Univ. Press.

 

Haggard, Stephan, Robert Kauﬁnan, and Steven Webb. 1992. Democracy, Dictatorship,
and Inﬂation in Middle-Income Countries. Harvard Univ., Rutgers Univ. and
World Bank, Washington, DC. Processed.

Haggard, Stephan and Steven Webb. 1994. Voting for Reform: Democracy,
Political Liberalization, and Economic Adjustment. Washington, DC: World
Bank.

Haggard, Stephan and Sylvia Maxﬁeld. 1993. “Political Explanation of Financial Policy
in Developing Countries.” in The Politics of Finance in Developing Countries,
eds. Stephan Haggard, Sylvia Maxﬁeld, and Chung H. Lee. Ithaca, NY: Cornell
Univ. Press.

. 1996. “The Political Economy of Financial
Internationalization in the Developing World.” International Organization 50,
no. 1 (Winter):35-68.

 

Hanson, James A. 1994. “An Open Capital Account: A Brief Survey of the Issues and the
Results.” in Financial Reform: Theory and Experience, eds. G. Caprio, Jr.,
1. Atiyas, and J. Hanson. Cambridge: Cambridge Univ. Press.

Hardy, Chandra. 1987. “Debt Negotiations and the North-South Dialogue, 1974-1980.”
in Positive Sum: Improving North-South Negotiations, ed. 1. William Zartman.
New Brunswick, NJ: Transaction Books.

Helleiner, Gerald K. 1987. “Stabilization, Adjustment, and the Poor.” World
Development 15, no. 12:1499-1513.

Hopmann, P. Terrence. 1978. “Asymmetrical Bargaining in the Conference on Security
and Cooperation in Europe.” International Organization 32, no. 1 (Winter):
141-177.

. 1995. “Two Paradigms of Negotiation: Bargaining and Problem Solving.”
The Annals of the American Academy of Political and Social Science 542
(November):24-47.

 

. 1996. The Negotiation Process and the Resolution of International
Conﬂicts. Columbia, SC: Univ. of South Carolina Press.

 

Iklé, Fred C. 1964. How Nations Negotiate. New York: Harper & Row.

ILO Yearbook of Labor Statistics. 1987-1994. Geneva: International Labor Organization.

300

IMF Articles of Agreement of the International Monetary Fund. [Online] Available
http://www.imf. org/extemal/pubs/ft/aa/index.htm.

---- Annual Reports. 1983-1998. Washington, DC: International Monetary Fund.

---- Direction of Trade Statistics Yearbook. 1982-1998. Washington, DC: International
Monetary Fund.

---- International Financial Statistics. 1983-1998. Washington, DC: International
Monetary Fund.

---- Survey. 1983-1998. Washington, DC: International Monetary Fund.

----- . 1987. Theoretical Aspects of the Design of F und-Supported Adjustment Programs.
IMF Occasional Paper, no. 55.

-----. 1995, 1998a. Oﬂicial Financing for Developing Countries. Washington, DC:
International Monetary Fund.

----- . 1997. The ESAF at Ten Years: Economic Adjustment and Reform in Law-Income
Countries. IMF Occasional Paper, no. 156.

----- . 1997-1998. Member Country Publication — by Country. [Online] Available
http://www.imf. org/extemal/np/loi/mempub.htm.

----- . 1998b. Toward a Framework for Financial Stability. Washington, DC: International
Monetary Fund.

----- . 1998c. World Economic Outlook (May and October). Washington, DC:
International Monetary Fund.

----- . 1998d. “Financial Facilities and Policies: IMF Financing Helps Members Pursue
Sound Policies.” IMF Survey Supplement on the Fund (September).

IMF Economic Forum. “Capital Account Liberalization: What’s the Best Stance.” IMF,
Washington, DC. [Online] Available
http://www.imforg/external/np/tr/ 1998/T R981002A.htm, October 2, 1998.

Jain, Arvind and Douglas Nigh. 1989. “Politics and the International Lending Decision of
Banks.” Journal of International Business Studies 20, no. 2 (Summer):349-359.

Johnston, R. Barry, Salim M. Darbar, and Claudia Echeverria. 1997. Sequencing Capital
Account Liberalization: Lessons from the Experiences in Chile, Indonesia, Korea,
and Thailand. IMF Working Paper 1997/157.

JOnsson, Christer. 1981. “Bargaining Power: Notes on an Elusive Concept.” Cooperation

301

and Conﬂict XVI:249-257.

Kahler, Miles. 1989. “International Financial Institutions and the Politics of Adjustment.”
in Fragile Coalitions: The Politics of Economic Adjustment, ed. Joan M. Nelson.

New Brunswick: Transaction Books.

 

. 1990. “Orthodoxy and Its Alternative: Explaining Approaches to
Stabilization and Adjustment.” in Economic Crisis and Policy Choice: The
Politics of Adjustment in the Third World, ed. Joan M. Nelson. Princeton, NJ:
Princeton Univ. Press.

Kang, Seok-Jae. “IMF to Provide First Portion of Bailout Funds for Seoul This Week;
Two Sides Fail to Narrow Gaps in Grth Range, Bank Liquidation.” [Online]
Available http://www.koreaherald.co.kr, December 1, 1997.

Kapur, Devesh. 1998. “The IMF: A Cure or a Curse?” Foreign Policy 111 (Summer):
114-126.

Keohane, Robert O. 1983. “The Demand for International Regime.” in International
Regimes, ed. Stephen D. Krasner. Ithaca, New York: Cornell University Press.

Keohane, Robert O. 1984. After Hegemony: Cooperation and Discord in the World
Political Economy. Princeton, NJ: Princeton Univ. Press.

Keohane, Robert O. and Joseph S. Nye. 1989. Power and Interdependence, 2nd ed.
Harper Collins Publishers.

Khan, Moshin. 1990. “The Macroeconomic Effects of Fund-supported Adjustment
Programs.” IMF Staﬂ Papers 37, no. 2 (June): 195-231.

Killick, Tony. 1984. The IMF and Stabilization: Developing Country Experiences.
London: Gower.

. 1990. Problems and Limitations of Adjustment Policies. Working Paper 36.
London: Overseas Development Institute.

 

. 1993. The Adaptive Economy: Adjustment Policies in Small, Low-Income
Countries. EDI Development Studies. Washington, DC: World Bank.

 

. 1995. IMF Programmes in Developing Countries: Design and Impact.
London: Routledge.

 

King, Gary. 1989a. “Event Count Models for International Relations: Generalizations
and Applications.” International Studies Quarterly 33 223-147.

302

------------- . 1989b. Unijying Political Methodology: The Likelihood Theory of Statistical
Inference. New York: Cambridge Univ. Press.

Knight, Malcolm. 1998. “Developing Countries and the Globalization of Financial
Markets.” World Development 26, no. 721185-1200.

Knorr, Klaus. 1975. The Power of Nations. New York: Basic Books.

Kristof, Nicholas D. 1997. “3 Seoul Candidates, Fearing Default, Bow to IMF. Plan.”
[Online] Available http://www.nytimes.com, December 14, 1997.

Krueger, Anne. 1984. “Problems of Liberalization.” in World Economic Growth, ed.
A. Harberger. San Francisco: ICS Press.

Lawler, Edward. 1992. “Power Processes in Bargaining.” The Sociological Quarterly 33,
no. 1:17-34.

Leech, Dennis. 1998. Power Relations in the International Monetary Fund: A Study of
the Political Economy of A Priori Voting Power Using the Theory of Simple
Games. University of Warwick CSGR Working Paper 06/98.

Lehman, Howard P. 1993. Indebted Development: Strategic Bargaining and Economic
Adjustment in the Third World. New York: St. Martin’s Press.

Lensink, Robert. 1996. Structural Adjustment in Sub-Saharan Africa. London: Longman.

Lister, Frederick K. 1984. Decision Making Strategies for International Organizations:
The IMF Model. Monograph Series in World Affairs. Graduate School of
International Studies, Univ. of Denver.

Lockhart, Charles. 1979. Bargaining in International Conﬂicts. New York: Columbia
Univ. Press.

McCarthy, Thomas. “Molecular Nanotechnology and the World System.” [Online]
Available http://mccarthy.cx, October 12, 1996.

McKeown, Timothy, Leona Pallansch, and Strom Thacker. 1999. “Political
Conditionality in US. Bilateral and Multilateral Foreign Assistance.” paper
presented at the 40th Annual Convention of the International Studies Association,
Washington, DC, February 16-20.

McLaughlin, Sara, Scott Gates, and Havard Hegre. 1998. “Timing the Changes in
Political Structures: A New Polity Databases.” The Journal of Conﬂict Resolution
42, no. 2 (April):231-242.

Mahler, Walter. 1990. The Taxation of the Windfall ﬁom Internationalization of the

303

Korean Capital Market. IMF Working Paper 1990/54.

Martin, Matthew. 1991. The Crumbling Fapade of Aﬁ'ican Debt Negotiations. New
York: MacMillan.

. 1994. “Negotiating External Finance for Adjustment.” in Negotiating
Structural Adjustment in Africa, ed. Willem van de Geest. New York: United
Nations Publications.

 

Mayer, Frederick. 1992. “Managing Domestic Differences in International Negotiations:
the Strategic Use of lntemal Side-payments.” International Organization 46,
no. 4 (Autumn):793-818.

Morgan, T. Clifton. 1990. “Issue Linkages in International Crisis Bargaining.” American
Journal of Political Science 34, no. 2 (May):311-333.

Morgenthau, Hans. 1973. Politics among Nations, 5th ed. New York: Alfred A. Knopf.

Morrow, James D. 1994. Game Theory for Political Scientists. Princeton, NJ: Princeton
Univ. Press.

Mosley, Paul. 1987. Conditionality as Bargaining Process: Structural Adjustment
Lending I 980-1986. Princeton Essays in International Finance, no. 168.

. 1988. “On Persuading a Leopard to Change His Spots: Optimal Strategies
for Donors and Recipients of Conditional Development Aid.” in Toward a
Political Economy of Development: A Rational Choice Perspective, ed. Roberts
Bates. Berkeley, CA: Univ. of California Press.

 

Mosley, Paul and Jane Harrigan. 1989. Evaluation of the Impact of ‘Program Finance’
for Development. Institute of Development Policy and Management, Univ. of
Manchester.

Mosley, Paul, Jane Harrigan, and John Toye. 1995. Aid and Power: The World Bank and
Policy- based Lending, Volume I Analysis and Policy Proposals, 2nd ed. London:
Routledge.

Most, Benjamin A. and Harvey Starr. 1989. Inquiry, Logic and International Politics.
Columbia, SC: Univ. of South Carolina Press.

Nachmias, Nitza. 1988. Transfer of Arms, Leverage, and Peace in the Middle East. New
York: Greenwood Press.

Nagel, Jack H. 1975. The Descriptive Analysis of Power. New Haven, CT: Yale Univ.
Press.

304

Nelson, Joan M. 1984. “The Politics of Stabilization.” in Adjustment Crisis in the Third
World, eds. R. E. Feinberg and V. Kallab. Washington, DC: Overseas
Development Council.

. 1988. “The Political Economy of Stabilization: Commitment, Capacity
and Public Response.” in Toward a Political Economy of Development: A
Rational Choice Perspective, ed. Robert Bates. Berkeley, CA: Univ. of California
Press.

 

. 1990. “Introduction: The Politics of Economic Adjustment in
Developing Nations.” in Economic Crisis and Policy Choice: The Politics of
Adjustment in the Third World, ed. Joan M. Nelson. Princeton, NJ: Princeton
Univ. Press.

 

 

. 1996. “Promoting Policy Reforms: The Twilight of Conditionality?”
World Development 24, no. 9:1551-1559.

Norris, Floyd. “Korean Crisis: Blame the Lenders.” [Online] Available
http://www.nytimes.com, December 14, 1997.

Nowzad, B. 1989. “The Debt Problem and the IMF’s Perspective.” in Third World Debt:
The Search for a Solution, ed. Graham Bird. London: Edward Elga.

O'Brien, Timothy L. “Sovereign Debt of Korea, Indonesia, Thailand Rated as 'Junk'.”
[Online] Available http://www.nytimes.com, December 23, 1997.

Odell, John. 1980. “Latin American Trade Negotiations with the United States.”
International Organization 34, no. 2 (Spring):207-228.

-------------- . 1985. “The Outcome of International Trade Conﬂicts: The US. and South
Korea, 1961-1981.” International Studies Quarterly 29 (September):263-286.

OECD. 1996. International Capital Market Statistics: [950-1995. Paris: Organization for
Economic Cooperation and Development.

Organski, APR and Jacek Kugler. 1980. The War Ledger. Chicago, The Univ. of
Chicago Press.

Oxfam Policy Department. 1995. A Case for Reform: Fiﬁy Years of the IMF and World
Bank. Oxford: Oxfam Publication.

Passell, Peter. “Economic Scene: South Korea Facing Difficult Economic Choices.”
[Online] Available http://www.nytimes.com, December 18, 1997.

Polak, Jacques J. 1991. The Changing Nature of IMF ’s Conditionality. OECD
Development Centre Technical Papers, no. 41.

305

Pollack, Andrew. “South Korea Suspends Operations of 9 Banks in Preparation for
I.M.F. Bailout.” [Online] Available http://www.nytimes.com, December 3, 1997.

Powell, Robert. 1991. “Absolute and Relative Gains in International Relations Theory.”
American Political Science Review 85, no. 4 (December): 1303-1320.

Pruitt, Dean G. 1981. Negotiation Behavior. New York: Academic Press.

Putnam, Robert. 1988. “Diplomacy and Domestic Politics: The Logic of Two-level
Games.” International Organization 42, no. 3 (Summer):427-460.

Rapopoort, Anatol. 1960. Fights, Games, and Debates. Ann Arbor, MI: Univ. of
Michigan Press.

 

. 1969. T wo-person Game Theory. Ann Arbor, MI: Univ. of Michigan
Press.

Rodrik, Daniel. 1990. “How Structural Adjustment Programs be Designed?” World
Development 18, no.7:933-947.

Rubinstein, Ariel. 1982. “Perfect Equilibrium in a Bargaining Model.” Econometrica 50:
97-107.

Sachs, Jeffrey D. 1987. “Trade and Exchange Rate Policies in Growth-Oriented
Adjustment Programs.” in Growth-Oriented Adjustment programs, eds. Vittorio
Corbo, Morris Goldstein, and Mohsin Kahn. Washington, DC: International
Monetary Fund and World Bank.

 

. 1989. “Conditionality, Debt Relief, and the Developing Country Debt
Crisis.” in Developing Country Debt and Economic Performance: The
International Financial System, ed. Jeffrey D. Sachs. Chicago: Univ. of Chicago
Press.

Sanger, David E. “Clinton's Personal Touch in Asia Rescue.” [Online] Available
http://www.nytimes.com, December 18, 1997.

Schadler, Susan, et al. 1995. WF Conditionality: Experience under Stand-By and

Extended Arrangements, Part I : Key Issues and Findings. IMF Occasional Paper,
no.128.

Schelling, Thomas. 1960. The Strategy of Conﬂict. Cambridge: Harvard Univ. Press.

Sidell, Scott R. 1988. The IMF and Third-World Political Instability: Is There a
Connection? New York: St. Martin Press.

SIPRI Yearbook of Arms Control andDisarmament. 1982-1998. Stockholm, Sweden:

306

Stockholm International Peach Research Institute.

Smith, Jr., Fred L. 1984. “The politics of IMF lending.” Cato Journal 4, no. 1
(Spring/Summer):21 1-241.

Snidal, Duncan. 1991. “Relative Gains and the Pattern of International Cooperation.”
American Political Science Review 85, no. 3 (September):701-726.

Stithl, Ingolf. 1994. “The Rubinstein and Stiihl Bargaining Models: A Comparison and an
Attempt at a Synthesis.” EFI Research Paper 6535. Stockholm School of
Economics.

Stein, Howard. 1992. “Deindustrialization, Adjustment, the World Bank and the IMF in
Africa.” World Development 20, no. 1:83-95.

Stevenson, Richard W. and David E. Sanger. “Asian Crisis Could Wreak Havoc on
Balance of Trade.” [Online] Available http://www.nytimes.com, December 20,
1997.

Stewart, Frances. 1987. “Alternative Macro Policies, Meso Policies, and Vulnerable
Group.” in Adjustment with a Human Face, Volume I Protecting the Vulnerable
and Promoting Growth, eds. Giovanni A. Comia, Richard Jolly, and Frances
Stewart. Oxford: Clarendon Press.

. 1995. Adjustment and Poverty: Options and Choices. London:
Routledge.

 

Stiles, Kendall W. 1990. “IMF Conditionality: Coercion or compromise?” World
Development 18, no. 7:959-974.

. 1991. Negotiating Debt: The IMF Lending Process. Boulder, CO:
Westview Press.

 

The Economist. 1997. “New Illness, Same Old Medicine.” (Dec. 13):65-66.

Taylor, Lance. 1987. Varieties of Stabilization Experience: Toward Sensible
Macroeconomics in the Third World. New York: Oxford Univ. Press.

Tollison, Robert D. and Thomas D. Willett. 1979. “An Economic Theory ofMutually
Advantageous Issue Linkages in International Negotiations.” International
Organization 33, no. 4 (Autumn):425-449.

Treasurer’s Department. 1998. Financial Organization and Operations of the NF.
Pamphlet Series, no. 45. Washington, DC: International Monetary Fund.

UN Statistical Yearbook. 1993, 1997. New York: United Nations.

307

----. 1998. World Economic and Social Survey: Trends and Policies in the World
Economy. New York: United Nations.

van de Geest, Willem. 1994. “The Bargaining Process of Policy-Based Lending.” in
Negotiating Structural Adjustment in Africa, ed. Willem van de Geest. New
York: United Nations Publications.

Va’tsquez, Ian. 1998. “The IMF’s Dubious Purpose: Long on Bailouts, Short on Success.”
The Freeman 48, no. 10 (October):600-601.

Vaubel, Roland. 1991. “The Political Economy of the International Monetary Fund: A
Public Choice Analysis.” in The Political Economy of International
Organizations, eds. R. Vaubel and T. D. Willett. Boulder, CO: Westview Press.

Vital, David. 1971. “The Analysis of Small Power Politics.” in Small States in
International Relations, eds. August Schou and Arne O. Brundtlan. Stockholm:
Almqvist & Wiskell.

Wagner, R. Harrison. 1988. “Economic Interdependence, Bargaining Power, and
Political Inﬂuence.” International Organization 42, no. 3 (Summer):
461-483.

Walton, John and Charles Ragin. 1990. “Global and National Sources of Political
Protest: Third World Responses to the Debt Crisis.” American Sociological
Review 55, no. 6:876-890.

Walton, Richard E. and Robert B. McKersie. 1991. A Behavioral Theory of Labor
Negotiations: An Analysis of a Social Interaction System, 2nd ed. Ithaca, NY:
ILR Press.

Waltz, Kenneth N. 1979. The Theory of International Politics. Reading, MA: Addison-
Wesley.

Wells, Robin. 1993. “Tolerance of Arrearages: How IMF Loan Policy Can Effect Debt
Reduction.” The American Economic Review 83, no. 3:621-633.

Werner, Suzanne. 1998. “Negotiating the Terms of Settlement: War Aims and
Bargaining Leverage.” Journal of Conﬂict Resolution 42, no. 3 (June):321-343.

White, Halbert. 1980. “A Heteroscedasticity-Consistent Covariance Matrix Estimator
and a Direct Test for Heteroscedasticity.” Econometrica 48, no. 42817-83 8.

Wiarda, Howard J. 1990. “The Politics of Third World Debt.” PS: Political Science &
Politics (September):41 1-4 1 8.

Williamson, John. 1983a. “On Judging the Success of IMF Policy Advice.” in IMF

308

Conditionality, ed. John Williamson. Washington, DC: Institute for International
Economics.

 

. 1983b. “The Lending Policies of the International Monetary Fund.”
in IMF Conditionality, ed. John Williamson. Washington, DC: Institute for
International Economics.

World Bank. 1988. Adjustment Lending: An Evaluation of Ten Years of Experience.
Washington, DC: World Bank.

----_-----——--, 1990, Adjustment Lending Policies for Sustainable Growth. Washington,
DC: World Bank.

------------- . 1992. Adjustment Lending and Mobilization of Private and Public Resources
for Growth. Washington, DC: World Bank.

........... .-, 1993a. Adjustment Lending: Lessons of Experience. Washington, DC: World
Bank.

----------- -. 1993b. Getting Results: The World Bank ’3 Agenda for Improving
Development Eﬂectiveness. Washington, DC: World Bank.

............. , 1994a. Trends in Developing Economies. Washington, DC: World Bank.
_--———-------, 1994b. World Development Reports. Washington, DC: World Bank.

-------------. 1994c. World Debt Table: External Finance for Developing Countries I 994-
95. Washington, DC: World Bank.

.............. , 1997a. Global Development Finance CD-ROM. Washington, DC: World
Bank.

-—————-—------, 1997b, World Development Indicators CD-ROM. Washington, DC3 World
Bank.

WuDunn, Shery with Nicholas D. Kristof. “Japan Seems Paralyzed by Asian Crisis.”
[Online] Available http://www.nytimes.com, December 17, 1997.

Yoo, Cheong—Mo. “Seoul Strongly Denies Rumors on IMF Bailout Fund.” [Online]
Available http://www.koreaherald.co.kr, November 17, 1997a.

. “Seoul to Seek IMF Loans; Formal Announcement Due in 2 to 3
Days.” [Online] Available http://www.koreaherald.co.kr, November 27, 1997b.

 

. “IMF Pressures Seoul to Revise Macroeconomic Targets.” [Online]
Available http://www.koreaherald.co.kr, November 28, 1997c.

 

309

 

. “IMF Talks Stalled on Key Issues; Seoul Sticks to Own Growth,
Restructuring Policies.” [Online] Available http://www.koreaherald.co.kr,
December 2, l997d.

 

. “Public Anger Mounting over Kim Government Economic Policy
Failure.” [Online] Available http://www.koreaherald.co.kr, December 6, l997e.

Young, Oran R. 1968. The Politics of F orce: Bargaining During International Crises.
Princeton, NJ: Princeton Univ. Press.

Zartman, I. William. 1971. The Politics of Trade Negotiation between Aﬁica and the
European Economic Community. Princeton, NJ: Princeton Univ. Press.

 

, ed. 1976. The 5 0% Solution. Garden City, NY: Doubleday Anchor.

 

. 1977. “Negotiation as a Joint Decision-Making Process.” in The
Negotiation Process, ed. 1. William Zartman. Beverly Hills, CA: Sage.

 

. 1987a. “Introduction: Explaining North-South Negotiations.” in
Positive Sum, Improving North South Neotiations, ed. 1. William Zartman. New
York: Transaction Books.

 

. 1987b. “Conclusion: Importance of North-South Negotiations.” in
Positive Sum, Improving North South Neotiations, ed. I. William Zartman. New
York: Transaction Books.

Zartman, I. William and Maureen Berman. 1982. The Practical Negotiator. New Haven:
Yale Univ. Press.

310

 

I‘IICHIGQN
I: ,n, n

I I III II
3 l 29