‘ :.-s.n-EE¥EE1WI1:IL.‘.

E-' L‘ :iE I; jg! Ez. .{E‘lh ”WES;- | ”NJ
.E EEEEEE Q-‘W‘EEEL .3] MY

W. ‘ In '0 ‘ EU

u ,n ‘i‘. '1": ‘

E31

C-

u
LE;

' l
W, “nan-.. _
M...
“3
V1.4 _
—. _

"a.

w

a.
“m-
k*".‘"
n
.—
mar -
u .. .
-M a
' .flu '
. .-.,.-. -
y.’
M
{II A.
. .

r-

 

‘ .
,.
,.
. . a
‘ . A
-W

n“ . . h f.
. , ' . v ’v
' .1. 1%. - . n): :4" ‘

E ‘71:: de- 53'1va Km"
' {LEE ‘ “We“ 5:3;
$1.?" IEAIE ~5§£31§ 1‘1 E:
mam. E: mu
‘ ~ ‘ ‘ - A j .
l *-“"“I"“‘ WE“, 3153.3 AW
1' gssgtdz.‘:!?-it*‘ez E ‘E z" E
E1. $1.543?! 2-31.51 ;E. ‘:::' . :4 z E
214:3t1fi11'1‘E‘5W“ . .. E‘ W WW %
31.:‘131-33, «111 xiii : ‘I‘EIE'Wgh‘W'b 1. Hi. ‘inl ‘.
”QR-(49.341 WEE“ '-§"‘a{u«3~ :"lE “1‘ 2:43. is“;
. 3E 'né‘vh‘WEEE Ej‘hE ' E,’ ': ‘ TEN In: EEIE a 9'1me
" ‘ ' :E 411.131? 55mg"?- LE“ “WWW
' 1 ".13- ‘s‘.;gEE:;=.ir';lF 1‘53 W

r~

.-
O.
" .A. .
7.. .. a
«E- °

35"

A-v.—r-

. ¢
»
.,__.-... ~. .
z‘dt'. v-

p.

Y

E. E
E .Ntf'E ,
, V

M‘
J. ‘
"‘ ‘cfx‘afl’
v».-
a,“

2‘“ ”r...

 

.-,A 1:1: 9
QIEEEZJE
1E 31}

." ‘5';
H“

‘1‘,

iElégi" "v '1“: .

I E‘ r E:-. l .

 

“v.9 (5:35"! .2“ 9'
. :' E
1 . z_v

 

‘EE MEEWEHH

. .. ‘EE‘E.
_ E E.) _
'EW

 
 

          

     
 

     

 

 

.E E _ .. .33 3E {hf .. ‘ ' . “MEN
". 5115'; 1: JE‘: I ‘l :19. ”El .31”; .'| .t; ‘EE' ‘ D a” 3‘.“ I F” : W!‘ Y'EEIL l'.‘ .
. ’1!le M' W4 3' .. ‘E' E E '.E‘.t..'i?-,E'EE;- ., {WEI ‘E gE-E- E 1; '.~.EE
‘ J E’E‘j.‘ "(-31" 534. l‘ 'J' ‘ "I. $1 I‘~“.“ (Iru E‘ I 4 'L-EI- :fl'E- IE; in: ‘I I E, "V.- . E" . l:_L, ‘SWE'Wf‘ ‘“ fill“
“it“ . EEEEEE “Ewing‘s“ m. ‘1" 1‘ .EEJEEEEE «E ‘EE:E‘! EELLEE E “1“" le EE'h E42: 1” BREW" EEEE: “EH 5?: EEEE: it ~ E‘EEEEEEE EEEE'EE A“ E:
. ’:1!l1.:"‘3:!1~fl‘1-E7E1f21 E: EEEEEEEE‘ “213;" . ..... {1“ EE'EE‘E ‘ETEEE' E SN“: .“'1EE|1'3"“ EEEEE IEIEEE‘EEE :IEE‘E: l ttv'Ezgfl -PET'I!E“I'I}"'L!¢ ‘Elfiklhly‘j‘Lfl! “kw E ‘EM l
- ~15Ehfi~hhinhm QM. {EMI‘V Hf m' h: ““1"“:in (““21'13kw. . .n . Ii'fi'vhm‘i‘mmlm'ilfil‘imf. E11 * .13L_____‘. WEI”

 

 

 

 

This is to certify that the

dissertation entitled

AN EMPIRICAL ANALYSIS OF THE EFFECT OF
POLITICAL RISK ON THE MODE OF OPERATIONS

IN FOREIGN MARKETS
presented by

SAMUEL CHIGOZIRI OKOROAFO

has been accepted towards fulfillment
of the requirements for

PhD nBUSINESS ADMINISTRATION

degree i

 

 

 

Date 7/? //K4

MS U i: an Affirmative Action/Equal Opportunity Institution 0-12771

 

 

 

IVIESI_J BEIURNING MATERIALS:
Place in book drop to
LJBRARJES remove this checkout from
Ala-23.3... your record. FINES will
——— a be charged if book is
returned after the date
stamped below.

 

 

 

 

 

 

1 - l.
-\ A a is“.
\J - ‘—

 

 

 

 

 

 

AN EMPIRICAL ANALYSIS OF THE EFFECT OF
POLITICAL RISK ON THE MODE OF

OPERATIONS IN FOREIGN MARKETS

BY

Samuel Chigoziri Okoroafo

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Marketing and Transportation

1986

Copyright by
Samuel Chigoziri Okoroafo
1986.

ABSTRACT
AN EMPIRICAL ANALYSIS OF THE EFFECT OF

POLITICAL RISK ON THE MODE OF OPERATIONS
IN FOREIGN MARKETS

BY

Samuel Chigoziri Okoroafo

The objective of this research was to examine the role
of political risk (compared to market size and market growth
rate) and its impact on the mode of operations decisions of
U.S. manufacturing firms. Mode of operations, such as
exporting, licensing, joint ventures and investing, was
viewed as a behavioral pattern of multinational firms. The
study was conceptualized within the framework of two schools
of thought. The incremental school sees modes of operations
as a passive, gradual and evolving process, while the
strategic school views them as dynamic alternatives to
access foreign markets.

Cross-sectional and time series data was pooled in
order to test hypothesized effects of political risk on the
substitution of exporting for investing behavior of U.S.
manufacturing firms. Personal interviews were also conducted
with executives of Michigan firms with foreign operations
from five industries: chemicals, computers, aircraft,
agricultural, motor vehicles and transportation equipment
industries.

The analysis suggests that the role of political risk

is different for developed and developing country-markets.

In developed country-markets, political risk plays a minimal
role in U.S. firms decisions to export or invest. For
developing' country-markets Ihowevery political risk: was. a
major factor in comparison to market size and market growth
rate in the investment decision only.

On the substitution of exporting for investing behavior
in high risk markets, this research showed that in
developing country-markets, political risk does not cause
U.S. firms to substitute exporting for investing, rather,
licensing and joint ventures were preferred. In developed
country-markets, market size was more closely associated
with exporting, not investment. This implies a preference by
U.S. manufacturing firms for exporting in developed country-
markets with relatively large market size.

It is suggested that mode of operations is a behavior
pattern and that the choice of either depends on some key
variables. Firms do not have to initially export, then
license, and finally invest to gain a foothold in a foreign
market.

Future research therefore should concentrate on factors
that determine the mode of entry and their relative
importance and not simply assume an evolution from exporting
to investing. Also, attention should be paid to key factors

that cause firms to switch modes to accomplish their goals.

To my parents, Abel Chikwendu & Gladys
Uchem for their love, encouragement, and
support. You have taught me a lot. You
are my idols.

ii

ACKNOWLEDGEMENTS

It is difficult to complete a PhD jprogram and
dissertation without placing reliance upon a number of
professors, friends, colleagues, and family. First of all, I
thank God with a deep sense of humility and gratitude for
giving me good health , strength and perseverance.

To 'my' Dissertation Committee :members, Professors
Gilbert Harrell, Donald Taylor and especially John L.
Hazard, my chairman, I am greatly indebted for improving the
ideas of this research and their presentation.

To the faculty in the Department of Marketing and
Transportation, particularly Dr. D. J. Bowersox, Dr. Forrest
"Sam" Carter, Dr. George Wagenheim, and Dr. David "Skip"
Smith who have at one time or the other provided valuable
assistance and advice, I thank very much.

To my friends/mentors, Dean Richard J. Lewis, Dr.& Mrs.
George Axinn, without their reposed confidence in me, this
would not have been possible.

To my Doctoral colleagues, Morris Perry, David
Amponsah, Masaaki "Mike" Kotabe, Haksik Lee and Walter Zinn,
it certainly has been a pleasure knowing and working with
them. The exchange of ideas and personal experiences was an

invaluable help.

iii

To the companies that agreed to participate in this
research and their executives, I am also appreciative.

I am. grateful to IMichigan State ‘University for
providing an enlightening and enriching atmosphere for
pursuit of academic knowledge.

Finally I am grateful to my family. My wife, Jane and
daughter, Katrina Uchem, thanks for the inspiration, concern
and faith. They have been pillars of strength.

While others assisted in various ways, I alone am
responsible for the interpretation of the information

presented in this dissertation.

iv

TABLE OF CONTENTS

List of Tables ........ ... ....... ................... vii
List of Figures... ....... .......................... ix
CHAPTER
I INTRODUCTION
The Problem..................... ...... ....... 1
Defining Mode of Operations.................. 7

Defining Political Risk...................... 12
Significance of the study.................... 18
Theoretical contribution................... 19
Practical contribution..................... 21
Limitations of the study..................... 22
Organization of the study.................... 23
Theoretical Framework........................ 24
Incremental school of thought.............. 24
Strategic school of thought................ 27
Cost based models.......................... 32
Business Reaction to Political Risk........ 36
Model Specification
Market size and Growth rate.............. 42
Political Risk........................... 45
Conceptual framework....................... 51

II LITERATURE REVIEW............................. 55
Incremental School.......................... 58
Cross-country studies..................... 58
Cross-firm studies........................ 63

Strategic School............................ 71
Cross-country studies..................... 71
Cross-firm studies........................ 84

III METHODOLOGY
Research design............................. 96
Rationale................................... 99
Theoretical Decision Rules.................. 102
Rules for selecting time periods.......... 105
Rules for selecting variables............. 106
Rules for selecting measures.............. 106
Hypothesis and their rationale.............. 107
The Variables and their Measures............ 115
Political Risk (RISK)..................... 115
Market Size (GDP)......................... 119
Market Growth Rate (MAGRO)................ 120
Exporting (EXPORT)........................ 121
v

IV

V

VI

Foreign Direct Investment (FODIV).........
Data Analysis Procedure.....................
Homogeneity test..........................
How to pool data..........................

THE SUBSTANTIVE FINDINGS......................

Lag zeroMOdeIOO0............OOOOOOOOOOOC...

Exports.........OIOOOOOOOOOO0.000.000.0000

Foreign Direct Investment.................
Lag One Model...............................
Exports...................................
Foreign Direct Investment.................
Lag Two Model...............................
Exports...................................
Foreign Direct Investment.................

DISCUSSION OF RESULTS... ..... .................
Selecting the Appropriate Lag Model..........

Synopsis of Findings.........................

Finding one...0....O......OOOOOOOOOOOOOOOOOO

Finding Two.................................
Finding Three...............................
Finding Four................................
Finding Five................................
Finding Six.................................

IMPLICATIONS OF RESULTS AND FUTURE
RESEARCH DIRECTION...C......OOOOOOOOOOOOOOOOOO

Glossary of Terms...................................

Appendices

A.
B.
C.
D.
E.
F.
G.
H.
I.
J.
K.
L.

List

EXPORT DATA: SOURCE,COVERAGE,DEFINITIONS .....
U.S. FDI DATA: SOURCE,COVERAGE,DEFINITION ....
NATIONAL ACCOUNTS STATISTICS,SOURCE,COVERAGE..
COUNTRIES SELECTED FOR THE STUDY..............
LETTER OF INVITATION SENT TO COMPANIES........
RESPONSE SHEET................................
FOLLOW-UP LETTER..............................
INTERVIEW SCHEDULE............................
DATA ON COMPANIES INTERVIEWED.................
PERSONAL INTERVIEW QUESTIONS..................
BERI DATA COLLECTION METHOD...................
LETTER OF THANKS............. ............ 2....

Of ReferenceSOOOOOOOOOOOOO0......0.00.00.00.00.

vi

121
122
123
123

131
136
136
145
156
156
162
173
173
173

183
183
187
188
194
198
202
207
213

218

226

230
234
237
241
242
243
244
245
246
247
249
256

257

LIST OF TABLES

1. Three Major Taxonomies of Modes of Foreign
Involvement.................................... 11
2. Conceptual Framework of the Political Process... 17
3. Comparison of Major Studies on Business
Reaction to Political Risk..................... 37
4. Political Risk: A Conceptual Framework.... ..... . 50
5. Summary of Relevant Literature Reviewed......... 59
6. Summary of FDI-Political Risk Studies........... 60
7. Taxonomy of TimeSeries, Crosssectional Models... 124
8. Initial Correlation Coefficients................ 133
9. Final Correlation Coefficients.................. 134
10. OLS Timeseries Estimates for Exports, Fixed

Intercept Model .......... ...................... 137
11. OLS Timeseries Estimates for Exports, Random

Intercept Model ....... . ............. ........... 140
12. OLS Cross-sectional Estimates for Exports,

Fixed Intercept Model ........ .................. 141
13. Export Regression Results with Lag = 0.......... 143
14. Anova Table for Exports (Lag = 0)........ ....... 144

15. OLS Timeseries Estimates for FDI, Fixed
Intercept ModeIOOOOCOO......COOOOCOOOO0.0000... 146
16. OLS Timeseries Estimates for FDI, Random

Intercept Model ..... . ...... ............. ....... 147
17. OLS Cross-sectional Estimates for FDI,
Fixed Intercept Model ..... ....... ....... . ...... 149

18. FDI Regression Results with Lag = 0............. 150
19. Anova Table for FDI (Lag = 0)................... 151
20. Partial Correlation Coefficients Controlling

for Market Size and Growth Rate (LDC).......... 154
21. Partial Correlation Coefficients Controlling

for Political Risk and Market Growth Rate (DC). 155
22. OLS Timeseries Estimates for Exports,

Fixed Intercept Model (Lag 1)........ ...... .... 157
23. OLS Timeseries Estimates for Exports,

Random Intercept Model (Lag 1)................. 159

24. Export Regression Results with Lag = 1 ........ .. 161
25. Anova Table for Exports (Lag = 1)........ ...... . 163
26. OLS Timeseries Estimates for FDI,

Fixed Intercept Model (Lag = 1). ....... ........ 164

27. OLS Timeseries Estimates for FDI,

Random Intercept Model (Lag = 1)............... 165
28. FDI Regression Results with Lag = 1............. 167
29. Anova Table for FDI (Lag = 1)................... 168
30. Partial Correlation Coefficient Controlling

for Market Size and Growth Rate (LDC).......... 171
31. Partial Correlation Coefficient Controlling

for Political Risk and Growth Rate (DC)........ 172
32. Export Equation Regressio? Results (Lag = 2).....174

v1

33. Anova Table for Exports (Lag = 2)............... 175
34. FDI Regression Results (Lag = 2)................ 176
35. Anova Table for FDI............................. 177
36. Partial Correlation Coefficients Controlling

for Market Size and Growth Rate (LDC).......... 180
37. Partial Correlation Coefficients Controlling

for Political Risk and Growth Rate (DC)........ 181
38. Alternate Hypothesis Results for All Three

Lag Models..................................... 184
39. Factors Motivating the Export Effort............ 215

viii

6.
7.

8.
9.
10.
11.
12.
13.

14.
15.

LIST OF FIGURES

Distinction between International Channel of

Distribution and Mode of Entry ........ .........
Distinction between Mode of Entry and Mode
of Operations. ................... ........ .....

Relationship between U. S. Manufacturing

Exports and FDI................................
Modes of Operations as a Function of Fixed

and Variable Costs of Operation................
Exporting and Overseas Production as a

Function of Demand.............................
Alternative Models to be Tested....... ..... .....
Firm and Market Specific Factors that

Determine Modes of Operations..................
Effect of Political Risk on Mode of Operations..
Order of Literature Review......................
Internal Determinants of Export Behavior........
IPLC from the Viewpoint of a Follower Nation....
Investment as a Function of Firm Size...........
Modes of Operations, Competitive Position,

and Country Attractiveness Grid................
Exporting as a Learning Experience..............
Revised Conceptual Framework....................

ix

3O

33

35
43

46
53
57
66
68
77

89

212
225

CHAPTER I

INTRODUCTION

THE PROBLEM

After a firm decides to seek profits and growth in
foreign markets, the next concern is which markets to
enter and what modes to use to access those markets. The
mode of entry and operations then becomes an important
determinant of the amount of resources to be committed
and other operational issues. Investment will require
commitment of large amounts of resources over an extended
time period while exporting on the other hand requires
relatively less resource commitment. What then
determines the mode of operations which a firm will use?
This study looks at the effects of political risk in
influencing U.S. manufacturing firms mode of operations
decisions. The objective is to develop and test some
hypothesis concerning the role of political risk in the
mode of operations decision.

The mode of involvement and operations has been the
subject of investigation for decades. Early theorists

have been concerned with why nations trade, and the

benefits from trade and the commodity composition of
trade, etc. While these early trade theorists made
significant contributions to the development of thought
in international trade, they have created a research
direction around which future studies in the area have
followed. The assumption has been that a firm interested
in foreign markets has to access the foreign market first
via exports, then licensing, then joint ventures and
finally full scale commitment of resources or foreign
investment in order to minimize risk and accrue

experience.

OTHER STUDIES

Researchers in international marketing have followed
the trails of the trade theorists. The first studies
have focused on initial involvement decisionsl: the
result has been a multiplicity of studies on exporting.2
The studies on exporting have been concerned with the
following questions: who initiates the export decision,
what are the factors affecting export involvement and
success, and how does a firm implement its decision?

Recently the mode of operations decisions has been
viewed as a strategic decision involving a choice among
alternatives. (Ayal and Ziff 1978,1979):3 The choice of
mode of operations therefore will depend on several

factors. Neilson (1980) has related entry decisions to a

firms differential advantage which he called a resource
mix. Ayal and Ziff (1978) related a firms' choice of
operations mode to the desire to pursue market expansion
versus market concentration strategies. In a later
study, Ayal and Ziff (1979) indicated that the choice of
a market expansion strategy will be affected by ten key
product-market factors.

Goodnow and Hansz (1972) using an environmental
temperature gradient concluded that the mode of entry
into foreign markets by U.S. firms depends on the
environmental attractiveness of the market.4 Their
conclusions were that U.S. investors tended to use entry
strategies involving the greatest/most commitment of
resources (i.e. foreign. direct investment) in. markets
that are environmentally attractive or hot, and exported
to markets that are environmentally unattractive or cold'
MaClayton, Smith & Hair (1980) arrived at the same
conclusions using a sample of firms in the. U.S. health
care products industry. According to MaClayton, Smith &
Hair (1980), the dimensions used by these firms to
evaluate foreign markets include: " market and marketing
opportunity, legal barriers and their economic
objectives, cultural unity' and physiographic ibarriers,
political stability, and the level of economic

development and performance of the industry."(p.45)

Davidson & Harrigan (1977) in their study on new product
introductions in foreign markets showed that licensing
and FDI were used at roughly the same rate as means of
introducing new products in foreign markets.

Unfortunately in all these studies on mode of entry
or operations, political risk has either not been
included or its effect not isolated and tested. Yet
surveys of international executives have ranked it a very
important factor in the foreign involvement decision. In
studies where political instability or risk has been
considered it has however been empirically tested
exclusively with one mode of operation primarily FDI as
if other modes are not affected by political risk. (for
instance see Bennett & Green 1972: Nigh 1985) In other
studies, political risk and/or its consequences have only
been considered with variables other than modes of
operations.

This study focuses on the U.S. manufacturing sector
only. Kobrin (1980) argues that the probability of forced
divestment (a consequence of political risk) is a
function of three interrelated characteristics of
investors: i.e. industrial sector, ownership structure,
and level and maturity of technology. The political
risks which a firm faces cannot be attributed to
nationalism alone. Kobrin's study showed that forced

divestment isn't simply a manifestation of "economic

nationalism", reflecting national pride, an anti-foreign
bias, or political opportunism.

Forced divestment according to Kobrin is a policy
instrument used to attempt to achieve national economic
objectives by increasing control over economic factors.
In the vast majority of countries in Kobrins' study
(approximately 89%), forced divestment was selective.
While political and/or ideological motives were certainly
not absent, only a portion of foreign-owned firms were
forced. to divest. Further' more, there, is sufficient
evidence to conclude that the selection of firms forced
to divest is far from a random process. Enterprise
vulnerability is 21 function of industrial sector,
ownership structure, and behavior.

It is apparent that U.S. firms cannot take FDI in
high risk environments for granted. As Bregsten, Horst,
& Moran (1978) have suggested; with time the net result
is a shift in power from multinationals to home countries
that is cumulative, irreversible, and speeding up with
time and thus resulting in unfavorable consequences.
This study postulates that political risk, along with
market size and market growth rate are major determinants
of the mode of operations (exporting, licensing, joint
venture, foreign direct investment-FDI) used by U.S.
manufacturing firms in accessing foreign ‘markets.

Moreover the process is dynamic. U.S. firms continue to

adapt their strategies to changing political risk factors

in markets.

RECENT TRENDS

As the world's leading supplier of foreign direct
investment capital, U.S. companies have been sensitive to
the shifts in the global investment climate. Feldman
(1978a) attributes this to the "shifting perception"(p.2;
and policies of host governments. Nation-States through
their policies may constitute a barrier to achievement of
the firms' goals . Between 1966 and 1976, the book value
of American direct investments abroad grew steadily at an
impressive average annual rate of about 10% per year:
however, the composition of the annual increase in total
U.S. foreign investment has undergone a gradual and
revealing change. The proportion financed through
capital outflows from U.S. parent companies have
declined, with larger increments supplied from reinvested
earnings of affiliates. In 1967, 67% of the added
increment to book value of U.S. direct investments abroad
was accounted for by net capital outflows from U.S.
parents. Outflows continued to dominate until 1972, when
for the first time reinvested earnings accounted for 64%
of the net increment in book value. In 1976, only 35% of

the increase in U.S. direct investments abroad came from

capital outflows, while reinvested earnings accounted for
nearly 60%. The remaining 5% was attributed to
"international and unallocated" source.5

One might attribute this to U.S. government action
in order to protect the balance of payments; however, the
controls have been lifted and capital outflows have only
rebounded moderately. The high political risk countries
have accounted for only 20% of total investment.6 This
suggests that American investors are evaluating their
investment decision, and adjusting to political risk
among other variables.

The shift in the climate for foreign investment in
most developing countries is more extreme than that of
the developed world. This suggest that U.S. firms may be

opting for other forms of investment such as rentals,

licensing, franchising, service contracts and exports.

DEFINING MODE OF OPERATIONS

The conceptual definition of mode of operations
starts with an understanding of a similar and sometimes
substituted term - mode of entry. A foreign market entry
mode according to Root (1982) is "an institutional
arrangement that makes possible the entry of a company's
products, technology, human skills, management, or other

resources into a foreign country." (p.5) This definition

distinguishes international channels of distribution from
an entry mode. International channel of distribution is a
broader term and deals with movement of products through
domestic and foreign intermediaries from. producers to
consumers within the foreign market. Mode of entry on the
other hand deals with movement of factors of production,
products, technology, and management across countrv
boundaries. Figure 1 uses an adaptation of Cateora's
international. channel of «distribution framework: 'to
illustrate this difference.

This study' also distinguishes between Wmode of
entry" and "mode of operation." The former refers to the
means used to initially access or enter a foreign market
while the latter to mode used to continue operations in
that market. A firm may decide to enter a market via
joint ventures (mode of entry) but subsequently increases
its' commitment of resources in the form of foreign
direct investment (FDI). Using FDI as an example, mode of
operations refers to initial investments, reinvestment
and divestments. Figure 2 shows this distinction. Th
feedback loop in Figure 2 represents a desire by
management to adjust modes.

While most authors refrain from offering formal
definitions of mode of entry, there are numerous
typologies or classifications. Table 1 compares three

major classifications. Export entry mode probably is the

.omm .m .A.ocH .chzuH .a enmnoflm
"AH .mooBoEom. mafiboxumz HmcoflumcuoucH .muooumo afiaflnm Eoum povmmcd "mumoom

>Hucm we
owe: can cofiwsnfluumflo mo Hoccmnu HmcoflumcuoucH cwoBumm cofluocwumwoll.fi wusmfih

 

 

 

mmomoommm mmmeo _

\\...I 4! Has LA

 

 

 

 

 

mgHsm>hsmow_

W

 

szmzmoazaz l—

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mmzomzoo mMH:EMm ezmo< - -

onmmom onmmom onmmom oszzmqu — wooqozmoma 11—

c c , _ I , - \ k

7...\\\ /I\\\ ‘, fl ozHemomxm _\ n F meoooomm 1—
wmezsoo onmmom smezm mo ago: smezsoo mzom

 

 

 

 

mm>HB¢2mMBA¢ ZOHBDmHmBmHQ ho AMZZ<ZU AdZOHB<ZmMBZH

 

 

10

.mCOHumuomo mo memo: mam muucm mo wmoz comBuom coauocflumfloll.m musmflm

 

 

ZMHh

— mBUDoomm—

 

 

 

 

 

 

m
meZDOU

 

 

'

 

 

.P

moamommm

 

.._

Ham {*g

 

l

T... i,

||II

.— UZHmZmuHA

ill

 

 

*_ OZHBmOmehL

 

 

 

 

 

TBUDQO&Q—

 

ZmHh

 

 

 

 

mZOHB<mmmO
m0

wmfizm ho mDOZ

 

 

 

Q
>MBZDOU

 

 

 

 

TABLE 1.

11

Three Major Taxonomies of Modes of Foreign Market
Involvement

 

S T U D Y

 

Mode of Operations or Entry

 

 

 

 

 

Mason, Miller, We1ge1 Root (1982) Cateora (1983)
(1975)
1. Export 1. Export Entry Modes 1. Export
Indirect
2. License, technical Direct Agent/ 2. Licensing
aid Distributor
Management contract Direct branch/ 3. Joint
Subsidiary Venture
3. Joint venture with Other
minority interest 4. Manufactur-
2. Contractual Entry Modes ing
4. Joint venture with Licensing
majority interest Franchising 5. Management
Technical agreement Contracts
5. Wholly owned Service contracts
subsidiary Management contracts
Construction/turnkey
Contract manufacture
Coproduction agreement
Other
3. Investment Entry Modes
Sole Venture: new
establishment
Sole Venture: acquisi-
tion
Joint Venture; new
establishment
Source: R. H. Mason, R. Miller, and R.HD. Weigel. (1975).

The Economics of International Business (New York:
John Wiley & Sons, Inc.

Franklin Root. (1982). Foreign Market Entry Strategies
Amacon.

 

Philip Cateoria. (1983). International Marketing.
Homewood, IL: Richard D. Irwin, Inc.

 

12

major classifications. Export entry mode probably is the
only distinguishable mode since the company's final or
intermediate product is nmnufactured outside the target
country and subsequently transferred to it. Investments
involves some percentage of ownership by an international
company of manufacturing plants in the target country.
Mason, Miller & Weigel (1975) have argued that
intermediate modes such as licensing could be considered
a form of exporting i.e. "export of know-how."(p225) Also
a joint venture agreement could be termed investment if
the foreign firm owns a high equity share. Thus although
these modes are conceptually different, they provide
measurement difficulties. We shall now turn our attention

to defining political risk.

DEFINING POLITICAL RISK
This study defines political risk as the "perception
of international business executives that host
governments will take actions that will affect the firms'
profit, earnings potential or competitive, position in
that market." That perception is based on sources of
information concerning the host environment. Information
can be obtained from internal or external sources.
The concept of risk implies uncertainty about the
outcome of an event, which is a function of the

interaction of variables affecting the event. Most

13

authors in defining political risk focus on the aspect of
the environment which interests them. Political risk has
thus been equated to political instability, non-business
risks and expropriation. What follows is a discussion of
the current vieWpoints on political risk, political
instability, political events, etc.

Lee Nehrt's study on the political climate for
private investment views the investment climate as a
composite of economic, social, administrative, and
political factors.7 The political climate encompasses
risks of nationalization or expropriation. Root (1968)
appropriately' defines political risk as those "beyond
normal business risks that are generated by the
attitudes, policies and overt behavior of host
governments and other local power centers such as rival
political parties, labor unions, and nationalist
groups."(p.72) Cracco (1972) defined political risk as
the "subjective probability that certain political
decisions will be taken (or certain political events will
occur) which will perceptibly change the present business
environment."(p.6) The implication of this definition is
that political risk can lead to favorable and unfavorable
outcomes. As Kobrin (1978) put it "one can only say that
political events may affect the firm: whether they do so
is a function of both environmental conditions and

industry and firm specific factors. A coup for example

14

may place a radical socialist government in power which
expropriates all foreign-owned firms (as in Ethiopia); or
it may result in a conservative government which actually
returns expropriated property (as in Chile in 1973) or
may simply replace governing elites without affecting
foreign investors at all."(p.114) Boddewyn and Cracco
(1972) say that the strategy used in multinational
markets is dependent on national policy and functional
areas within the organization, a viewpoint not shared by
Kobrin. The conclusion can only be that the
consequences of any political risk for the foreign
investor depends upon its nature, the conditions under
which it occurs and the characteristics of the specific

investment in question. (Kobrin, 1976:p.69)

Political Instability

Whether political instability is equated to
political risk depends upon a number of considerations.
Theoretically political instability does not equate
political risk. Thus Kobrin (1978) claims that many,
perhaps most, attempts to control foreign investment are
motivated by economic factors. However a closer
evaluation of management views of political risk points
to the fact they tend to equate the occurrence of

specific events with political instability and

15

consequently political risk. This leads to the issue of
managerial assessment and evaluation of the political

environment.

MANAGEMENT VIEWS

Managers have consistently rated political
instability and political risk as one of the major
influences on the foreign investment decision. Further,
managers have been observed: 1) to consider political
instability or political risk, typically loosely defined,
2) to be an important factor in the foreign investment
decision, 3) and they seldom do any rigorous and
systematic assessment and evaluation of the political
environment. Most analysis is superficial and subjective,
not integrated formally into the decision-making process
so they assume instability and risk to be one and the
same, and 4) to rely primarily on internal sources for
environmental information. When. they look for outside
data they are most likely to go to their banks or the
general business media. Cracco (1972) added to these
conclusions by stating that "at the organizational level,
headquarters have a more conceptual, principle oriented
view of political risk, while the subsidiary views it
more from an operational vantage point."(p.145) In other

words, headquarter's perception of the risk of certain

16

types of political events is higher than the perception
of indigenous executives. He further determined that the
perception of political risk varies by nationality.

The heart of political risk seems to depend on
events which culminate in these major business obstacles.
When an unfavorable consequence such as expropriation
occurs, it is too late to react. Political risk studies
have focused on sources of political influence and on the
channeling institutions which. translate influence into
decisions. Cracco (1972) has attributed the sources of
political influence to three factors: ideological,
social, and economic. In his conceptual framework (Table
2) these influences are channeled through institutions
such as government and national legislatures to influence
the business environment. The business environment in
turn affects different firm's functional areas such as
marketing, financial, management and production.

The relationship between political risk and
political instability could depend on whether the analyst

seeks a reactive or proactive stance (offensive or

defensive strategies). A proactive stance focuses on
events that have the probability of resulting in
potentially unfavorable consequences to business. A

proactive stance equates political risk with political

instability. A reactive stance on the other hand focuses

17

TABLE 2.--Conceptual Framework of the Political Process

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sources of Institutions Types of Firm's
Political Channeling influence Dependent
Influences Political on Variables
Influence BUSiPeSS
Political Environment
Structure
1. Ideological Marketing Variables
Socialism _*National '_‘ '—_fil. Price
Communism *fiLegislature k“ R-—-2. Product
Nationalism (Legislative 3. Promotion
Capitalism body) 4. P. Distribution
5. Environment
2. Social Government
Institutional (Executive Financial V
Groups body) Regulate 4
Army 1. Capital Structu
Church 2. Remittance
Bureaucracy _* Agencies h* policy
‘ . . . PF' (adminis- n-iMaintain 3. Investment
Non1nst1tut1onal tration) P011CY
Groups 4. Environment
Middle Class 5. Accounting
Elites Legal 6. Taxation
Peasants Legislature
. (regional Mold Management V
3. Economic body
GNP/capital level City 1. Foreign pers.
Balance of government} policy
payments ..J _4 2. Local pers.
GNP Growth Rate ,_J r— policy
Unemployment 3. Labor
level legislation
4. Taxation
Prodgction V
1. Inputs
2. Plant layout
3. Machinery
4. Prod. design
Source: Etienne Cracco. (1972). "The Nature and Perception of

Political Risk for the International Corporation:

An

Exploratory Analysis with Special Reference to Brazil."
Unpublishied Ph.D. Dissertation, Michigan State University.

 

 

 

 

 

 

18

on outcomes due to changes in the political process, such
as expropriation. Under such posture political risk is
not equated to political instability. Robock (1971) has
suggested that the decision maker, in order to anticipate
changes in the business environment, should recognize the
evolutionary path along which changes are occurring,
identify the principal motivating forces behind the

change and make judgments as to timing.

SIGNIFICANCE OF THE STUDY

Firms that decide to seek profits and competitive
advantage by entering foreign markets are faced. with
political as well as normal business risks. Normal
business risks address issues such as, will consumers buy
the products, What are the effects of marketing programs,
What 'will be the competitive response, etc.‘ Root
(1968) has suggested three responses to political risk by
firms he surveyed. The three responses are avoidance,
adaptation and risk transfer. (p.73)

U.S. businesses through the decade of the 1980's
have been facing increasing global competition from
emerging industrialized nations such as Japan, Korea,
Taiwan, Brazil (to mention a few) and cannot afford to
completely avoid rather than adapt in foreign markets

where political risks are high. As Shapiro (1981) put

19

it: "The real issue is the degree of political risk a
company is willing to tolerate and the return required to
bear it. A policy of staying away from countries
considered to be politically unstable ignores the
potentially’ high returns available and the extent to
which a firm can control these risks."(p.64)

Firms can 'avoid' a market before involvement , but
once it commits itself it has to 'adapt' to changes in
political risk. This study suggests that the common mode
of adaptation is via mode of operations. As risk
increases, firms can firms would opt for licensing, joint
venture or exporting strategies rather' than FDI. The
substitutability of exporting for investment is used to
test the adaptation strategy of U.S. manufacturing firms.

The connection between risk resulting form political
influences and investment has significance for
international executives. The consequences of political
risk can be substantial. It could range from. mild
inconvenience to confiscation and violence. What follows
is a delineation of the theoretical and practical

contributions of this study.

THEORETICAL CONTRIBUTION
First, this study would help resolve some of the
contradiction in findings8 between survey studies and

cross-country studies.9 Also by looking at the

20

relationship of political risk to modes of operation from
a strategic choice decision process and using the same
methodology as the cross-country studies, some of the
interesting findings of cross-country studies can be
explained. This conceptual approach in itself is unique
to this area of study.

Secondly, this study contributes to the
understanding of the role of political risk in the
strategic choice decision of mode of operation when
compared to market size and market growth rate (both
significant influencers on the decision to invest).
While international executives of multinational firms
consider political risk before entering into foreign
markets, the role of political risk has not been shown as
influencing the choice of mode of operations after entry.
The contention of the author is that political risk plays
a significant role in the strategic choice decision of
mode of operations. The survey studies have indeed shown
that political risk is a major consideration in the
choice of nations to invest in. The author further
suggests that political risk results in adaptation
policies through modes of operations in that nation
rather than complete avoidance.

Thirdly, the consideration of modes of operation as
an incremental process Ihas for' a long time. been the

10

dominant approach or conceptualization. Following the

21

works of Adam Smith, Bertil Orlin and subsequent
economists and marketers, the multinational firm is seen
as going through a gradual and incremental process in
internationalization, ‘Very little attention ihas been
paid to the fact that mode of operation are and could be
dynamic strategies or actively pursued alternatives.
This study therefore tries to show that multinationals
have for a long time used these modes of operations as
strategies to access their chosen foreign markets. This
suggests a need to reconsider mode of operation as active
alternatives rather than as passive incremental
processes.

Finally, this study provides a comprehensive review
of current literature on political risk and mode of

operations.

PRACTICAL CONTRIBUTION
First, the role of political risk in the mode of

operation decision affects a host of business decisions
relating to the following areas:

1) intermediaries used

2) production levels

3) horizontal and vertical integration

4) desired locus of control
So the findings will be relevant to multinational firms

planning.

‘ ~ .- M

'22

Secondly this study has practical implications for
public policy makers. It suggests policies to pursue to
attract investment and achieve economic objectives. The
success of such policies may depend on the political risk

in that nation as perceived by international investors.

LIMITATIONS AND FUTURE RESEARCH DIRECTION

There are a number of limitations of the research.
First, the focus of the study is on mode of "operation
rather thanq initial involvement -decision. Mode of
operations refers to initial involvement as well as
expansion decision. Using FDI as an example, mode of
operations ‘will include new investments, reinvestments
and divestments.

Secondly, conclusions from the study is limited to
the actions of U.S. manufacturing firms at the aggregate
rather than for select industries or individual firms.
Further research direction involves investigating the
same phenomena for Canadian firms, Japanese firms, etc.
Inclusion of intermediate modes of operations such as
joint ventures, licensing (if measurable) can add
significant insights to the strategic choice process. The
behavior of Third World multinationals needs to be
researched before overall theory relating to this
decision process can be developed. This study begins

that process of theory building in this area.

23

Thirdly, a sophisticated model would have to include
firm specific and country specific variables that would
affect mode of operations choice decision process. This
model will not follow the current popular approach in
theory' development (the eclectic approach courtesy of

John Dunning).11

The author does not deny that the mode
of operations has to be a function of country specific
and firm specific variables - especially firm size.
However, available data suggest that 80% of exports and
definitely a greater percentage for FDI is carried out by
large firms - U.S. multinationals. Therefore adjustments
using mode of operations may be a phenomena of large
firms.

Finally, the conclusions are only as good as the

secondary data used. Errors in collection and reporting

data could bias the conclusions (See appendix A).

ORGANIZATION OF THE STUDY

In Chapter I, the first task is to fully develop the
conceptual framework by bringing together the dependent
and independent variables. Therefore the language of the
research will be clarified by discussing the key
concepts.

Chapter II reviews the relevant empirical research

from the perspective of different schools of thought. Its

24

objective is to show how the major variables of the
research are derived.

Chapter III discusses the research design, hypothesis,
operationalization. of the ‘variables and. data analysis
procedure.

Chapter IV reports the results of all lag models
tested.

Chapter V is a discussion of the results and
conclusions.

Finally in Chapter VI, the implications of the
findings for theory and policy are discussed. Future

research direction is also outlined.

THEORETICAL FRAMEWORK
'Incremental School

The theoretical framework can be built around two
streams of research. The first school of thought focuses
on initial involvement in foreign markets.
Conceptualization of the internationalization process can
be traced back to economic theory. Bertil Ohlin (1933)
in his works for instance postulated that fbreign trade
was only a special case of domestic trade.

Exporting has been identified as the first mode of
operations and its adoption is said to be passive and

gradual. This school is adequately represented by the

25

research works of Johanson and Wiedersheim-Paul, Finn,
Olson and Welch (1978); and Bilkey and Tesar (1978).

Johanson & Wiedersheim-Paul (1975), who traced the
complete internationalization process for Atlas-Copco,
Facit Volvo, and Sandik (all Swedish firms), had observed
that each stage in the process represented successively
larger commitment of resources. They concluded that the
pattern was a gradual internationalization, rather than
large spectacular investments.

Wiedersheim-Paul, Finn, Oslon & Welch (1979) later
studied small firms in Australia and found that they
tended to expand interstate before their first export
sale. Bilkey and Tesar (1977) proposed a stage model for
examining export behavior. The International Product
Life Cycle (IPLC) which was developed by Wells (1966) and
R. Vernon (1966) focused on innovation/adoption process
in explaining export movement to foreign markets. The
theory acknowledges that at the maturity stage, foreign
production starts, however production in foreign markets
is by indigenous investors rather than by foreign
investors. Root (1982) has argued that a company will
gradually change its entry mode decision in a fairly
predictable fashion. Increasingly, it will choose entry
modes that provide control over foreign marketing

operations. But to gain greater control, the company

26

will have to commit more resources to foreign markets and
thereby assume greater market and political risks.
Two conclusions can be derived from these studies:

1) Export marketing is usually considered to
be a first step in the process of internationalization.
(Cavusgil & Nevin, 1980). Root (1982) attributes this to
the desire to gain experience; "the general belief is
that prior involvement in foreign markets in the form of
exporting increases experience and consequently reduces
uncertainty."(p.90) But this may prove to be fallacious
since political risks can, and does result in substantial
losses even in the face of accumulated experience.
Exporting experience may not cut down on certain types of
risks such as expropriation. A firm such as Shell (Nig)
Ltd. with large accumulated experience is still
vulnerable to political risks unless that firm assumes
such a position of size and importance to its host
economy and consequently becomes untouchable. Maclayton,
Smith & Hair (1980) also did not substantiate the
experience explanation.

2) The initial involvement decision in
international marketing is conceived as a gradual and
sequential process. The gradual process is thought to be
the consequence of greater uncertainty, higher cost of
gathering information, and the lack of experiential

knowledge in international marketing activities.

27

Further, two implications can be drawn from these
conclusions:

1) The concern for risk of exporting exhibited
by the firms studied can be extended to other modes of
operation. At the very least, the same level of
perceived risk in foreign market operations will be shown
for the other modes of operation. If firms are concerned
about riSk in exporting, greater concern will exist for
other modes that involve more resource commitment. The
significance of risk in foreign markets involvement is
thus emphasized.

2) The "gradual and sequential" process can be
extended to other modes of operation. In other words,
licensing, joint ventures and FDI will similarly follow
in a gradual and sequential manner. If that is the case
the modes of entry substitute rather than complement each
other. This leads us to the question of substitutability
or complementarity of exports for FDI and the views of

the second school of thought.

Strategic School

The second school of thought sees entry or expansion
into foreign markets as a strategy or dynamic
alternative. Studies under this school of thought have
looked at two dimensions. The first sees the choice of

modes of operations as a function of firm specific

28

decision variables (see Grosse, 1985) while the others
see the decision choice depending on country specific
variables. (see Goodnow and Hansz, 1972) U.S.
manufacturing firms in trying to access foreign markets
have a choice among adaptation strategies. One mode or a
combination of modes can be chosen. The choice depends
on several factors.

Neilson (1980) says that the decision is based on a
firm's differential advantages or resource mix. Goodnow
and Hansz (1972) say that the mode of operation decision
depends on environmental factors. Environmental factors
such as market opportunity, economic development and
performance, political stability, cultural unity, legal
barriers, physiographic barriers and geocultural distance
determined the environmental attractiveness of a fereign
market and invariably the mode of entry strategy that is
used.

Grosse (1985) tried to develop an imperfect
competition theory of the MNE. He concluded that MNE
decision makers can consider four alternatives for
serving national markets which lead to optimal decisions
on production level, internal and external pricing, and
horizontal and vertical integration under 'various
conditions. No study to date has investigated the

relationship between political risk and the mode of

29

operation despite overwhelming evidence that political
risk. is a major consideration in mode of operations
choice decisions in foreign markets. Root (1968) in
discussing a firms' adaptation policies to political risk
notes joint venture as a means of adaptation to political
risk. Root, suggesting joint venture as a mode of
adaptation implies that a less risky mode is used in
situations of high political risk. This implies that
firms strategically adjust in politically unstable
environments using mode of operations.

Horst (1974) investigated the issue of
complementarity or substitutability of exports for FDI.
This topic had. been previously analyzed by empirical
researchers but emphasis was related to the policy
question of the impact of multinational firms on the
balance of payments. Horst applied regression analysis
to twenty-three manufacturing industries and eight
countries and the relationship in Figure 3 was observed.

As long as foreign subsidiary net sales as a
percentage of domestic shipments were small, an increase
in sales was accompanied by a rise in U.S. exports. Also
Lipsey & weiss (1976) study which investigated the same
issue produced mixed results. They examined the
relationship between the exports and FDI of the U.S. and
thirteen other major exporting countries. Using

regression analysis, they found that for U.S.

30

TM(%)

    
  

.7 Ran e o
‘ g f Range of

substitutability

complementarity

 

.-----------

_A

 

v v r 1 T I 1 U I 1 l I I I '

1 2 3 4 5 6 7 8 9 10 11121314 15 PM(%)

Figure 3.--Re1ationship Between U.S. Manufacturing Exports
and FDI

SOURCE: T. Horst, "American Exports and Foreign Direct
Investment," Harvard Institute of Economic
Research, Discussion Paper 362.

NOTE: TM = Estimated relationship between US exports as
a percentage of domestic shipment
PM = Subsidiary net sales as a %age of domestic

net sales.

31

manufacturing affiliates their level of activity was
positively related to U.S. exports. Also, U.S.
manufacturing affiliate activity was negatively related
to exports by the thirteen other countries.

Although results seem to be mixed, Horst (1974)
concludes that the choice between exporting and overseas
production depends on whether savings from exporting are
greater or less than the costs of control and
coordination of foreign production. However correlation
of U.S. exports and FDI to a particular foreign market
using macro data may not be enough to suggest that a
firm's exports and FDI to a market are complementary.
David Aaker (1984) has suggested that exporting would be
a logical choice when "heavy commitment is inadvisable
because of political risks or because the market is
either not sufficiently attractive or its prospects
uncertain."(p.285) The question is, does political risk
however measured constitute a significant influence on
corporate mode of operations strategy or is it a
consideration among many as suggested by the Goodnow and
Hansz (1972) study and by Maclayton, Smith and Hair
(1980) studies. If political risk affects foreign
direct investment, then it should also have some effect
on other modes if they are alternatives.

One might argue that foreign governments are

dictating the mode of operation of firms and therefore,

32

that, it really isn't a decision variable of the firm.
Examples of Brazil, Mexico and Nigeria with local content
laws come to mind. It is important to note that these
countries are trying to build an economic base by
attracting investment. The success of such policies is
ix: doubt. or' can. be ‘jeopardized. if political risk is
prevalent in that market. This study tests the effect of
political risk on exporting and investment. The findings
will help determine why some of the host government
policies of encouraging/discouraging investments or

exports maybe offset by political instability.

COST BASED MODELS
Buckley and Casson (1981) were concerned with timing
of entry and developed a cost based model. The authors
state that fixed production costs for exporting is
negligible while variable costs are almost certain to be
high. The converse is true for FDI (low variable costs
and high fixed costs). Their rationale is that
exporting, licensing, and FDI are in ascending order of
fixed costs and decreasing order of variable costs. The
relationship is diagramed as in Figure 4.
The illustration shows that as a nmrket grows, it
may be efficient to switch directly from exporting to FDI
and forego the use of licensing. If the potential size

of the market is small then the firm will export

33

 

 

Cost
Exporting
Licensing
I
I
I
I
I
I
I
I
I
I
I
+— Exporting -—.fl— FDI —.""
- ' ‘r Quantity

Switch

Figure 4.--Mode of Operation as a Function of Fixed and
Variable Costs of Operation.

SOURCE: Adapted from,P. Buckley and M. Casson, The Future
of the Multinational Entegprise (London:
MacMillan, 1976).

 

 

34

indefinitely. If the potential market is only of
moderate size, the firm may switch from exporting to
licensing, but not from licensing to FDI. Alternatively,
if the market is large to begin with, the firm may omit
the exporting stage and begin with licensing: if the
market is very large it may even commence servicing with
FDI.

Hazard (1977) was concerned with location decisions
of firms. According to his model a firm will locate in a
foreign market if the costs of production abroad are
less than the costs of production at home plus the
transportation costs to the foreign market. He utilized
the automobile industry to illustrate how foreign nations
have restricted United States exports to their markets
making FDI necessary to preserve market access and
tending to lower foreign production costs to the point
that they were able to export to the U.S. market.

In another cost related explanation, Horst (1974b)
explained that the choice between exporting and overseas
production depends on whether savings from overseas
production is greater or less than the costs of control
and coordination. Figure 5 shows the relationship. The
shaded portion represents the savings from overseas

production.

35

PRICE MARGINAL REVENUE FROM FOREIGN SALES

 

 

 

 

FOREIGN DEMAND CURVE
/// Marginal COSt
gggéa7’ cthmgnming
Mamfinalcnst
ofcwenflms
‘\\\\\\\\\ Exbdmnion
Quantity

Figure 5.--Exporting and Overseas Production as a Function
of Foreign Demand.

SOURCE: T. Horst, “The Theory of the Firm" Chapter 2 in
J. H. Dunning, ed., Economic Analysis and the
Multinational Enterprise (London: George Allen
& Unwin, 1974b):31-46.

 

 

36

BUSINESS REACTION TO POLITICAL RISK

There have been conceptual and empirical articles
detailing business reaction to political risk.
Unfortunately there isn't universal agreement on business
reaction to political risk. This section discusses those
responses. Table 3 provides a comparison of the studies.

Root (1968) surveyed (by mail) and interviewed (in
person) top executives of companies listed in the Fortune
Directory of the 500 largest industrial corporations.
Three major responses to political risk were identified:

1) Avoidance. The response pattern denotes staying
out of foreign market without respect to the modes of
operation. Under such circumstances the decision not to
invest implies that perceived opportunities are more than
offset by perceived political risk costs. In determining
how management weighs opportunities and risk factors the
author attributed it to the aggressiveness or passiveness
of the management. An aggressive management will be
reluctant to refrain from investing in high political
risk markets while a passive management will not. A
defensive management might opt to exploit the new
opportunity by relying on direct exports or licensing.

2) Adaptation. This policy consists of a range of
dynamic alternative policies encompassing one or :more
aspects of the company's operation abroad. Adaptation

could be passive (minimal) or active. Active adaptation

13'7

.ooocououoa uo soda com.

 

.ouo- uuagm ..
use usauflom .n
ocasnnou "unquaaom .«

ucuauuo>:«
Hufiwcut .d

 

cauuouncot .n

uuuoaonoxouu
Macon weano~v>on .v

oduuu uwuocon
\uuoo 0:» ocamcosu .n

noduouaawnoa
uauoaa snou-uuonm .m

unsuno>uo Quandam .n

 

codunuoaooo .«

coup-acoaucou .n

 

 

 

Sauna haunt

 

£5533:
oval >uucu .n

856392 . H

 

ucuacouaoco

as»
acquaauoooz .n
005.32; . a

oucucwo>£ .n

 

 

unasnnoa aoca«au-
ocean no

acol850I I undouaomli
>uouuuu.

anaconousn «Icons:
Que-Io can
unocducw uo coauQOOH

any ocaaflouucoouu
muououu-

Rhonuoc .ooun ~.ucnou
huubauuo uofiuauvuu

Ins- osu ocuuouu¢lu
nouuouauu-

have. can acaIAUII
oooueauoau

unoluooacu «0 coal:

00 oOHuaGOI .n
oo o: .n
00 .H

 

nounsoua
load .n

:o«uaunaa« .A

006.040»‘ .n

 

snug» ououon

 

coma

 

Ouaouoxo

 

mama
oudnosm

 

who"
Hounuo¢

 

.uha~_
xoonou

 

 

coma
cu

 

scum «coauuuoa o» ecuuoaloz gonad-9a so couvaum Mann: «0 eouuudnl90ll.n Hausa

38

seeks. to ensure ‘the long-run survival of ‘the foreign
enterprise by making it an essential element of the
national economy in the eyes of the host government
officials and other political persons. Specific policies
to minimize political risk include staffing foreign
operations with local managers and workers, using local
suppliers and participating in joint ventures.

3) Risk Transfer. Companies seek to transfer risks
to third parties. True risk transfer does not alter risk
exposure: it merely shifts the incidence of loss.
Favorite third parties that assume these risks are
private insurance companies and government
agencies/programs such as AID.

Robock's (1971) conceptualization had three choices
- go, modified go, and no go. According to Robock "the
international enterprise is not helpless in the face of
political risk."(p.18) Robock's risk minimization
strategy or 'modified 90' strategy are as outlined below:

1) the use of investment guarantees

2) timing and entry strategy

3) altering the subsidiary's activity

4) international production network strategy

5) controlling the location of intangible assets

6) local purchasing strategy

7) sourcing and movement of funds

8) direct lobbying

 

39

Antoine W. Van Agtmael (1976), a bank executive,
relied on past experience and probably contacts with
businessmen. to decry their lack. of sophistication in
political risk analysis. He suggested a framework to
systematically analyze political stability.12 Van
Agtmael's strategies of confrontation or cooperation is
for firms faced. with. political risk: in. their' foreign
markets. The determination of the policy to pursue will
depend on "candid evaluation of one's bargaining
position."(p.27)

Shapiro (1981) did not use any empirical data to
support his framework. He was concerned about findings
that suggested that managers are not involved in any
systematic analysis of political risk. So his approach
was to discuss policy options available to international
decision makers after determining that a target market
represents considerable political risk, Shapiro thus
suggests "four separate, though not necessarily mutually
exclusive"(p.64) policies.

.A policy of avoidance simply implies staying away
from the market in question. But Shapiro argues thus:

"The real issue is the degree of political
risk a company is willing to tolerate and
the return required to bear it. A policy
of staying away from countries considered
to be politically unstable ignores the

potentially high returns available and the

extent to which a firm can control these
risks."(p.65)

40

.A policy of Insurance involves insuring assets in
politically risky areas. Coverage provided by U.S.
government through the Overseas Private Investment
Corporation (OPIC) is typical. Shapiro states that this
approach is inadequate since the economic value of its
future cash flows is not covered, only the investment in
assets is covered by insurance.

Negotiating the environment involves 'concession
agreements' prior to undertaking the investment. This
policy’ however is not reliable since neW’ governments
could repudiate such agreements.

Finally, a policy of structuring the investment
involves adjusting; 1) corporate operating policies in
the areas of production, logistics, export, and
technology transfer, and 2) its financial policies.

After the multinational has invested in a project
its ability to influence its susceptibility to political
risk is greatly diminished but not ended. Policies
available to the firm under such circumstances - planned
divestiture, short-term profit maximization, changing the
benefit/cost ratio of expropriation, developing local
stakeholders, and adaptation.

Planned. divestiture implies phasing' out their
ownership of foreign investments over a fixed time period
by selling all or a majority of their equity interest to

local investors.

41

Short-term profit maximization is manifest in such
practices as "defining maintenance expenditures, cutting
investment to the minimum necessary to sustain the
desired level of production, curtailing' marketing
expenditures, producing lower quality merchandise,
setting higher prices, and eliminating training programs,
cash generation maximized for the short-term."(p.66)

Changing the benefit/cost ratio is manifest by such
practices as establishing local research and development
facilities, developing export markets for the affiliates
output, training local workers and managers, expanding
production facilities, and manufacturing a wider range of
products locally as substitutes for imports.

Developing local stakeholders involves incorporating
individuals and groups who have a stake in the affiliates
continued existence as a unit of the parent
multinational.

Adaptation policies involves "trying to earn profits
on the firm's resources by entering into licensing and
management agreements."(p.68)

This section shows that firms have a wide range of
reactions to political risk before and after they enter
the foreign market. In a sense one can say that data for
investments and exports applies to firms that have
already entered the foreign markets. Data for firms that

have avoided high risk foreign markets is not recorded.

 

42

How effective are these strategies? Beeman (1979)
surveyed 132 firm-nation pairs in order to test six
hypothesis and developed a paradigm to explain
differences in effectiveness of risk reduction
strategies. Executives believe that adaptation
strategies have the greatest risk reduction effect.
Dependency strategies have little or no effect. Hedging

strategies increase risk.

MODEL FOR MODE OF OPERATION ANALYSIS

A firm's mode of entry could be by FDI (when
political risk is low), but reduces FDI (i.e.
reinvestments) as political risk increases. In ‘what
appears to be a contradiction of his earlier position,
Root (1982) suggests 'that. international. executives (are
apt to consider political risk in a target country, such
as general political instability or the threat of
expropriation, they favor’ entry modes that limit the
commitment of company resources. Conversely, low
political risks encourage equity investment in a target
country. Three major factors are encompassed in the

model shown in Figure 6.

Market Size and Growth Rate
This is one variable that the survey studies and

cross-country studies have universal agreement as to its

MODEL ONE

.————. 10w

FDI mOd

L._(

 

 

igh

MODEL TWO
“low

1 —

FDI mod

.L__..

 

 

hi

MODEL THREE

{ high
—_l

FDI mod

 

 

Figure 6.--Alternative Models to be Tested.

 

 

 

 

 

 

 

 

 

 

 

arg large
ed MARKET med
SIZE
small small
arg arge
MARKET
med
med GROWTH
RATE
small small
high high
med POLITICAL med
RISK
low low

high

mod

low

high

low

19W

high

 

EXPORTING

 

 

 

 

EXPORTING

 

 

EXPORTING

 

 

 

 

 

44

effect. Market size is definitely a variable that
influences investment decisions. Indicators of market
size were used by Aharoni (1966) and Goodnow & Hansz
(1972) and. was identified as significant influencers.
Basi (1963) surveyed international executives in an
attempt to determine the criteria they use in deciding
upon a foreign investment and found that the two most
important factors in foreign market investment were
political instability and market potential. Aharoni
showed the same results through in-depth interviews with
international personnel in thirty-eight firms. Bennett
and Green (1972) recognizing the significance of market
size and potential controlled for both variables in
determining the effect of political instability on FDI.
Market size is an important variable because sales volume
helps achieve economies of scale. Also, Davidson (1980)
says that FDI makes economic sense when volume exceeds a
level at which the average cost of serving the market
through exports exceeds the average cost of production
within the market.The significance of' market size is
evidenced (by its inclusion in studies of political
instability and foreign. direct investment. (Green and
Cunningham 1975; Kobrin 1976; Thunell 1977) Grosse
(1985) in developing an imperfect competition theory of
the MNE concluded that the choice between alternative

modes of operation is "dependent upon the costs and

45

government policy constraints involved in each
alternative, the size of the market, and changes over
time."(p.74)

The results are therefore impressively consistent
that market size and potential are major determinants of
entry into foreign markets. Although market size is not
the only variable of interest, any model explaining modes
of operations should include market size since it is a
significant explanatory variable. The Mason, Miller &
Weigel (1975) model (Figure 7) summarizes the factors

affecting mode of operations.

Political Risk

There has been attempts at theory-building in
political risk. The theorists have tried to look at
factors that increase the risks of a market. Most have
focused on the host government as the major actor and the
MNC as the prime target. Boddewyn & Cracco (1971) have
attributed an increase in political risk to national
interest, national sovereignty, and national identity.
Kobrin (1980) however refutes the assertion that forced
divestment is simply a response to economic
nationalism , reflecting national pride, an anti-foreign
bias, or political opportunism. It is however incomplete
to explain an increase in investment risk or "forced

divestment" as Kobrin refers to it, to only firm specific

46

License Joint Joint
Size technical venture venture Wholly-owned
fin ' aid mgt. with with subsidiary
I ’ u o o -
mana- EXPORT contract minority majority

gerial investment investment -
techni~ ,///;'
cal

strengtfih

and /
int'l

expe- ,I’II'

rience

2:... /
/
/

 

/'

 

 

J/

 

Market size, level of develop-
ment, availability of natural
resources, permissiveness of
the environment

Figure 7. Firm and Foreign Market Specific Factors
Determining Modes of Operation.

SOURCE: Mason, Miller, and Weigel, The Economics of
International Business (Toronto: John Wiley
& Sons, Inc., ), p. 252.

47

factors. ILf his findings also are tenable then why is
there more forced divestment in certain countries? Also
given the three interrelated factors, if true, one would
expect a vulnerable firm to experience the same level of
forced divestment across all countries.

Gurr's (1969) theory of relative deprivation points
to a high level of national frustration as the key
determinant of expropriations. The host government uses
the IMNE as a scape-goat for the country's problems.
Green (1974) tried to attribute political risk to
political structures. His theses ‘was that political
structures can predict radical political changes.

Simon (1980) has used a composite model. According
to Simon, the MNEs vulnerability is a function of it
environments consisting of host country environment, home
country' environment, international environment and. the
global environment. Authors such as Bennett & Green
(1972) and Nigh (1985) implicitly believe that political
risk is a function of political instability and political
events respectively - a view the author shares.

Political instability has been identified as one
form of political risk. A vital element of the
environmental analysis is political instability. There
are other types of political risk, but the political
instability can arguably be the most significant.

Political risk consists of political instability and

48

political events which are forms of political risks. The
notion of political events as a determinant of political
instability has been investigated by Brewer (1984). He
looked at the association between specific governmental
policy changes and other forms of political instability.
His analysis focused on three types of governmental
instability: personnel change, factional change, and
systemic change. His findings showed moderately strong
relationship between policy instability and factional
change. The other two governmental instability variables
were positively related to the instability of foreign
exchange policies, but not statistically significant at
the 0.05 level. Together, the forms of governmental
instability are more strongly associated with political
instability among industrial nations than among
developing nations. This piece of research ties the knot
between a relationship researchers have been interested
in for so long - namely, can firms use political events
to predict policy changes?

Political instability also has been proven to
influence international executives in their decision to
invest albeit no systematic evaluation and analysis is
done before commitment to invest. (Root 1968a; Root
1968b: Rummel & Heenan 1978) A Conference Board study of
international executives cites political instability as

the most frequently mentioned political obstacle. This

49

could be due to the fact that a nations' level of
political instability affects the rate of return that
investors expect to receive on their foreign ventures.
In general, the greater the instability a nation
exhibits, lower profits can be realized from
manufacturing FDI. (Green & Smith, 1972) While Kobrin
(1978) concludes that political instability is neither a
necessary nor sufficient condition for political risk,
Green & Korth (1974) succinctly put the relationship as
such: "a vital element of the environmental analysis is
the question of political instability .....instability is
the most significant form of political risk."(p.23)
Robock (1971) notes that political instability
depending on how it is defined is a separate although
related phenomenon from that of political risk. He says
that political risk exists when; 1) discontinuities or
radical changes occur in the business environment, 2)
which are difficult to anticipate, and 3) which result
from political fluctuations and disturbances.(p.7) It is
probably due to Robock's definition that studies trying
to measure political instability have used political
events. His conceptual framework as shown in Table 4
suggests that consequences (expropriation, confiscation,

etc.) result from sources (political events).

5(3

 

aocuaoo ownssaou s.u:oammomo< one coduou«u«ucoou

.ONuo "Aymams<nmanov moonwmsm oHuoz no
.xoonom :duoum

”Mada Houduaaom:

 

..Hbmav

«mumbow

 

.num: one
occausao>ou .ncowuoouusmcu .nuOAu
souu Amaconuom no huuomoum ou confide

ocwuoouucoonsm >u0mazm
taco moxou am noon coauocdsquonao

mucoSOouoo one nucouucoo
:« ocona>ou Honouoawcn uo nonoooum

.munmwu manmuoczo uo Hoccomuom .moooo
.Amucoexmm unououca .mocoo«>«o ..u.o.
Hoaococau ”Boooouu uowucouu mo mmoq

.ouo .mflnmuocao

oouocm SHHMUOH .noaoHHoa ucoe>onEo
.aoqunauouoouoso uosooua .mouonu
uoxnofi uncowuoHuumou Hoseauouomo

.ououomo 00 Eoooouu uo mmoH
"sewuomcomeoo nu“: newuowumoumxm

acquomcomsou woo
Icuw3 muommm no mmoa acofiumomaucou

cowaaonou

«unwound uuommsn

ou no uowaucoo oosuo
oucw uouco ou mafiaaat
mucoscuo>oo :uaouom

0mm 0:» no £09m
.mowocoqo Hmucoacuo>oo
no mucoacuo>oo coaouom

.ouo .moauwuocaa
.uuconmom .muoxuOI
.nucoosum “unsouo one

showed canoo oouwcomuocoz

“Suucsoo uo oodnuso
no canvas Bonn ocaxuoz

nucoao>oa oaawuoov 2.2qm:

cowuooa<v masoum ceaum
Iommo >Houcoaowauomcoz

museum coda
:«mommo auoucoaewauom

moaocomo mcwumuomo mua

one Meson ca ucoscuo>oo

oucoocomooca dooauaaom
ocuocomfia one ucooom

mmsoum unocwmsn
HonoH no mumuouca oounob

noouomao one amends Howoom

AEmansEEoo .amaaoaoom .Bmaao:0aumcv
mownmomoHazm auowuwaom ocwuodsoo

 

woodwouomo unocwmom

ooumuocoo on

HocoeuocuoucH co mucosa—CH ' coo xmfim Hoowugom

to mouse "muumuum swam Hobgugaoa

50%;: noaouaa mmsouu

t

xnwm ”coaufiaoa no manusom

 

xuo3o5duh Hooumoucou t “Roam Hufiuwaom

.v mqm<fi

51

CONCEPTUAL FRAMEWORK

In a sense this study tries to determine the
variance explained by location-specific factors which are
deemed essential in determining the mode of operation in
a foreign country-market. My conceptual framework builds
from the theoretical framework and literature review in
the area of political risk analysis. It is partial, not
complete since it omits firm/industry specific factors.

The conclusions of the survey studies about the
behavior of international executives are namely:
1) country-market entry decisions are not
based on systematic evaluation of
political risk,
2) executives rely on sources internal to

the firm, and external sources such as

the general news media.

Cracco (1972) has indicated that there's a difference in
the perception of headquarters executives and subsidiary
executives.

Based on these reports the executive forms a
perception about each country-market and a decision as to
the mode of entry is made in consideration of other
factors such as market size and market growth rate. If
the export mode is chosen, the firm then organizes itself
for at least minimum control of marketing activities.
The opposite analysis applies for FDI. In host nation

markets where political risk is low, market size is high

and market growth rate is high, investors can be expected

52

to commit as much resources as possible to maximize their
presence in that market. The converse is true for
exports. Exporting should be attracted to markets with
high risk, low market size, low nmrket growth rate. The
framework in Figure 8 illustrate the sources of
information used, decision variables and strategic
action. It was diagramed based on the theoretical and

empirical studies discussed earlier.

NOTES

1. See Cavusgil, S.T. and J.R. Nevin (1980), "A
conceptualization of the initial involvement in
International Marketing", Theoretical Developments
in Marketing, C.W. Lamb and P.M. Dunne, eds.,
(Chicago: American Marketing Association), 68-71, for
a summary of the studies and their findings.

2. See Glossary of terms (p.226) for definitions of these
terms.

3. All references will be listed in LIST OF REFERENCES on
page 257. Only the authors, & date of publication and
page numbers (if quotation) will be used in the text
to avoid excessive repetition of common
bibliographical information necessitated by extensive
basic referencing of basic documents in separate
chapter. Chapter footnotes, in turn, will be confined
to explanatory documentation rather than bibliographic
references.

4. Goodnow and Hansz (1980) use the term environmental
attractiveness or hot to mean country-markets that
have a high degree of political stability, ample
market opportunities, high level of economic
development and performance, cultural unity, limited
legal barriers, low physiological barriers etc.
Environmentally unattractive or cold countries have
high political instability, few market opportunities,
low economic development, cultural disunity, numerous
legal barriers, high physiological barriers etc.

53

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

SOURCES or DECISION ACTION
INFORMATION VARIABLES EXPORT IF'
INTERNAL E: i: EggH
ex
Past 1 MGR IS LOW
IDGERUEKE
IUXETHJJSRHHFFCFACKXB 1
Perception of PoliticaI
Risk
EXTEENAL —->—Market Size & GrOWth 1
w Rate INVEST IF:
MAGAZIEHSPZEEESRS Competition PR IS LOW
Cost MS IS HIGH
Other Environmental MGR IS HIGH
Factors

 

 

 

 

 

 

 

FIRM SPECIFIC FACTORS

 

Firm size

R&D Intensity
Advertising Intensity
Mgt. Percpetions

 

 

 

Figure 8.--Conceptual Framework of the Effect of
Political Risk on Modes of Operations.

54

5. Survev of Current Business,U.S. Department of Commerce
(Washington D.C., U.S. Government Printing Office,
1980.

6. Ibid p. 20

7. Nehrt, L.C. (1980), The Political Environment for
Foreign Investment,(New York: Praeger Publishers.)

8. Studies using survey methods have concluded that
political risk is a major consideration in foreign
market investment, while studies using cross-country
sample conclude that political risk does not influence
foreign investment.

9. See 'Glossary of terms', (p. 226) for explanation.

10. The study of theory in international trade usually
starts with issues such as classical, neoclassical and
neofactor to explain why nations trade. Theorists
interested in firm behavior have subsequently
concentrated on exporting first, then licensing, then
joint ventures etc. in that order.

11. See Dunning,J.H. (1981), International Production and

the Multinational Enterprise, (London: george Allen &
Irwin)

12. This framework was originally published in "Country

Risk Analysis for Medium Lending, by Antoine W. Van
Agtmael.

CHAPTER II

LITERATURE REVIEW

A difficult task facing a researcher is how to
organize the conceptual and empirical studies which have
made positive contributions in his area of interest. In
this case however there has been paucity of research on
the relationship in question. While many studies have
dealt with the issue of political risk and others have
focused on each of the modes of operation, few studies
have tried to relate the two. In fact, only one

empirical study has focused on political risk and mode of

involvement. It is therefore necessary to explain the
procedure to be used to organize the 'outside'
literature.

Initially, the studies belonging to the incremental
school will be reviewed. Then, those studies that view
modes of operations as strategies will be reviewed.
Within each school of thought, an attempt has been made
to classify studies according to the sampling frame used
i.e. cross-country and cross-firm. Cross-country studies
have typically used secondary sources and macro measures
of the variables studied. Their sampling frame is usually

countries, hence the name cross-country. Secondly, the

55

56

cross—firm studies will be reviewed. Cross-firm studies
typically use survey methods (primary data) to interview
a sample of firms. They can be defined as micro level
studies. So the typology is divided into "SCHOOL OF
THOUGHT-SAMPLING FRAME USED." The schema for the

literature review is as in Figure 9.

The basic objectives of this chapter are three-fold:
1) To show how research has progressed,

2) To identify basic ideas and the state of
knowledge,

3) To show where concepts, variables and

propositions to be measured are coming

from.
The author seeks to be selective rather than exhaustive.
Therefore, in addition to developing a taxonomy of
studies some criteria has to be used to eliminate non-
essential studies. The criteria to be used are as
follows:

1) Only empirical studies will be included.
Empirical studies are those that rely on or are based
solely on experiment and observation rather than pure
theory or speculation. Also these studies must have
significant findings.

2) Major studies that have considered some

aspect of political risk; for example, political

instability, will be referred to although some of the

57

 

moflosum
Euflmlmmouu

 

 

 

 

 

 

AOOmUm
UHOflBdMBm

 

.Bofl>om ousvmuwuflq mo HmoHOI|.m madman

 

mowosum
moaosum hupcooo
Euflmnmmouu Immouo

 

 

 

 

AOOEUm
AdfizmzmmUZH

a

 

 

mmoeammqu-mone<mmmo m4
moo: oza mem oaoHquo

 

58

studies belonging to the incremental school do not
include political risk. Only studies that relate
political risk to investment will be reviewed.

3) Only studies that have investigated one or
more modes of entry with the specific objective of
identifying the strategic use of those modes, will be
reviewed. A number of studies on export marketing
strategies (Bilkey 1978; Czinkota & Johnston 1980; etc.)
are not reviewed here although their conclusions will be
used. However, studies that relate to the choice between
exporting and investment are included whether or not they
deal with. political risk; For instance a :number of
studies by economists investigating determinants of entry
are included.

4) The literature will be presented in
chronological order.

Table 5 shows a listing of the studies reviewed

while Table 6 shows the studies on political risk, other

variables used, and their findings.

INCREMENTAL SCHOOL

Cross-country studies

Kobrin (1976) tried to explain the variation of FDI
in terms of economic, political, social and cultural

1

variables. His hypothesis tested for the amount of

59

TABLE 5.--Summary of Relevant Literature Reviewed by Research
Methodology and School of Thought

 

School of Thought

 

Incremental

Strategic

 

Research Methodolgy

 

Cross Firm

 

Snavely, Weiner, Ulrich
& Enright (1974)

Cavusgil (1976)

Vesper (1979)

Cavusgil & Nevin (1980)
Cavusgil & Nevin (1981)

Axinn (1985)

Buckley & Pearce (1979)
Harrell & O‘Keifer (1981)
Kirkpatrick & Yamin (1981)

Kravis & Lipsey (1982)

 

Cross Country

 

 

Kobrin (1976)
Kobrin (1978)

Ayal & Ziff (1981)

 

Bain (1956)

Mann (1966)

Scaperlanda & Mauer (1969)
Comanor & Wilson (1967)
Miller (1969)

Orr (1974)

Bennett & Green (1972)
Green & Smith (1972)
Horst (1972a)

Horst (1972b)

Rock (1973)

Situmeang (1978)
Baldwin (1978)

Nigh (1985a)

Nigh (1985b)

 

 

60

 

 

 

 

oumu nvsoum umxuofi .ouflm poxuofi .Ham no» Qmmma .nqwz
oumu suzoum poxnmfi .mnam uoxume .th no» mmmma .nmflz
>o>uom moanwflum> ooumHou uoxuma .muommcouu .ooxo Hmaomom mow mbmfl omen<
a poem
mama >uoocooom How .Hmwucmpom uoxnoE .ouflm uoxumz mom puma
mung Haocona
mucofio>ao>cfl uuomxo
mama whoocooom .m.D uoflua .Hom .HMflucouom umxumE .ONfim umxumz no» mhma .cflunox
mama Smoocooom Ham .Hmwucouom poxuoe .mnam uoxumz mow mumH
Eosmcwccou a
comma
coaumuucoocoo muucm no» mumfi
uoxoonuoxOASM
mama Sumocooom Ham «0 >uaafioouamoum no» whoa
.nuflsm a comma
mama Snoocooom :ofluoeum> HmGOAqou
.ucoEQOHo>oo mo Ho>oa .Hmflucouom poxumz .Ham mm» mhmH .cmouo
ocm uuoccom
>o>usm no» mood .uoom
Hmsumoocoo oz moma .com:floom
mama Sumocooow open mmeumu .ONHm uoxuoz oz momH .Hosmz
w ocmauommom
>m>nsm Ham mow mama .Rmmm
Hmzuomocoo Ham mow momH .socmMmae
>m>usm +Hom coma .acoumna
>m>nsm +Hom mama .moHox
>o>usm +Ham homfl .QOflz
cmwmoo noumomom oouoowmcoo moafiofiumfi MOnmz Txmfim oumo ocm MOSHE

 

 

 

i

.mocaocwm Hflone ocm oonoowmcou moaomflum> Home: mcwsogm mmflosum xmfim HMOfiuflHomnHom mo mumfifismlll.o wands

61

variance explained by market size and potential and for a
linear relationship between FDI and political event data.
His first conclusion was that a reasonable percentage of
the variance in the flow of manufacturing FDI is
explained by market related variables. A second
conclusion was that no significant relationship is
apparent between flows of FDI and any of the indicators
of political instability. The author used previous
export involvement as a measure of market familiarity to
explain an incremental 10% of variance for all countries
and 11% for LDCs only.

Kobrin's study has limitations. As the author
admitted, "there are limits on.IhOW’ one can. take the
results: any conclusions must be viewed as
tentative"(p.37), due to the cross sectional design used
to investigate what is obviously a longitudinal
phenomena. Kobrin's article is suggestive that exports
is a 'step' in the ultimate decision to invest in a
foreign country. He therefore would argue that modes of
operation are an incremental process, and ultimate
success with export marketing along with market related
variables, not. political risk, determines the foreign
direct investment decision. His findings with respect to
political risk agree with that of Bennett & Green

(1972).2

62

Kobrin (1978)3 looked at a different angle of
political risk. This time he was interested in the
relationship between political events and the investment
climate. Prior to his study, other research had equated
perceived instability’ with. a poor investment climate.
Kobrin used statistical analysis of forty-eight
countries, all of whom had received FDI of at least one
million dollars as of year end 1967.

His results showed market size, and growth, previous
export involvement and the regional dummies for Latin
America and Europe are significant at the .05 level with
signs as expected.

Kobrin sought to show the relationship between
political instability and political risk. ID1 his model
he only focused on anti-regime violence which was fOund
to be significant. In his summary he concluded that the
relationship between conflict and flows of manufacturing
FDI is complex and indirect. It depends on both the
nature of the conflict and the socio-economic conditions
under which it occurs. Kobrin could have been influenced
by' Robockls statement that "political instability
depending on how it is defined is a separate although
related phenomena from political risk."(Robock, 1971 p.8)
The author cites examples of political instability in
countries that did not necessarily result in increased

political risk. He also notes that political risk is not

"a homogeneous phenomena: vulnerability is clearly
industry, firm and even project specific." (p. 114)
However the issue really is how correlated are
political instability and political risk rather than the
use of isolated examples to support a position. Thus
Brewer (1984) tested for and showed a fairly strong
relationship between political events and policy changes.
This dissertation deals with political risk (as
perceived by investors) and its effect on modes of
operations decisions. Kobrin however admits that
managers equate political instability and political risk.
(Kobrin 1980, p. 74)
The econometrics studies 'which. belong' to the
incremental school are limited. The incremental-cross

firm studies will be discussed next.

Cross-firm studies

As mentioned earlier, cross-firm studies have used
survey method (questionnaires or personal interviews) to
investigate :modes of operation. Their conclusion. or
position is usually that modes of entry or operations
are incremental rather than dynamic alternatives. What

follows is further explanation of their contributions.

64

Export Marketing Studies.

Initial export studies (prior to 1978) have tried to
discover types of firms engaged in exports and why. More
recent research has focused on the process of developing
an exporting strategy. Specifically current research has
sought to answer the following questions:

1) How did a firm arrive at the decision
to export?

2) Having decided on export as a strategy,
how did a firm implement its decision?

3) What factors caused a firm to move from
non-exporting to the initial stages of
exporting, and beyond?

4) Can a theory of exporting be developed?

Two studies [Snavely, Weiner, Ulrich & Enright
(1964) and Cevusgil (1976)]4 have identified some
distinctive characteristics of currently exporting firms.
These characteristics are:
1) Managers should have favorable expectations
regarding the effect of that activity on
firms growth, market development and

profits,

2) Managers who have systematically planned
and executed the export activity,

3) Firms have annual sales of $1,000,000 or more,
4) Firm has unique product or distinctive
competitive advantage and have sole rights
or patents.

These studies have also identified two key

characteristics that affect a firm's involvement in

65

export; managers characteristics and firm characteristics
(size and unique products). Home country specific
factors usually have not been included as explanatory
variables.

Other studies have added richness to the model by
considering change agents. Change agents help stimulate
export development, but do not necessarily result in the
firms adoption of exporting. Change agents can be
external or internal. External change stimuli to the

decision whether to export include banks, industrial

associations, government agencies, and other firms.
(Johanson & Weidersheim 1975)5 Vesper (1979)6 has
identified other external change agents. These are

unsolicited overseas orders, attractive foreign markets
and availability of patents.

Internal change agents can be classified into four
areas according to Cavusgil & Nevin (1981):?

1) differential firms advantages,

2) strength of managerial aspirations for the
firm's growth, profitability and market
development,

3) management decisions,

4) organizational commitment.

Cavusgil and Nevin's model are diagramed in Figure 10.

In trying to identify the key internal change agent
Bilkey & Tesar (1979)8 have attributed this to a key

decision-maker within the organization.

.afiaueflfl ".HSmH sumsunmm.
soummmmm.mmwuoxuoz mo Hmcusoo =.:O«ummflumm>cH HMOHHHQEN c< "H0fi>msmm
mcflpoxuwz uuomxm mo mucocaEHmpoo Hmcuoch: .:H>mz .o w Hamms>mu .m "mumaom

 

.H0fl>mnmm uuomxm mo mucmcHEuovmn HmcumucHil.oa musmwm

 

 

mA¢OD mmmZHmDm ZO mA<

 

UZHBmOmxm m0 mBUMhhm mmMZHmDm mom mZOH
MEB UZHZMMUZOU Idemmd fldemwd

.802 m0 mZOHB<BummNm

a

 

 

 

 

mOH><mmm
UZHBmmm<S
Emomxm

 

 

 

 

- LT

 

 

 

OZHBMMmmz BmOmNm OB mm0¢82<>afi

IBZMEBHZSOU ho AH>HQ ZMHh AflHBmemhhHQ

 

 

 

 

mmqdem¢> Bzmazmmmo mmqdem<> 02Hzm>mmBZH mmqmfiHm4> QZDOfiUMUdm

67

Cavusgil & Nevin (1980)11 have a 'stage' process of
export marketing involvement. The five stages are
domestic marketing stage, pre-export stage, experimental
involvement stage, active involvement stage, and
committed involvement stage.

The export studies represent a specialized area of
study but provide some insights from which implications
can be drawn. They highlight significant factors and
behaviors. Exports is based on, a comparative product
advantage as economists claim. But involvement in
foreign markets is a cautious process. Secondly,
managerial perceptions and motivations are important. As
shown in the FDI studies, these factors also affect
investment decisions. Export studies would argue that
the activity could be a process not a strategy.

The international product life cycle theory can be
classified to this school. The diagrammatic relationship
of the international product life cycle theory (IPLC) is
shown in Figure 11. According to the hypothesized cycle,
in phase one, U.S. exports dominate the world market
while in the next three phases, producers from other
developed countries become increasingly competitive,
first in their own markets, then in the third country
markets, and finally in the U.S. market. The initial U.S.
strength derives from market size, ready acceptance of

innovation, R & D resources, well developed marketing

68

 

< ”oaozo OMHA uooooum HmcofluocuoucH:

UD

aAammHmHv

m:..

004

.m.D oou Mo was» ooHOHHmumm mocoEuOmHmm unomxo m.HoonH omBonm measmou m..am>¢

 

..HmmH flame. me mahumxumz
mo Hmcuooo :.mcofluoOHHaEH >OHHom uooooum ocm ucofimmwmmmom

.Hmsm .H soot omuamam

"mBOZ«

«WUmDOm

.coflumz umBOaaom o mo uCHOQBOH> on» Eoum UAmHII.HH ousmflm

 

 

 

 

""L

 

>H
mm<mm

 

 

HHH
mmflmm

 

HH
mmdmm

 

mmfimm

 

HSLHOdWI IHN

HHIHOdXS LEN

69

information systems. Later shifts are based on factor
endowments and costs.

Ayal (1981)12 tested the IPLC theory from the
viewpoint of a follower nation. Ayal demonstrated from
his findings that Israeli exports were highest in phases
one and two of the cycle. He used data at the industry
level and for 2 cross-sections. (1971 and 1975) His
findings was that best overall performance was achieved
by Israeli industries in the early stages of IPLC while
the worst performance was shown by industries in the
middle stage. His study challenges the IPLC theory. IPLC
implies that a country like Israel can achieve export
success only after the U.S. initial relative advantage
has eroded. His data showed that frequently successful
Israeli export coincided. with successful U.S. export.
Furthermore, the data showed that narrow range
specialization is an important determinant of such
success.

The relevance of the IPLC implication to this study
is diminished since it does not state whether domestic
production (in stage two) is by foreign firms or

indigenous firms.

70

Axinn (1985)9 in a study conceptualized within the
framework of innovation. theory considered. two factors
that were expected to influence export involvement. Her
findings showed that the combination of perceived
innovation .attributes and. adopter characteristics Ibest
explained export involvement.

A firms success or failure in initial export
involvement is key to continued export marketing
development. Managerial expectations plays a key role in
export success or failure.

The choice. of an. export strategy (direct or
indirect) is dependent upon firm size and level of
experience. On the choice of export destinations, firms
are known to export to psychologically close countries.

Perceived obstacles of exporting include
insufficient financial resources, insufficient knowledge
regarding selling opportunities, and lack of foreign
market connection.

Finally, several concepts have been postulated for
explaining export marketing behavior. Export marketing,
first of all, has been conceptualized as a gradual and

10 Secondly, the adoption of export

sequential process.
has been conceptualized as an innovative strategy.
(Axinn, 1985) Initial involvement. of a firm. in the
process of developing exports is similar to the adoption

of innovation.

71

STRATEGIC SCHOOL
Cross-countrv,§tudies

The cross-country (macro) studies can be rightly
traced back to the trade theorists. Several studies by
economists have looked at the determinants of entry.
Entry, however, is not defined as a choice between modes
of operation, rather it represents entry into the foreign
market through establishment of production facilities
only. These studies have been concerned with independent
variables such as economies of scale and barriers to
entry and have sometimes used indirect measures for the
entry variable.

Three studies illustrate this point. Bain (1956)13
examined. twenty (mostly four-digit) United. States
manufacturing industries and concluded that the most
significant barriers to entry were product
differentiation, economies of scale in plant or firm, and
control of patents or scarce resources respectively.

Mann (1966)14 used thirty United States
manufacturing industries to investigate entry
determinants but did not examine the relative importance
of the various barriers to entry.
Comanor and Wilson (1967)15 regressed profit rate
rather than entry on structural characteristics

considered to be barriers to entry. These studies try

to determine the point at which entry defined as the

72

number of firms or profit rate will stop. For instance
Orr (1974)16 states "entry continues until the industry
profit rate is driven to the point where the entrant's
expected rate of return on capital is equal to the
opportunity cost of the capital." (p.59) Orr concluded
that "capital requirements, advertising intensity and
high concentration are significant barriers to
entry."(p.65) Research and development intensity and
risk are modest barriers to entry, while past profit
rates and past industry growth rate had a positive but
weak impact on entry.

Bennett & Green's (1972)17 objective was to
determine the extent to which countries' political
instability influences allocation. of foreign :marketing
investments. Their study is innovative in two ways.
First, it parted ways with the conclusions of previous
studies on political risk (mostly surveys) in concluding
that political instability was not a determinant of
direct marketing investments. Secondly, it was the first
study to use an econometric approach to investigate the
relationship between political risk and foreign direct
investment. Prior studies have used surveys (cross-
sectional) to investigate what apparently is a
longitudinal phenomena. So this piece of research is
significant for its' innovative methodology and its'

surprising (at that time) conclusions. This kind of non-

73

conventional conclusion Zaltman, Heffring and IeMasters
(1981) called a 'generative theory'.

The authors used Feierbend and Feierbend political
risk index to measure political instability which they
define as "the major type and number of politically
relevant, aggressive behaviors occurring within a nation
over a stipulated time period."(p.183) The measure of
FDI used was mean per capita marketing investment per
country in 1965.

Bennett & Green (1972) tested hypothesis relating to
marketing investment throughout the world and the level
of national political instability. A negative
relationship was tested for. The authors also tested
hypothesis relating to level of development and regional
preference as controlling or intervening variables.

The authors concluded that political instability did
not affect the overall allocation of U.S. foreign direct
investment decisions throughout the world. No
significant relationship between the variables for either
the developed or less developed nations was found. The
level of political instability was not a significant
factor in the allocation of marketing investment, even
when per capita GNP and stage of economic development
were both controlled for. For the Latin American nations
whose special relationship with the U.S. has been the

reason for some authors postulating a 'regional

74

preference' on the part of U.S. firms, no relationship
existed between per capita investment and political
instability.

Bennett & Green (1972) study is important because of
its non-conventional conclusions; political instability
does not affect allocation of FDI. Prior conclusions
pointed.tx> 'the significance of political instability in
investment decisions. The index used seems appropriate.
A criticism is the relatively small number of
observations. Another shortcoming is the lack of
consideration of a lag period to test the sensitivity of
the analysis. The assumption was made that adjustments
to country political risk factors are quick (1-12
months). Investment is a high involvement decision and
requires time to plan and organize. At the very least a
one year lag should exist between political instability
in a country and investment in that country. Finally
this study neglects the longitudinal nature of the
decisions. Only 1965 data was used to test the
relationship.

Bennett & Green (1972) served as a springboard for
Green & Smith (1972) . In the latter study, an attempt
was made to derive the relationship which exists between
a nation's level of political instability and the returns
the multinationals expect to receive on their

investments.

75

Again the Feirebend and Feirebend political risk
index was used. The 1965 investment data from Survey of
Current Business was used. Four aspects of U.S. foreign
direct investment was used for each nation. Cross-
sectional data for twenty-three nations was used.

The authors hypothesized that a positive
relationship exists between political instability and
foreign investment profitability; The higher the
political instability of a nation, the higher the rate of
return on the foreign investment located in that nation.

The hypothesis was accepted for investment behavior
in mining and petroleum, but rejected for manufacturing
data. Higher profits are derived from investments in
politically stable countries or stated differently
manufacturing investment was less profitable in
politically unstable countries.

The results for manufacturing is not surprising
since the concept of political risk and its results
suggests substantial losses due to instability and
expropriation. The results can be explained due to
faster reaction time for manufacturing investment
compared to mining and petroleum investment. It may
explain why U.S. manufacturing firms are less apt to
invest in less developed countries.

Horst (1972b)18 was concerned with the differences

between investors and non-investors within the same

76

industry. He commented that "many of the determinants of
the decision to invest abroad (e.g. tariff, transport
cost, dependence on natural resource) are the same for
any two firms in the same industry."(p.259) His
conclusions were that larger firms were more likely to
invest abroad than smaller firms. Once inter-industry
differences are washed out, the only difference of any
separate significance is firm size. He also found that
industries in which the economies of size are important
tend to have fewer foreign investors controlling a larger
share of the foreign market.

Horst made use of 187 multinational firms making up
the Harvard Business School sample at that time.
Graphically the relationship is illustrated as in Figure
12. Graph 1 shows partial relationships between firm
(holding industry influences constant) and the
probability that it would be multinational and the
probability that it would have a Canadian subsidiary.

The later observation lead the author to suggest that a
Canadian subsidiary may be a stepping stone to investing
in other countries. Graph 2 shows a plot of the relative
heights of the two functions in Graph 1 showing how the
conditional probability of being multinational.

While holding firm size constant, Horst concludes
that inter-industry differences - research and

development level, resource concentration and estimate of

77

 

   

 

 

 

 

 

 

P
1.0.
'9‘ Probability of being a [—-
Canadian investor IT '
.8-
.7.
.6‘

5. Probability of being
' a multinational
.4 ‘ investor.

.3‘
.2‘
.11
v U v I V GRAPH 1
50 125 250 500 1000
Firm Sales (106 Dollars)
P
1.0 .
.9 q
- . Combined probability of
.7 being a multinational
.
O 6 ‘
.5 ~
0 4 1
.3 .
.2 4
.1 d
I I l I GRAPH 2

50 125 250 500 1,000
Figure 12.--Investment as a Function of Firm Size.

SOURCE: T. Horst, "Firm and Industry Determinants of
the Decision to Invest Abroad: An Empirical
Study," Review of Economics and Statistics 54,
279-294.

 

78

minimal efficient plant size as further variables
influencing investment.

Horsts first conclusion agrees with Mason, Miller &
Weigels (1975) conceptual development of the effect of
firm size. However the later are more explicit in saying
that firm size affects the firm's chosen mode of
operations in foreign markets. Smaller firms will tend
to export while larger firms will invest when all things
are equal. Firm size is not included in this
dissertation because only location specific factors are
being investigated. However a composite model of modes
of operation should include firm specific and location
specific factors.

The Goodnow & Hansz (1972)19 study is the only one
that specifically looked at the relationship between
modes of entry and environmental factors [political
stability, market opportunity, economic development and
performance, cultural unity, legal barriers,
physiographic barriers, geocultural distance]. The
hypothesis tested was originally developed by Litvak and
Banting and read "a firm will tend to pursue an entry
strategy involving greater control over and greater
investment in marketing channel activities as the
country's environment becomes 'hotter'."(p.33) The
authors used multiple macro indicators of the environment

to classify nation-states as either' hot, moderate or

79

cold. A cross-sectional survey of managers was used to
find out their entry mode for their main product in their
main foreign market. Goodnow et. al. found that U.S.
firms used exporting strategies most often in
environmentally 'cold' countries, and invested in
environmentally 'hot' countries. While their study is
insightful it is riddled with methodological problems.

The 'entry' variable was measured by a self-
administered questionnaire which was sent to directors of
international divisions requesting them to identify for
each country in whiCh the firm's MAJOR product is sold,
the type of market entry strategy used by the firm. The
data collected is erroneous partly because it eliminates
data for 'minor' products.

Conclusions were based on reporting percentages
only. No statistical relationship was sought for the
data and no significance tests were done. Finally the
measure of political risk was based on estimates by the
U.S. Department of Commerce country specialists. This
problem while not peculiar to this study is serious. How
can we correlate the actions of managers who make
decisions about modes of operation with perceptions of
the political risk by country specialists? A better
analysis would be to have these managers evaluate these

countries based on their perceptions of their riskiness.

80

This study will improves on the Goodnow and Hansz
study by using macro aggregate data for exports and FDI.
Also explicit criteria will be used in selecting the
periods and nation-states in accordance with the
hypothesis.

Rock (1973)20 was interested in investment
guarantees as a spurious variable affecting the
relationship between political risk and FDI flows. It
had been postulated that developing countries offer
guarantees on FDI against political risks. Omission of
those guarantees in some FDI studies have resulted in
surprising conclusions. Rock considered two periods for
analyzing the effect on political instability in host
developing countries on U.S. FDI in manufacturing
industries. In the first period when an investment
guarantee did not exist, U.S. FDI was found to be
negatively correlated with political instability while in
the second period (when a guarantee was introduced) the
correlation disappeared.

Situmeang (1978)21 concluded that political
instability was statistically unrelated to the flow of
FDI in all sectors. (non-extractive, manufacturing, high
technology and low technology industries)

Baldwin (1979)22 tried to determine whether trade

and direct investment can be explained by the same

81

variables - capital and labor content. His reason and

conclusion was as follows:

"Because of the capital intensity of natural
resource imports, U.S. import trade is capital
intensive relative to export trade, taking into
account all industries. When only trade in
manufactured goods is concerned this result
does not hold. On the other hand, U.S. firms
investing abroad in manufacturing apparently
are attracted by the relative abundance of
unskilled labor and bias their activities
toward labor intensive industries. However
the heavy direct investment in capital
intensive - natural resource industries, such
as petroleum and mining that is designed to
facilitate imports of these products prevents
the sign of the capital/labor coefficient in
all-industry direct investment equation from
also being significantly negative."(p.46)

The cross-country, strategic school studies have
clearly shown that certain factors would affect a firms
choice of operations or entry mode. The authors seem to
have focused on those factors that interest them or that
are most relevant to their disciplines. There doesn't
seem to be any attempt to build a body of knOwledge. The
authors have looked at location specific and firm
specific factors. It is not clear from the studies if
one or the other influences the mode of operations entry
decision. The major contribution of the school is in
identifying factors relevant to the mode of operations
decision.

The most recent study on political risk and foreign

direct investment was Nigh (1985a) and Nigh (1985b).23

82

Although Nigh examined the relationship between political
events and manufacturing direct foreign investment (MDFI)
decisions, he was interested in political risk. The
study pooled time series (21 years) and cross sectional
(24 countries) data. The Conflict and Peace Data Bank
(COPDAB) was used to measure political events. This
index records events that are deemed political for 148
countries obtained from seventy news and media sources.
Political events are classified as either inter-nation
conflictive, inter-national cooperative, intra-nation
conflictive or intra-nation cooperative. Nigh felt that
separation of political events as conflictive and
cooperative would reveal the true effect of political
risk on flows of U.S. FDI. Prior studies had not
separated the two, and one reason for the surprising
findings of the econometrics studies has been the nature
of the political risk phenomena. Robock (1971)24
suggested that political instability does not always
enhance political risk for FDI while Thunell (1977)25 has
suggested that the degree of political risk emanating
from political instability in a country is likely to vary
for FDI of different origins and in different industries.

Nigh tested the relationship for less developed
countries and developed countries separately. For less
developed countries (for one-year lag model), his

conclusion was that "political events affects the

83

manufacturing direct foreign investment of U.S.
multinational corporations when less developed countries
are being concerned. Both conflictive and cooperative
inter-nation. political events, as well as conflictive
intra-nation political events, appear to have an affect
on MDFI in less developed countries."(p.9)

For developed countries regression analysis for the
one-year lag models indicated that the conflictive and
cooperative inter-nation (but not intra-nation) political
events affect the MDFI decisions of U.S. multinational
corporations when developed countries are being
considered. 1&3 also found that market size affects the
MDFI decisions.

Nigh's findings agree with the conclusions of the
survey studies and are contrary to the findings of
Bennett & Green (1972), Green & Cunningham (1975) and
Kobrin (1976). His methodology is superior since he uses
a cross-sectional, time series design. The use of the
COPDAB is innovative.

Nigh (1985b) followed up his earlier study be
investigating the effect of political events on German
MDFI flows. He was influenced by the criticism of
Thunell that the relationship varies for MDFI from
different origins. His conclusion was similar to that of

the U.S. MDFI flows and thus validated that study.

84

The role of tariffs in the export, foreign direct
investment substitution issue is not clear-cut. Two
studies highlight this contradiction. Horst (1972a)26
examined the relationship between U.S. exports to the
Canadian. market and. the sales of U.S. affiliates in
Canada. He concluded that the most significant factor
influencing the decision whether or not to invest in
Canada was the height of Canadian tariffs. As the tariff
increased, U.S. subsidiary production in Canada
increased. In an earlier study Scaperlanda and Mauer
(1969)27 arrived at diametrically opposing results. The
authors used least squares regression to test three
hypothesis relating to the size of market, economic
growth and tariff discrimination. Data on U.S. direct
investment in the European Economic Community for the
1952-1966 period was used. The authors were interested
in explaining the rapid growth of U.S. investment in
Europe during the 1959-1966 period. Empirical tests led
to conclusions that only the market size hypothesis was
supported statistically. Tariff discrimination did not

explain the switch from exports to production.

Cross-firm studies
Cross-firm studies typically use survey methods to
determine factors influencing choice of mode of entry or

operations. These factors determine all modes used not

85

simply the export mode. Buckley and Pearce (1979)28
focused on sourcing policy. Their main thesis was that a
firms sourcing policy is a function of firm and industry
characteristics. Their hypothesis was to examine the
determinants of sourcing policy and to show that a firms
international imports are strongly related to its
research and development intensity.

Buckley' and. Pearce's findings 'were 'three-fold.
First, even among the largest firms, size of the firm is
a major influence on the type of sourcing policy adopted:
ceteris paribus, the larger the firm, the more likely it
is to service its foreign market by production in those
markets. Second, there are significant variations in the
sourcing policy of firms when grouped not only according
to industry but also nationality.

Thirdly, the size of internal flows from parent to
subsidiary is very clearly connected with the degree of
research intensity of the industry in which the firm is
competing . The more research intensive the industry,
the more likely are internal transactions to be of
importance. The conclusions of this study are not
surprising. The importance of firm size has been
established in the literature.

MaClayton, Smith and Hair (1980)29 used health care
firms to test hypothesis relating to foreign market

entry. The specific objectives were two-fold: 1) to

86

identify those foreign market environmental factors which
companies within the health care products industry
perceive as relevant in evaluating foreign markets and,
2) to investigate the relationship between these factors
and the internal characteristics of the companies. They
concluded that firms in the U.S. based health care
products industry seemed to evaluate foreign markets on
the basis of the following home country specific
dimensions: market and marketing opportunity, legal
barriers and their economic objectives, cultural unity
and physiographic barriers, political stability, level of
economic development and performance of the country. An
interesting conclusion was that no statistically
significant relationship existed between the companies
experience in overseas business, number of employees,
research and development expenditure, total sales nor
total investment and home country specific dimensions.
This finding is fascinating because it goes against
theoretical speculations. Although the mode of
operations was not specifically addressed, the authors
recognize that "it makes a difference if a given market
is entered through exporting from the parent country, or
one of its subsidiaries abroad, through a direct

investment in a manufacturing facility or through

87

licensing arrangements. Obviously direct investment
represents a greater risk and requires more analysis."
(P- 40)

MaClayton, Smith & Hair (1980) have speculated that
erroneous reporting of data for sales, investment and
research and development expenditures may be the cause of
findings. However one significant implication is that it
destroys the myth of companies seeking to familiarize
themselves with the home country market thereafter opting
for exporting or licensing in preference to full scale
investment. Further research on this important
relationship is necessary as a strong relationship has
been ‘theorizeda This finding, although inconclusive,
lends credence that a firm might enter directly into a
foreign market via foreign direct investment if the
factors are right.

Harrell & Kiefer (1981)30 used the product portfolio
concept to postulate strategies U.S. firms should pursue
in international markets. Using the portfolio matrix
developed by George Steiner, et al., the authors
suggested that harvest/divest/combine/license strategies
should be used in product markets in which the firm has
low competitive strength and where country attractiveness
is low meaning modest in market size and growth rate.

Since the underlying concept behind the competitive

strength variable is the experience curve, that variable

88

can be interpreted as market share. The competitive
strength variable was measured as a function of market
share, product fit, contribution margin and market
support. Therefore the authors suggest that the mode of
operations will be a function of market attractiveness
and firm size. Thus a multinational such as Ford Tractor
could invest in Kenya or Pakistan where country
attractiveness is high and where it [Ford] has a high
competitive strength while a smaller firm such as Snell
Environmental Group, Lansing, will be content with a
joint ‘venture strategy in those markets. Figure 13
illustrates the thrust of the relationship between mode
of operation, competitive strength and country’ market
attractiveness.

This article is insightful in its application of
portfolio analysis to the development of international
marketing strategy.

Kirkpatrick & Yamin (1981)31 were concerned with the
determinants of export subsidiary formation by U.S.
transnationals in developing countries. It has been
suggested by Kobrin et. al. that political risk may be an
industry specific phenomena, hence the use of U.S.
manufacturing FDI rather than mining, petroleum or
finance. This empirical analysis lends support to that

theory in showing that U.S. transnational's propensity to

89

 

 

High
Country ”XXII'IIII° , ” .14
Attract-L . - . Li JOINT VENTURE
iveness f .. . DIVEST
I Z-INVEST';
v .1 + + t T ‘l
' + + 4 v T 3
° : -§- 1 .' fiv 'I’ I T
. 'f 1 w t + +-+ w
1» + at 't 1 w ~t
+ + -: 1 1 * ‘1

twt++**"

s+++1~o4+~o

 

 

 

+ +-§+'++-01-14
744**4f+++

+4+1+1++i1
1* 4+ d-id‘id 1

+ 1+ ++-+1++Ht+
4 111++11+i1
4"! 4+11-I-I-I
4 1 41-+ 44-:

 

 

 

‘11 11 4 «'11 4 -4 HérveSt q 4 4
w + 1 + 4 +-t-+ 4 diveFt '** *
combine
1" T * 3 ** " I license I I I
4+ +4 4‘14+-t+*.,,,+« 4
*+--t~-***'~$}.++*1*
Low
High LOW
COMPETITIVE
STRENGTH

Figure l3.--Modes of Operation, Competitive Strength and
Country Attractiveness Grid

SOURCE:

G. D. Harrell & R. Kiefer,

"Multinational

Strategic Market Portfolios," MSU Business

Topics (Winter 1981): 7-15.

 

90

invest in export-oriented manufacturing subsidiaries in
LDCs is a function of industry characteristics.

The authors used information from the Harvard
University Business Data Bank. Two subsets of data - for
all LDCs and for five LDCs was used. The later performed
better than the former (i.e. all independent variables
had the expected sign and significance at the 10% or 5%
level). The authors thus concluded that export
subsidiary formation is a function of value added per
production worker, research and development personnel as
a percentage of total employees, advertising expenditure
as a percentage of sales and expenditure on after sales
service as a percentage of sales.

The focus of this study was on export seeking
rather than market seeking FDI. It would be interesting
to determine if the same relationship holds for market
seeking investment.

Kravis & Lipsey (1982)32 tried to explain location
decisions from the economic consideration implied by the
market scanning hypothesis. They assumed that production
in a host country is a result of interaction of several
sets of factors. One has to do with country and
commodity characteristics while the other is firm
characteristics. In trying to determine who invests
where, the authors were concerned with whether there is

some relation between the characteristics of a country

91

and the type of firm that sets up manufacturing.
Following a tabulation of the number of firms investing
in. certain countries per industry' group, the authors
hypothesized that location of production in a foreign
country is a function of proximity to the U;S., the use
of English and the size of market. These findings
suggest a regular country order for each product/product
line that a firm decides to market abroad, not a regular
mode of entry pattern.

Their results showed large differences in size
between U.S. investors in Canada and U.S. investors in
other country-markets. U.S. firms invest only in Canada
or Canada first.(p. 205) Within each industry low wage
firms tend to have investments in high wage countries,
while U.S. companies invest more frequently in low wage
destinations (p.223), such as developing countries.

Davidson (1980)33 investigated the role of host
country characteristics as determinants of FDI location
patterns. The main variable was corporate experiences.
Prior experience in a host country is found to increase
the firms priority for projects in that country relative
to other investment options. In addition, the experience
level of the firm influences the relative importance of
different country characteristics in determining location

patterns. Inexperienced firms exhibit greater preference

92

for near, similar markets than firms with broader

international operating experience.

CONCLUSION

The literature reviewed shows that many studies have
been interested factors that affect changes in mode of
operations. Few studies have considered political risk
although political risk is a major determinant of the
foreign market involvement decision. This dissertation

addresses that missing link.

NOTES

1. See, Kobrin, S. (1976), "The Environmental
Determinants of Foreign Direct Manufacturing
Investment: An Ex-Post Empirical Analysis", Journal
of International Business Studies, (Fall/Winter),
29-42.

2. See, Bennett, P.D. and R.T. Green (1972), "Political
Instability as a Determinant of Direct Foreign
Investment in Marketing", Journal of Marketing
Research, Vol. 9 (May), 182-186.

3. See, Kobrin, S. (1978), "When Does Political
Instability Result in Increased Investment Risk?",
Columbia Journal of Business, (October).

4. See, Snavely, W.P., P. Weiner, K. Ulrich and E.
Enright, (1964), Export Survey pf the Greater
Hartford Area, Vols. 1 & 2, University of
Connecticut.;. Cavusgil, S.T. (1976),
Organizational Determinants of Firms Export
Behavior: An Empirical Analysis, PhD. dissertation,
The University of Wisconsin, Madison, Wisconsin.

10.

11.

12.

13.

14.

15.

16.

93

See, Johanson & Weidersheim-Paul, (1975), "The
Internationalization of the Firm: Four Swedish Case

Studies", Journal of Management Studies, (October),
305-332.

See, Vesper (1979), "Entrepreneurship in Foreign
Trade", Journal of Small Business Management,
(April), 5-11.

See, Cavusgil & Nevin (1981), "Internal Determinants
of Export Marketing Behavior: An Empirical

Investigation", Journal of Marketing Research,
(February), 114-119.

See, Bilkey & Tesar (1978), "The Export Behavior of
Smaller Sized Wisconsin Manufacturing Firms",
Journal of International Business Studies, (Spring),
93-98.

See, Axinn, C. (1985), "An Examination of Factors
that Influence Export Involvement", Unpublished PhD.
dissertation; Michigan State University.

See, Cavusgil & Nevin (1980), "A Conceptualization of
the Initial Involvement of International Marketing",
Theoretical Developments in Marketin , C.W. Lamb and
P.M. Dunne, eds., Chicago: American Marketing
Association, 68-71.

Cavusgil & Nevin, op. cit.

See, Ayal, Igal (1981), "International Product Life
Cycle: A Reassessment and Product Policy
Implications", Journal of Marketing, Vol. 45,
(Fall).

See, Bain (1956), Barriers to New Competition,
(Cambridge, Mass: Harvard University Press).

See, Mann (1966), "Seller Concentration, Barriers to
Entry and Rates of Return in Thirty Industries 1950-
1960", Review, (August), 296-307.

See, Comanor & Wilson (1967), "Advertising, Marketing
Structure and Performance", Reyiey, (November), 423-
440.

See, Orr, D. (1984), "The Determinants of Entry: A
Study of the Canadian Manufacturing Industries",

Review of Economics and Statistigs, LVI,
(February),.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

94

See, Bennett, P.D. and R.T. Green (1972), "Political
Instability as a Determinant of Direct Foreign
Investment in Marketing", Journal of Marketing
Research, Vol. 9, (May), 182-186.

See, Horst (1972b), "Firm and Industry Determinants
of the Decision to Invest Abroad: An Empirical
Study", Review of Economics and Statistics, 54, 258-
266.

See, Goodnow, J. an J. Hansz (1972), "Environmental
Determinants of Overseas Market Entry Strategies",
Journal of International Business Studies, (Spring),
33-50.

See, Rock, M. (1973), "Cross-Country Analysis of the
Determinants of U.S. Foreign Investment in
Manufacturing Less Developed Countries", unpublished
PhD. dissertation, University of Pittsburgh.

See, Situmeang, B. J. (1978), "The Environmental
Correlates of Foreign Direct Investment with
Reference to Southeast Asia", Unpublished PhD.
dissertation, University of Oregon.

See, Baldwin, R. (1979), "Determinants of Trade and
Foreign Investment: Further Evidence", The Review
of Economigs and Statistics, Vol. 6, Amsterdam, 40-
48.

See, Nigh (1985), "The Effects of Political Events on
U.S. Direct Foreign Investment: A Pooled Time-
Series Cross Sectional Analysis", qurnal of

International Business Studies, Vol. XVI, No. 1
(Spring), 1-17.

See, Robock (1971), "Political Risk: Identification
and Assessment", Columbia Journal of World Business,
(July/August), 6-20.

See, Thunell, L. (1977), Political Risks in

International Business: Investmen; Behavior of

Multinational Corporations, New York: Praeger
Publishers.

See, Horst, T. (1972a), "The Industrial Composition
of U.S. Exports and Subsidiary Sales to the Canadian

Market", American Economic Review, 62, 37-45.

27.

28.

29.

30.

31.

32.

33.

95

See, Scaperlanda, A. and L. Mauer (1969), "The
Determinants of U.S. Direct Investment in the EEC",
The American Economic Reviey, Vol. 59, Evanston,
558-568.

See, Buckley & Pearce (1979), "Overseas Production
and Exporting by the World's Largest Enterprises: A
Study in Sourcing Policy", Journal of International
Business Studies,10, (Spring/Summer),9-20.

See, Maclayton, D.,M. Smith, and J. Hair (1980),
"Determinants of Foreign Market Entry:A multivariate
Analysis of Corporate Behavior" Management

Ipternational Review,20 (3),40-52

See, Harrell & Keifer (1981), "Multinational
Strategic Market Portfolios," MSU Business TOpics,
(Winter),5-15

 

See, Kirkpatrick & Yamin (1981), "The Determinant of
Export Subsidiary Formation by U.S. Transnationals
in Developing Countries", World Develo ment, 9, 373-
382.

The market scanning hypothesis states that MNCs use
their superior knowledge to locate manufacturing
activities in countries that are most advantageous
from the standpoint of market and cost
considerations. See, Kravis & Lipsey (1982), "The
Location of Overseas Production for Export by U.S.
Multinational Firms", Journal 0: Intepnatipnai
Economics, 12, 201-223.

See, Davidson, W.H. (1980), "The Location of Foreign
Direct Investment Activity: Country Characteristics
and Experience Effect", Journai of international
Business Studies, 11 (Fall), 9-22.

CHAPTER I I I
METHODOLOGY

In Chapters one and two, the variables of interest
were discussed. In this chapter, the research design, the
research hypothesis and their operationalization, and
issues of sampling will be discussed.

The research design involves the use of both
secondary and primary data. The hypothesis will be
examined using the secondary data which will be pooled.l
The testing of the hypothesis constitutes the first stage
of the research. The second stage involves validation of
the findings through personal interviews with eight
Michigan corporations involved in foreign markets.

The secondary data consists of pooled cross-
sectional (20 countries) and time-series (11 years) data.
Pooling of data offers the advantages of increased
reliability although the variance is reduced (Wallace,
1972). Bass & ‘Wittink (1975) say' that "the use of
multiple observations gives rise to opportunities through
the gain in degrees of freedom"(p.414) provided pooling
is appropriate. This study is concerned with the role of

political risk in influencing the mode of operations

96

97

decision. In research jargon we are interested in the
amount of variance which is explained by political risk
(and indeed other identified variables) in the
export/investment decision.

Although the independent variable of interest is
political risk, market size and growth rate are included
to avoid biased and inconsistent estimates of the
constant and slope coefficients by violation of
assumptions concerning the disturbance term. (Kelejian &
Oates 1981,p. 233) Since we are interested in dependency
of the variables and all variables are measured as
interval or ratio data, the appropriate technique is
multiple regression. Multiple Regresssion Analysis is a
statistical technique which can be used to analyze the
relationship between a single dependent (criterion)
variable and several independent (predictor) variables.
The objective is to use several independent variables
whose values are known to predict the dependent value the
researcher wishes to know. (Hair et. al.,p. 35) A series
of multiple regressions will be run on the dependent
variables of interest.2 The selection of the sample
(countries and years) will be based on sampling theory.3

Studies using secondary data tend to use pg
available data to test relationships. For example, Nigh
(1985) pools and regresses all available data (countries)

for the years 1954-1975 to test the hypothesis of his

98

study. This will tend to inhibit our ability to make
inferential decisions since the sample is a convenient
one not a probability sample. Therefore in this study the
selection of the sample size, sampling element, time
period, variables and measures will be based on decision
rules or theory in order to isolate the effects
(variance) of the variables under investigation. (Hage,
1975) Table 6 shows research methodologies that have
been used to investigate political risk. The objective of
this research is to explain a phenomena.

The primary data will be used to validate the
results of the secondary analysis. Data will be obtained
through personal interviews with some Michigan firms
involved in foreign markets. Although personal interviews
would allow a very limited number of firms to be studied,
it provides significant insights and maintains a real-
world context for data collection. A mail survey will not
capture the dynamic processes involved in this decision
area. The duration of the interviews were between one
and one-half and two hours and they were conducted in the
months of April and May, 1986. An attempt was made to
structure the interviews such that the same questions
were asked of each individual respondent. Thirty-five
letters of solicitation were initially sent out. Of
these, twenty replies were received (after follow-up

letters were sent). Of the twenty replies, eight were

99

positive while twelve were negative. Reasons cited for
refusal were mostly the length of interview time
requested. The respondents felt it was too long. A few
replies indicated that the month of April was a
particularly busy one for them and could not therefore
make out time for the interview. Appendix J shows a list
of the questions used for the interviews. The
responses/conversations were recorded using tape
recorders in order to ensure that valuable information
was not omitted. The interviewees were aware that the
conversation was being recorded. Later the tapes were
transcribed4 and edited for inclusion in this study.
Appendix I provides data about the firms used for the
interviews. Appendices E, F and G show documents used to
solicit these firms. Appendix H shows the interview
schedule. The factors taken into consideration in
selecting firms in the sample were size (sales and
employees), nature of business, years .of foreign

involvement and commitment to foreign markets.5

RATIONALE FOR RESEARCH DESIGN

Kobrin (1980) has criticized the research in this
area for using cross-sectional techniques: "all the
studies entail major data and methodological problems
ranging from the use of composites indices of instability

to the almost universal use (with one exception) of

100

Hensoumem ..onH .mneummm

 

nOHueEnomnH ..n> .Haenmuez =.meamoHooonuoz noneemem hufiaflneumnH HeowuHHom mo unoEmmemmn
mm. .HmmmHv .eHmmom .3 .O one eHofluonn .o .m mo monfipflnz on» Scum ooumeon ”monnom

 

HabmH .Hmoumoms

 

 

 

 

 

 

 

 

 

noHHOHoenm one nOHueHnEHm .Q
n
Amhma xoeASOHH one “Home mnflumeoeHOm neflmexem .0
nowuowoenm oemen Hoooz Hemneu .U
mHmS ene oemEHImmou .
HommH ..He um . H u . O o
anuemaonnv mnoueOHonw mnfioeeq .m AHBmmmv mneuweoeHOE finmaea .m
nofiumwoonm oomen eueo une>m .n o>wuowoonm o>flu0floonm uneecono mxnw nOHuwnan .n
HommH ..He um numbness
Emwuonuoe one wuHOHnnum
.uneEm>Oz Hewoom .Q
HvsmH .Honmoxmv
nOHunHO>om
one HeHOOm oeoum .0
mnOH ene ax one I o O emne .
.u H m e m H e 2 H O O HmsmH .nomnwnnHues
mnoflueneamxo umHuHHm .Heo nowunao>em one mouHHm .m
IHOO o o>m . eOH H O . e OHOO .
. H n m H .u.H n H u . m m HmomH.maozv,
nOHuenemem OHEOnOom .< Snouenemem Shoueneamxm mnOflunHo>emmwmunemeom .<
. Hmoma Nunenom oH>eQ
OHHwonm noHnenuHm HennounH O eoma nOmnnOb mnefiaenu
mcoHuHcHeme ,msmH .qsmH.sHHan
mnemou oomemneueo uno>m .m SpfiawneumnH HeOHUHHOm mo
wnoH eOH Home ne m e .m
momH Honenm .u .m. HO o H o0:
mnOHanHmoo HeOHHHQEm .n o>Humwnomeo o>HumHHOmeo moaonum emeo Henow>fionH .n
m>Huumnno m>Huownno
noneomem noueemem

 

m>HueuHucena

 

o>HumuHHmno

 

peacoHooonuez noenemem SuflawneumnH HeowuHHomII.m manna

101

cross-sectional techniques to investigate what is
obviously a longitudinal phenomena."(p.74)

The implication is that a design that captures the
dynamic nature of the relationship would be more
appropriate. Kimberly (1976) defines a longitudinal
organizational research as one that "consists of those
techniques, methodologies and activities which permit the
observation, description and/or classification of
organizational phenomena in such a way that processes can
be identified and empirically documented."(p.329) Some of
the advantages of using longitudinal data to study
organizations as suggested by Kimberly (1976) are:

1) Longitudinal research facilitates attempts to
establish causality (i.e. temporal precedence can be
established).

2) Longitudinal research can help minimize the
problems encountered when process is inferred from cross-
sectional data.

3) Longitudinal research facilitates the development
of better models of organizational growth and change.

4) Longitudinal research permits one to take
contextual constraints into consideration in the research
design.

5) Longitudinal research ultimately enhances the
effectiveness. of 'various. strategies for’ organizational

intervention.

102

THEORETICAL DECISION RULES

Certain decision rules have to be made with respect
to the data which is available. Hage (1975) puts it this
way, "since nation-states are complex, they require
special methodological rules for sampling."(p.133) The
suggested implication is that societies should be
selected on the basis of their score on some independent
variable and studied over long periods of time so as to
maximize information gain and minimize cost. (Hage
1975,p.131) The fact that most of the macro variables do
not. have normal distributions can raise a number of
statistical problems regarding inference. It is thus
necessary to follow rules of purposive sampling in order
to approximate an experiment that will build on present
theoretical knowledge.

The decisions necessary to develop an appropriate
design are:

1) What should the number of countries in
the sample be?

2) What countries should be chosen (sampling
element)?

3) What time periods should be chosen?
4) How many variables should be measured?

5) How many indicators for each variable
need to be measured?

103

Determination of Sample Size and Element

The upper limit is the number of countries available
in the political risk index. Since BERI Index6 will be
used. that upper limit is forty-eight (the number of
countries monitored by the service). It is not necessary

to use all forty-eight countries.

Decision Rule 1”. Choose countries that have high
and low scores on the independent ‘variable and. where
possible, find instances where the same country has high
and low scores. (Hage 1975, p. 138) With respect to this
study this means to study countries that have high and
IOW' scores on. the independent ‘variable: of interest -
political risk. The reason is to maximize variation, and
thus information gained and allow for an approximation to

an experimental design.

pecision Rule 2. Choose high and low scores on the
independent variable of alternative explanations and,
where possible, find instances where the same society has
both high and low scores. (Hage 1975, p. 140) This is
essential to test a theory when we can safely eliminate
alternative explanations. Therefore countries with high
and low scores on all our independent variables need to

be chosen.

104

Decision Rule 3. Select societies and time periods
that allow for the control of particular variables, which
do :not necessarily’ represent alternative «explanations,
but which do confound or contaminate the results. (Hage
1975, p. 141) This is a particularly useful rule when we
are not sure what the alternative explanations are.
Examples of these variables are population size and level
of development because they are related to a large number
of other variables in a number of ways that are not
necessarily understood.

Hage (1975) suggests that if there is any conflict
between this decision rule and the first, the third
should take precedence because we always need
experimental manipulation or variation in over
independent variables before we try to eliminate or
control for confounding factors. We therefore need to
take countries that have low and high scores on the
independent variables from the geographic area (e.g.
African nations). Choosing nation-states that are much
alike in terms of economic development or political
structure is counter productive. Thus this third
decision rule implies the use of control variables in
conjunction with the selection of societies that are high

and low on independent variables.

105

Decision Rule 4. When searching for the cause of
societal events, select instances of absence and presence
or relatively high and low scores. (Hage 1975, p. 144)
If the researcher also chooses time periods within the
same societies when events such as revolutions did not
occur; then they have a powerful kind of research design
because we can partially control for various biographical

factors.

Determination of Specific Time Periggs

The selection of specific time periods presents a
problem: however, Moore (1963) has suggested that it may
not. be a problem because "most variables are always
changing either to higher or lower scores."(p.23)

Decision Rule 5. Choose a time period that is one-
half the amount of time needed to represent a meaningful
change in the independent variable. (Hage 1975 ,p.145)
Since variables have their typical change rates,
naturally we would want to measure variables at least as
often as their change rates . The determination of a
'meaningful change' depends on the information being
collected. Such information as cultural trends,
distribution of wealth, and urbanization are the slowest

to change. (Hage 1975,p.146)

106

Decision Rule 6. Choose a duration that is longer
than the causal process we are interested in
studying.(Hage l975,p. 148) Therefore an eleven year
time-span provides an adequate time period for the

phenomena under investigation.

Determination of the Variables

The implication of the first two decision rules is
that we want to include not only the independent
variables from our own theoretical model, but also, the
independent variables that have been isolated by others.
Two variables that have been consistently isolated in
research studies are market growth and market size. Hage
(1975) suggest inclusion of variables from other

disciplines for statistical reasons.

Decision Rule 7. Systematically include the key
variables of several disciplines if relevant to the

problem. (Hage 1975, p. 153)

Determination of Measures

Where possible alternative measures for the same
variables should be included. This is because using only
one indicator tends to favor certain countries and not
others. In this regard the care that Cutright (1963)

took in combining two indicators to measure variation

107

between societies on the dimensions that concerned him

were exemplary.

CONCLUSIONS

Based on these decision rules relating to the
sample; time period variables and measures, the following
countries have been selected. Twenty countries (ten LDC
and ten DC) were selected (Appendix D). The time period
selected for the analysis is the period from 1973 through
1983.Yearly measurements within this time period is
deemed appropriate to observe and measure variables and

their affects.

HYPOTHESIS AND THEIR RATIONALE
H1: A positive relationship exists between U.S.
manufacturing firm's exports to a host country
in year t and international decision makers
perception of political risk in that host
country in year t-k : ceteris paribus.

This hypothesis investigates the relationship
between exporting behavior to a host country as political
risk is perceived to increase and decrease over time. The
incremental school postulates that firms usually export
first in the internationalization. process to increase
learning (Root, 1982), reduce risks.(Cavusgil and Nevin,

1980) This implies that exporting is an effective

strategy in high risk markets. Therefore as risk

108

increases firms would resort to exporting and therefore a
positive relationship is expected. The hypothesis tests
the theory of the incremental school of thought. Basi
(1968) surveyed international firms and concluded that
political risk was a significant determinant of mode of
operation. The contention is that even when firms are.
accessing foreign markets through investing (when
political risk is low), they would refrain from further
investing as the level of political risk increases. To
continue to seek profits and maintain a presence in that
market, the firm would then switch and opt for less risky
modes of operation. Exporting then becomes the logical
choice (least risky mode), if the firm chooses to adapt.
Hence, a positive relationship can be expected between
exporting and political risk. Observation of U.S.
manufacturing exports and FDI to high risk country
markets lends support to this hypothesis. Peru, for
instance, received $75 million dollars worth of foreign
direct investment in 1973 and $415.2 million in exports.
But as political risk increased, the proportions shifted
dramatically. By 1982, U.S. manufacturing firms FDI had
declined to $2 million and exports increased to $1,116.9
million. While these figures8 do not adjust for
inflation, they do not negate the conclusions since any
deflation will be used for both FDI and exports and

therefore cross-out. This phenomena holds for most

109

developed countries. Other factors such as tariff and
non-tariff barriers can influence exports, hence there is
a need to assume all other factors equal (ceteris
paribus).

It should also be noted that the relevant decision
isn't IF to enter the host market or not, but WHAT mode
of operation to use given the level of perceived
political risk. One would therefore expect a significant

positive relationship between exports and political risk.

H2: A negative relationship exists between

the flow of U.S. manufacturing firms

foreign direct investment to a host

country in year t and international

executive's perception of political

risk in that country in year t-k,

ceteris paribus.

This hypothesis tries to investigate the opposite

(i.e. negative) effect of political risk using FDI as the
dependent variable. Root (1982) has conceptualized that
firms stop investment if political risk is high.
Kassicieh, Suleiman & Nassar Jamar (1984) investigated
the effect on MNCs sales due to the war in the Persian
Gulf and found that the level of sales, contracts, and
investments by MNCs reflect the perceived risk of losses
from political instability brought about by the Iran-Iraq
war. Sales by MNCs from FDI from June 1977 to December
1981 decreased from $5.256 million in 1977-78 to $1.888

million in 1980-81 (Nassar Jamar, 1984). That political

110

risk reduces FDI cannot be denied. Foreign direct
investment by definition involves tremendous commitment
of resources. Markets with higher political risk
increase the probability of intervention (see definition
of political risk) and therefore firms would want to
guard against losses by minimizing FDI commitments. That
is, decrease the increase in the flow of FDI funds. A
positive relationship is therefore expected between FDI
and stability. Hypothesis 1 and 2 jointly test the theory
of the strategic school with respect to substitutability
of exporting for FDI in high risk markets.
H3a: In developed host countries, market size

in year t-k explains greater variance

than political risk and market growth rate

in year t-k for the exporting decisions of

U.S. manufacturing firms in year t,

ceteris paribus.
H3b: In developed host countries, market size in year

t-k explains greater variance than political risk

and market growth rate in year t-k for foreign

direct investment decisions of U.S. manufacturing

firms in year t, ceteris paribus.

The objective of this hypothesis is to indirectly
determine the effect of political risk on the export-
invest decision by using the data for developed
countries. An assumption is that the perceived political
risk in developed countries is low or stability is high.
The international decision maker interested in developed

country markets realizes that they do not have to worry

as much about expropriation, nationalization or

111

intervention. Consequently the amount of variance
explained by political risk will be minimal. As Bennett
& Green (1972) put it, "it is possible that the
allocation of investment within the developed nations is
not made on the basis of their level of
stability."(p.186) Market size then becomes a very
important variable explaining a greater share of
variance. In a step-wise forward regression market size
will enter the equation. Market size will be significant
for exports decisions as well as investment decisions.
However a firm would like to maximize its presence in
large markets and therefore would seek to invest in those
markets. The amount of variance explained for foreign
direct investment by market size would be greater than

that explained for exports.

H4a: International executive's perception of
political risk in less developed countries
in year t-k does significantly explain
greater variance than market size in the
exporting decision of U.S. manufacturing
firms in year t, ceteris paribus.

H4b: International executives's perception of
political risk in less developed countries
in year t-k does significantly explain greater
variance than market size in the fioreign direct
investment decisions of U.S. manufacturing
firms in year t, ceteris paribus.

Again this hypothesis indirectly tests for the
effect of political risk. As a group of countries, LDC

show a high degree of political risk. Investors would be

112

determining the political risk level of the markets they
operates in. Political risk will therefore explain a
significant amount of variance in the decision to export
or invest. The amount of variance explained by political
risk for FDI will be greater than that for exports.

Also there is theoretical support for considering
the relationship different for less developed nations and
developed nations. Fer instance Ievi (1979) has argued
that including developed nations in his sample would hide
the actual relationship between political instability and
FDI in less developed countries. (p. 60) Bennett & Green
(1972) and Nigh (1985a) have shown that level of
development is negatively related to political
instability. Lewis (1979) empirically tested the
relationship between political instability and the flow
of foreign direct investment in developing countries
(LDCs). His conclusions suggested that political
instability is an important albeit not the prime
determinant of foreign direct investment decision. run.
Burton (1984) examined expropriation of foreign assets
for the period 1960-1977 by reference to a matrix of the
nationality of expropriated firms, industrial sector,
time period, and the regional and income levels of host
countries. The data, consisting of 1,857 cases, forced
sales in whole or in part and contract renegociation. He

found that low income countries accounted for a high

113

proportion of confiscation. The southern African region
accounted for 44.5% while Latin America were responsible

for 36.2% of the expropriated firms. (p.132)

H5: In LDCs the amount of variance explained by
political risk in year t-k for FDI in year t
is greater or equal to the variance for exports
in year t, ceteris paribus.

This hypothesis tries to compare the contribution or
explanatory power of political risk for the dependent
variables - foreign direct investment and exporting.
This hypothesis is designed to determine if political
risk is more closely associated with exporting or
investing. Goodnow & Hansz (1972) has concluded that
environmental factors determine the mode of entry used by
U.S. firms to access foreign markets. This study tries
to isolate elements of what Goodnow & Hansz (1972) calls
environmental factors - political risk versus market
size. The choice of market size is due to the conclusion
of many survey and econometrics studies about the effect
of market size. The author hypothesizes that in nations
of high political risk, i.e. LDCs, it is a major
consideration in the choice of modes. Firms would prefer
to access risky markets via exports or joint ventures or
licensing (depending on local regulation) rather than
commit resources in the fOrm of FDI; hence, the greater

amount of variance accounted for by political risk. So

firms that decides to adjust for political risk and

114

maintain a presence in the market would opt for
exporting. This mode of adaptation is fully supported
by the studies of Root (1968), Shapiro (1981), and Robock
(1971).

Root's (1968) early work is particularly useful in
supporting this hypothesis. "Once a company invests in a
foreign country it has no choice but to adapt in some
fashion to the political, social, and economic condition
it finds there." (p. 76) He further stated that joint
venture is an effective way to reduce political risks of
foreign operations, especially in developing countries.
Maclayton, Smith & Hair (1980) did not separate the
determinants of foreign markets for developed and less
developed countries, but concluded that market
opportunity and political risk were two major
determinants.

Verification/refutation of this hypothesis will be
based on comparison of the partial correlations (two-

order partials) and testing for their significance.

H6: In DCs, the amount of variance explained
by market size in year t-k for FDI in year
t is greater or equal to the variance
explained for exports in year t, ceteris
paribus.

This hypothesis tests to see if the authors
rationale is supported. Market size explains greater

variance in the FDI-Export decision than other

115

independent variables in DCs. Political stability is
usually taken for granted and therefore isn't a major
determinant. of the :mode of' operations. IMost of the
studies that have surveyed U.S. international executives
overwhelmingly conclude that it is a major consideration.
My contention is that significance is applicable only for
DCs and not for LDCs. A ‘U.S. firm ‘will export to
developed countries with low relative market size.
Verification/refutation of this hypothesis will be
based on comparison of the partial correlations (two-

order partials) and testing for their significance.

THE VARIABLES AND THEIR MEASURES
Independent Variables
Political Risk (Risk)
Political risk is operationalized as the yearly
perceptions of international executives on three

dimensions:
a) political risk index (PRI)

b) operations risk index (ORI)
c) R-factor (R-FACTO)
Political scientists cannot agree on a definition for
political risk and their focus of interest is not likely
to produce the answers needed for international
businesses.
Political risk indices is measured on a 100-point

scale indicating the degree of risk. While some studies

116

have used political events to ‘measure political risk
(e.g. Nigh 1985), the use of international executive's
perceptions provides a more sensitive measure since it is
the same perceptions which the executives used before
deciding on the mode of operation to choose. Studies by
Bauer de Sola Pool & Dexter, Van Agtmael (1976) and Zink
(1973) have arrived at the same conclusions with respect
to how managers evaluate political risk. Kobrin (1976)
aptly summarizes:

First it is clear that managers consider
political instability or political risk,
typically quite loosely defined, to be an
important factor in the foreign investment
decisions. Second it is just as clear that
rigorous and systematic assessment and
evaluation of the political environment is
exceptional. Most political analysis is
superficial and subjective, not integrated
formally into the decision making process
and assumes that instability and risk are one
and the same. The response frequency is
avoidance: firms simply do not get involved
in countries or even regions that they
perceive to be risky. Last, managers appear
to rely for environmental information primarily
on sources internal to the firm. When they
look for outside data, they are most likely to
go to their banks or the general and business
media. (p.75)

The data for political risk; is obtained from
Business Environment Risk Index (BERI). This index is
preferred over others available primarily for two
reasons.

First the index is based on ratings by executives

from companies, banks, governments and institutions with

117

extensive international experience. One criticism of
other indexes such as Conflict and Peace Data Bank
(COPDAB) is that weights are assigned to political
indicators In! political scientists. Unfortunately
international executives do not consult. political
scientists before deciding on where to invest or the mode
to use. Even though political scientists are in a better
position. to assess country-markets and. political risk
level, international executives do not rely on them in
forming their perception of political risk. Bennett &
Green (1972) have stated the need to use an index
validated by international executives. "It is possible,
however, that future research could be directed toward
developing: an index: of’ political instability: designed
for, and ‘validated. with, investment. decision ‘makers."
(p.186)

Secondly the BERI index is very current - providing
data as recent as 1985. Given the rapidly changing
nature of the international environment, the more current
the more reliable. Also part of the methodology calls
for personal interviews with international executives in
1986.

Lastly the BERI index measures a broad range of
variables [PRI, ORI, R factor] and the sensitivity of
these measures to each other and other independent and

dependent variables.

118

What follows is a definition of the sub-indices of
risk:

1) Political Risk Index (PRI) measures socio-
political changes based on eight causes [regional
factors: global factors; political fractionalization:
religious/ethnic/language fractionalization; coercive
measures to retain power; etc.] and two symptoms
[societal conflict; assassinations, guerrilla warfare,
coup d'etats]. The BERI panel consists of more than
sixty political scientists from around the world.

Scale: 0 I ----------------- I ---------------- I 100
Most stable 50 Most risky
2) Operations Risk Index (ORI) is the degree to
which complex operating conditions affect profits earned
in the local currency by a foreign firm. One-hundred
plus business executives from around the world use
fifteen factors - labor conditions; bureaucracy; attitude
toward foreign businesses: nationalization potential;
infrastructure: credit availability; economic factors:
contract. enforcability: local partners and. management:
and other factors to rank each country on a scale.
Scale: 0 I ------------------ I ------------------ I 100
Very Attractive 50 Least attractive
3) R-Factor is the risk associated with remittance

of profits and repatriation of capital in a convertible

119

currency. Commonly referred to as remittance and
repatriation risk is measured by computer analysis of
data based on four sub-indices:

a) legal framework - laws as written and
actual practices

b) foreign exchange generation current
and capital accounts

c) accumulated international reserves
analysis

d) foreign debt assessment

Scale: 0 I ---------------- I --------------- I 100
Very Favorable 50 Least Favorable

Market Size (GDP)

This study relies on indirect measures of market
size mostly economic measures . Reliance on this type of
measures is due to the paucity of multi-country data on
product sales and the historical bias thereof. Market
size is operationalized as the gross domestic product10
of each nation and is measured in U.S. dollars. Although
the collection and reliability of data is suspect, it is
still the best available measure of market size. One
might question the use of a single indicator rather than
multiple indicators, however in many cross-sectional
studies multicollinearity has been found between multiple
indicators of market size for instance per capita GDP,

GNP, GNP per capita, energy consumption, roads,

120

advertising expenditures, etc. However: data for two
indicators GDP and. GDP per capita will initially be
entered into the equation.

GDP is to be denominated in U.S. dollars. This is
the GDP of each market measured at factor cost rather
than market price (this helps deal with the inflation
problem). The original GDP figure is first obtained in
the countries currency then converted to U.S. dollars by
using the appropriate exchange rate. GDP is to be
obtained from two separate sources; a) Statistical
Abstract of the ‘United States published by the U.S.
Department of Commerce and, b) The Yearbook of National
Accounts Statistics published by the Statistical Office
of the United Nations. This data will be interval in

measure .

Market Growth Rate (MAGROl

The variable market growth is operationalized as the
rate of yearly change of gross domestic product. Since
GDP is obtained as a yearly statistic, market growth rate
will also be yearly. The data for the previous year
becomes the standard. used to measure an increase or

decrease in the rate of market growth.

MAGROt = GDPt - GDPt_1 / GDPt ; where t = year

This data will be collected as ratio data.

121

Dependent Variables

Exporting (EXPORTl

Exports of domestic and foreign merchandise include
commodities which are grown, produced, or manufactured in
the ‘United States, and commodities of foreign origin
which have been changed in the United States from the
form in which they were imported, or which have been
enhanced in value by further manufacture in United States
(for further explanation of source and coverage of
exports see Appendix A). The dollar value of cumulative
yearly balance of exports to each foreign market is used.
That value is denominated in current U.S. dollars. Data
on exports is obtained from the Department of Commerce's

export schedule E.

Foreign Direct Investment (FODIVl

The value of direct investments abroad rather than
direct investment earnings or direct investment capital
flows is used. Within this category only manufacturing
FDI is used. Since U.S. DOC data measures outstanding
book value, the flows of FDI will be the change in value
of FDI. That value is reported in U.S. dollars and
represents investments, reinvestments and divestments.
According to Mrs. Patricia Walker of the Department of
Commerce an increase or decrease in the book value of FDI

results from the interaction of these three measures:

122

1) Equity and Intercompany Account Outflows

2) Reinvested Earnings of Incorporated
Affiliates

3) Net Outstanding Loans to Foreign Affiliates
The only source used in this case is the Survey of
Current Business published. by the Department of
11

Commerce .

The data to be obtained will be interval.

DATA ANALYSIS PROCEDURE

The testing of the hypothesis will be done using the
secondary data exclusively. The primary data collected
will be used for the sole purpose of ‘validating the
results.

Since the methodology involves pooling of time
series (eleven years) and cross-sectional data (twenty
nations), the first step is to ensure appropriateness of
pooling of the observations. Bass & Wittink (1975) have
demonstrated that pooling of the observations has to be
appropriate if reliability resulting from the increased
degrees of freedom. Also since cross-sectional variation
is ordinarily substantially greater than time-series
variation, the pooled estimates would have the desirable
property of being derived from a wider space of variation

than estimates based solely on time-series data. However

123

pooled, the estimates should not be accepted uncritically
therefore a homogeneity test has to be performed.

A homogeneity test involves a comparison of the
error of sum of squares from the separate regressions
(i.e. each country and each year) with the error sum of
squares from the pooled ordinary least squares estimates.
Two issues related to pooling are; a) whether or not
pooling is appropriate, and b) how should the data be
pooled?

1) Should Pooling Be Done?

As mentioned earlier the test of appropriateness
involves comparison of the error sum of squares from the
separate regressions with the sum of squares from the
pooled ordinary least squares estimates. If there are
wide departures from homogeneity then pooling will
distort the conclusions about the nature of the
relationship. If, however, differences exist and are
thought to be significant, but small, Wallace (1972)
suggests accepting some bias in order to reduce
variances.

2) How is Data Pooled?

Table 7 shows the models that can be used to pool
cross-sectional time series data. The model used depends
on the assumptions about the slopes and intercepts.
Three common statistical tests for homogeneity are:

(Johnston,1984)

124

TABLE 7.--Taxonomy of Time Series, Crosssectional Models

 

 

Model Intercept Vector of Slope Disturbance Term
a Coefficients U.
1t
8
. . 2
1(a) common for all 1,t common for all 1,t E(uu;) = O uIn
I(b) common for all i,t common for all i,t E(uu') = v
IIa varying over i common for all i,t fixed effects model
IIb varying over 1 common for all i,t random effects model
IIIa varying over i,t common for all i,t fixed effects model
IIIb varying over i,t common for all i,t random effects model
2
IV varying over i varying over i E(pu) = O uIcrE(uu') = V

 

Source: J. Johnston, EconometriCMethod, 3rd ed. New York: McGraw-
Hill Book Company, 1984.

 

125

a) tests for overall homogeneity
(slopes and intercepts)

b) tests for partial homogeneity
(slopes)

c) tests for partial homogeneity
(intercepts)

If we assume regression coefficients are fixed
parameters, the pooling procedure is represented by the

equation:

 

 

IYI le El
Y2 = X2 P + F2
= XB + E ...... 1
Y x E
b Nd L Nd L NJ

 

 

 

 

Y1 = X181 + E1 1 = 1,2, ... N cross-sections ... 2
where Y1 = Ti X 1
X1 = Ti X k

fixed k x 1 vector

III
II

E- = Ti x 1 random vector

Under the assumption of fixed and common slope
coefficients and intercepts which are not fixed, but
random, variance components is the appropriate method.

If, however, the assumption is that the intercepts
are fixed then ordinary least squares with dummy variable
intercepts (OLSDV) is the appropriate method for
estimation. In this case the test of the hypothesis of

equality of slope coefficients (Bi = BJ) is accomplished

126

on the basis of an F-Test with degrees of freedom K(n -1)
and N(T - K) where:

N number of cross-sections

T number of time-series observations

K = number of estimation parameters

Bass & Wittink (1975) say that "it is wise when
analyzing cross-sectional and time series data to: 1)
test the homogeneity hypothesis before pooling the data,
and 2) estimate pooled data with a variety of methods in
order to determine the sensitivity of the estimates to
the underlying assumptions."(p.415)

The model to be tested is stated as below:

EXPORT

f(RISK, GDP, MAGRO)

FODIV

f(RISK, GDP, MAGRO)

TESTING FOR MULTICOLLINEARITY
Initially six independent variables will be entered
and a correlation. matrix used to determine if
multicollinearity exists between the following sets of
variables:
PRI, ORI, RFACTO [Political Risk Variables]

GDP, GDP/capita [Market Size Variables]

To the extent that multicollinearity exists, the
variables will either be combined or dropped from the

model.

127

The final independent variables will then be entered
into the equation via backward elimination. Three Lag
Models can be used; zero, one year and and two year lag
models. Green & Smith (1972) have suggested that
manufacturing investment reacts/responds more quickly to
changes such as political instability than investment in
mining and petroleum. So One and Two year lag models
seem appropriate for manufacturing data.

Equations for the lag models are:

EXPORTit = BO+B1RISKi(t_k)-B3GDPi(t_k)-B3MAGROi(t_k)+....

EI(t-k)

FODIVit = B9_B1RISKi(t_k)+B2GDPi(t_k)+B3MAGROi(t_k)+....
EI(t-k)
where: i = country

t = time of year

lag period

TESTING FOR RELIABILITY OF THE MODELS

The reliability of the model will be tested through
inspection of the residuals to ensure that what we have
is a linear relationship. Durbin-Watson statistics will
also be used to test for autocorrelation. Interpretation
of the results will then be based on the multiple

regression coefficients and significance levels.

128

The statistical package to be used for the analysis
is SPSS' version 9 which is the current package
available. The advantage of the regression procedure
available in SPSS is the plotting of various residual
variables. Thus verification of the appropriateness of
the model through observation and analysis of the
residuals is possible. It is reasonable to expect cross-
section effects between the equations within the same
country and serial correlation within an equation. Using
pooled time-series and cross-section data model is then
adequate and used for this study since it can properly
handle statistical problems.

Testing for serial correlation will involve the use
of the Durbin-Watson (1951) test which is computed from

the vector of OLS residuals e = y -xB and defined as:

Ema.

cl

“

Z

LCétt " ale-IL)-
A 2
6“

th

‘ L
The hypothesis of no autocorrelation is tested. using

Durbin-Watson tables.

CORRECTING FOR SERIAL CORRELATION

The generalized differences method as detailed by
Cochrane-Orcutt is used to correct for serial
correlation. The procedure is as below:

STEP 1: Regress Yi t-k) on Xi(t-k) to obtain
residuals 1(t-k)

129

STEP 2: Estimate the relationship among the
disturbance terms. We assume this
relationship is of the form:

eit = V8 Mat-I) + M*

whereaue satisfies all the assumptions of
OLS error terms.

STEP 3: We then use the expression below to
estimate P(rho):

 

% ‘ a

c . ,

I) H 1::2 "t “it” __
~32 614*")

STEP 4: Estimate the parameters (taking care of the
time-series correlations) from the
transformed model:

\1: = 71-: ‘ hydra)

L‘I’.‘ A
XL: = Xtt "‘ PeXiLe—U

STEP 5: This procedure continues via a series of
Cochrane-Orcutt estimations until
successive values of rho converge
(for instance,p = .6791, .6795, .6795 . . )

NOTES

 

Cross-sectional and time series data will be pooled.

A minimum of sixty regression is run for each
dependent variable in each lag model.

See "Decision Rules" on page 103.
Recordings are available from the author.

This is determined by the percentage of sales
resulting from foreign operations.

10.

11.

130

BERI stands for "Business Environmental Risk Index".
Description of the index is provided in Appendix K.

t=year : k=lag period

Exports and FDI data are from Department of Commerce
documents. See Appendices A and B.

See Appendix K for description of BERI methodology.

Appendix C provides definitions and explanations of
National Accounts Statistics.

Survey of Current Business (1974-83)

CHAPTER IV

THE SUBSTANTIVE FINDINGS

In this chapter the results of the data analyzed will
be presented. The author aims to present empirical
findings rather than discuss the findings - the latter is
left for the next chapter. As mentioned in Chapter 3 the
results for all three Lag Models (Zero, One and Two Year
Lags) will be presented.

An attempt has been made to maintain consistency in
the format used for presenting each lag model in order to
ensure comparability. Data on the following variables was
collected and analyzed:

1) Political Risk Index (PRI)

2) Operations Risk Index (ORI)

3) Remissions & Repatriation Index (R-FACTO)

4) Absolute Market Size (GDP)
5) Per Capita Market Size (GPCAP)

6) Market Growth Rate (MAGRO)

It is necessary first to determine if all or a few of
the six variables should be used in the regression model.
When independent variables are correlated among
themselves, intercorrelation or multicollinearity is
said to exist. If a high degree of multicollinearity
exists between independent variables they will be

131

132

combined or one of the co-linear variables eliminated.
When multicollinearity among the independent variables is
minimal (or removed by factor analysis), then we can
identify: the extent. to ‘which. each. of ‘the independent
variables is related to the dependent variable. (Hair et.
al. 1979, p.36)

The process of eliminating/combining intercorrelated
independent variables ultimately results in a reduction
to three independent variables [Political Risk (RISK),
Market Size (GDP), Market Growth Rate (MAGRO)] from the
original six identified on the previous page. The
computer results are shown in Tables 8 and 9. Table 8
shows the initial correlation coefficients of all
variables. Correlation is a statistical technique that
measures the degree of association or intercorrelation
between variables. The numbers in Tables 8 and 9
represent coefficients of correlation or the degree of
association between each pair of variables.

The starred coefficients represent points where
multicollinearity is suspected. As a rule of thumb,
multicollinearity exists if the coefficient of
correlation between the two independent variables is
greater than the sum of the coefficients resulting from
the dependent variable and each independent variable.

Observations of the coefficients reveals the

following:

133

TABLE 8.--Correlation Coefficients

 

 

 

 

 

 

 

 

Export PRI ORI RFACTO GDP GPCAP
Dependent Variable = Export

Exports

PRI .17585

ORI .29684 *.77766

RFACTO .19754 *.52764 *.75538

GOP .60574 .23013 .29008 .32925

GPCAP .16359 *.81452 *.78708 *.64427 .32623

MAGRO -.00576 .00410 —.02997 -.00509 -.12330 -.10255
FODIV PRI ORI RFACTO GDP GPCAP

Dependent Variable = FDI

Foreign

Direct

Investment

PRI .17715

ORI .16193 *.77766

RFACTO .12676 *.52764 *.75538

GOP .22241 .23013 .29008 .32925

GPCAP .00332 *.81452 *.78708 *.64427 .32623

MAGRO .26058 .00410 -.02997 -.00509 -.12330 -.10255

Key

*: Implies variables with high correlation coefficients.
PRI: Political Risk Index
ORI: Operations Risk Index
RFACTO: Repatriation Index
GDP: Market Size
GPCAP: Per Capita Market Size
FODIV: Foreign Direct Investment
MAGRO: Market Growth Rate

134

TABLE 9.--Correlation Coefficients

 

Export Risk RFACTO GDP

 

Dependent Variable = Export

 

 

 

 

 

 

Exports
RISK .17455
RFACTO .18715 *.71213
GDP .65918 .21541 .31738
MAGRO -.07397 .06311 .03464 -.15202
FODIV Risk ’ I RFACTO
Dependent Variable-mFDI
Foreign
Direct
Investment
RISK .24461
RFACTO .18098 *.71152
GDP .28927 .21724 .31726
MAGRO .19461 .05296 .02802 -.15330
Key
* : Implies variables with high correlation
coefficients
PRI : Political Risk Index
ORI : Operations R1sk Index
RFACTO : Repatriation Index
GDP : Market Size
GPCAP : Per Capita Market Size
FODIV : Foreign Direct Investment

MAGRO

Market Growth Rate

135

a) PRI and ORI and RFACTOR are highly
correlated to each other, and

b) PRI and ORI and RFACTOR are highly correlated
with GDP per capita.

As a result of these observations, GDCAP will be
eliminated while PRI and ORI will be combined and new
correlation coefficients calculated. Table 9 shows the
results of the new analysis. PRI and ORI have been
combined to form the new variable RISK while GDCAP has
been eliminated. The correlation coefficients seem to be
satisfactory except for variable risk an RFACTOR.
Therefore, RISK and RFACTOR will be combined.

Beyond this point analysis proceeds using three
independent variables: Political Risk (RISK), Market Size
(GDP), Market Growth. Rate (MAGRO), and two dependent
variables - Foreign Direct Investment (FODIV) and
Exporting (EXPORT).

The subsequent reporting of results for each lag
model starts with the test for homogeneity 'to ascertain
the appropriate method to pool the data and ends with
confirmation/rejection of the alternate hypothesis. The
four steps are outlined below:

1) Test For Homogeneity
a) Complete homogeneity (slopes and intercepts)
b) Partial homogeneity (slopes)

c) Partial homogeneity (intercepts)

2) Results for EXPORT obtained through the
appropriate pooling method

3) Results for FDI obtained through the
appropriate pooling method

136

4) Confirmation or Rejection of Alternate Hypothesis

LAG ZERO MODEL1

The Tables 10 through 18 provide a summary of the
regressions run on the data allowing for no lag time. The
implication is that investors react to the independent
variables: political risk, market size, and market
growth within months (1-12 months).

As indicated in the 'Data Analysis Procedure'
section of Chapter 3, the first step in analyzing the
results is to determine if it is appropriate to pool all
data for the purposes of estimation of one parameter.
Since there are two dependent variables (Foreign direct
investment and Exports) the homogeneity tests will treat
each separately. In examining the appropriateness of
pooling subsets the study uses three formal statistical
tests - tests of overall homogeneity, partial homogeneity
(slopes) and partial homogeneity (intercepts). The
results for the first dependent variable - EXPORTS will

be discussed.

Exports

Table 10 shows ordinary least squares time-series

estimates by countries reflecting the relation between

137

Honéh ouoon n. u.“ unenduunuHm

36A 3....» n .3 3323630

 

 

 

n
no. v a.
o-~.nn ha.oa oNS.H- ea.nH- nn.H m~.on nano.n vooo. Hoo.Hu mm.mamHu «HucH
omn6.HH vm.~m ham. . mmo.n- Hmm.- an.ma- mo.m oHsHo. oHo. ~m.omev oHuocoecn
umv.~o «H.5m Hoo.H vsuo.q nw~.- Hso.«- ann.6H avqo. no”. omH.sm nocHnHHHHna
ono.nH an.mo mom. mas.~ HHn.H men.no vo~.Hu Hsowo. vw~.Hu sm.ooa~u ouHuu: auaom
uHoH.o cm.mp moo.H- vvm.v- om>.H- os.~v . n~n.v «ano. mHm.H co.mnn~ aHuuon
cana.HH co.na n-. mama. nsvm.~- mm.va I nva.m aHsmo. ene.“ vo.o~om anon
ammo.a~ ma.~a Hoo.H oo.mm nao.n- Ho.mno- nvwo.m aamao. oeo.n o.m~nen oono:
unv.oH em.ao oo~.H amH.6H HHE. GHH.HH nanm.m cosmo. mmo.H- EH.mo~Hs .HnsoHoo
Hem." ".mm amo.H va.mm oHv.H- mH.Hv- n~m6.~ mmoHo. HH.H o~.ma- HHuuua
oan.h so.Ho Ame. Hoo.~ HHv.H- em.o~- n~o6.n aHHHo. am~.H om.o~s ucHucoou<
uam.nv nH.ma on. v-.~H Hoe. . e~.mn- n~vn.m mmnwo. man. oo.vnHH aoeucHu uuuHca
coo.an ao.va «mm.H- eHm.HH- QHN. mno.h naHs.o mmHHo. mam.- mm.omm- ocean»
mo.m~ ~H.~o amom.n- -s.m~- «Ho.n- an.wo- anm.m HHHHo. maa.n H».~mne cHuam
ons.HH H..Ho Hem. - smH.HH- mmH.- man.ea nHao.m oHoHo. moo. voH.o~vH sHauH
omv.nn ov.no ana.~- som.H~- mam. has.mH nNHm.m Hummo. ammn.~- sm.~soH- acaHuuuqum
o~n.a~ Ho.~m n~e~.v- mno.m- nHao.~ Hm~a.~ naom.o «ooHo. HHS. . ooo.o~H- xuaacoo
omo~.H~ RHH.om avn.- ovsa. n mm. . oHo.m . nooa.a oHeHo. can. amo.eo~ cooozm
uHm.~o mHv.oa noH~.m- moH.vo- was. von.o~ nsHh.o anvo. va.H «Hs.nm~H munaHuonuoz
umoH.~n n~.na mav.- ~He.H- env.H- Hom.m - naqv.o emoHo. away." coo.mon snauoz
ova. «HamH. Hoa. HH.¢H vmv.- n-.mo- vo~.H Hoeoo. com. moo.aoon scasnoo
ouMom acoHUHunooo oumom ucoHuHuuoou oumom ucoHuHuuooo oumom ucoHoHuuooo
632 n ...m >358
86:. .32 a8 2328

 

' I

Hoop: uncouounn oouwh 39396 new noueawunm defiance—.3. 30:14: 393.

138

exports and political risk, market size and market growth
rate for a fixed intercept model.2

The initial test is the test for complete
homogeneity (equality of intercepts and slopes) for all
cross-sections and is accomplished on the basis of a F-
Test with degrees of freedom of 76 and 140.3 The
critical value of F .01,4 76, 140 is approximately 1.00.
The computed value of F under the fixed intercept
assumption is 43.81. Thus the assumption of overall
homogeneity is conclusively rejected.

The test of the hypothesis of equality of the slope
coefficients alone is based on a thest with degrees of
freedom of 57 and 140. The critical value of F .01, 57,
1405 is approximately 1.00. The computed value of F in
testing for equality of slope coefficients is 21.0.
Therefore the hypothesis of equality of slope
coefficients is also rejected. In determining the
appropriate method to pool, evidence indicates that the
intercepts for the different cross-sections are not
equal. The question is whether to treat the intercepts
as fixed or random. If we assume the intercepts are
fixed, then ordinary least squares dummy variable (OLSDV)
regression is the procedure to use. However, if the
intercepts are random testing of the hypothesis, slope

coefficients should proceed on the basis of that

139

assumption. This assumption implies that a zero
intercept model should be employed in estimation. (Bass &
Wittink, 1975)

Table 11 shows the OLS estimates for each country on
the eleven time-series observations for a random
intercept model.6 The test of the hypothesis of the
equality of slope coefficients is accomplished on the
basis of an F-Test with degrees of freedom 57 and 140.
The critical value of F .01, 57, 140 is approximately
1.00. The computed F for the EXPORT equation is 8.7.
Thus the homogeneity hypothesis for slope coefficients is
clearly rejected if an assumption of random intercepts is
employed. The evidence indicates that if a fixed
intercept assumption is made, pooling is appropriate and
OLSDV is the correct method of estimation. If the
intercepts are assumed to be random and slopes fixed,
pooling is questionable and should not be done at all
unless, using Wallace's argument, one is willing to
accept a great deal of bias in exchange for reduced
variance. Also there is some indication of systematic
variation. Systematic variation in the intercepts would
suggest that the assumption of random coefficient is
inappropriate and thus on balance, pooled estimates are
probably most appropriately developed with OLSDV. Table

12 shows OLS cross-sectional estimates. An inspection

1.4!)

mm." A Gave... h .3. uneoaqnmimo

H24. A Boo. ... 3 £333.53

 

 

 

 

n
no. v on
umm.m- ~mm.mm nho~.nu ~o~.oHu HHo.H mHm.h nomn.m momoo. uu uu «HecH
uam.no ~m.em mow. honos.H owe. van.~ anne.m meHo. uu uu «Hmocoucn
omm.mvn hm.mm nH~.H ovH.v ho~.u aonom. u nv6~.~H vmvvo. uu uu nocHaHHHHza
u~m.mo ~n.oa mno.Hu mvH.oHu omN. pva.~ neem.m ammHo. uu uu auHuun zusom
o~mo.ev mm.va «Ha. cosmo. mHo.u Noan. u n~Ho.n «ano. uu uu «Huoon
uom.am om.ma mom.H nsov.~ «em. mHom.~ mam. «memo. uu uu anon
uevo.ho eo~.ea nvmo.~ who.oHH oHv.u heHv.vHu ammo.v mama. uu uu ouonx
unso.mp we.em mam. mvv.e vnm.u vHHao.v u nmm~.~ mmovo. uu uu aHnaoHoo
ueo.m~H so.>m ammo.~ mmH.os vom.u moHv.mHu nooH.v HnnHo. uu uu HHuonn
oaHH.on nan.nm o~o. HmH.n m~n.u mn~H.v u nvoH.m vomHo. uu uu ocHucooun
oo.omn v~.om HHm. oon.~H oma.u Homo.eHu neoH.mH n~v~o. uu uu aoeocHu voana
coa.omn H~.am om~.~u www.mnu Hmo.u omaen.H u nHma.a onHHo. uu uu vacuum
uoos.oo mo.am mom.Hu HHm.-u GVH.H mee.~H nnmm.v ~mnHo. uu uu :Hoam
uoo.moH Ho.mm Nae. u oo~.n u Naa.H nnan.o~ nno.e «HoHo. uu uu sHoun
uoo.ooH oo.hm oem.Hu nee.mHu noa.Hu ovo.H~u neon.oH mno. uu uu ecoHuouuHam
coo.mnv mn.oo Hmm.vu mHeo.e u nnHv.n mom.m nHao.m «mace. uu uu xueaeoo
cam.va~ no.am «mm. u «Ham. u omo.u mnam.~ u nmvH.oH HSVHo. uu uu covoam
one.aHo sm.am nHHo.mu an.~eu nHvo.~ Hnn.mv nEOH.VH moevo. uu uu neccHnonuoz
umv.Hm~ mo.mm Geo. amHmH. com. oomsm.H anH.m mmmHo. uu uu scxuoz
comm.o~ mo.mm «on. ooo.o~ HoH.u awnv.m u noon.n mmoHo. uu uu sausage
onoum unoHuHunooo onoom ucoHoHuuoou ouoom ucoHoHuoooo ouoom ucoHoHuuooo
oHuou n _ «a a a e a shunsoo
Once: qum moo uneunnou

 

Hoop: uooouounn aoonen .Hoh uOu noueeHu-u «OHuonoaHa mAOIu.HH mnm<9

141

26A ofiaunu 3 ufiozéfimu

omo.~ A.ouoo. a u“ unuoauwcuam

 

 

 

n
no. v m a
6H.N mm.mn awn. vao.hn NHN... mom.0fil fiwNn.N meOO. mmwu mfifloanwH nmmn
Oman," mumN HMO. nv~.~n Has. mmnw.0 nomN.N vnmoo. Hmv. vn.QQMH NQOH
00N0.w mo.nm ¢a2n vv.nnu ¢on.Hl wn.v.nHl Qfimh.n FNOHO. hho.n mvudv~nw Hmmu
omho.n wu.nm Jpn.~ nom.~m Hmo.an so.mmu nflov.n vm~ao. coo.“ mn.vvmn omma
Dvw.v mm.0v JNBN.NI ¢VO.MI nmm.l nominml nwo>.N vnmoo. NON...“ mn.Omov whmu
owno.n n~.mm 5mm.~ nna.mo~ an. som.m~ nomm.n naooo. a~m. n o~.pnoun ahma
owns.» n~.oo aan.flu on~.mvn «me. u ono.na u nmmm.n mmmoo. mwv.n hm.oo- shag
omo.w ma.hm mum. n moo.onn Hm~. n meme.» u n~ov.n Hmofio. swo.n woo.nHmH who“
oven.o ~>~.vm mmm. noo.m~ nmow.n onmn.~ nNow.n vomoo. «mm. mm.vo~ mnmfl
.Hno.m mn.~m mwo. u onnom. u nw. u mm-~.m n nuv~v.n vo-o. «mm. «on.s-a vnma
ovmn.n oo.mm «ac. namm. mfim. u nwa.oH n nsH.v m¢oo. mNH.H aon.mao~ aha”

«a 8mg. 230238 ouMum 232880 ouMom 5.3338... SMum 23032.8
canoauu>\udov
ouwox xnwa moo ucouucou

 

I

gone: unannoucH voxau .nuuoaxm new aouqaauum macaquoomnuuouo m40ln.- manta

142

indicates that the magnitude of the coefficients appear
to change more systematically with time. Pooling all the
data will not provide any insight into trends and long-
term forecasts.

The above statistical tests provide evidence that
the relationship between exports and political risk,
market size and market growth rate are different for less
developed countries versus developed countries. Table 13
shows results of the OLSDV corrected for serial
correlation or the autoregressive least squares (ALS)
model for both developed and less developed countries. It
should be pointed out that the results of the two models
(OLSDV and ALS) are very consistent in terms of which
coefficients are significant. The sensitivity of the
parameters estimates to the underlying assumptions of the
model is low.

Table 13 shows that market size explains
significantly the variance in export behavior of U.S.
manufacturing firms. For LDCs an increase in market size
of one-billion dollars results in a flow of exports worth
$2.361 billion. The amount of variance explained by
market size is 35.733%. For DCs an increase in market
size of one-billion dollars results in an increase in
exports of $1.059 billion dollars for a variance

explained of 53.59%. Table 14 shows the analysis of

143

TABLE 13.--Export Regression Results with LAG = 0

 

 

 

Constant RISKQ MKTSIZE MKTGROWTH R2
812.115 -18.49 .02361* 24.406
35.733* LDC
(1976.39)+ (35.355) (.00316) (14.004)
2576.84* -32.969 .01059* —9.252
53.59* DC

(774.074) (19.1474) (.009967) (11.944)

 

 

*Significant at the .05 level (1 tail)

+Figures in parenthesis are standard errors.

 

144

TABLE 14 .--ANOVA Table for Exports (LAG = 0)

 

 

 

DF Sum of Squares Mean Square

LDC
Regression 3 285572265.93435 95190755.31145
Residual 103 513600944.99986 9986416.94175

F = 19.09001

Signif F = .0000

 

 

Multiple R .59777
R Square .35733
ADJ. R SQ .33862
Std. Error 2233.02865
DC
Regression 3 522673342.66464 174224447.55488
Residual 106 452488317.17899 4268757.70924

F = 40.81385
Multiple R

R Square

Adj. R. Square

Std. Error

Signif F = .0000
.73211
.53211
.52285

2066.09722

 

145

variance table.
What follows is an analysis of the results for U.S.
foreign direct investment equation before determining if

the hypotheses are confirmed or rejected.

Foreign Direct Investment

Table 15 shows ordinary least squares time-series
estimates by countries of the relationship between
foreign direct investment and political risk, market size
and market growth rate for a fixed intercept model. The
test of complete homogeneity (slopes and intercepts) is
rejected based on an F-Test with degrees of freedom of 76
and 140. The critical F for the fixed intercept model is
approximately 1.00. The computed value of F with the
fixed intercept model is 43.81.

The test of equality of slope coefficients alone
also fails to confirm the hypothesis of equality of slope
coefficients. In determining what assumption to make
about the intercepts we test for equality of slope
coefficients for the random intercept model. Table 16
shows the OLS estimates for the random intercept model.
The F-Test with 57 and 140 degrees of freedom [F .01, 57,
140 = 1.00] is far below the computed value of 12.79 for
the FDI equation. The evidence thus indicates that if a

fixed intercept assumption is made, pooling is

146

35A 0.3.: u 3 0533.530

~24. A 88. a 3 9533.58

 

 

 

n

no. v no
awn". flnmno. can. moomv. mam. c~s.~ «n. u nhoooo.n v-.n om.onv queen
066.6 «.mp ~no. Hoosw. hau.uu ~on.v u ~oo.au «Hooo.u ~««.~ ha.Hsn aduocoocn
o-p.n mo.mo ma.” oo.~ vom.au uno.o u “o. noNoooo. mam." coo.as~ nocaauaaasa
mmm.fl m~.vv mNm. nnHH.~ now. u onv.v u use.” amfloo. who. mpo.omfl ousuu¢ zusom
mos." -.~v now.” meow“. nmo. homo. mod.” mouooo. con.u mmp.¢u cauooaz
comm. o~.H« mac.~n ammoo.~u n.Hu cm.o~n uva.fi nonoo. m~n.u o-.o~o suom
usuoo.m nm.«n awn.“ ~H¢~.o~ smS. u onn.~ou ”on. oeuoo. moo. nnu.~os~ coax»:
nvv.H -.on now.” omm.~ cm.” ~ov.a oov.fi negoo. nnm.~u n«.onan «anecdoo
name.” ano.~n sou. om~.n vmv.uu no.6Hu wag. u mn~ooo.n omo.u am.mn- adsoun
usov.n mu.po pom." mama.“ ~n~.~u wom.m u mum.” «coco. moo. av.w~« «cauco6u4
~m«.~ am.om spam.” cov.bm men. n ooo.o~u can. u ”Naoo.u ~vn. oo.-a conned: venue:
om~o.o om.nh abnv.n h~.m~ nwmq.~u «no.6mn on”. n o-ooo.u .mq~.~ ouv.~ovfl oocnum
sea.” oo.ov new. mNm. mso.~u man.mfln nme.au mflfioo.u ova." vo.oonn canon
acme. mov.mn ova. oo.n nmo.u voo.vflu vac. n vooooo.u nfio. «65.nma saaua
“aov. v5.6“ awn. vnhmm. vnm. mm~.n ham. u «HHoo.u nh~. vooo.~h vcuauonuuam
~noau. no.v «aw. u oomo.u an". mane”. hon. ovaooo. «on. mmH.a xuaacoo
om~o.~ mm.~m new." Hoop. «on. n nvpo.d u n¢am.~u amaoo.u amv.u 'NH.~H« coc¢3m
co.“ o~.~v «fl.~ ns~.m moo. u oao.p u ago.” vaoo. vov. mnv.mm~ aucuauocuoz
umo.v mam.hw .Jnmm.~u mmq.mu mac. mom.” sham.nn oomoo.u moo.” nn~.~on >a3uoz
hovofl. oswho.. «no. mop.HH om“. u men.» n and. u nouooo.n «av. hs.~v~ scqsuoo

oumom acoaofiuuooo oumom ucodoauuoou canon ucoaoauuoou oumum ucoaoauuooo

o«u~¢ a «a auucsoo

anon: moo unduncoo - .

 

dove: uoouuoucu vox«u .Hom

uOu nouns—«vow acauouoauu. 30:64 392.

o

147

36A 038 u u“ 05021530

8~.~A 2.8. ... 3 £333.63

 

 

 

 

n
no. v QC
ooum.m «n.0r oHv.. nvnvu. ova. nvoan. hNo. n avoooo°.n nn nn navcn
anam.0n ooo.~m Amao.n «corn. «ammo. ”name. «00. I naaooo.n nl nn unaccoucu
mmo.u -.nm mno.u m~nv~.~ 0mm. n «mann.n no. I Huoo.n nn nn nocaouuaaza
um>.o o.mh mn.n mv~m0.u no. I thoo.mn mva.n cmaoo. nn nn sauna: nunom
uvH.o ~n.m> Hh.m «bong. ooH.Hn houuu.n 0n.~ ovmaooo. nn nn dunno“:
oooo. on.¢a non. n vnna. n nan. wuahu. awn. «noooo. nn nn anon
omen.o Np.uh namo.n hum.m« «no. «Nana. o-.an mnnoo.n nn nn acqxot
uwvm.o vo.nh an~.« ~o~.n nmn. mhhnn. hon. abnooo. nn nn cabanaoo
unmo.~n na.uo wom.n m~o-.o~ now. n have.“ n Ono. «Naoo. nn nn awuaum
ovoa.m 0H.Nh mvh.« hhvoa.u owo.~n 15m.“ n nhvo.n vanoo. nn nn anuucoou<
nonv.v sv.no Amos.n mov.sm and. n ammo.“ n one. I mvwooo.n nn nn accusqu vouwco
o~>~.v «o.no Annn.n vav.nn nvo.an vno.oHn mam. omvooo. nn nn vacuum
mn~5.~ «n.0n vv. mans.” «mm. vonn.n «mm. n unmooo.n nn nn canon
hnuo.u ov.ov owe." acua.o hum. onwa.n «go. voooooo. nn nn huoun
ono.n no.0n non. canon. ~vv.« «vmv.v mo. n nvwooo.n nn nn unannouuaam
005.0 ow.oh no“. n homo. n Hum. noaan. wwh. honooo. nn nn xudacoa
orb.” oo.ov nflo.u ovvo. 0mm.“ mvn.~ hmh.~n moHoo.n nn nn cooozm
ownw.a nn.mh 00nn.n nnmv.o vmm. n wen.v n mou.~ hcwoo. In nn uvcdnuonuoz
mmo.« am.ov o~m~.~n mowo.on 0000.0 wvm~.n noon.nn Vnoo.n nn nn s03u02
nvo.n m~.mn mam. Nom.n.n awn. wvan.h 0mm. mnvooo. I... nl r3030
0000 oaoom
Swan 003043000 ouMOm 0:33:08 9 m ucoaouuuoou u. 00303000
0‘00: u «a suucsoo
0000: god“ 000 ucuuacoo

 

2.00: 008035 9005.. .3» 08 .333.» 63088.2. 303.3 305.

148

appropriate.

Table 17 shows OLS cross-sectional estimate. The
magnitude of the coefficients appears to change 'more
systematically over time ending at approximately one-half
the 1973 level for Bo.

Again the relationship is better estimated by
pooling data for LDCs separate from that for DCs. Table
18 shows the results obtained for foreign direct
investment for the LAG=0 model.

Table 18 shows that for LDCs, market size, market
growth rate and political risk significantly explain the
variance in foreign direct investment. An increase of
political risk by one percentage point in LDCs decreases
foreign direct investment decision by $7.818 million
dollars. An increase of one-billion dollars in market
size results in an increase of $1.01 billion dollars of
FDI activityu For DCs, only' market size and. market
growth rate are significant. An increase of one-billion
dollars in market size results in an increase of $633
billion dollars in foreign direct investment to DCs. The
amount of variance explained market size for FDI in DCs
is 13.27%.

Table 19 provides the analysis of variance table for

LDC and DC.

2..” A 033 ... 3 0533.630

08.... A 88. a 3 0:83:56

 

14459

 

 

 

 

n
no. v 0a
was.“ as.on mem.u awn.mu neh.u mo.nn nmo.~u nmooo.n fine.” ma.ovu mama
NAN.” «n.m~ ova. nom.n oom.u nano.vn novm.~n cocoo.u 0mm. ~o~.ao~ ~omH
«Non. no.0 ova. o~mo.m amo.u vamm.vu «mo.u ooooo.u mom. omud.~flm Hood
0 am.o -.no omm.~ o~o.v~ onm.u amo.mn nono.e onnoo. 53¢.n aoo.no~n coma
eon.” v0.0a «om. nomoq.~ mmo. va~.m mom.“ smooo. oov.u mn0.vo~n oboe
.uofi.n~ H~.Hp on”. aommn.~ vn0.u mom.eu .0vn~.m ooaoo. man. ~00.ov~ arms
um.H 00.nu . mon.u ham.~u oHo.u humo.u mm.“ cocoa. mw. noo.oo sha—
..n.o ~v.~o ”we.” moo.v Hav.u v~.«u nmflm.q unwoo. o~.n onn.mqu chad
afimo.n mv.Hm s~v. oo-.d n-. Hoa~v. nHoo.n n~floo. 0H~.u vnm.o~u mam"
ovno.m wo.om 00m.u homo. u mm".u hmovv. naen.n ”afloo. one. mnm.mofi who"
umm~.o~ on.m~ vss.n 0mm.~u mmn.u mqumv.fiu naom.o avnoo. memo. oom.ova npufi

oumom accouauucoo oumom acoauauuooo oumom unusuquuoou oumom ucoauduuooo
Danna u «a canduuc>xuqo»
008:: xuam moo 0:000:00

 

H000: uaouuoucn voxuu .HOh now nouaaduna accowuoouanouo muonn.h~ mqm<a

150

TABLE 18.--FDI Regression Results with LAG = 0

 

 

 

Constant RISKO MKTSIZE MKTGROWTH R2
357.83 -7.818* .00101* 3.7198
21.52* LDC
(184.079)+ (3.414) (.000296) (1.76617)
127.875 -4.503 .000633* 6.713*
13.26* DC
(159.98) (3.957) (.000206) (2.468)

 

*Significant at the

.05 level (1 tail).

+Figures in parenthesis are standard errors.

151

TABLE l9.--ANOVA Table for FDI (Lag = O)

 

Analysis of Variance

 

 

 

 

 

DF Sum of Squares Mean Square

LDC
Regression 3 1187770.98898 395923.66299
Residual 101 4329829.63959 42869.60039
F = 9.23553 Signif F = .0000
Multiple R .46397
R Square .2152?
Adj. R Square .19196
Std. Error 207.04975

DC
Regression 3 2956786.14512 985595.38171
Residual 106 111077401.81852 104790.86621
F = 9.40538 Signif F - .0017
Multiple R .36426
R Square .13268
Adj. R. Square 10814

Std- Error 427.00804

 

152

CONFIRMATION/REJECTION OF HYPOTHESIS

H1: Hypothesis one tries to test for a positive
relationship between exports and political risk for LDCs
and DOS. As Table 13 shows, we do not have significant
results. Therefore we accept the null hypothesis and
reject the alternative since a significant positive

relationship was not found for both LDCs and DCs.

H2: The second hypothesis raises the issue of
whether a negative relationship exists between political
risk and FDI in both LDCs and DCs. This hypothesis holds
true for only LDC nations with a significant negative
relationship. Political risk explains most of the

variance in FDI flows for LDCs.

H3: The third hypothesis tries to compare the
variances explained by market size and political risk for
each dependent variable in DCs. Observation of the
results clearly shows that the alternate of the third
hypothesis should be accepted and the null rejected. The
results show market size significantly explains the most
variance in export and FDI flows to developed countries
(Tables 13 and 18). Political risk shows negative signs

albeit not significant.

H4: The fourth hypothesis tries to compare the

variance explained by market size and political risk for

153

each dependent variable in LDCs. For FDI political risk
explains a greater ‘variance than market size. With
exports; however, market size explains a greater variance
than political risk. So the null hypothesis holds for
exports but not for FDI. Beta coefficients for political
risk and market size are .30363 and .25970 respectively

in the FDI equation.

H5: The fifth hypothesis tests to see if RISK
explains more variance for FDI than for EXPORTS in LDCs
only, while controlling for market size and market growth
rate. The Semi-Partial Correlation Package of SPSS was
used to obtain the coefficients. Table 20 shows the
results obtained.

A comparison of the coefficients suggests that we can
accept the null hypothesis and reject the alternate. For
LDCs risk does not explain more variance for FDI than

EXPORTS.

H6: This hypothesis tests to see if market size
explains more variance for FDI than for EXPORTS in DOS.
What variance does market size explain for exports
compared to FDI. Table 21 shows the partial correlation
coefficients obtained through SPSS.

The data suggests that the null hypothesis should be

accepted. When the effects of political risk and market

154

TABLE 20.--Partial Correlation Coefficients Controlling for
Market Size and Growth Rate (LDC)

 

Controlling for GDP

Controlling For MAGRO

EXPORT FODIV
Export 1.0000 FODIV 1.0000
(0) (0)
P = xxxx P = xxxx
Risk -.0515 Risk -.2296
(103) (101)
P = .602 P = .020

 

 

(Coefficient/D.F./Significance)

 

155

TABLE 21.--Partial Correlation Coefficients Controlling
for Political Risk and Market Growth Rate (DC)

 

 

Controlling for RISK
Controlling for MAGRO
Export
Export 1.0000 FODIV
(0)
p = xxxx
GDP .7181 GDP
(106)
p < .001

(Coefficient/D.F./Significance)

Export

1.0000

(0)

p = xxxx

 

.2862
(106)
p = .003

 

156

growth. rate are removed, market. size Ihas more of an
influence on exports than on foreign investment . U. S .
manufacturing firms tend to export to countries with a

large market and not invest.

LAG ONE MODEL

The Tables 22 through 31 provide a summary of the
regressions run on the data with the independent
variables lagged one period or one year. As stated
earlier, theory does not provide an adequate
justification of one model or the other thus the need to
consider more than a single lag period. With a one year
lag, we test to see if decision's adjust their mode of
operations strategy based on past information, i.e.
information about the target market for the previous
year.

Again the homogeneity test for the appropriateness
of pooling the data will be reviewed along with the
results of the three models. The results for the Export

equation for the 1 year lag is presented next.

Exports

Table 22 shows ordinary least squares time-series
estimates by countries of the relationship between

exports and political risk, market size and market growth

157

36.4 033 a 3 2.33530

8...." A88. 9 3 2:33.53

 

 

 

n
no. v an
onw.m hm.n mvo.~u nu.-- svo.n n.o« a Ho.n mmao. n¢h.~ nw.o~nv caucn
uoo~.ofl «Nam. nnm.~u av>.oHu nmvo.~u «an.o-- nmv~.v fimvflo. nono.~ ouv.nomn aquocovcu
no.“ oo.nn are. apnea. hwo.~ ov.q~ nu no msc.u no.~0n~n accumaaaanm
ov«.ofl No.mm omn.u voa.mu nnH.vu ov.nuH- nmvn.o namno. non.n ~o.mamv auuuu< nuaom
sflm.n ”mm.mn nv~. n mpg.” n wms.ao ao.nn . nmmm.n mango. mnp.~ ama.mman cannon:
n~.~ «.mv nom~.~ -~.m~ aov. hm.m~ oh. I mo~oo.- om~.s «o.naon anon
uo-.m mo.~h ma.“ hn.mva ma. u oh~.av a omm.H “none. mam. ~o.vm¢n ooaxu:
awn. om.m~ «no. ao.H ~mm.~u oovnn.ov u «No. . mmaoo.n can.” sn.vvsn ouncodoo
nv~.~ oh.on cam." v-.mn on~. can.o HNm.n cameo. oov. «oh.n~m aauuun
amen. an.mfl an". . hflvwo. . vmo.~u noo.on u one. u manooo. mam.” mm.w~m~ «caucoou<
okom.s o~.mh nmn~.n «mo.oH~ no.nu HHa.anu sown.~ ooHo. amnfl.n mm.cmmmH sococqu vouqca
ommn.~n s~.nm mm. «v0.9 «mm. . mmv.o u anom.o anuo. mHH. u o~a.mn~u oceans
nvfiwv. unm.w~ v. a ono.ou amo. . oo.~mn «mm. - venoo.u mwm.~ mn.«~mn edema
unn>.ofi v~.~o nvm. mn.nH obs. wvo.n an~.m nooao. how. m~.mom sauna
o~oe.~n mo.no ammo.“ noo.n~ How. - noHn.mu amo~.~ «ova. mno.~u mav.asflan ocuduouudzm
«Hm. o~.m~ ovm. . mhv.v - won. I mom.ou mac. mvoooo. mn».~ ~h~.~mm xuoacoo
«fiv.u s.hn ”on.“ ~n.v «mo. «no.m~ 5mm.” Hosea. moo. n hmm.nh u couozm
usvn.mm mH.oo mmm.~- nom.~m- onn.~ hvm.mn n-o.- buamo. we. u ooo.vovu aucufiuoguoz
vow.~ ~m.~v h~.H- moo.H~- aoo. “moon. anv~.~- osoooo.u o~o«.~ mmm.mvm sosuoz
ummm.w Ho.Hm mac. ~osau. m-.Hn vow.ov u Ammo.v nonao. vs~.~ 5H.~o~a scuauoo
OHOUW OHOUm ououm Guam

osuum m mm a ucoauauuoou p ucoaoquuoou a unoauuuuoou a acoaoquuooo

>uucsou
anon: xnuz moo ucsuucoo .

 

.fi I 0‘4. umouuoucu

ponds .uuuoaxu uOu nounaaunu noauouoawe mAOIn.~n mqm<a

158

rate for the fixed intercept. model. The initial step
tests for complete homogeneity (equality of intercepts
and slopes) for all cross-sections. That test is
accomplished on the basis of an F-Test with degrees of
freedom 76 and 140. The critical value of F.01, 76,140
is approximately 1.00. The computed value of F with the
fixed intercept assumption is 30.557. Thus the
assumption of overall homogeneity is completely rejected.

The test of the hypothesis of equality of the slope
coefficients alone is based on an F-Test with degrees of
freedom of 57 and 140. The critical value of F.01, 57,
140 is approximately 1.00. The computed value of F in
testing for equality of slope coefficients is 108.26.
Therefore the hypothesis of equality of slope
coefficients is also rejected. In determining the
appropriate method to pool, evidence indicates that the
intercepts for the different cross-sections are not
equal. The issue then is whether to treat the intercepts
as fixed or random. If we assume the intercepts are
fixed, OLSDV is the procedure to use. However if the
intercepts are random then random testing of the
hypothesis of slope coefficients should proceed based on
that assumption. Therefore a zero intercept model needs
to be observed.

Table 23 shows OLS estimates for each country on

159

36 A2»: a 3 5.33.530

83.. A 88. ... 3 283263..

 

 

 

mo. v Q.
unmnadn mm.~a vo.~ ammo.a mmv. n ann.~n nowh.v whoHo. In In cavcH
u on.hn v.nm oaN. neon." our. ownm.n bwm.n moono. In In uauocDuCH
«mmo.sn can.vo hrs. navav. _n~mo.h mnwa.nn nn In nn nl nocanaaauzm
amom.n~ awn.flm mmh.a ova.vu HnH.Hn mvo.man nwmv.n hhwno. In In unauuC suaom
«moo.ov om.no vaw.a momm.v ohw. n mama.nn boom.v nnwuo. nn nn auuoouz
c hh.mn wo.nm gnaw." nmv.Vn nooh.v «05.0n mh.l «hoooo.n nn nn anon
o vm.hn nv.na £vom.~ nnm.~ma Nam. mum.vn Down.~ «ammo. nn nn Oowxo:
..Oon.o am.mh moo. o-.o~ no." ~mm.n~ o~o.n anooooo.n nn nn canao~oo
o mm.ma m~.bm aa.~ mvm.vv on." mow.un ovum.” awoo. nn nn Hauuum

a oun.n o.oo vmn.l onnun.n Abov.n o~moo.- Ono. NfiHoooo. II II unducoou<
o oo.on n.uo amon.n m.m- h-. n on.-n amo«.n chuo. In nn Eovocau vouuca
o mo.~on man.am own. bn.m bvh. n nan.on Aoao.~« Nonno. nn nn oucuum
u mov.w~ ho.om on". n Hh~.wn abom.~ nana.hm owe. I no~oo.n nn nn :«oam
.ouvo.asa hv.om hso.a Hov.o~ flan.“ nmm.on nmo~.n oomo. nn nn auouu
u on.mvn m~.oo bmm.~ mn.- Amn.nn mmo.h~n nwwv.«H mono. nn nn nauuuonuasm
u Hm.o~ om.o¢ no.n o~m.vn Dm~.n mhn.v~ onh. shaoo. nn nn xudscoo
o hvh.nv m~.vm ohm.a acm.v v~o.~ HnN.v~ numN.N anhoo. nn nn copozm
o h.owo o.mo hm.an www.mwn Nvm.~ 5n.m~ bao.o~ omvo. nn nn nvcoauonuoz
q.oma.on ooo.~a vmm.un vov.v~n Jvnw.v www.mm mh~.~n whoooo.n nn nn xasuoz
u mo.o~ mnm.nm Anas.~ Ho.nha Nod.an o~m.mwn Avo~.v wwuno. In nn >cdfiuoo
. uoo ouoom
cauam m m oumom unmanauuoou ouMum occaoquuoou o a m ucoauwuuoou a ucoaoduuoou
w . auucsoo
ouoax xaum moo acuuncou

 

AA 044v H000: unwououCH Banana .uuuoaxu

uOu uouuflwuwu noduowflfiah maOlinnN Hand?

160

eleven time-series observations for a random intercept
model. The test of the hypothesis of the equality of
slope coefficients is accomplished on the basis of an F-
Test with degrees of freedom 57 and 140. The critical
value of F.01, 57, 140 is approximately 1.00. The
computed F for the EXPORT equation is 108.26. Thus the
homogeneity hypothesis for slope coefficients is clearly
rejected if an assumption of random intercepts is
employed. The evidence indicates that if a fixed
intercept assumption is made, pooling is appropriate and
OLSDV is the correct method of estimation. If the
intercepts are assumed to be random and slopes fixed,
pooling is questionable unless we are willing to accept a
great deal of bias in exchange for reduced variance.
Such variation suggests also that the assumption of
random intercepts is inappropriate and therefore pooling
under such assumptions will not provide any insights.
Therefore data for LDCs has to be pooled separately
from 'that of DCs. 'Table 24 shows that. market size
explains significantly the export behavior of U.S.
manufacturing firms. For LDCs an increase in market size
of one billion dollars results in a flow of exports of
$5.74 billion. The amount of variance explained by the
two significant independent variables -:market size and

market growth rate is 7.74%. However the majority of

161

TABLE 24.--Export Regression Results with LAG = 1

 

 

 

Constant RISKl MKTSIZEl MKTGROWTHl R2
21894.06 -61.1947 .00574 4.095* .”
7.74* LDC
(2102.14)+ (39.46) (.00962) 1.811
2857.08* -39.56* .00962* 21.548
43.02* DC
(803.22) 18.141 (.0011) (13.22)

 

*Significant at the .05 level (1 tail)

+Figures in parenthesis are standard errors.

162

that is accounted for by market size as the high T-score
implies. Market growth rate adds an incremental variance
of only 2.2%. For DCs an increase in market size of one-
billion dollars results in an increase of $9.62 billion
in export flows for an explained variance of 43.02%.
Table 25 shows the analysis of variance results. The
results for the foreign direct investment equation for

the Lag 1 model will be discussed next.

Foreign Direct Investment

Table 26 shows ordinary least squares time-series
estimates by countries of the foreign direct investment
and political risk , market size, market growth rate for
a fixed intercept mode. The test of complete homogeneity
(slopes and intercepts) is rejected based on an F-Test
with the degrees of freedom of 76 and 140. The critical
F.01, 76, 140 from the tables is 1.00. The computed
value of F with the fixed intercept model is 3.667.

The test of equality of slope coefficients also
fails to confirm the hypothesis of equality of slope
coefficients. we therefore test for a random intercept
model. Table 27 shows OLS estimates for the random
intercept model.

The F-Test with 57 and 140 degrees of freedom [F.Ol, 57,

163

TABLE 25.—-ANOVA Table for Exports (Lag = 1)

 

Analysis of Variance

 

 

 

 

 

DF SUM OF SQUARES MEAN SQUARE
LDC
Regression 3 621854ll.11293 20728470.37098
Regidual 106 275870101.46l99 2602509.83455
P 7.96485 Signif E = .0354
Multiple R .27824
R Square .07741
Adj. R. Square .05130
Standard Error 2644.12534
DC
Regression 3 419588079.35353 139862693.11784
Residual 106 555573580.49009 5241260.19330
F = 26.68494 Signif F = .0000
Multiple R .65595
R Square .43028
Adj. R. Square -.4l415
Std. Error 2289.37987

 

1.654

an." A 3...: 3 £83583".
8a.... A 88. a 3 ufiodgcoqma

 

 

 

no. v an
88¢. o.m Hhumun n.~n -.Hn om.un oHH.~ oaoo. 86¢." Hmo.a- nave"
aoao. nu.nn «a.n mvunn.n mum. n mow.an mov.nn cacao.) mum. hmv.o~« aauocoucH
000.. oo.om ha." vmomo. n~a.~n wn.hn In It namm.~ omn.~nn nocamaaaanm
coo." n.Ho Han.~ nov~.n wov.un pmon.mn vo.fl vvaoo. mm~.« 5mm.o«~ noduu< zunom
“mo.“ o.~v vao.~n some". n .nmo.~n «cavn.n coo. oooooooo. can.“ ocm.om aauvmaz
uoao.mo mn.nm nomn.na can.“ amnm.n vo~.m nvo.ou unnooo. n~p~.on ~m.onnn anon
ammo.” no.vv onv.a mam.m~ gov. wo~.o~ nao.an ooo.n nan. n mn.oonn ocuxnx
vnv. ow.m~ «no. ooo.~ onw. omen. wnn.u moaooo. «no. n ~n~.m n nansoaoo
coo. m.mn ouv. nfluv.n on». n ov~.o n ova. n mmooo.n can.“ no.~nm Masque
oovwn. na.vu on". nwvno. mac. n vo~.v n ono. n uofiooo.n and.” woo.nun acaucoou<
naov. nom.o~ may. ”mo.na «an. n omv.nan cam. n m-oo.n vac. «v.moma eooocqx conga:
oa.v o~.vo sown.“ s~.v~ “mo.~n nvo.nan “Ho.~n NmHoo.n omm.~ ~n.nm- oucuuu
an." mo.nv on.” ufim.n ano. n 5H-.mn can. n voooo.n ono.a 8mm.omv :«adm
vso.~ doo.nv noon.“ ~o~.un vvv. omm.n mac. n manooo.n awn. n o¢.~mn Sana“
umum.o~ o~.~o n~o~.mn ooo.sn mov.H abo~.n «mm. n ncooo.n as». omn.~m ccaauouuaam
n-v. m~.n~ was. moanm. «mm. n noflp. n mmv. mosoooo. flea.” mona.ac xuuacoo
was.” mm.nv nnm.an anvnm. n «ma. von~.n nvo.~n vapooo.n Ame. n Nom.vdn covosm
H~w.a boo.” coo.“ moo.« new.Hn Ho~.mn ooH.m «00. ”Ha. n «mo.nn nvcuauozuoz
nmvo. com.” o~. omvn.H nnfl.n anuom. n nvm. nooooo. moo. n ooH~6.Hn suzuoz
mvsov. ~m.o~ vmm. mvwas. «an. as.“ no. n ~muoooo.n Hem. n ~oq.von scoauoo
035. u a Snow 233398 8M6 332380 oumom 233380 Snow £883.38
N >uucaoo
cumo: xnau moo unauncoo .

 

An Q¢A. deco: uncououcn noxwh .Hoh nan ooaaduum nowuonOBAH muonn.o~ fiance

165

mm.n h Oaucu m u« £393.30me

SN." 5863 a 3 £333.53“.

 

 

 

no. v A.
0N.w vm.ow nah. hnvwv. mom. wqua. mvn. voooo. II II uavcw
ammo.n No.5m owv.I ovvou.I nva.n vu~m. nov.aI mnnooo.I II II nwuocovcn
who'd. amu.n wam. oneao. mom. aummn. II II II II nocandanazm
umv.NH vh.mh awow.n ouh.n mmv. I. voH.HI nnm.« «moo. II II auauud £090»
omnm.v an.av «on. I Hmo. I van. «mane. haw. muaoooo. II II awuouaz
ow.nn mn.~m Avm~.h muo.~ nmvm.vI moovv. I nub.w nanooo.I II II auom
has.“ onén vow.” nonén m2..." marina Jn~n.~l «mmoo; II II 00:0:
Unnn.o o~.mm Vnh. wmmm. new. mwvu. wm~.u waooo. II II canEOHOU
ommo.nn mo.nm NN.H 500.0 an.H nNn.m nmu. I HmHooo.I II II ddfldun
umvm.v« vwh.oo who. hfiaao. mam.u hnmmm.n wvv. I cmmoooo.l II II unaucuout
mmn.u ~v.nn nvh. owv.ou mam. nvo.ma coo. I unHoo.I II II Eovocdx ooudca
omnv.n Hm.om 85mm.“ nav.wN «mm. vom.vI hh~.HI ammooo.I II II conduh
Unna.n o.am Qmm.n nmn. annmv.n NN.o owo.I v0noooo.l II II edema
umH~.m nh.mw Doha.” unm.mn mvm. ~mH.N omn.I abnooo. II II hauuu
oo~.o~ vo.mo nhmn.mn mmm.on nwb.n ovo.v ovw. moaooo. II II vcaduuauASm
Dona.o o.wo ovo. FAAN. Nam. omnn. no." mmaooo. II II xudacvo
mum.“ vn.vv hwo.un memo. I nomo.N aha.“ flaw.~I mflmooo.I II II cousim
osm~.oH an.wh ANH.N ham.o omm. maa.n mas. muwooo. II II nucdduunuoz
«woo. an.n mom. anmn.u “ov.I «moa. I Hon. mwmoooo. II II >csuoz
~wv.~ N.cv on.H mmm.v~ vvn.I hoh.hI mam. onhoo. II In >cnluoo
cauam h «a ouMow uco«u«uuoou ouMom unoAUauuoou ouMom uco«u«uuooo ouMUm uncaowuuoou
E58
anus: xoum moo ucuuocou . :

 

3 I 05.

Hove: unwoucucu Bovcdu .Hah uOu ouqaauau nowuow aide mAOII.sN ounce

166

140] is below the computed value of 11.383 for the FDI
equation. The evidence thus indicates that a fixed
intercept model is an appropriate assumption for pooling.
Also, the magnitude of the coefficients appear to change
systematically over time. Thus the relationship is best
estimated. using ordinary least squares dummy' variable
(OLSDV) regression.

Table 28 shows the results obtained from foreign
direct investment data. This table shows that for LDCs
only political risk significantly explains the variance
in foreign direct investment. An increase in political
risk by one percentage point results in a reduction of
$6.539 billion in FDI flows. The coefficients for DCs
does not significantly vary from zero as the non-
significance of the T-ratios implies.

Table 29 displays the analysis of variance tables

for LDCs and DOS.

CONFIRMATIONZREJECTION OF HYPOTHESIS

H1: Hypothesis one tries to test for a significant
positive relationship between exports and political risk
for LDCs and DCs. Table 24 shows that the postulated
relationship does not hold for DOS or LDCs. LDCs does

show a negative relationship although not significant.

167

TABLE 28.--FDI Regression Results with LAG = 1

 

 

 

Constant Risk 1 MKTSIZEl MKTGROWTHI R2
434.935* -6.539* .000103 -.09545 ,
3.867* LDC
(180.45)+ (3.39) (.000369) (.18156)
237.24 -5.0759 .00369* 4.373
5.838 DC
(156.10) (3.9144) (.000214) (2.569)

 

*Significant at the .05 level (1 tail).

+Figures in parenthesis are standard errors.

168

TABLE 29.--ANOVA Table for FDI (Lag = 1)

 

Analysis of Variance

 

 

 

 

 

DF SUM OF SQUARES
LDC
Regression 3 213678.98297 71226.32766
Residual 104 309366.89521 8484.964
F = 8.39442 Signif F = .0038
Multiple R .19664
R Square .03867
Adj. R Square .01094
STD. Error 226.00816
DC
Regression 3 1299607.15459 433202.38486
Residual 106 20984780.80904 197969.63027
F = 2.18823 Signif F = .0937
Multiple R .24149
R Square .05832
Adj. R Square .03167

Std. Error 444.93778

 

169

It is important to note that the directionality is not
consistent for the relationship between exports and

political risk.

Hz: This hypothesis tries to test for a negative
Significant relationship between pol itical risk and FDI
in both LDCs and DOS. Table 28 shows political risk is
significant and negatively related to foreign direct
investment for LDCs, but not for DOS. A negative

relationship (not significant) holds for DCs. The amount

of variance explained is rather small - 2.394% but is
Significant.
H3: Hypothesis three tries to compare the

contributions of the independent variables for each
equation. It is postulated that for developed countries
market size explains greater variance for both EXPORTS
and FDI than other independent variables. Table 24 shows
market size with a T-Score of 8.745 compared to 1.964 for
political risk; therefore, it certainly contributes more
and the hypothesis is accepted for the EXPORT equation.
Table 28 shows that market size has a higher T-Score
(1.724) than political risk although significant at the

.05 level (one-tail test).

170

H4: Does political risk explain more variance than
other variables for the each dependent variable (EXPORT
and FDI) in LDCs. Table 24 shows market size claims the
greatest contribution. Therefore the alternate
hypothesis is rejected for the EXPORT equation. Table 28
however, shows that risk explains greater variance than
market size. Therefore H4 is accepted for FDI but

rejected for EXPORTS.

H5: Table 30 provides the results to determine the
variance explained for EXPORTS as compared to FDI by

political risk.

Based on the results we can accept the null hypothesis
and reject the alternate. For LDCs risk doesn't explain
greater variance for FDI than exports. The results are

not significant.

H6: Does market size explain more variance for the
EXPORT equation or the FDI equation in DCS? This
hypothesis is designed to test this question. Table 31

shows semi-partial coefficients.

We accept the null hypothesis and reject the alternate.
U.S. manufacturing firms tend to export to DC countries

with a large market size rather than invest.

171

TABLE 30.--Partial Correlation Coefficient Controlling for
Market Size and Market Growth Rate (LDC)

 

Controlling for GDPl
Controlling for MAGROl
Export FODIV
EXport 1.0000 FODIV 1.0000
(0) (O)
p = xxxx p = xxxx
Riskl -.1254 RiSkl -.l314
(106) (106)

p = .196 p = .045

(Coefficient/D.F./Significance)

 

172

TABLE 31.--Partial Correlation Coefficients Controlling
for Political Risk and Market Growth Rate (DC)

 

Controlling for RISKl

Controlling for MAGROl

Export
Export 1.0000 FODIV
(0)

p = xxxx

GDPl .6461 GDPl
(106)

p < .001

(Coefficient/D.F.)/Significance)

FODIV

1.0000

(0)

p = xxxx

.1650
(106)

p = .058

 

173

LAG YEAR TWO

The three formal statistical tests: overall
homogeneity, partial homogeneity (slopes) and partial
homogeneity (intercepts) suggested that a random
coefficients assumption is inappropriate7. The tests
indicate that the relationship is different for LDCs and

DCs.

Exports

Table 32 shows results of the export equation. In
this case, market size explains significantly the
variance in exports for DOS. The LDC model was not
significant.0ne percentage point increase in political
risk results in a negative flow of only $44.19 billion in
DCs. Also in DOS an increase in. market size. by one
billion dollars results in a positive flow of $7.8
billion worth of exports. Table 33 shows the analysis of
variance table .

The FDI equation results will be discussed next.

Foreign Direct Investment

These results are suspicious and no meaningful
interpretation can be deduced. The coefficients are in
the anticipated direction but the T-Scores and R2 are not
significant. Tables 34 and 35 show the results and

analysis of variance respectively.

174

TABLE 32.--Export Equation Regression Results [LAG = 2]

 

 

 

Constant RISK 2 GDP 2 MAGRO 2 R2
6145.55* -80.128* .00151 -1.134
1.678 LDC
(2.904)+ (-2.023) (.672) (-.543)
3165.01* -44.19* .0078* 27.28

14.817* DC
(3.754) (-2.064) (6.321) (1.857)

 

*Significant at the .05 level (1 tail)

+Figures in parenthesis are T scores.

175

TABLE 33.--ANOVA Table for Exports (Lag = 2)

 

Analysis of Variance

 

 

 

 

 

DF SUM OF SQUARES MEAN SQUARE
LDC
Regression 3 36431971.46009 12143990.48670
Residual 66 477470112.2367 7234355.81240
F = 1-67865 Signif F = .1760
Multiple R .21297
R Square .04535
Adj. R Square .01834
Std. Error 2689.67578
DC
Regression 3 28811696.71718 96039232.23906
Residual 66 427781010.0012 6481546.82195
F = 14.81733 Signif F = .0000
Multiple R .54356
R Square .29546
Adj. R. Square .27552
Std. Error 2545.88822

 

176

TABLE 34.--FDI Regression Results [LAG = 2]

 

 

 

Constant RISK 2 GDP 2 MAGRO 2 R2
272.19 -3.405 -.000062 -.01319
1.14 LDC
(1.502) (“1.003) (-.325) (“.074)
248.37 -3.44 .000193 1.84
1 65 DC

(1.65) (-.899) (.874) (.701)

 

TABLE 35. ANOVA Table for FDI

177

 

Analysis of Variance

 

 

 

 

 

DF SUM OF SQUARES MEAN SQUARE
LDC
Regression 3 63048.46624 21016.155461
Residual 66 3466852.01000 52528.06060
F = .40009 Signif F = .7532
Multiple R .10682
R Square .01141
Adj. R. Square .01711
Std. Error 229.190
DC
Regression 3 368683.88757 122894.62919
Residual 66 13645627.1010 206751.92525
F .59441 Signif F = .6200
Multiple R .12864
R Square .01654
Adj. R. Square -.01129
Std. Error 454.69982

 

178

CONFIRMATION/REJECTION OF HYPOTHESIS

H1: This alternate hypothesis tests to see if a
positive relationship exists between exports and
political risk for LDCS and DCs. Table 32 shows we
can reject this alternative and accept the null
hypothesis. Exports is not positively related to

political risk in LDCS and DCS.

H2: This hypothesis tries to determine if a
significant. negative relationship exists for' political
risk and FDI in LDCS and DCS. Table 34 Shows a negative
relationship between FDI and political risk for both LDCS
and DCS although not Significant. Therefore we accept
the null and reject the alternate. According to this
model, FDI is not significantly negatively related to

political risk.

H3: Hypothesis three tests to see if market size
accounts for the greatest amount of variance in exports
and FDI in DCS compared to political risk and market
growth rate. Tables 32 and 34 show that we accept the

alternative for the export equation only.

H4: Hypothesis four tests the amount of variance

179

significantly explained by' political risk. compared to
market size and market growth rate in LDCs for the EXPORT
and FDI equations. Tables 32 and 34 show that we accept
the alternate for exports and accept the null hypothesis

for the FDI equation.

H5: The fifth hypothesis which compares the
variance explained by risk for the EXPORT equation and
FDI equation controlling for market size and market
growth rate cannot be substantiated because of the
results below in Table 3 6 . Table 3 6 shows the

coefficients could not be calculated.

H6: The sixth hypothesis compares the variance
explained by market size for the EXPORT equation and FDI
equation controlling for market size and market growth
rate. This claim cannot be substantiated based on

available results - Table 37.

CONCLUSIONS

In this chapter the results of the three Lag Models
have been presented. In the next chapter a rationale is
developed for the appropriate model chosen. A discussion
of the results along with the validation of results via
personal interviews conducted with international business

executives will be presented.

180

TABLE 36.--Partial Coefficients Controlling for Market
Size and Market Growth Rate

 

Controlling for

Controlling for

Export
Export

(0)

p = xxxx

Risk 99.0000

(66)

p = xxxx

1.0000

GDP

MAGRO

FODIV

Risk

 

FODIV
1.000
(0)

p = xxxx

99.0000
(66)

p = xxxx

(Coefficient/D.F. ISignificance)

99.00 implies a coefficient cannot be

calculated.

 

181

TABLE 37. Partial Correlation Coefficients Controlling
for Political Risk and Market Growth Rate

 

Controlling for Political Risk
Controlling for MAGRO
Export FODIV
Export 1.0000 FODIV 1.0000
(0) (0)
p = xxxx p = xxxx
GDP 99.00 GDP 99.00
(66) (66)
p = xxxx p = xxxx

(Coefficient/D.F./Significance)

99.00 implies a coefficient cannot be
calculated

 

182

NOTES

1. Residuals for each lag model was examined to determine
the appropriateness of the predictive model in terms

of: (a)
(b)
(C)
(d)

Based on
detected

Cochrane-

page 128

A fixed

the linearity of the phenomena measured

the constant variance of the error terms

the independence of the error terms

the normality of the error term's
distribution.

the examination, serial correlation which was
in some instances was corrected using

Orcutt method. The procedure is explained on

intercept model allows for computation of an

intercept for each cross-section (or country). Each
cross-section is computed based on eleven data points
(11 years).

Based on degrees of freedom of K(N—l) and N(T-K)

This significance level is conservative and ensures

that we

are 99% confident of our findings.

Based on degrees of freedom of (K-1)(N-1) and N(T-K).

A random intercept model does not allow for an
intercept for each cross-section (or country). The
regression line is thus forced through the origin.

The actual results of each cross-section and time-
series for the fixed and random models are not
included here. ’

CHAPTER V

DISCUSSION OF RESULTS

The purpose of this chapter is to provide rationale,
explanation and validation of the findings. To this
point, the place to start is with a summary of the
hypothesis results derived in the earlier chapters as
listed in Table 38.

This chapter is organized into three sections.
Initially a justification will be provided for choosing
one of the lag models. Secondly, a written summary of
the findings will be presented, and finally each finding
will be examined. The results of personal interviews
conducted with international business executives will be
integrated in the explanation. Collectively they provide

some validation for the findings.

WHICH LAG MODEL IS APPROPRIATE?

A lag model is appropriate because there is reason
to believe that all the variables do not relate to the
same time period. Thus FDI in year t could be due to

perception of political risk formed in time t-l . In

183

184

 

 

 

N v v omummooa dance

manomxm spas can»

Ham spflz meMHUOmmm whoa
copooflmm oouomhmm omuomflmm ma mnem umxume moon :H "mm

musedxm Baas can»

Ham nuw3 UUUMHUOMMM
owuooflmm Umuomnmm cmpowmmm UHOE ma xmflu .muoa CH "mm

mmuoq cw

Ham Mow cmuomflmm monomxm Mom omuomnom muuomxm How cwuomflmm Ham cam manomxm mo ucm
muuomxm How owumooofi Ham How woummoofi Ham How pmumwoofl Infleumuwp Momma xmwu mH new

no m\OQ :a How cam

Ham Mom omuomflwm Ham ROM pmummooa Ham new omummoo¢ muuomxm mo ucmcHEHmumc
munomxm HON Umummoofl monomxm How pmummood muuomxm How pwummood HoflmE wuwm uoxumfi mH "mm

moo “0w pmuomnwm moo How pmuomflmm on How Umpomflmm mMmam new Ham
moon How owuownmm mung you cwumooofi Dog How cwummoofl comzuwm coflumaom m>aummwz "mm

moo new pmuowmmm mood HOm cmuownmm 00 Mom Umuommom wxwfim cam manomxm
moon you pmuowflmm moo How pouowmwm Dog Mom cwuoommm :mm3umm cowumHmm m>fluamom "Hm
whom» 039 mo moo Hmww mco mo mmq mumww 0 mo can mammzuomhm

 

mambo: and mouse Ham How mufismwm mammnuommm m>aumcumuacll.mm mqmca

185

deciding to use three lag periods to investigate the
postulated relationship, the author was cognizant of the
fact that reason existed about why one lag period should
be preferred over any other. The basic question here is
"how long does it take an international executive to
plan, organize and execute entry decisions or shift mode
of operations in a foreign market either by exporting,
joint ventures, or foreign direct investment?" Certain
factors such as: 1) difficulty of obtaining information
about the foreign market, 2) nature of the business, 3)
company's operational policies, and 4) problem of
obtaining the necessary permits from the host country as
well as the home country, will definitely result in some
delay in entering the target foreign market or shifting
mode of operation.

Exporting logically as a lower involved decision
would require less time than FDI. As is shown in Table
38, the results for model lagged zero and one year are
similar. For purposes of interpretation; therefore, it
should really make much difference whether the Lag Zero
or Lag One model is used. The personal interviews reveal
that twenty-four months or less is usually time enough to
prepare for and enter a foreign market. One executive
commented that the trend was to "build smaller and more

efficient" plants that do not require much time to

186

operationalize.1

Another executive of a transportation
equipment firm said it was necessary to react or respond
quickly due to competitive response. "We have to act
quickly in entering foreign markets . . . take China for
instance, whom ever gets there first can establish a
strong foothold."2

The conclusion therefore is that manufacturing firms
usually can respond to information within a year. The
lag’ period used. is therefore the One Year Lag. In
studies where lag periods have been considered, a one
year lag was deemed appropriate. Nigh (1985a) for
instance tested three lag period (0, 1, and 2) and
concluded that the one year model performed best
according to his model prediction. (p. 9) In a later
study Nigh (1985b) also used the one year lag as the
appropriate model to investigate the relationship between
flow of German FDI and political risk.

The lack of a difference in results between the zero
and one period models suggests that past studies that
have not lagged the independent variables may be making
correct interpretations. In Nigh (1985a), Lag One and
Lag Two models supported the same hypothesis. But one
year lag models are more adviseable for future studies
utilizing more precise corporate data.

What follows is a summary of this study's findings.

187

It should be noted that this study does not infer firm
behavior from aggregate data. All interpretations are at
the aggregate level. Therefore its conclusions are only
tenable for all or collective U.S. manufacturing firms

behavior at the aggregate rather than individual firms.

SYNOPSIS OF FINDINGS

1) Exporting behavior of U.S. manufacturing firms
is not positively related to political risk. No
relationship or a slightly negative relationship tends to
exist between the exporting and political risk. The
relationship is the same for less developed as well as

developed countries.

2) In less developed foreign markets a significant
negative relationship holds between U.S. manufacturing
firms investment activities (FDI) and their perception of
political risk. For developed nations however, a
significant negative relationship) does not exist, but
there is, never- the-less, a negative relationship which

is not significant.

3) In developed countries market size is the most

significant explanatory variable for exporting and

188

foreign direct investment behavior of U.S. manufacturing
firms when compared to political risk and market growth

rate.

40 In. less. developed. countries international
executive's perception of political risk explains more
variance for foreign direct investment behavior of U.S.
manufacturing firms . Political Risk does not explain as

much variance exporting decisions.

5) In less developed countries political risk is
not. 'more closely associated with fOreign direct
investment than exporting decisions of U.S. manufacturing

firms.

6) In developed countries market size has more of
an influence on exporting than on fOreign direct

investment by U.S. manufacturing firms.

DISCUSSION OF FINDING
Finding One

This study had expected to find a positive
relationship between U.S. manufacturers exports to
country-markets with a high level of political risk as a

substitute for the more risky FDI. However the

189

conclusion based on the hypothesis test is that no
relationship existed or a slightly negative relationship
for LDC markets with exports dropping slightly as
political risk increased.

Intuitively one would expect that as the perception
of political risk in a foreign market rises,
manufacturing investors would shift to modes less
vulnerable to risk. This conclusion does imply that
investors do not prefer exporting activity as political
risk increases. Indication is that firms cut down on both
FDI and Exporting as political risk increases. There is
some indication 'that. joint ‘venture and licensing
arrangements are common responses to political risk. The
international executives interviewed overwhelmingly
agreed that they would operate in high risk markets via
joint ventures or licensing.

The reviewed adaptation framework. (Root 1982,
Agtamel, Robock 1974) also suggests that joint venture
will be a common mode of operation. If that is the case,
then exporting as a less risky mode of operation (as
theory' will suggest) will increase as political risk
increases. It. could. therefore :mean. that ‘while
theoretically exporting' is considered the least risky
mode of operation, (ie. less to lose) there may be other

factors blocking it as an adaptation strategy-for

190

instance changes in exchange rates,etc.

The term joint venture also is vague. What really
constitutes a joint venture? If a U.S. investor and a
foreign investor agree to enter a joint venture
arrangement whereby the U.S. partner is responsible for
only 5% of operations costs while the foreign partner
covers the remaining 95%, that theoretically is a joint
venture. But the scenario just described is definitely
involves less risk, than exporting to a foreign
market.Exporting typically involves loss of the cost of
items shipped (exported) if funds are not repatriated.
Companies that maintain elaborate foreign-based marketing
departments are more vulnerable than. those that rely
exclusively on export middlemen to market their products
in the foreign counrty. In such a case joint venture
rather than exporting becomes the least risky
alternative.

It thus may be that U.S. manufacturers prefer and
indeed solicit this kind of joint venture arrangement as
a means to avoid political risk. The U.S. Department of
Commerce3 definition of foreign direct investment also
lends credence to the suggestion that joint venture may
now be a term meaning minimal commitment in the host
market. The Commerce Department3 defines direct

investment as "ownership interest by a single U.S.

191

investor of 10% and above." This percentage strikes this
author as quite low given the definition of FDI as "full-
scale commitment of resources." Also joint venture
attempts to insulate operations from typical political
risks (ie. takeover) by involving local capital and
attaining local identity.

A framework thus may exist whereby as political risk
increases, U.S. manufacturing firms who export and those
that invest opt for joint ventures and licensing
agreements. As will be discussed next, exports is not a
risk free activity.

Exporting generally involves producing products at
home (modified or unmodified) and transporting it to the
foreign market. As goods move from home country to host
country, payment, rewards, compensation flows in the
reverse order.

Recently the monetary flow has not been reliable nor
predictable due to foreign exchange shortages and host
government economic policies in the case of DCs.

Repatriation of earnings may not be a major obstacle
for FDI since the earnings, if restricted, can be
reinvested. The economic objective of increasing the
'value of the firm' is still achieved when funds are
reinvested rather than repatriated as earnings.

Another consequence of high political risk on

192

exporting activity of U.S. manufacturers is the
incurrence of unusually high costs of operations. These
costs, as earlier mentioned, include costs of obtaining
information, transfer costs, costs of product re-design
due to change in product standards, costs of unsalable
inventory, etc. Some may be costs imposed by restrictive
govrnment policies such as tariffs and barriers that crop
up frequently in a plitical crises or periods of foreign
exchange shortage. These costs could rise dramatically in
high political risk countries to make exporting a high
cost activity - more so than joint ventures.

The exports statistics used to measure exporting
activity could distort the results in some cases. Those

statistic4

includes government and non-government exports
of domestic and foreign merchandise whether it involved a
commercial transaction or not. Thus U.S. government food
aid to a country such as Bangladesh is counted as
commercial exports. Unfortunately the commerce
department does not tell us what percentage of exports
can be considered as government originated.

Kobrin (1978) regressed an index of previous export
involvement on LOG manufacturing FDI and concluded that
previous export involvement was a prerequisite for

investment in a foreign market. FDI, in fact, may be a

reaction to "maintain existing markets after barriers to

193

continued import are raised? (p.117) If, according to
Kobrin, previous export involvement is really a
determinant of foreign direct investment, then a positive
relationship should exist between exports and political
risk, since exporting will be chosen mode of operation to
reduce political risk.

One aspect of the conclusions of the export studies
is confirmed by finding number one. Involvement in
exporting activity is not a hasty affair. It is preceded
by cautious planning and execution. It is "gradual
acquisition, integration and use of knowledge about
foreign markets and operations, and on its successively
increasing' commitment to foreign. markets." (Johnson &
Vahle, 1977, p.23)

This study does not however, confirm Cavusgil &
Nevin's statement that "export marketing is usually
considered to be the first step in the process of
internationalization." (p.68) There is ample evidence
from personal interviews that licensing and joint venture
are at least equally used as modes of initial market
involvement and as fallback position as political risk

increases.

194

Findinq Two

This study expected to find a negative relationship
between foreign direct investment and political risk. A
significant negative relationship was found for LDC while
no relationship was found for DC markets. The
homogeneity test showed that conclusions are most
reliabLe if LDC and DC data are pooled and investigated
separately. Also Levi (1979) excluded data for developed
countries so as not to bias the relationship between
political instability and FDI in less developed markets.

That LDC markets are riskier cannot be doubted. The
majority of nationalizations and takeovers have been

5 Recent studies that have

reported in these markets.
investigated political risk and Manufacturing FDI have
observed some relationship. Nigh (1985a & 1985b) showed
that conflictive political events was negatively related
to FDI in both LDCS and DCS. This study does not make
any distinction (as Nigh did) between conflictive and
cooperative political events. However it serves to note
that earlier studies did not find any relationship
between the two. Finding number two does confirm Nigh's
finding with respect to LDC markets.

It is also not surprising that no relationship was
found between political risk and U.S. manufacturing FDI

flows in DCS. Interviews with executives in Michigan

reveal that a different set of criteria might be used in

195

accessing investment decisions in DCS and LDCs.
Political riSk was rarely mentioned when the executives
were asked, "what factors will affect your mode of
operations in your DC markets?"6 However, when the same
question was posed for LDC markets, the response was
different. For LDC markets, political risk was usually
mentioned first, while maintaining a competitive
presence, market size, and 'fOllowing the J0meses' were
the most highly ranked influencing factors in DC markets.

The conclusion from the empirical analysis and the
personal interviews is that political stability is taken
for granted in DCS while political risk is a major factor
in LDC markets.

Political instability (if it is perceived to exist
in a host market) affects profitability. Green & Smith's
(1972) conclusions were that while Mining and Petroleum
FDI profitability' was. positively related. to political
risk, Manufacturing FDI's profitability wasn't related.
This conclusion supports the negative relationship
between U.S. manufacturers FDI flows and political risk.

U.S. manufacturing firms operating or investing in
Western Europe know that the probability of expropriation
and domestication occurring is very low. A shift in

concern therefore occurs away from. political risk. to

196

business risk. The U.S. MNCS therefore is concerned with
risks of producing and marketing a competitive product in
an unfamiliar market.

Market size then is a significant factor for DCs as
the T—Scores suggest. All studies have attested to the
importance of market potential. (Goodnow & Hansz 1972,
Bennett & Green 1972, Kobrin 1978, Nigh 1985) Therefore
finding number two is not surprising.

This study, however, does take issue with those
studies that have concluded that no relationship exists
between political risk and U.S. manufacturing FDI.
Difference between this study and those in the areas of
research design, sampling selection and political risk
index used, could partially explain the contradictory
conclusions.

By pooling cross-sectional and time-series data
reliability is increased. (Bass & Wittink 1975, p. 414)
Bennett & Green, 1972 and Kobrin, 1976 used cross-
sectional data, a design which he noted thus, "problems
are exacerbated by the need to evaluate a historical
phenomena cross-sectionally."(p.37) Bennett & Green also
acknowledged this weakness in their research design.

Whether the perception of political risk by U.S.
investors in less developed country markets is a true

measure of actual political risk in those markets does

197

not really bias the conclusions. What is important is
that these perceptions which are sometimes based on
'soft' data rather' than 'hard! data. do influence 'the
actions or the decision makers.

Ample evidence on in-house assessment of political
risk was found during the personal interviews. In
response to the question, "Is your firm involved in any
systematic analysis of political risk before entry and
after entry?", most responses were positive. When the
next question, "How is the political risk factor in your
country-market determined?", was asked the majority of
the responses were that it was in-house. Much reliance
was placed on information by executives of subsidiary
firms already in the host market. Those firms not
already in the target market sought outside source such
as publications, other firms executives, etc.7

Executives do base their decisions on their findings
which. bothers on 'gut feelings' rather used rigorous
political risk analysis. It is therefore unlikely that
the conclusions was due to the political risk index used.

The conceptual development is fully supported by the
findings. Most authors (e.g. Root, 1982) admit that some
relationship can be expected between political risk and

FDI flows. Cateora (1983) uses a broad definition for

198

political risk and admits that it will affect investment.
Finding number two is also supported by the survey
studies. Basi (1963) and Aharoni (1966) survey's reveal
that political risk along with market potential are the
two most important considerations for investments in
foreign markets. The NICB 1969 study listed political
risk as a major obstacle to private foreign investment.
The conclusion is that political risk is a major

factor only for LDCs.

Findinq Three

The survey studies (Basi 1963 , Aharoni 1966) have
concluded that two major factors affecting the investment
behavior of U.S. firms are market size and political
risk. Some studies (Bennett & Green, 1972) have
concluded that market size is the overwhelming
determinant of investment behavior. There has also been
suggestions that the influencing factors may vary for
LDCS and DCS. In fact, finding two supports that
contention.

Finding number three tests to see if market size
plays a more important role than political risk in the
exporting and foreign direct investment behavior of U.S.
manufacturing firms. Market size definitely contributes

more than political risk in explaining exporting behavior

199

of U.S. manufacturing firms. It is not surprising that
exporting is attracted to large markets of developing
countries. To understand this finding, one has to
understand the motivations for exporting.

The personal interviews reveal that reasons for
engaging in exporting are profitability potential and to
dispose of excess inventory. It is possible the U.S.
exporters associate a large market size with
profitability potential. Green & Smith (1972) failed to
establish any relationship between U.S. manufacturing
profitability and political instability. It is
possible, however, that a positive relationship exists
between profitability and stability and market size.

After reviewing the export literature Glover & Cady
(1983) have commented on the motivation to export as

thus:

"Many firms began exporting with no definite
corporate objective for the exporting program.
Rather, exporting is a response to external
stimuli such as an unsolicited order.

However, among those firms which did plan an
export development strategy, short-term
profit was NOT a key motive for exporting.

Rather, market diversification and long-term

200

growth - expected contribution to long-term

profitability - were the principal factors

motivating experimental exporting."(p. 8)

Other studies have indicated that motivations for

exporting are; adverse domestic market conditions,
example, loss of domestic market share, or market
saturation. The studies certainly support the position
that market size and exporting should be strongly
associated. We have already discussed why political risk
is not a significant factor in entering developed country
markets.

For foreign direct investment behavior, again,
market size explains more variance than political risk.
This is indicative of the minimal role of political risk
in developed country' market investment. decisions.
Political risk can thus be a function of level of
development. Its importance depends on investors
perceptions about the level of development of the market.
Political risk may not therefore be a discrete country by
country phenomena. That is, LDC markets are perceived to
be high risk while DC markets are stable.

Piper's (1971) conclusions that political variables
are of minimal concern to investors could be applicable
only to DCS. Further the cross-country studies that have

included all nations in their sample may be biasing the

201

insignificance of political risk in developed markets.
Situmeang's (1975) conclusions could be interpreted in
that light.

Bandera & White (1968) found a statistically
significant correlation between the U.S. FDI in EEC
countries and national incomes measured by GNP. EEC
countries constitute the developed country block used in
this study. Behrman (1962) attributed this to a desire
to penetrate the growing market which is defined in terms
of level and growth of GNP in host countries.
Scaperlanda & Mauer (1969) using U.S. data on FDI in the
EEC for the period 1952-1966 collaborated the market size
hypothesis. They showed that tariff rates was not a
statistically significant explanatory variable for
exports. Goldberg (1972) contradicted the findings
somewhat by attributing investment propensity not to the
size of the EEC market but to its growth rate.

Unlike the contradiction which exists in the role of
political risk, the studies on the relationship between
market size and FDI are unanimous as to its role.
Another explanation is that developed countries offer
guarantees on FDI against political risks which are
availed by their investors in many cases (Agarwal,

Jamuna). This may mask the true relationship.

202

Finding Four

So far the conclusions from findings suggest a
negative relationship between U.S. FDI flows and
political risk only in less developed countries and a
positive relationship between U.S. FDI flows and market
size in DCs. Finding four tries to compare the relative
importance of the contributions of political risk and
market size in the exporting and foreign direct
investment behavior of U.S. manufacturing firms. It
shows that political risk explains more variance than
market size for foreign direct investment decisions of
U.S. manufacturing firms in LDCS. However for exporting
behavior of U.S. manufacturers in LDCs no significant
explanatory power existed for political risk and market
size but one did for market growth rate. This may
indicate some LDCS in take-off stage are becoming
"developing" countries.

That less developed countries exhibit a high level
of political risk is not suSrprising. More
expropriations have been recorded for the LDCS than DCS.
(Weekly, 1977) A causal observation of the political

risk indexes show that LDC nations score systematically

203

higher than DCS. For instance the average in 1985 for
LDCs using BERI index is 67 while DCs is 38 on a 100-
point scale.

Schwartz (1976) found that both EEC and LAFTA
investments were significantly related to sales of U.S.
owned foreign subsidiaries. However, with regard to
market size hypothesis there were some differences
between the two regions. Whereas the absolute size of
the market of the host countries emerged as the primary
external determinant of FDI, in the case of the EEC, it
was the growth of the market in the LAFTA which took the
position of the most important determinant of U.S.
investments. Most Latin American Free Trade Association
members can be classified as LDCS.

Bennett & Green (1972) however controlled for level
,of development, but did not establish any relationship
between political risk and FDI flows . A negative and
non-negative relationship was tested for between FDI and
political risk in LDCS and DCs respectively. Their

hypothesis was rejected for both. They commented thus:

"There is no significant relationship
between the variables for either the
developed or less developed nations in

either condition. While this was

204

expected for the developed nations, the
lack of a relationship for LDCS
necessitates the rejection of the
hypothesis." (p.185)

There is conceptual evidence about the effect of
political risk in LDCS. Since FDI can be riskier than
licensing and joint ventures, it isn't surprising then
that political risk explains the greatest amount of
variance in FDI activity. In fact, the incidence of
foreign exchange and repatriation of earnings
difficulties is more pronounced for LDCS than DCS.
Brazil, for instance, has a long list of items that
cannot be imported. The need to conserve foreign
exchange resources has prompted many developing nations
in recent years to tighten controls over profit and
capital repatriation by foreign owned enterprises.

The constant shift in policy is a possible cause for
the finding. Gerald M. Feldman of the Industry and Trade
Administration's Investment Policy' division. has
commented, "the shift in the climate for investment in
most of the developing world was more extreme than that
of the developed world, and the liberalization in many
LDCS have been even more dramatic." (p.115) Such changes
could have adverse or favorable consequences for

investment. However it occurs too dramatically to be

205

believable. Peru, for instance, has made substantial
modifications in it laws governing repatriation of
profits, labor relations, petroleum exploration, etc.
Zaire has reversed its earlier policy of widespread
nationalization of foreign-owned enterprises, and has
returned most of the holdings to their previous owners.
(Feldman, 1978)

Such dramatic changes do not convince U.S. investors
of the non-vulnerability of their operations. Another
issue is that. of market size. 'The 'market size for
individual LDC countries (except for Brazil, Argentina,
Mexico, India and. Nigeria) are relatively small thus
discouraging' major commitments either' by exporting or
foreign direct investment.

Finding four also shows that political risk and
market size did not provide significant explanations of
exporting behavior in LDC markets. Market growth rate in
this case was the major factor. Yet market growth rate
has not played any significant role in explaining the
behaviors in question. Some studies have not mentioned
that market growth rate or market potential affects
foreign direct investment decisions, but none have
conclusively shown it affects exporting.

Although the conclusions about the motivations for

exporting (i.e. market diversification, market

206

development, profits) could rightly be argued to be
highly correlated to market growth rate in LDCS.
Absolute market size in those markets are relatively low,
hence the emphasis shifts to the next best measure -
market potential.

The export studies have concluded that a significant
factor determining the level of commitment, and
interpretation of the outcome of experimental exporting
was management's expectation of the value of exporting.
Firm growth is a key determinant of export behavior for
those that systematically plan for exporting. For those
that started exporting based on an 'unsolicited order',
continuation of the exporting activity could be based on
the rate of market growth. A case in point is Nigeria
that exhibited tremendous growth in the 19705 due to rise
in oil prices (which accounts for the majority of foreign
exchange revenue). During this period exports grew
tremendously even to the extent of congesting the ports.
During the same period FDI flows did not grow nearly as
much. Although not all of Nigeria's exports and FDI were
from U.S. sources it does show that market growth and
foreign exchange availability could be sufficient to
increase exports received by a host country. However

when it comes to FDI, market growth does not play a major

207

role; but rather, perception of political risk does.

The personal interviews did not isolate the
differential impact of market growth rate and political
risk on exporting and investment respectively. The
smaller firms involved in exporting indicated that they
were concerned about political risk factors even in the
markets they exported to. Earlier on a case had been
built for why exporting is not and should not be a risk-
free activity.

The conclusions here are that the findings can only
be tentative. Perhaps using a second order partial
correlation that removes the effects of political risk
and market size from market growth would provide more
insightful and conclusive evidence of the factor
determining exporting behavior of U.S. manufacturing

firms in LDCS.

Finding Five

In trying to determine the true variance explained
by political risk for exporting versus FDI behavior it is
necessary first to control or eliminate the effects of
market size and market growth rate. A casual examination
of the data indicates that country markets that have a
high level of political risk also have relatively small

market size while those with low political risk have high

208

market size. For instance in 1983, the combined market
size of the LDC nations in the data set was
$1,221,050,000r0008 (and this included nations such as
Nigeria, India , Brazil, Mexico) while that for DCs is
$2,786,800,000,000 or about 69.54% of the total.9 The
average GNP/capita for 1983 for LDCS was $1,914 while
that for DCS averaged $11,079.10 During the same year
the ratio of political risk in the ten LDC nations
averaged 61.4% and DCs 38.7%.11

Further evidence that a correlation could exist can
be obtained from the studies that have controlled for
market related variables in determining the relationship
between FDI and some notion of political risk. Kobrin
(1978) controlled for market related factors in comparing
flows of FDI and focused anti-regime violence (one
dimension of interstate conflict). Bennett & Green
(1972) also controlled for market size (measured as per
capita GNP).

Although finding four shows that political risk
significantly explains FDI behavior in LDCS, market size
and market growth rate could be distorting the findings.
The hypothesis about the substitutability of exports for
FDI in high political risk markets is rejected if either
or' both relationships are found to be insignificant.

Although the null hypothesis was accepted the fact that a

209

significant negative relationship was found between FDI
and political risk strengthens the conclusions of finding
four.

This study had expected political risk to cause U.S.
manufacturers to substitute exports for FDI. The results
show that exports do not substitute for FDI even in LDC
markets (where it had earlier been shown to be a
significant explanatory variable for investment). Thus
exports and FDI as modes of operation do not substitute
for each other in LDCS.

The findings from the personal interviews are
particularly relevant here. The international executives
interviewed indicated a preference for joint ventures and
licensing if political risk was determined to be high.
The joint venture arrangement usually was with a local
investor who was deemed to have political clout. The
selection criteria for local partners sometimes includes
intangible factors such as tribal origin or accessability
to rulers.

This finding cast doubts on the incremental method
of involvement developed by Cavusgil and Nevin (1980).
That framework identifies exporting as the first
activity in foreign. market involvement. Increasingly

executives favor joint ventures or licensing as modes of

210

initial involvement and adjustment to political risk.

Although Goodnow & Hansz (1972) used a multi-
indicator model, the findings of this study casts doubt
on their conclusions. According to Goodnow & Hansz
(1972), "As firms move from 'Hot' to 'Cold' country
clusters, they move away from the use of licenses and
joint ventures partners while they make significant
greater ‘use of strategies involving' decreased control
over sales such as overseas agents and distributors as
well as U.S. based intermediaries." (p.45) A cold
country in the Goodnow & Hansz study is one that is low
on market opportunitylz, low on political stability'13
low on cultural unity, high on legal barriers and high on
geocultural distance.

Root (1982) relied on Goodnow & Hansz's study in
stating that "a company ordinarily initiates
international business with a low-risk entry mode, which
is almost always some form of export . . . , alternative
modes are considered only if exporting is not possible."
(p.179) Root describes this approach as the "pragmatic
rule". This study suggests that the pragmatic rule
doesn't always seem to apply to U.S. manufacturing firm's
strategies in LDC markets for exports and FDI. It is not
clear what Root means by low risk, but he makes

references to political risk (p.65) and states that

211

exporting is a learning experience. Learning experience
is supposed to reduce vulnerability to political risk.
The relationship between perceived risk and experience is
shown in Figure 14. Learning could conceivably reduce
business risks, but not political risk. The contractual
entry modes accordingly should be used "when exports were
no longer possible to a target market with the sudden
imposition of tariffs or quotas or when exports were no
longer profitable with the appearance of more intense
competition." (p.97) This finding suggests that
licensing and other intermediate modes may be used more
often as entry and operation modes than the literature
suggests. Since licensing and joint ventures may involve
lower political risk than foreign investment or
exporting. Also licensing and joint venture has been
used as low risk vehicle to serve U.S. firms in foreign
markets during periods of instability.

The executives interviewed indicated that the
primary disadvantage of licensing (i.e. loss of patent
rights) after a number of years is not a major
consideration. To license, a company has to have a
technology, trademark, or a company name that is
attractive to potential foreign users. (Root 1982, p.68)
The executives said that usually they enter into

licensing agreements for periods that are considered to

212

 

 

 

 

 

time—I

Perceived
rietas

camxued
to<kmes-
tic risks

0 Knowledge of foreign
country market
Figure 14.--Exporting as a Learning Experience

SOURCE: F.Root, Foreign Market Entry Strategies
(Amacon, 1982).

213

be the useful life of the technology. Since technology
is changing very quickly, and competitive response
faster, product or technology obsolescence is occurring
in less time. Hence a foreign licensee that hangs on to
a patent for a long time may find himself with obsolete
technology.

In conclusion this finding shows that in country
markets with high political risks exports are not
substituted for foreign direct investment. Generally both

fall but FDI more than exports to LDCS.

Findinq,Six

Although not central to the study, it was necessary
to show the substitution effect between exports and
foreign investment attributable to market size. Since
market size significantly explains exporting and FDI
behavior only for DCS, data for DCS was used.

The finding suggests that a substitution effect
between exports and FDI exists. However contrary to
expectations, exports was more correlated to market size
than FDI. The model (Figure 6) had expected that in
large markets, FDI would be the preferred mode of
operation, hence a strong positive relationship. Very
strong ~empirical and conceptual evidence exists for

expecting a strong positive relationship between FDI and

214

market size. The survey studies indicated that market
size ‘was a. primary determinant of foreign investment
behavior. (see Basi, 1963; Aharoni, 1966) All the
empirical studies on political risk and FDI have obtained
strong relationships between FDI and market size.

On the export side, market size is not mentioned as
a motivating factor. However support for the finding can
be found in Czinkota & Johnston (1981). They outlined
proactive and reactive motivations for exporting behavior.
(Table 39)

Particularly interesting is the reactive
motivations. Motivations such as declining domestic
sales, excess capacity, and saturated domestic markets
could be indicative of U.S. investors search for large
markets as export targets. Since the DC markets have
large markets and convertible currencies exporters are
attracted to their markets. Another plausible
explanation for the suprising relationship could be the
short-term orientation policies of U.S. investors. Short
term orientation would lead decision makers to look for
short term benefits obtainable from exporting.

Bruce R. Scott (1984) and Hays 8: Abernathy (1980)
have decried the emphasis on short-term financial returns
rather than long-run competitiveness. Blame has been

placed on results oriented criteria for evaluation of

215

TABLE 39.--Factors Motivating the Export Effort

 

 

Proactive Motivations Reactive Motivations
Exclusive Information Competitive pressures
Managerial urge Over production

Unique products Declining domestic sales
Profit advantage Excess capacity

Marketing advantage Saturated domestic market
Technological advantage Proximity to ports

DISC

Tax advantages

 

Source: M. R. Czinkota and W. J. Johnston, "Segmenting U.S.
Firms for Export Development." Journal of Business
Research 9 (1981): 353-365.

 

216

performance. Since U.S. exporters typically wait until
domestic markets are mature or saturated it is
conceivable that only markets where there is the highest
probability of increased profits and sales are selected.
Exporting behavior tends to ensure short-term financial
results, while foreign direct investment ensures long-run
competitiveness.

The personal interview results indicated that global
competitiveness was a major motivation for FDI. A
surprising number of firms (mostly' parts and. machine
manufacturers) followed their major domestic customers
into their foreign markets.

Perlmutter's (1969) behavioral framework
(Ethnocentric, Polycentric, Geocentric Framework) which
was later expanded to ethnocentric, polycentric,
regiocentric and geocentric (E.R.P.G) (Wind, Douglas &
Perlmutter, 1973) can provide some explanation. This
framework attributes a company's internationalization to
management attitudes. An ethnocentric company especially
views an overseas operation as a means to "dispose of
surplus domestic production." It is possible that such
attitudes still exist thus foreign markets are viewed as
a 'dumping ground'.

In conclusion therefore, the significance of market

217

size in exporting behavior is outlined. U.S.
manufacturing firms would rather export to large-sized

country markets than invest.

NOTES

 

1. Mr. John T. Gallagher,President, Barker Perkins,
4-16-86

2. Mr. Joe Mack, II, President,Fruehauf Int'l, 4-25-86
3. See Appendix B for explanation
4. See Appendices A & B

5. See, "A profile of recent cases of nationalization &
take-overs of foreign enterprises", United Nations
Economic & Social Council (New York: United Nations
A19716, 1974. or Weekly, James, (1977),
"Expropriation of U.S. Multinational Investments,"
MSU Business Topics, 25, (Winter),27-36.

6. See Appendix J for complete questionnaire.

7. See Kobrin, S., (1980),"Political Risk:A Review and
Reconsideration", Journal of International Business
Studies, 67-79.

8. This was computed from the data used for analysis in
this study. This GDP at factor cost obtained from
National Accounts Statistics (Appendix B)

9. This is GDP at factor cost obtained from National
Accounts Staatistics.

10. This is GDP at factor cost obtained from National
Accounts Statistics.

11. This is obtained from BERI index (see Appendix K)

12. Market opportunity was measured as total population,
percent adult male literacy, televisions/1000 people
radios/IOOOpeople, GNP annual growth rate, GNP/capita
etc.

13. Political stability was measured as: # of years of
independence, type of government, # of pressure
groups major wars etc.

CHAPTER VI

IMPLI CATIONS OF RESULTS

In this chapter the author intends to examine the

contributions of this study. It is organized into two
sections - theoretical implications and policy
implications.

THEORETICAL IMPLICATIONS

Concerning the role of political risk on the mode of
operations, this study has shown that U.S. executive's
perception of political risk plays a major role in the
mode of operations decisions in developing countries
only. It has resulted in lower U.S. manufacturing FDI
flows to LDCS. On the other hand, political risk does
not significantly explain exporting behavior in LDCS. No
significant relationship was established between
exporting behavior and U.S. executive's perception of
political risk.

For DC markets, political risk played a minimal role
in explaining exporting or FDI behavior of U.S.
manufacturing firms. Market size proved to be a strong
explanatory factor for both exporting and FDI.
Therefore, the importance of political risk depends on

218

219

the level of development of the country market in
question. The findings of the survey studies is
partially substantiated (for LDCs only). On the other
hand, the cross country studies that claim that political
risk has no effect on FDI flows also are partially
substantiated (for DCS only).

The differing effects of political risk on LDCS
versus DCS suggests that U.S. investors perception of
political risk may be a group phenomena. Since other
studies have shown that firms do not engage in any
systematic analysis of’ political risk (Kobrin, 1980).
The implication is that. political risk could be a
regional rather than country phenomena. Firms that
export as a first entry strategy do so primarily due to
company policy and the unique nature of their product.
For instance, firms in highly technological fields tend
to export and are reluctant to engage in joint ventures,
licensing or FDI, due to lack of foreign expertise in
some areas.

Licensing and joint ventures seem to be increasingly
preferred. as modes of operations in foreign. markets.
Exporting may not provide the learning experience
necessary to eventually maintain a strong competitive
presence in a foreign market. Licensing and joint
ventures are preferred. because they' are faster, they

provide learning experience and they ensure against

220

political as well as business risk. Feldman (1978)
observed that "the number of joint ventures involving
American capital overseas has grown substantially across
a wide range of industries, and the vast majority of
established U.S. subsidiaries abroad are moving toward
compliance with host government requirements for foreign
equity holdings."(p.114) Also some of the indirect
foreign operations are a source of imports to serve U.S.
markets. When the regression models are considered it is
surprising' the small amount of ‘variance explained by
political risk, market size, and market growth rate for
FDI compared to exporting. For all models the variance
explained ranged from 3.87% (Table 28) to 21.52% (Table
18) compared to 7.74% (Table 24) to 53.59% (Table 13).
The significance of this lies in the importance that both
empirical and conceptual studies have placed on the three
variables in explaining the foreign investment decision.
Alternatively few studies have considered the explanatory
power of the three for exporting behavior. The very
strong relationship between exporting and market size is
particularly noteworthy for DC markets. Plausible
explanations can be found in conceptual frameworks in
international marketing such. as the Ethnocentric,
Regiocentric, Polycentric, Geocentric (E.R.P.G)

framework.

221

An implication of the findings is the importance of
home environmental factors in U.S. firms' decisions to
invest in a foreign market. U.S. government support is
important. An executive noted that although Mexico was a
risky market, his company would still maintain operations
there as long as the U.S. government was not willing to
let that country go down.1

This suggests that the decision to invest is more

complex than theory postulates. Each determinant plays a

relatively minor role in the ultimate decision. FDI
theory which has been fragmented and needs to be
integrated. Possibly a multi-factor model would be
appropriate. .John Dunning's Eclectic. Theory (Dunning

1981) of international production can be seen as an
appropriate start. By enlarging the theoretical
framework of explanatory factors into home and host
country characteristics, all the determinants are
captured in on framework.

This study shows that a relatively small amount of
variance is explained by certain host country
characteristics such as political risk, market size and
market growth rate. The findings also point to the
importance of behavioral dimensions and attitudes of
decision makers. Apparently their perceptions,
impressions, management style and negotiation play

significant roles.

222

Finally, theory suggests that desire for control
plays a significant role in the mode of operations
decision. Channel literature has pointed out the
advantages of maintaining control within the channel
Root (1982) says "channel control is significant because
with control a company can more nearly develop a channel
that meets its specifications."(p.70) The role of
control in international marketing is not clearly
outlined in the literature. The preference for licensing
and joint venture could signal a decline in the
significance of channel and source control in
international marketing.

The consideration of modes of operations as a
strategy is significant for understanding investors
actions. Axinn (1985) has argued that the key question
is, "Why do firms choose exporting as a means of
obtaining growth or other goods instead of alternative
available strategies?." (p.147) Likewise the study
suggests that the basic question is, "What factors affect
a firms choice of mode of entry or operations?" rather
than, "Which mode of entry or operation is the first, or

last?".

Not much is known about how firms integrate the
mode of involvement/ entry/ operations decision into the
decision to go abroad . Do firms select a

country/region/area ,then decide the mode to use? Or do

223

firms first decide the mode before selecting suitable
foreign markets? For instance a firm that has a policy of
exporting first to access a foreign market. Will it then
choose foreign markets that are conducive to exporting?
Or does it choose the market first and then determine
whether it is more accessible by exporting, investment or

joint venture mode of operation.

POLICY IMPLICATIONS

For U.S. multinationals, exporting does not have to
be the first mode of entry used to access foreign
markets. Failure to develop an export program does not
imply that other options cannot be explored. Other modes
of operations could be used to reduce vulnerability and
achieve firm objectives.

Diversified firms can use a different mode of
operation for different products or product lines. MNCS
can fully integrate their mode of operations strategy
into their strategic marketing or global sourcing plans.

LDCS have been making concerted efforts to attract
foreign investment, through relaxation of restrictive
policies to no avail. Some of the liberalization has
been too drastic to be real. A policy that focuses on
other modes of operations could be more rewarding in the
short run. A policy of stability of economic policies

will be desirable in the long run.

224

FUTURE RESEARCH DIRECTION

Now that the role of political risk has been
empirically classified and its practical value imputed,
future research can be directed towards testing the
revised conceptual model (Figure 15). Model testing will
require the empirical examination of the potential
strategies suggested by the research. Breakdown by
industry, firm and commodity' will provide significant

insights.

Notes

1. Mr. Joe Mack, II, President, Fruehauf Int'l, 4-25-86

225

.xHOBOEmuh Hmsummosoo pomw>mm .mH wusmfim

 

 

 

 

 

 

 

 

 

nus-Iona!— vacuum
9530.. 3.01600 «0 5331.393—
conuluesq «33.33250 Inauoauu-ou tog
a!“ can Eng-anus: unsung
00.150- 313 :31 :8 33:029-2— 8.3.4538
sedanguuusu i509 noun-Juan cogs-in. .
30938 onus- uxl 03009:» nus-Infidnvou acoucoo 303 J
.2638» condo £2.95 a on: ax: 3: 3.338 ..- uuagoo
39598 a .3033». ex: undone 00934 395.50 «aqua
0503.— E #0.» no ouch :53: conga.“ wanna A
fiasco-um.” 9:95.25.“ H335" noun-nu Aaouufiom .

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

m 62323de 263‘s and.
I can; .

F?.....Bu..I 22.52: 'W) :3 . - - — um:— *
r as. as: 5...... (.1

_...I§§ .... 42...... If

i) 9.68:: .292. 6a.. 1 3483*. _I|I|I_§. )_n
w ..fl aim M, ..-: T... u.

 

 

 

 

 

 

1

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

T
P UZHHROAXH Egg __ ad

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mzozéuno nngmux m: 23.3.30.— — firms 8 a8: — L Ego ..
6252328 “Gaussian .3 g
.8 ago:

 

 

 

 

 

 

 

 

'1

 

 

 

 

GLOSSARY OF TERMS

FOREIGN DIRECT INVESTMENT

Refers to acquisition by domestic companies of
foreign-based operating facilities, such as factories,
hotels, banks. The investor buys specific tangible
assets that possess the capacity to produce either goods
or services outside the investors home country. It
represents a commitment that is not only long term but
that also implies managerial control over assets or
enterprise (Kolde, ). The Department of Commerce

defines FDI as equity investment over ten-percent.

MANUFACTURING FOREIGN DIRECT INVESTMENT

FDI that relates to manufacturing only. It includes
the following industries: food products, chemicals and
allied products, primary and fabricated metals,

machinery, transportation equipment.

MODE OF ENTRY

Any institutional arrangement that makes possible
the entry of a company's products, technology, human
skills management, or other resources into a foreign

country. (Root, 1982 p. 5)

226

227

MODE OF OPERATIONS
Subsequent modes used to maintain presence in a

foreign market after initial entry.

MODE OF INVOLVEMENT

Same as mode of entry.

EXPORTING
A behavior or process by which a firms products is

moved across national boundaries.

EXPORTS
The product(s) a firm moves across boundaries. It
is a result of exporting, and typically measured in

dollars.

LESS DEVELOPED AND DEVELOPED COUNTRIES

A general term used to classify nations based on
such measures as income per capita, literacy rate,
productivity levels, cultural unity, etc. The United
Nations 'Group of 77' is often referred to as less
developed while the Western nations are referred to as

developed.

228

POLITICAL RISK

The subjective probability that certain political
decisions will be taken (or certain political events will
occur) which will perceptibly change the business

environment. (Cracco, 1972)

POLITICAL INSTABILITY
Destabilizing events or actions in a foreign

country.

POLITICAL EVENTS
An occurrence or action usually government initiated

that affects firms operation in a foreign market.

INCREMENTAL SCHOOL

An academic school of thought that views modes of
entry as involving gradual and incremental commitment of
resources. They postulate that a firm will enter a

foreign market via exporting and ultimately invest.

STRATEGIC SCHOOL

An academic school of thought that views modes of
entry as a function of certain factors. According to the
strategic school, a firm will enter a foreign market via

either mode.

229

CROSS-COUNTRY
Studies that have used countries rather than firms
as their sample. Usually aggregate data is collected and

analyzed.

CROSS-FIRM
Studies that use firms, rather than countries as
their sample. Usually data is collected through personal

interviews and/or questionnaires.

BUSINESS RISK

Risks associated with the consumers and operations.

COUNTRY-MARKET
A term used when a country is treated as a separate

market for analytical purposes.

APPENDICES

230

APPENDIX A

EXPORT DATA

Coverage

The export statistics reflect, in general, both
government and non-government exports of domestic and
foreign merchandise including non-monetary gold and silver
from the U.S. Customs territory (includes the fifty states,
the District of Columbia, and Puerto Rico) and the U.S.
Virgin Islands to foreign countries, 'whether' the export
involved a commercial transaction or not. Excluded from the
statistic are:

1) shipments to U.S. Armed Forces and diplomatic
missions abroad for their own use;

2) shipments between the United States and Puerto Rico
and. the U.S. and its possessions (including’ the ‘Virgin
Islands), and between these outlying areas:

3) exports from U.S. possessions;

4) intransit shipments through the United States when
documented as such with customs, and

5) transactions not considered to be of statistical
importance such as personal and household effects, temporary
exports, low-valued or non-commercial exports by mail and

issued monetary coins of all component metals.

231

APPENDIX A (cont'd.)

Exports of domestic merchandise refers to
commodities which are grown, produced or manufactured in the
United States, and commodities of foreign origin which have
been changed in the United States from the form in which
they were imported, or which have been enhanced in value by
further manufacture in the United States.

Exports of foreign merchandise consists of commodities
of foreign origin which have entered the United States as
imports and which, at the time of exportation, are in

substantially the same condition as when imported.

Source

The official U.S. export statistics are compiled by the
Bureau of the Census primarily from copies of shipper's
export declarations. The month of exportation is generally
based on the date when shipments leave the United States.
(For vessel or air shipments it is the date when the carrier
departs or is cleared from the port of export). However due
to processing problems there is an average carryover of
about 3 to 4 percent (in terms of value) of the shipments
from the actual month of exportation to a subsequent month,

usually the succeeding month.

232

APPENDIX A (cont'd.)

Valuation

The value reported is the equivalent to a f.a.s. (free
alongside ship) at the U.S. port of export, based on the
transactions price, including inland freight, insurance and
other charges in placing the merchandise alongside the

carrier at the U.S. port of exportation.

Country of Destination

The country of destination is defined as the country of
ultimate destination or the country where the goods are to
be consumed, further processed, or manufactured, as known to

the shipper at the time of exportation.

Errors

Effective from March 1979 statistics, data for shipments
value $501 - $999 (formerly $251 - $999) are estimated based
on a 50% sample of such shipments. Sampling errors can
generally be expected to be less than 2% fOr value totals
over ten-million dollars and less than 2% for value totals
between three-million and ten-million dollars. The sampling
error will almost always be less than possible errors due to

rounding.

233
APPENDIX A (cont'd.)

Sources of Errors

Sources of error include the reporting and/or
processing of information as to commodity classification.
However, the procedures used to compile both the import and
export statistics include clinical and computer processing
checks designed to protect the accuracy of the statistics to

the fullest practicable extent.

Citation

Ft 410, U.S. Exports, Schedule E, Commodity by Country,
U.S. Dept. of Commerce, Bureau of Economic Analysis
(Washington D.C.: U.S. Government Printing Office)
years 1973-83

234

APPENDIX B

U.S. INVESTMENT ABROAD

According to the Bureau of Economic Analysis, DOC
investment is of two kinds: direct and portfolio. Direct
investment is defined strictly from the viewpoint of a
primary or controlling equity owner. An interest of a U.S.
person of less than 10% is not counted as part of direct

investment but rather as portfolio investment.

DEFINITION

From 1950 to 1961 direct investment was defined as; 1)
ownership by a single U.S. investor (including an associated
group of investors) or at least 25% of the voting interest
in a foreign business enterprise, or 2) ownership, by
several U.S. investors collectively, of at least 50% of the
voting interest in a publicly held foreign business
enterprise in which no one U.S. investor owned as much as
25%. In a few instances, interests of slightly less than
25% are included where important management relationships
were known to be associated with the interests. Beginning

in 1962, all ownership

235
APPENDIX B (cont'd.)

interests by a single U.S. investor of 10-25% were also
included in direct investment. In 1970-71, the only two
publicly held affiliates of significance were dropped from
the direct investment universe because the 50 percent
criteria was no longer met. Since then, that criteria has

not been used in defining direct investment.

COVERAGE

Collection of direct investment, unlike exports, is not
based on estimate; but rather, it is a census intended to
cover the investment universe. Reports are obtained form
U.S. parent companies and their foreign affiliates.

Universe estimates were computed for each country-
industry cell shown in the published tables. Universe
estimates of interest, dividends, earnings of unincorporated
affiliates, royalties and license fees, service changes and
rentals are usually collected.

The presentation format shows data on the U.S. direct
investment position abroad, equity and intercompany account
outflows, and reinvested earnings of incorporated foreign
affiliates. Data on direct investment position is further
broken down into:

1) Manufacturing, which has recently been
divided into six industry groups,

a. food products

2)
3)
4)
5)
6)

7)

236

APPENDIX B (cont'd.)

b. chemicals and allied products
c. primary and fabricated metals
d. machinery
e. transportation
f. equipment
Mining and Smelting
Petroleum
Transportation
Communication and Public Utilities
Trade

Etc.

The primary document utilized in this dissertation is:

Survey of Current Business, U.S. Dept. of Commerce
(Washington, D.C.: U.S. Government Printing Office, 1974-83)

237

APPENDIX C

NATIONAL ACCOUNTS STATISTICS

DEFINITION

GDP at market prices is the market value of the
product, before deduction of provisions for the consumption
of fixed capital, attributable to factor services rendered
to resident producers of the given country. It is
identically equal to the sum of consumption expenditure and
gross domestic capital formation, private and public, and
the net exports of goods and services of the given country.
It differs from the gross national product at market prices
by the exclusion of net factor incomes received from abroad.

GDP at factor cost is the value at factor cost of the
product, before deduction of provisions for the consumption
of fixed capital, attributable to factor services rendered
to resident producers of the given country. It differs from
the gross domestic product at market prices by the exclusion
of the excess of indirect taxes over subsidies.

The form and concepts of the statistical tables

generally conform, for countries with market economies,

238
APPENDIX c (cont'd.)

to the recommendations in A SYSTEM OF NATIONAL ACCOUNTS,
STUDIES IN METHODS, Series F, No. 2, Rev. 3.

To compile the large volume of national accounts data,
the statistical office of the United Nations each year sends
to countries and areas with market economies a national
accounts questionnaire requesting recipients to indicate
where the scope and coverage of the country estimates differ
for conceptual and statistical reasons from the definitions
and classifications recommended in the SNA.

The basic data is obtained at current price series and
is converted into U.S. dollars and per capita levels by
respectively applying the corresponding exchange rates and

population figures.

THE UNITED NATIONS SYSTEM OF ACCOUNTS IN PRACTICE

In A System of National Accounts and Supporting Tables
the hypothetical system of nine standard accounts obtained
by setting up three accounts for each of these sectors is
for practical reasons modified in a number of ways. These
modifications both simplify the presentation and concentrate
attention on the formation of the more important national
aggregates. As a result, the system of accounts presented
in the report appears as follows:

Account 1. Domestic product. This account is

essentially the consolidated production account
of the economy.

239

APPENDIX c (cont'd.)

Account 2. National income. This account shows
the national income as a redistribution of the
value added in production and is largely a
consolidation of the appropriation accounts of
enterprises.

Account 3. Domestic capital formation. This
account represents the consolidated capital
reconciliation account of enterprises. In
addition to the capital formation of enterprises
proper, it includes that undertaken on behalf of
households, private non-profit institutions and
general government agencies.

Account 4. Households and non-profit institutions.
This account is divided into two parts, a current
account, and a capital reconciliation account.
Capital formation undertaken on behalf of households
and non—profit institutions together with that of
unincorporated enterprises is recorded as capital
formation taking place in the enterprise sector
(Account 3) financed by a transfer from the present
capital reconciliation account.

Account 5. General government. As with households
this account is divided into two parts, a current
account corresponding to an appropriation account,
and a capital reconciliation account. The treatment
of capital formation, in this case on behalf of
general government, is also similar.

Account 6. External transactions. This is a
supplementary account bringing together all
transactions with non-residents. It is divided
into a current account and a capital reconciliation
account for recording separately external
transactions appearing on the current and capital
reconciliation accounts of the individual sectors.

THE STANDARD ABLES

The system of supporting tables outlined in the report

constitutes a tabular rearrangement of the system of

240
APPENDIX C (cont'd.)

accounts itself, supplemented by detailed classifications of
the major flows. The tables presented for various countries

in this publication follow closely this system of supporting

tables.

241

APPENDIX D

COUNTRIES SELECTED FOR THE STUDY

 

COUNTRY LESS DEVELOPED DEVELOPED
Argentina X

Brazil X

Colombia X

Denmark X
France x
Germany X
India X

Indonesia X

Italy X
Mexico X

Netherlands X
Nigeria X

Norway X
Peru X

Phillipines X

South Africa X

Spain X
Sweden X
Switzerland X
United Kingdom X

242

APPENDIX E

LETTER OF INVITATION SENT TO COMPANIES

Department of Marketing and

Transportation Admin.

East Lansing, MI 48824-1121
January 22, 1986

ADDRESS
Dear Mr. :

I am currently researching the degree to which exports,
licensing, joint ventures, and foreign direct investments
are emphasized by U.S. manufacturing firms in foreign
markets with high political risk. This is part of my
dissertation for the doctorate degree in Marketing at
Michigan State University. I am also interested in other
factors that will determine the mode of operations used
in your foreign markets.

My research design involves conducting personal
interviews with Michigan businesses active in the
international marketplace and your firm has been chosen
based on a set of criteria as part of the sample. I would
therefore like to ascertain your willingness to
participate in this survey. The personal interviews will
last between 60-90 minutes and will be conducted in the
months of April and May, 1986. The specific days will be
Monday, Wednesday, or Friday.

This research promises to contribute to theory and
practice in this important strategic choice decision
area. You will NOT be required to disclose any
confidential company information as I am concerned mainly
with your perceptions. Please find enclosed a response
sheet and a self addressed stamped return envelope. Your
immediate response will be appreciated.

Thank you.
Yours sincerely,

Sam Okoroafo

243

APPENDIX F

RESPONSE SHEET ENCLOSED WITH LETTER
A. YES, I am interested and will be available in
APRIL for a personal interview. Call me

to set up an appointment.

NAME OF INTERVIEWEE

 

TELEPHONE NUMBER

* Please enclose any background information if

available.

B. NO, I am not interested and will not be

available in APRIL for a personal interview.

COMMENTS

 

 

 

 

 

244

APPENDIX G

FOLLOW-UP LETTER

Department of Marketing and
Transportation Admin.

315 Eppley Center

Michigan State University
East Lansing, MI 48824-1121

February 15, 1986

ADDRESS

Dear Mr.

Last month I wrote you a letter requesting a personal
interview in April. Unfortunately I have not received any
reply. To refresh you, I am investigating the degree to
which exports, licensing, joint ventures and foreign
direct investment are emphasized by U.S. manufacturing
firms in high versus low risk country-markets. This is
part of my dissertation for the doctorate degree in
Business Administration. The interview will last between
60-90 minutes and will be conducted in the month of April
on a Monday, Wednesday or Friday.

I would kindly appreciate your letting me know your
decision one way or the other in order for me to plan
ahead. Please disregard this letter if you have already
responded.

Thanks very much for your time.

Yours sincerely

Sam Okoroafo

 

245

APPENDIX H.--Interview Schedule

 

 

Name and Location of No. of Name and Position of Date, Time of

Business Employees Interviewee Appointment

1. Asgrow Seed 1,000 Earl L. Burkybile 4/8/86
Corporation Product Manager 1 p.m.
Sub. of Upjohn Int'l Division
Company

2. Motor Wheel Inc. 1,200 A. N. McCotter 4/10/86
Lansing, MI. Vice President 2 p.m.
Sub. of Goodyear Marketing & Sales
Tire & Rubber Co.

3. Baker Perkins Inc. 750 John T. Gallagher 4/16/86
Saginaw President 10 a.m.

4. Duramettalic Corp. 563 James S. Ware 4/21/86
Kalamazoo Chairman & CEO 9 a.m.

5. Durametallic Corp. 563 4/21/86
Kalamazoo Foreign OperatiOHS 12:30 p.m.

Manager

6. Fruehauf Int'l Ltd. 36,000 Joe Mack, II 4/25/86
Detroit President & CEO 9:30 a.m.

7. Ex-Cell-O Corp. 16,000 David Riechard 5/6/86
Troy Vice President 2 p.m.

Export Operations

8. Newcor, Inc. 700 Clifford 0. Bath, Jr. 5/6/86

Troy President 3:30 p.m.

 

246

.osH .COADMOAanm mmmcflmsm pauo3eco .sowuwom sum

 

.Hmmsd .A Hms0>sh >0 UDHAQEOU .mwfiuucsoo smflmuom ca meuwh snowH0E< m0 uouomuflo .m

 

 

 

.ufiouumo ..UcH mcowumoflansm xowm .>uouomuwo muonsuommscmz Goodnow: .H “condom
cowaafin MHOOB osfinomz MD .mocmHumsuoz .MHUCH >0HB
ovH.Hm woodmcm ammuouea ooo.VH mcmeumw .mosmum .momsmu .muoo OIHHDOIxm w
mfimumxm means»
cofiHHHE nommscms consummucH hone
mmH.mmw mfiwumwm uwusefioo oom Esflmaom .momcmo .osH .HOO3oz h
ooumsmamx
mHooe kumum .MD .Esfimamm .QHOU
mafixomm new mammm mom muommmsfim .>smshmo .mcmsmo oaaamuumfimuso o
mcfimmmooum Eoceswx copes: 3mswmmm
Hmowsmno a 000m omn Monaco .UcH .sflxumm umxmm m
commemme
mommm UHECODmm scones m0 05m
0:0 0HD0u000> ooo.H >HmuH .xcmEHDO .mosmuh .QHOU pmmml3oumm4 v
mmcwumon
umHHou soeuofiuw cosmnoq «Show “maosumc0>
Iflucm .muume paws «unwaumupw3m «momsmo «mucoum awouuwo.tmuoo
Iwumamwu smasowsw> oom.mH “Edemamm “meamuumsfl «deflusomuc Homoznamumomm m
mmflHOmmwoom Hz .OCflmcmq
0cm muumm uouoz oom.H momsmo .00 H0053 uouoz N
mocmaumnuoz
coflHHHn unusemucoo “SsmEumo .MAHmuumsm ameuumsa DHOHDDQ
mvm.mm .mumaemuu .mxosue ooo.wm “momsmo «Heumum monommmcwm ..muou wsmnmsum H
mm mm mmmsams 0 who 6 momonmEm m 0 no cqawuo wsmmeoo mo
H . m m u 2 mo .oz 0 x z . m soHumooq com 0802 .02

 

H xHazmmm¢

10.

11.

12.

13.

14.

15.

247

APPENDIX J

PERSONAL INTERVIEW QUESTIONS

What is the nature of your business-mftg,dist,etc?

How long has your firm been involved in international
markets?

What was your motivation to seek foreign markets?

What products do you market abroad?

Sales level in foreign markets- estimate only?

What organizational structure is used- regional,
International,subsidiary, functional etc.

What mode(s) was used to enter your foreign markets?*

What factors determined the mode of entry used?*
What mode(s) was used for continued operations?*
What factors determine your mode of operations?*

Did you feel a need to change modes at any time?
If so, when and why?*

Is your firm involved in any systematic analysis of
political risk before entry into country-markets?*
after entry?

How is the political risk factor in your country-
markets determined?*

Suppose you came across some reliable information
about the political risk factor in your country-
market, how long would it take you to adjust your
strategy if at all?

In LDC markets, what factors would affect your
decision to export or invest?

248

APPENDIX J (cont'd)

16. In DC markets, what factors would affect your
decision to export or invest?

17. What overall strategy do you use- global, regional or
country?

18. Are the majority of your exports final products?
intermediate products?

19. Do you make use of foreign nationals in subsidiary
management?

249

APPENDIX K

BUSINESS ENVIRONMENT RISK INDEX DESCRIPTION*

Introduction

The Business Risk Service is specifically designed to
meet the needs of banks and companies with operations in
foreign countries. Executives responsible for profit
performance, strategic planning, cash management, and other
functions use the service on a day-to-day basis. Each of
the forty-eight countries listed below is analyzed
qualitatively in every issue. The factors behind the
ratings are discussed and are the bases for +5 Years
forecasts. The quantitative ratings included in the country
forecasts are based on a perfect country achieving a rating
of 100. The three measures are as follows:

1) Political Risk Index—A panel of approximately
75 permanent members submit their opinions on
socio-political conditions according to a
structured system which underwent trials during
1975-1978 before being published in the 1978-III
Issue.

2) Operations Risk Index-A panel of approximately
105 executives submit their ratings according
to a system conceived in 1965 and refined in the
following years until the ratings were first
published in 1972.

3) R Factor-A unique computer program has been
conceived to measure a country's capacity to
meet its obligations in convertible currency.
Each country rating utilizes more than 14,000
cells of data.

Countries Covered

AMERICAS EA§I_A§IA £23223
Argentina Australia Belgium
Bolivia Indonesia Denmark
Brazil Japan France

Canada Korea (South) Germany

250

APPENDIX K (cont'd.)

Chile Malaysia Greece
Colombia Phillipines Ireland
Ecuador Singapore Italy
Mexico Taiwan (R.O.C.) Netherlands
Peru Thailand Norway
United States Portugal
Venezuela Spain
Switzerland

United Kingdom

 

MIDDLE EAST & SUB-SAHARA
NORTH AFRICA AFRICA
Egypt Ivory Coast
India Kenya

Iran Nigeria
Iraq South Africa
Israel Zaire
Morocco

Pakistan

Saudi Arabia

Turkey

The Business Risk Service is published 31 March,

31 July, and 30 November each year.

Political Risk Index (PRI)

The concept focuses wholly on socio-political
conditions in a country by:

1)
2)
3)

4)

Creating a multicomponent system with flexibility
to weigh key factors.

Utilizing a permanent panel of experts with a
political science orientation rather than business.
Providing data which can move independently of
other BERI S.A. risk measures.

Three time periods are involved:

a. Present Conditions

b. +5 Years Conditions

c. +10 Years Conditions

Eipst Step in fine Sysgem. The expert rates the present

conditions for each of the eight causes shown below from
seven (no problems) to zero (prohibitive problems). Then,
the two symptoms are rated on the same scale in the present.

 

251
APPENDIX K (cont'd.)

The perspective is from the viewpoint of an international
corporation rather than private enterprise owned by
nationals. This subtotal involves a maximum of 70 for the
perfect country.

1) Six Internal Causes of Political Risk:

a. Fractionalization of the political spectrum
and the power of the factions.

b. Fractionalization by language, ethnic and/or
religious groups and the power of these
factions.

c. Restrictive (coercive) measures required to
retain power.

d. Mentality, including xenophobia, nationalism,
corruption, nepotism, willingness to
compromise.

e. Social conditions including population
density and wealth distribution.

f. Organization and strength of forces for a
radical left government.

2) Two External Causes of Political Risk:

a. Dependence on and/or importance to a hostile
major power.

b. Negative influences of regional political
forces.

3) Two symptoms of Political Risk:

a. Societal conflict involving demonstrations,
strikes, and street violence.

b. Instability as perceived by non-constitutional
changes, assasinations, and guerilla wars.

Second Step in the System. One or more of the causes
may have a very positive impact on the overall political
risk. The second subtotal of the system permits
discretionary use of 30 points. Experts typically allocate
approximately 20 to a low-risk country, approximately 10
for moderate risk, etc., and usually opt to allocate no
additional points to a prohibitively risky country. The
perfect country would receive a rating of 100 as a result of
steps one and two.

Forecast of 2R1. Steps one and two are repeated for
each of the +5 years and +10 years period. The points
awarded to a country under present conditions serve as a
basis for changes in the future.

Interpretatipn pf the Bapings, Four categories of

political risk have become apparent from usage:
1) 70-100 Low Risk. Political changes will not
lead to conditions seriously adverse
to business. No major socio-political

252
APPENDIX K (cont'd)

disturbances are expected.

2) 55-69 Moderate Risk. Political changes
seriously adverse to business have
occurred, but governments in power
during the forecast period have a low
probability of introducing such changes.
Some disturbances will take place.

3) 40-54 High Risk. Political developments
seriously advers to business exist or
could happen in the near future. Major
socio-political disturbances are
occurring periodically.

4) 0-39 Prohibitive Risk. Political conditions
severely restrict business operations.
Loss of assets is possible. Disturbances
are part of daily life.

Operations Risk IndeinORIl

The objective of ORI is to gauge the business
operations climate. There are two variables being measured:
(1) The degree to which nationals are given preferential
treatment, and (2) the general quality of the business
climate, including the political environment for day-to-day
business.

Definition of the Index. A permanent panel of
approximately 105 experts around the world rate present
conditions for the 15 criteria which make up a cross-section
of the country's business climate from 0 (unacceptable
conditions) to 4 (superior conditions). The criteria are
weighted to emphasize critical success factors, and this
expands the 15 wo a weighted total of 25. A rating of 4 on
each criterion gives a perfect environment of 100.

a. The quality of the panel members is the foundation
of the concept. Executives in companies, banks,
governments, and institutions volunteer their
ratings. All have extensive international
experience. Geographic distribution is worldwide.

b. A version of the Delphi method is used. Data are
from a permanent panel. The first reply prepared
by a panel member requires research and care in
matching the rating with the definitions of the
criteria. A panelist is supplied with his
previous reply and the overall panel average per
criterion as input for decisions on current
ratings.

253

APPENDIX K (cont'd.)

Forecast of ORI. Each panel member rates a country for
a +5 years period by giving a whole number for the overall
rating such as 70, 62, etc. The high and low ORI forecasts
are discarded, and a panelwide average is used.

Guidelines on Riskiness. The categories below have
been developed to assist in interpreting the ratings.

1) 70-100 Stable environment typical of an
advanced industrialized economy.
Problems for foreign businesses are
offset by the country's efficiency,
market opportunities, financial
infrastructure, etc.

2) 55-69 Moderate-risk countries with complications
in day-to-day operations. Usually the
political structure is sufficiently stable
to permit consistent operations without
serious disruption.

3) 40-54 High risk for foreign-owned businesses.
Only special situations should be
considered, e.g., scarce raw materials
or unusual profits.

4) 0-39 Unacceptable business conditions for
foreign-owned businesses.

Qpitepia,

Weighting
Political Continuity 3
Attitude: Foreign Investors & Profits 1 1/2
Nationalization 1 1/2
Monetary Inflation 1 1/2
Balance of Payments 1 1/2
Bureaucratic Delays 1
Economic Growth 2 1/2
Currency Convertability 2 1/2
Enforeceability of Contracts 1 1/2
Labor/Cost Productivity 2
Professional Services and Contractors 1/2
Communications and Transportation 1
Local Management and Partners 1
Short Term Credit 1
Long Term Loans & Venture Capital 2

254
APPENDIX K (cont'd.)

Factor R (B-FACTOR)

The purpose of R (for remittances and repatriation of
capital) is to estimate a country's capacity and willingness
for private foreign companies to convert profits and capital
in the local currency to "hard" currency and transfer the
funds. The computer program manipulates over 14,000 cells
of data and makes hundreds of calculations for each R rating
and the four subidicies described below.

Legal Framework Subindex (20% of R11 Each of the six
criteria are rated from 5 (best case) to zero (worst case)
and weighted by either four or three. The weighted total of
20 times 5 equals the perfect legal framework for R.

1) Laws as Written: Weighting
a. Profit and Dividend Remittances 4
b. Royalties, Fees, and Renumeration

for Non-Dividend Cash Flow Services 3
c. Repatriation of Capital _3
10
2) Actual Practices: Weighting
a. Practices on Dividends, Royalties,
and Other Periodic Compensation 4
b. Practices on Repatriation of Capital 3
c. Hedging Opportunities Against a
Devaluing Currency _;
10
Foreign Exchange Genezation Subindex (30% pf B), IMF

data published in International Financial Statistics (IFS)
are used. Certain statistics were converted to "standard
normal variates." This technique makes meaningful
comparisons possible despite an immense range in the data
across countries.

a. Current account performance is half of the
subindex, or 50 points. Breakeven receives 25
points. A rolling average is used to dampen
the impact of unusual years.

b. Three measures are used to award the 50 points
for capital flows. First, capital account
points up to a maximum of 30 are awarded by
establishing breakeven as 15. Second, capital
flows attracted by high interest rates adjusted
for risk earn a maximum of 10 points. Third,
"safe haven" currencies also attract capital,
and a maximum of 10 points are awarded.

255

APPENDIX K (cont'd.)

Accumulated lntepnational Reserves Subindex (39% pt 3).

a. First, months of coverage for imports of
merchandise and services are used as a means of
relating hard currency reserves to scope of needs.
The most coverage each year receives 50 points; the
fewest months earns zero.

b. Second, the international reserve total is added to
the London valuation for gold holdings to give a
complete total on reserves. A ratio is then
calculated using total public foreign debt as the

numerator.
Foreign Debt Assessment Subindex (20% of B), The

developing country data for public foreign debt published by
the World Bank are the basic sources. Industrialized
countries required several sources.

a. A ratio using gross domestic product converted
to U.S. s as the denominator is used to put the
debt into perspective with the economy. Creditor
nations receive all 40 points; the largest result
receives zero.

b. Capacity to service debt is measured by a ratio
of annual public foreign loan obligations
(numerator) and foreign exchange earned as the
denominator. 40 points go to the creditor
nations.

c. The "saturation factor" is a theoretical
calculation. A country reaches saturation when
debt service plus imports of petroleum equal
foreign exchange earned: this situation receives
zero, but five points or fewer is critical.

Biskipess and pne +5 gears Egzepasp, Risk categories

are the same groupings as for ORI. Data used for the
forecast is the result of regression analyses, research
staff input based on scenarios in the FORCE reports, trends
in the ratings, and senior staff judgment. Wholly
quantitative forecasts proved unrealistic.

* Printed with permission from: BERI SERVICE, S.A.
1355 Redondo Ave., Suite 10
Long Beach, CA 90804

256

APPENDIX L

LETTER OF THANKS

Dept. of Marketing and
Transportation Admin.

Michigan State University
East Lansing, MI 48824

July 7, 1986

ADDRESS

Dear Mr. :

I am writing to thank you for granting me audience for my
dissertation research. My interview with you was very
informative and is an important part of my research
design.

As I promised, at the conclusion of the study, I will be
sending you a synopsis of the major findings and
implications (about 5 pages). Please expect the abstract
in November/December- 1986.

Thanks again.

Yours Sincerely,

Samuel C. Okoroafo.

LIST OF REFERENCES

LI ST OF REFERENCES

Aharoni, Yair (1966), The Foreign Investment Decision
Process, (Boston: Harvard University Press).

Aldrich, Howard A. (1979), Organizations and Environments,
Englewood Cliffs, NJ: Prentice-Hall, Inc.

Andriole, S.J. and G.W. Hopple (1982), An Assessment of U.S.
Political Instability Research Methodologies, Marshall,
VA; Information Systems, inc., (September).

Axinn, Catherine (1985), An Examination of Factors that
influence Export Involvement, Unpublished PhD.

dissertation, Michigan State University.

Ayal, Igal (1981), "International Product Life Cycle: A
Reassessment and Product Policy Implications," Journal
of Marketing, Vol 45, (Fall), 91-96

Ayal, Igal and Ziff, Jehiel (1978), "Competitive Market
Choice Strategies in Multinational Marketing", Columbia
Journal of World Business, 13 (Fall), 72-81.

------ and--------(1979), "Market Expansion Strategies in
Multinational Marketing", Journal of Marketing, 43
(Spring), 84-94.

Azar, Edward and Swan, T. (1975), Dimensions of

Interactions: A Source Book to; the Study Q: Behavior

of 31 Nations from 1948 through 1973, Pittsburgh:
International Studies Association.

Bain, J. (1956), Barriers to New Compepition, (Cambridge,

Mass.: Harvard University Press).

Baldwin, Robert E. (1979), "Determinants of Trade and
Foreign Investment: Further Evidence", The Review of
Economics and Statist'cs, Vol. 6, Amsterdam, 40-48.

Bandera, V.N. and J.T. White (1977), "U.S. Direct

Investments and Domestic Markets in Europe", Egonomia
Internazionale, Vol 21, Genova, 117-133.

257

258

Basi, R.S. (1968), Determinants of United States Fnivate
Direct investment in Foreign Countries, (Kent, Ohio:
Kent State University, Bureau of Economic and Business
Research).

Beeman, Don R. (1978), "Business Policies and Strategies for
Reducing Political Risks in Developing Nations: A
Theoretical Paradigm for Empirical Testing", Paper
presented at 38th Annual Meeting of the Academy of
Management, San Francisco, California.

Behrman, Jack N. (1962), "Foreign Associates and Their
Financing" in Raymond F. Mikesell (Ed), U.S.;Private
and Government Investment Abroad, Eugene, 77-113.

Bennett, Peter D. and R.T. Green (1972), "Political
Instability as a Determinant of Direct Foreign
Investment in Marketing", Journa o eti
Research, Vol. 9 (May), 182-186.

Bergsten, C.F., Horst, T. and Moran, T.H., (1978), American
Multinationals and American Interests,(Washington DC:
Brookings Institution)

Bilkey, W.J. and G. Tesar (1978), "The Export Behavior of
Smaller Sized Wisconsin Manufacturing Firms", Journa;
of International Business Studien, (Spring), 93-98.

Boddewyn, J. and E. Cracco (1972), "The Political Game in
World Business, Columbia ourna 0 W0 s' ess,
January-February

Buckley, P.J. and Casson, M., (1976), he Futu f e
Multinational Enterprise, (London:MacMillan)

Buckley, P. J. and R. D. Pearce (1979), "Overseas Production
and Exporting by the World's Largest Enterprises: A
Study in Sourcing Policy", our a1 t t a

Dusiness §tndies, (Spring/Summer), 9- 20

Cavusgil, S.T. (1976), Onganizational Determinants of Finns
Export Behavior: An Empirical Analysis, PhD.

dissertation, The University of Wisconsin, Madison,
Wisconsin.

Cavusgil, S.T. and J.R. Nevin (1981), "Internal Determinants
of Export Marketing Behavior: An Empirical
Investigation", Journal of a et'n ese ,
(February), 114-119.

259

--------------- &---------------(1980)," A Conceptualization
of the Initial Involvement in International Marketing",
Theoretical Developments in Marketing, C.W. Lamb and
P.M. Dunne, eds., Chicago: American Marketing
Association, 68-71.

Cohen, J. and Cohen, P.(1983), Applied Multiple
Regression/Correlation Analysis for the Behavioral
Sciences, 2nd ed., (LondonzLawrence Erlbaum Associates,
Publishers).

Comanor, W.S. and T.A. Wilson (1967), "Advertising,
Marketing Structure, and Performance," this REVIEW
(Nov), 423-440

Cracco, E. (1972), "The Nature and Perception of Political
Risk for the International Corporation: An Exploratory
Analysis with Special Reference to Brazil", Unpublished
PhD. dissertation, Michigan State University.

Cutright, P. (1965), "Political Structure, Economic
Development and National Social Programs", American
Journal of Sociology, 70, (March), 537-550.

Czinkota, M.R. and W.J. Johnston (1981), "Segmenting U.S.

Firms for Export Development", gounnni pf Dusiness
Research, 9, 353-365.

Davidson, W.H. (1980), "The Location of Foreign Direct
Investment Activity: Country Characteristics and
Experience Effects", urna of te at'o us'n ss
Studies, 11 (Fall), 9-22.

--------------- and R. Harrigan (1977), "Key Decisions in
International Marketing: Introducing New Products

Abroad", golumbia gounnal of Wonld Dusiness, 12
(Winter), 15-23.

Dunning, J.H. (1981), International Produgtion nnd the

Rultinational Enterprise (London: George Allen &
Irwin), 72-108.

Durbin, J. and Watson, G.S.(1951), "Testing for Serial
Correlation in Least Squares Regression",
DiometFika,Vol 37, 1950,409-428, vol 38, 1951, 159-178

Feldman, G. M. (1978a), "Coping with New Challenges to
Investments Abroad",§omme;ge Amerign,Vol. III,No. 14,
(July 3), 2-5

Feldman, G. M. (1978b), "How U.S. Firms are Responding to
New Investment Controls Abroad", Dommence Amepipa, Vol.
III, No. 15, (July 17), 8-11.

260

Galtung, T. (1970), "Diachronic Correlation, Process

Analysis and Causal Analysis", Quaiity and Quanpipy, 4,
55-94.

Glover, K.A. and Cady, J.F. (1983), "Developing Foreign

Export as a Market Entry Strategy", MBS gese Senies,
Harvard Business School.

Goldberg, M.A. (1972), "Determinants of U.S. Direct
Investments in the EEC: Comment", The American Economic
Review, 62, Evanston, 692-699.

Goodnow, J. and J. Hansz (1972), "Environmental Determinants
of Overseas Market Entry Strategies", Journa; 0;

International Business Studies, (Spring), 33-50.

Green, R. and Cunningham, W.H. (1975), "The Determinants of
U.S. Foreign Investment: An Empirical Examination",

Management International Review, 15, 113-120.

Green, R. and Korth, C.M. (1974), "Political Instability and

the Foreign Investor", Qalifornia Management Review,
(Fall), 23-31.

Green, R.T. and Smith, C.M. (1972), "Multinational
Profitability as a Function of Political Instability",

Management International Review, 23-29.

Hage, J., (1975)," Theoretical Decisions Rules for Selecting
Research Decision: The Study of Nation-States or
Societies", Sociolo ' Me hods Resea ch, Vol 4 No.
2 (november),13l-l65

Grosse, R. (1985), "An imperfect competition theory of the
MNE", urnal of er tio a s' ess t, 'es, Vol
16, No. 1, (Spring),57-80.

Hair, J.F., Anderson, R.E., Tatham, R.L. and B.J. Glablowsky

(1979), Multivariate Data Analysis wipn Readings,
Petroleum Publishing Company, Oklahoma

Harrell, 6.0. and R.O. Keifer (1981), "Multinational

Strategic Market Portfolios", MSU aniness Ionics,
(Winter), 5-15L

Hayes, R.H. and W.J. Abernathy (1980), "Managing Our Way to

Economic Decline", Manyeng Dusiness Review, (July-
August), 67.

Hazard, J.L. (1977), Inansponpation; Management Fconenice
Roliey, (Maryland: Cornell Maritime Press, Inc.

261

Hirsh, S. (1976), "An International Trade and Investment

Theory of the Firm", Oxford Economic Fapers, 28, 258-
270.

Hood, N. and Young, S. (1981), The Economics e: The
multinational enterprise, (London: Longman Group Ltd).

Horst, T. (1972a), "The Industrial Composition of U.S.
Exports and Subsidiary Sales to the Canadian Market",
American Economic Review, 62, 37-45.

Horst, T. (1972b), "Firm and Industry Determinants of the
Decision to Invest Abroad: An Empirical Study", Review
of Economic and Statistics, 54, 258-266.

Horst, T. (1974a), "American Exports and Foreign Direct
Investments", Harvard Institute of Economic Research,
discussion paper. no. 362.

Horst, T. (1974b), The Theory of the Firm", Chapter 2 in

Dunning, J.H. (ed), Economic Analysis end phe
Multinational Enterprise, (London: George Allen and
Irwin).

Hull, H. C. and Nie, N. (1981), SPSS Update 2-2, (McGraw
Hill Book Company)

Johanson, J. and J. Uahlne (1977), "The Internationalization
Process in the Firm: A Model of Knowledge Developments
and Increasing Foreign Commitments", Qounnei of
International Business Studies, (Spring/Summer), 23-32.

Johanson, J. and Wiedersheim-Paul (1975), "The
Internationalization of the Firm: Four Swedish Case
Studies", ournal of a a eme tu ' 5, October, 305-
332.

Johnston,J. (1983), Fconometnic Methods, 3rd ed.,(New York:
McGraw-Hill Book Company).

Kelejian, H. H and Oates, W.E. (1981), Introduction to
Econometrics,2n edition,(New York: Harper & Row
Publishers).

Kirkpatrick, C. and M. Yamin (1981), "The Determinants of
Export Subsidiary Formation by U.S. Transnationals in
Developing Countries: An Interindustry Analysis", World

Development, 9, 373-382.

Kimberly, J.R. (1976), "Issues in the Design of Longitudinal

Organization Research", SocioIogicaI Metnogs eng
Research, Vol. 4, No. 3, (February), 321-347.

262

Kobrin, S. (1976), "The Environmental Determinants of
Foreign Direct Manufacturing Investment: An Ex-Post
Empirical Analysis", ournal of ntern ' ess
Stugies, (Fall/Winter), 29-42.

Kobrin, S. (1978), "When Does Political Instability Result
in Increased Investment Risk?", Columbia Senpnal o:
Dusiness, (October).

Kobrin, S. (1980), "Political Risk: A Review and

Reconsideration", Sournal of InternationaI Dusiness
Studies, 67-79.

Kravis, 1.8. and R.E. Lipsey (1982), "The Location of
Overseas Production and Production for Export by U.S.

Multinational firms",Sournal o: Internapional Economies
12, 201-223.

Levis, M. (1979), "Does Political Instability in Developing
Countries Affect Foreign Investment Flow? An Empirical

Examination.", Management International Reviey, 19, 59-
68.

Lipsey, R.E. and Weiss, M.Y. (1976), "Exports and Foreign
Direct Investment in Manufacturing Industries", (New
York: National Bureau of Economic Research, working
paper No. 131 (May).

MaClayton, D. M. Smith, and J. Hair (1980), "Determinants of
Foreign Market Entry: A Multivariate Analysis of

Corporate Behavior", Menegement Inpepnatienei Review,
20 (3), 40-52.

Mason,R.H., Miller,R., and D.R. Weigel (1975), Tne_Fepnenie_

of Internapionai Businessl John Wiley & Sons, Inc., New
York

Mann, M., (1966) 'Seller Concentration, Barriers to
Entry,and Rates of Return in tqirtygipdustries 1950-
1960", this Review. (August). 90’.0

Moore,W. (1963), Social Onange, (Englewood Cliffs,
N.J:Prentice-Hall)

Nehrt, L.C. (1970), The Rolitical Environment fie: Fopeign
Investment. New York: Praeger Publishers.

Nielson, A.M. (1980),"Global Marketing Strategy: A resource
mix concept", paper presented at the Acagemy Q:

Inpepnationai Susiness Annual Meeting

263

Nigh, D. (1985a), "The Effects of Political Events on U.S.
Direct Foreign Investment: A Pooled Time-Series Cross-

Sectional Analysis", qurnal of Internationai Dusiness
Studies, Vol. XVI, No. 1 (Spring), 1-17.

Ohlin, B., (1983) Interregional and International Trade
(Cambridge:Harvard University Press, Harvard Economic
Studies, Vol. XXXIX)

Orr, Dale (1984), "The Determinants of Entry: A Study of the
Canadian Manufacturing Industries", Review of Economics
end Statistics, LVI, (February), 58-66.

Perlmutter, H.V. (1969), "The Tortuous Evolution of the

Multinational Corporation", gqumbia SounnaI e: WorIg
Dusiness, (January/February), 9-18.

Piper, J.R. (1971), "How U.S. Firms evaluate foreign
investment opportunities", MSU Business Topics,
(Summer),ll-ZO.

Robinson, H.J. (1961), The Motivation and FIow o; Frivepe

Foreign Investment, International Development Center,
Standard Research Institute, Merlo Park, California.

Robock, S.H. (1971), "Political Risk: Identification and

Assessment", Columbia JounnaI pf World Dusiness,
(July/August), 6-20.

Rock, M.T. (1973), Cross-Country Analysis of the
Determinanps o: D.S, Foreign Investment in
Manufacturing in Less Developed Countries, unpublished

PhD. dissertation, University of Pittsburgh.

Root, F. (1968a), "U.S. Business Abroad and Political
Risks", MSU Business Tepics, Winter), 73-80.

264

--------------- (1968b), "Attitudes of American Executives
Towards Government Investment Opportunities", Economics

end Business Dulietin, Temple University, (January),
14-23.

--------------- (1982), LOW:

Amaco.

Rummel, R.J. and Heenan, D.A. (1978), "How Multinationals
Analyze Political Risks", Harvard Business Review,
(January/February), 67-78.

Scaperlanda, A.E. and L.J. Mauer (1969), "The Determinants
of U.S. Direct Investment in the EEC", The American
Economic Review, Vol. 59, Evanston, 558-568.

Schwartz, R.H. (1976), The Determinants of U.S. Direct
Investment Abnoad, PhD. dissertation, Texas Technical
University.

Scott, B.R. (1984), "National Strategy for Stronger U.S.

Competitiveness", Harvard Business Review,
(March/April), 77-91.

Shapiro, A. (1981), "Managing Political Risk: A Policy
Approach", golumbia Joupnal of Dnsiness, (Fall), 63-70.

Snavely, W.P., P. Weiner, K. Ulrich and E. Enright, (1964),
Export Survey 0: the Greater Rentgepe Anees, Vols. 1 &

2, University of Connecticut.

Situmeang, B.J. (1978), The Environmentei QonneIetes pf

Fepeign Direct Investment wipn Reference p0 Sonpneesp
Asia. Unpublished PhD. dissertation, University of

Oregon.

Truitt, J.F. (1970), "Summary of the Post World War II
Experience of American and British Investors in the
Less Developed Countries", on 81 e ' a

Studies, (Fall).

Thunell, Lars (1977), Rolitical RisRs in Inpennepionel

Dusiness: Investment Dehavior o; MuItinapionaI
Qopporepions. New York: Praeger Publishers.

Vesper (1979), "Entrepreneurship in Foreign Trade," on n

e: SmaIl Dusiness Menagement, (April) 5-11

Wallace,T.D. (1972), "Weaker criteria and tests for linear
restrictions in regression", Feenenep;iee,40,
(July),689-698.

Wells, L.T. (1966), "A Product Life Cycle for International
Trade?", Spurnal o: MarReting, Vol. 32, (July), 1-6.

265

Weekly, James,(1977), "Expropriation of U.S. Multinational
Investments, MSU Business Topics,25 (Winter), 27-36

Weidersheim-Paul, Finn, H.C., Olson and L.S. Welch (1978),
"Pre-Export Activity: The First Step in
Internationalization", Journal 0 Internat o
Dusiness Studies, (Spring/Summer), 47-58.

Wind, Y., S.P. Douglas, H.V. Perlmutter (1973), "Guidelines
for Developing International Marketing Strategies",

gournal of Marketing, Vol. 37, (April), 14-23.

Zaltman, G., LeMasters, K., and Heffring, M. (1982), Theory
Construction in Marketing: Some Thoughts on Thinking, (New
YorE: John Wiley & Sons, Inc.)