‘39 mumllluurllmlllmllrmlmmlmnl

1293 01810 3824

This is to certify that the
dissertation entitled

"A Single Finger Cannot Lift a Stone":
Local Organizations and Democracy in Mali
presented by

John Uniack Davis

has been accepted towards fulfillment
of the requirements for

PH) degreeinIPO/"Ah/Slane‘u

WM,

Major professor

 

Date 3/IS" 9?

MSU is an Affirmative ActtOn/Equal Opportunity Institution 0- 12771

 

LIBRARY
Michigan State
University

 

 

 

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

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

I

w“ “F9807?“

 

 

 

 

 

 

 

 

 

 

 

 

we mumps-Q14

“A SINGLE FINGER CANNOT LIFT A STONE”:
LOCAL ORGANIZATIONS AND DEMOCRACY IN MALI

By

John Uniack Davis

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Political Science

1999

ABSTRACT

“A SINGLE FINGER CANNOT LIFT A STONE":
LOCAL ORGANIZATIONS AND DEMOCRACY IN MALI

By

John Uniack Dayis

During the recent democratic transitions occurring in much of Africa. significant
attention has been devoted by academics and practitioners alike to the purported
democratic effects of civil society. Civil society is hypothesized to have both interest
aggregation and socialization effects that contribute to the consolidation of democratic
gains. Until now, however, relatively little attention has been paid to the effects of
associational life at the rural, local level where most Africans live. Using data from West
Africa’s Republic of Mali, this dissertation does three things to rectify this omission.

First, it uses 52 in-depth interviews with organizational leaders to compare local
organizations of southern Mali to Western notions of civil society. This qualitative
analysis examines the appropriateness of the term “civil society” to rural Africa.

Second, it uses regression analysis of a USAID data set of 173 community
organizations to establish the extent to which local organizations in southern Mali engage

the state. Moreover, it seeks to establish the determinants of both civic action behavior in

general as well as civic action success.

Third. regression analysis of individual-level political attitudes data (n = 252)
gathered among organizational leaders. village elders. and simple heads of household is
used to test hypotheses regarding the effects oforganizational participation on individual
attitudes.

The results of this dissertation can be summarized as follows. First. though rural
Malian organizations provide a host of useful services to their communities. they bear
little resemblance to idealized groups replete with democratic effects. Second. even the
relatively well-connected community organizations in the USAID data set engaged only
sporadically in civic action behavior with local or national officials. Moreover. nearly
half of those who did engage in such activity reported no success. Sound management
practices, systematic financial systems, and effective representation by a regional or
national non-governmental organization or federation were important to both civic action
behavior as well as to success in such an endeavor. W omen‘s groups were both less
likely to engage in civic action as well as less likely to report civic action success.

Finally, simple associational membership was found not to contribute to individual levels
of civic values and organizational leadership only had modest effects. In contrast. wealth
and village council membership had a strongly negative effect on civic values. These
results make sense given the relatively uninclusive nature of the organizations studied as
well as the fact that founding families and other established village interests have much to

lose if procedural democracy trickles down to the local level.

Copyright © by
John Uniack Davis
1999

To the farmers of Southern Mali

ACKNOWLEDGMENTS

This dissertation has been researched and written over a period of several years. and
many friends, colleagues, and mentors have contributed to the process. I could not have done
it without the assistance of the people and institutions listed below, though any errors and
opinions are mine alone.

First and foremost, I owe a profound debt to the residents of all the villages who
provided hospitality to me and several research teams during the course of this project.

The family of Nicolas Sidibé, Assistant Principal of the Lycée Technique of Bamako
and his wife, Mme. Sidibé Marie Diarra. provided me with a place to live and learn Bambara,
as well as a base from which to conduct my research. Seydou and Karim Coulibaly and their
eXtended family and the residents of Sanambélé in the Koulikoro Region demonstrated great
Warmth and hospitality in providing me a research base and teaching me about associational
life and cotton production.

My enumerators, F anta Diallo. Mohamed ag Kaltanana, F ablan Doumbia, Amadou
K0dio, and Idrissa Diakité endured difficult conditions and meager pay to conduct the 21-

Village political attitudes survey. Nicolas Sidibé II (nephew of his namesake, listed above)

vi

went above and beyond the call of duty time and again in serving as head enumerator and
research assistant.

Fadiala Kamisoko. Elisée Sidibé. and the field agents ofthe Cooperative League of
the United States ofAmerica (CLUSA) proved to be a great reservoir ofinfomiation on rural
Mali and assisted me in getting established in Sanambele. Bino Témé and Makan F ofana of
the Institut d’Economie Rurale (IER) provided logistical support in Mali in 1994. Che'ibane
Coulibaly, the director of the Institut Malien de Recherches Appliquées au Développement
(I MRAD) provided great intellectual and logistical support during 1994 and 1995. and has
remained a friend and mentor. Mamadou Niang and Francois Giraudy of the Compagnie
Malienne du Développement des Textiles (CMDT) served as great resource people on the
ins and outs of the cotton zone.

The Social Science Research Council (SSRC) furnished generous financial support
in the form of an International Predissertation Fellowship for the 1994 calendar year.
David Atteberry and USAID/Mali provided financial support for my data collection
in 1995. Anna Diallo, Erin Soto and the Democratic Governance Strategic Objective
(DGSO) Team at USAID were helpful colleagues during 1997 and 1998 and permitted me
10 use DGSO monitoring and evaluation data. Michelle Kuenzi and Bakary Doumbia were
Stimulating collaborators in the collection and analysis of the USAID data. Lynn Carter and
Management Systems International (MSI) recruited me to do the Mali work and have been
a Source of much logistical support for my work in Mali.
Several friends served as a support network during my fieldwork in Mali. Among
these are Christine Kolars, Moulaye Ely Diarra, Salif Diarra, Georges Dimithé, and Zeric

Kay Smith. Zeric has also been an insightful colleague in the study of Malian civil society.

vii

Sugato Dasgupta and Brandon Prins provided valuable advice on data analysis.
Kimberly Butler, Kimberly Ludwig. Gina Lambright. and Jeffrey Coupe offered useful
insights into African politics.

My dissertation chair, Professor Nicolas van de Walle. has been generous with his
time and feedback and has always shown great interest in my intellectual and professional
development. My committee. Professors Michael Bratton. Jeff Riedinger. and John Staatz.
in spite of full schedules and peripatetic ways, have shown enthusiasm for the project and
offered useful feedback. Other professors who showed interest in my work and provided
helpful advice included Darren Davis and Jim Bingen.

Many friends over the years have been sources of support in my career and personal
life. In particular, I should mention Nancy Mezey and Martial Dembe’lé. Both are moved
by issues of social justice and African development and have been sources of inspiration to
me.

My family, particularly my mother and father, taught me the importance of education
and have offered much encouragement over the years. My brother. Andy, is a model of

inSpired social action and his example always spurs me to evaluate the relevance of what I

am doing.

Finally, I would like to thank Jennifer for her love, encouragement. support, sense

of humor and for inspiring me to be a better person.

viii

LIST OF TABLES

CHAPTER 1

CHAPTER 2

CHAPTER 3

TABLE OF CONTENTS

Introduction
1.1 Problem Statement
1.2 Theoretical Issues: Civil Society and Democratization
1.2.1 Civil Society
1.2.2 The Interest Aggregation Function
1.2.3 Political Socialization
1.2.4 Research Questions
1.3 The Malian Democratic Transition
1.4 Politico-Administrative Context: Decentralization
1.5 Legal Context: Cooperative Law
1.6 Data and Methodology
1.7 Dissertation Structure

Characteristics of Local Organizations in Southern Mali
2.] Introduction
2.2 Background on the Cotton Zone
2.3 Organizational Characteristics in the Cotton Zone
2.3.1 Methodology
2.3.2 Organizational Characteristics
A. Associations Villageoises
B. Women’s Groups
C. Youth Groups
D. Hunters’ Groups
E. Summary
2.4 Characteristics of USAID Community Organizations
2.5 Conclusion

The Interest Aggregation Function of Local Organizations
3.1 Introduction

ix

xi

AHb—d

17
20
21
26
32
4O
43

46
46
49
52
52
59
63
72
79
83
87
90
101

104
104

CHAPTER 4

CHAPTER 5

APPENDIX A
APPENDIX B

BIBLIOGRAPHY

Data and Methodology
3 The Sampling Frame and Design
3.4 Results
3.4.1 Description of Civic Action Behaviors
3.4.2 Determinants of Civic Action Behavior
3.5 Findings

The Educational Function of Local Organizations
4.1 Introduction
4.2 Theoretical Issues
4.2.1 Culture and Institutions
4.2.2 Political Socialization and Political Learning
4.2.3 Organizational Participation and
Political Attitudes
4.3 Hypotheses
4.4 Methodology
4.4.1 Methodology Selection
4.4.2 Sampling
4.4.3 Operationalization
4.5 Descriptive Statistics on Organizational Participation
And Political Attitudes
4.5.1 Sample Demographics
4.5.2 Orientation Toward Politics
4.5.3 Political Knowledge
4.5.4 Political Efficacy
4.5.5 Political Tolerance
4.5.6 Political Trust
4.6 Results of Regression Analysis

Conclusion

5.] Introduction

5.2 The Fit Between the Notion of Civil Society and
the Cases Studied

5.3 Local Organizations and Interest Aggregation

5.4 Local Organizations and Their Educational Effects
5.5 Recommendations

5.6 Future Research Directions

5.7 Concluding Remarks

USAID Community Organization Questionnaire

Political Attitudes Questionnaire

LIST OF TABLES

Table 1 - Village Characteristics
Table 2 - Characteristics ofLocal-Level Associations in Southern Mali
Table 3 - USAID Internal Democracy Criterion 1: Membership Type
Table 4 - USAID Internal Democracy Criterion 2: Leadership Selection
Table 5 -USAID Internal Democracy Criterion 3: Existence of By-Laws
Table 6 -USAID Internal Democracy Criterion 4: General Assembly Attendance
Table 7 - USAID Internal Democracy Index
Table 8 - The CO Sample, By Gender and PVO Partner
Table 9 -Distribution of Target CO Sample by Region and PVC
Table 10 - Comparison of Target C05 and Control Group COS. by Sector
Table 1 1 - COs Volunteering Evidence of Legal Recognition
Table 12 - COS Showing Evidence of Rank-and-File Participation

Table 13 - Participation in Civic Action at the Arrondissement
or Commune Level

Table 14 - CO-Initiated Contact with Their National Assembly Representative
Table 15 - CO-Initiated Contact with a Development Parastatal

Table 16 - COS Reporting Civic Action Activity below the Regional Level
Table 17 - COs Reporting Civic Action Success

Table 18 - COs Reporting any Kind of Development Partnership

Table 19 - COs Reporting Developmental Partnership with State Entities

Table 20 - COS Reporting Collaboration with Intermediary Organizations

xi

57

94

95

97

100

124

125

137

I39

141

142

145

146

147

Table 21 - Correlation Matrix for USAID Democratic Governance Data Set 153
Table 22 - Ordered Logit Estimates: Determinants of Civic Action 157
Table 23 - Ordered Logit Estimates: Determinants of Civic Action 158

Table 24 - Ordered Logit Estimates: Determinants of Successful Civic Action 159

Table 25 - Survey Questions Used in Value Scales 188
Table 26 - Regression Results: Political Efficacy and Simple Membership 201
Table 27 - Regression Results: Political Tolerance and Simple Membership 204

Table 28 - Regression Results: Political Tolerance and Organizational Leadership 205

xii

Chapter 1

Introduction

Chaque heure qui passe apporte un supplement d'ignition au creuset ou fusionne le monde.

Nous n'avons pas eu 1e meme passe, vous et nous. mais nous aurons 1e méme avenir.
rigoureusement. L'ere des destinees singulieres est reyolue.l

Cheikh Hamidou Kane. L'aventure ambiguc‘ (p. 92)

1 .1 Problem Statement

Dramatic changes have swept across a shrinking world in the past decade. Primary
among these is the advent of democracy in many contexts previously deemed inhospitable
due to a lack of certain prerequisites. I ndeed, democratic transitions ofvarying degrees have
been undertaken in over half of Africa's countries. While the routinization of competitive
electoral processes gives reason to be hopeful regarding prospects for stable democratic
governance, the long-term consolidation of democratic rule is an even more challenging
undertaking. In this vein, many in the international donor community have chosen to
Promote and support local associations as the building blocks of democracy. For instance.
the United States Agency for International Development (USAID) is promoting democratic
governance in twenty-four sub-Saharan African countries, and most of these programs

include civil society capacity-building as an essential element of their programming. In fact,

in the Republic of Mali, the promotion of such capacity-building among community

 

 

“Every hour that passes brings a supplement of ignition to the crucible in which the world is being
fused. We have not had the same past. you and ourselves, but we shall have, strictly. the same
future. The era of separate destinies has run its course.”

1

2

organizations (C03) and local non-governmental organizations (NGOs) is the sole focus of

USAID'S "Democratic Governance Strategic Objective" (DGSO). In Mali alone. USAID is
spending nearly four million dollars a year on such programs.2 With such resources being
committed. examining the assumptions behind the supposed democratic effects of civil
society is essential. The present study can make an immediate contribution to debates on
donor policy while in the long term contributing to theories on interest aggregation and
political socialization in agrarian. non-Western contexts.

The potential beneficial effects of civil society can be roughly categorized as of two
types. Having an associational sphere distinct from the state can serve an interest
aggregation, or articulation, function. It does this by serving as a counterweight to the state,
by representing member interests, or by engaging in development partnerships with the state
that respond to community needs. Civil society participation can also serve to socialize
members to hold political values such as compromise, trust, tolerance, and efficacy, which

are believed desirable for long-term democratic stability. In this dissertation, I examine local
Organizations in the Republic of Mali in order to evaluate that country's nascent civil society
in three principal respects: First, how do local rural associations compare to Western
Conceptions of "civil society"? Second, how effective are they in performing the interest
aggregation function? Third, do these groups in fact serve the role of socializing their
members to hold values consistent with democracy?

This dissertation contributes to the literatures on both civil society and

democratization in at least three ways. First, it demonstrates the diversity and complexity

\

2 This includes all direct DGSO funding. plus additional resources earmarked for civil society support in Mali‘s
three northern regions.

3
of some of the organizations that make up the "third sector" in Mali. In particular. it
demonstrates the local level associational heterogeneity that exists in rural areas. While rural
areas are generally less heterogeneous than urban areas. much research has ignored the
wealth of associational structures existing in rural Africa. Second. the extent of the
connection between local organizations and the national political system is established.
While many previous studies examine national NGOs in the political process or local
Organizations in development programming and implementation. this study is innovative in
examining the link between truly local associations and the national political process.
Finally, the relationship between participation in the types of associations most familiar to
Mali’s rural citizens and their political values is described. Not only are these values
important in what they contribute to the national political system, but also in what they say
about the inclusiveness and social justice of the milieux where most Malians spend their
lives. The conclusions can be cautiously generalized to state-society relations in other
emerging African democracies. Insights are offered which bridge the gap between Western
civil society theory and the reality of the exciting stories unfolding on the ground in Africa.
AS the world becomes smaller, political scientists have an obligation to assess the fit between
Western assumptions and institutions with grassroots realities into which policymakers often
Wander uninformed.
This introductory chapter is organized in the following manner. This section is
immediately followed by a theoretical discussion of the key issues motivating the
dissertation. Next, the general case of the Malian democratic transition is introduced, along

With background information concerning the politico-administrative, legal. and

4

socioeconomic context of the study. I then describe the data and methodology to be used in
the course of subsequent chapters. I conclude the first chapter with a section describing the
organization of the remainder ofthe dissertation.

1.2 Theoretical Issues: Civil Society and Democratization

During the "third wave" ofdemocratization that has unfolded since the 1970's there

has been renewed interest in factors which facilitate democratic transitions as well as those
that contribute to a subsequent consolidation of democracy (Huntington 1991). Indeed. in
recent years. elites in most of the world's nation-states have come to profess that the best

form of government is a democracy. "Democracy“ is generally conceived as describing a

govemment which is responsive to all of its citizens. These citizens are viewed as political
equals, enjoy a broad array of civil and political liberties, and participate in the change of
governments through periodic. competitive elections (Dahl 1989).

Coupled with the interest generated by the "third wave." the end ofthe Cold War has
engendered an unprecedented level of American ardor for spreading the gospel of democracy
throughout the world.3 Recently, this evangelism has often come to focus on the purported
Salubrious effects of "civil society." For example. the United States Agency for International
Development (USAID) has placed a great deal of faith in the possibility of civil society
Organizations serving a vital role in promoting democratic governance in Africa. both in their
direct contribution to democratic socioeconomic development, as well as by inculcating
democratic habits and values in individual citizens (USAID 1994). Many other donors,

irlcluding bilateral aid agencies from Canada, Denmark, the Netherlands, and Switzerland,

k

3For a practitioner's perspective. see Talbott 1996. For a sympathetic academic point of view. see Allison and
Beschel 1992.

3

among others. explicitly promote democratization through civil society support in Africa.
Multilateral agencies such as the UNDI’ and the World Bank have become preoccupied with
the political context of development. showing great interest in such issues as governance.
decentralization. and civil society.

While procedural democracy has begun to take hold in many countries across the

continent, the long-term consolidation of these gains is dicey in contexts without a
democratic past and lacking pluralist institutions. The transition from a non-democratic
regime to one that is democratic involves multiple phases and can stretch over a long time.
One must not be excessively impatient with current transitions -- Great Britain’s transition
took over two centuries! A prerequisite of national unity is necessary, along with a
preparatory phase including a crisis and breakdown of a nondemocratic regime. followed by
a decision phase where the polity moves in the direction of the establishment of a democratic
Order. This is barely the beginning. however, it must be followed by a long process of
Consolidation (Rustow 1970, Sorensen 1993, pp. 40-43). Consolidation is

the process by which democracy becomes so broadly and profoundly

legitimate among its citizens that it is very unlikely to break down. It

involves behavioral and institutional changes that normalize democratic
politics and narrow its uncertainty (Diamond 1996. p. 238).
The key institutions of a consolidated democracy include a strong, well-structured executive
With the capacity to make and implement coherent policies, "settled and aggregative" party
SVStems, effective legislatures. and autonomous. competent. and professional judicial
SYStems (Diamond 1996. p. 239). A truly "civil" society can contribute to consolidation as

Well. It ideally constitutes a counter weight to state power while nevertheless engaging and

providing legitimacy to the state. In so doing, it plays an interest aggregation role that

6

strengthens democratic institutions. It also provides a venue for political learning and the
concomitant behavioral change that Diamond describes as essential for consolidation.
Nonetheless. Bratton and van de Walle have concluded that "in general. the institutions of
civil society perform less well in the early stages of democratic consolidation than they do
during transition." This is because the conclusion of a transition can have demobilizing
consequences for the voluntary sector. It is thus imperative that civil society be
reinvigorated during consolidation and given an important role (Bratton and van de Walle
1997, p. 255). The functions described by Diamond are critical to this phase and are
fundamental to this dissertation.

This dissertation examines the role of organizations in the ”voluntary sector" or "civil
society" in democratization. Because these terms are often used interchangeably to mean a
Variety of things, there is considerable analytical imprecision in theoretical discussions of
these issues. Therefore. before determining whether the much-vaunted positive effects of
civil society really exist or are nothing more than so much snake oil. we must be clear about
What is meant by the term.

1.2.1 Civil Society

Most analysts agree that civil society is an arena beyond the household in which

Organizations interact with the state and influence the state. but are distinct from the state and
do not vie for state power, as "political society" does. This is deemed to be a space where
Profit is not the primary objective and where a moral code espouses values more public than

thOse of what Banfield called "amoral familism" (Banfield 195 8, Azarya 1994). Durkheim

7

noted the acute sense of unease -- "anomie" -- experienced when human beings are not bound

to others beyond the family through a sense ofcommunity (Durkheim I933 .
For many, inclusion of an organization under the rubric of "civil society" depends
upon its relationship with the state. Many definitions hinge on identifying a public space that
is distinct from the state and political society but is linked to the household and the market.
Consequently. many object to including organizations that do not directly interact with the
state in the category of civil society (Bratton 1989, p. 428; Chazan 1992. p. 281; Ignatieff
1995, p. 135). Others employ a broader definition. seeing the essential democratic
characteristic of civil society in its being an alternative center of social organization,
regardless of the thematic content of its constituent organizations' activities (Putnam 1993.
1995; Fukuyama 1995).4 In any case. describing it as the space between the household and
the state without imbuing the definition with greater precision relegates it to being a residual

category.

Disillusionment with the African state as a vehicle for socioeconomic and political
development has increased the focus on the potential of societal actors. Because of the high
expectations placed upon the space between the household and the state, it has been easy to
slip into the use of language that reflects more the prevailing optimism for associational life
tham it does an empirical reality. For example, many authors use the term "civil society" as
synonymous with ”society" without making any explicit effort to make clear just what it is

that makes it "civil" (Azarya 1994, p. 88). The following analogy permits greater precision

4 . . . . . . .
In this vein. Fukuyama defines ClVll socrety extremely broadly, as "a complex welter of intermediate

Institutions. including businesses. voluntary associations. educational institutions. clubs. unions. media. charities. and
Lhurches” (Fukuyama 1995, p. 4)-

8

-- Civil society is to society as "democracy" is to "polity" -- each of the first terms in the
analogy refer to a restrictive subset ofthe latter terms. As Callaghy says. much of the recent
literature on Africa has "mistaken resurgent societies for 'civil societies.‘ with very
misleading consequences" (Callaghy 1994. p. 238). While Callaghy sees little theoretical
usefulness in the term "civil society." it can offer a useful normative and empirical
distinction from society at large.

Much of the confusion inherent in varying definitions of civil society stems from
conflation of positive Characteristics with a normative role with which many wish to entrust
it. In other words, some authors who have studied civil society treat it as being virtually
synonymous with "the voluntary sector," or the "third sector" which is made up of a wide
range of not-for—profit organizations between the household and the state. Laura Renshaw
at once illustrates both the carelessness with which the term is used as well as its normative
implications when she says that "the exact definition ofcivil society becomes less of an issue
once we acknowledge that we are not supporting civil society, per se. but those elements of
civil society that most reflect our vision and values" (Renshaw 1994).

In order to have a definition which is empirically useful, civil society must not only
be described by what it is not -- the household, market, or the state, but also by what it is.
COrrespondingly, and consistent with the positive connotation in the very label "civil
SOCiety," some classify only a democratic subset of the voluntary sector as belonging. For
eXample, Azarya describes a "civil" society as one in which there is a reciprocal respect for
rigl‘tts and "the social contract [is] entered upon to ensure the optimization of individual and

group needs and desires" (Azarya 1994, p. 90). In order to avoid relegating the term to the

 

'-"I‘_1I'l

usuu‘b It

.
I
“~ ohm
I." a. o
7",” g).
' Ti 1
"In >.lh \
.1. i,
p .
.‘ \

 

 

 

«4 , 'V'”
d y-
\ .
““' Al \
. r
P V 7.,
l‘.- ‘
N“A.‘-
~ \
_.
.- ins”
“ K.. .‘
- .
>1.
I .
...
. . ._
"'~..‘
.1.
.
'c
. ._ ‘
'\.A~ . ‘.
“~‘
..
. , .
e.‘ t, ‘
._“
.
A
I u
l '1
\,v
‘4
u.”
a
«.

 

 

\.

.\

. -.

s, \ ‘
‘. . .

u“
‘»

b.

h ‘l

u x.
l“
,_ .n

u"

.

. .1
._ -p
5

1»

‘u b

n ‘ ‘

,

~\‘ .

9

status of "a metaphor masquerading as a player" (Young 1994. p. 43). an effort must be made
to define it with greater precision and in a way that can be operationallzed empirically.

For example. Naomi Chazan has high aspirations for civil society that are only

surpassed in degree by the exclusivity with which she uses the term (C hazan 1992). While
acknowledging the blossoming of associational life which has occurred in Africa in recent
years, most of these nascent groups do not meet her criteria for inclusion under the rubric of
"civil society." In fact, many local groups that were founded in order to protect the
particularistic interests of their members can have distinctively illiberal consequences. she
argues. The dramatic increase in voluntary associations, self-help groups, and the like. is
encouraging, but "there is no axiomatic connection between the expansion of the voluntary
sector and the consolidation of civil society in Africa" (Chazan 1992, p. 283). Nolutshungu
agrees, pointing to the "lack of democracy within and among civil organizations"
(Nolutshungu 1992, p. 327). For this reason, Peter Lewis offers the precision that only
Participation in certain types oforganizations contributes to "the construction of a democratic
political space" (Lewis 1992, p. 34).

Civil society has been described as, at its best, having both pluralist and educational
functions in the stability and performance of a democratic polity (Hadenius and U ggla 1996).
In Serving as the locus oforganized collective action, civil society performs its pluralist (or
"interest aggregation") function and "forms a bulwark against despotic tendencies in political
life" (Hadenius and Uggla 1996, p. 1622). Since the days of Tocqueville. political
Philosophers have spoken of this potential for autonomous associations to serve as a

Safeguard against tyranny. In this aspect, the external relations of organizations are of

 

 

 

 

”v...“

u.“

"we;

1.
. D

., '. .‘ ‘
. d . ‘ J
“‘ '5‘,‘

 

.
rag‘. .J
M“
. 'u
'~-. H
.._

u,
‘ t
hl' v‘»_
I .
. -_..”
t
\
I ,
\ ,-
ll‘e‘\\ .
‘V“
0..
..
\‘..-_. ‘.
. w.“
.1 q -.
.j ‘H
i
u s x n.
M"; x .4.
‘.~v.
'-
's
‘ t‘y-v‘
“I

10

concern. particularly their autonomy. strength. and openness of membership. Organizations
must necessarily be autonomous from the state or its local representatives and membership
must be voluntary in nature. Civil society also can serve an educational function. socializing
support of democracy through participation in democratic structures. In this function. the
inclusiveness and transparency of internal procedures of leadership selection and
decisionmaking are most important (Bratton 1994, Diamond 1994). These purported
democratic functions merit closer examination.
1.2.2 The Interest Aggregation Function
The very identity of civil society as most conceive it is based upon its relationship
to the state. For example, most definitions of the phenomenon agree in applying it to
intermediary organizations that engage the state but do not seek state power for themselves.
In this conception, civil society cannot exist in a stateless society. Consequently, the state
and the voluntary sector need each other in order to accomplish any meaningful political
liberalization. Thus, by such a conception. a large part of the normative definition of civil
Society springs from its role in aggregating the interests of the organizations that constitute
it- This assumes, therefore, that they have a commitment to norms of reciprocity and mutual
engagement for interests that are not exclusively private.

A difficulty facing those evaluating this so-called pluralist function of civil society
in Africa stems from the fact that the frameworks devised for understanding state-society
relations were developed in contexts far removed from that continent in culture and space
(Bratton 1994b). Whether speaking in terms of classic pluralism (e.g., Hodgkin 1956,

Wallerstein 1964, or more recently, Bratton 1989. Widner 1994), neo-Marxism (e.g., Leys

11

1975. Sklar 1979). or corporatism (Shaw 1982. Nyang‘oro 1989). imported models of state-
society interactions have fallen short in the analytic leverage which they offered for African
cases (Bratton 1994b. Callaghy 1994). Civil society itselfcomes with its share ofcultural
baggage (Wood 1991). It is thus necessary to develop analytic approaches to the role ofthe
voluntary sector and a nascent civil society in the African cultural and political context.
Nevertheless, the very purpose of theory is to discern broad patterns across a variety of
contexts. Therefore. especially considering the importance of the Western conceptualization
of liberal democracy and its current prominence in the African context. civil society provides
a useful approach to understanding popular participation and the integration of society into
the larger political system.

At the same time. when assessing the democratic potential of civil society, one must
be conscious of the cultural and political context in which it evolves. While "modern" civil
society has preceded democratization in many cases. "the development of civil society and
democracy are occurring in much of Africa as parallel and related processes" (Holm.
Molutsi, and Somolekae. 1996. p. 43). Indeed. regions of the world that have historically
been populated by authoritarian regimes pose interesting issues for students of participation.
In Africa, informal organizational activity flourished even where authoritarian control was
most evident (Chazan 1982). State-society relations have been suggested to be mutually
beneficial and as such, civil society could be an important factor in democratic political
development (Bratton 1989).

African societies differ from the networks which constitute civil society in the West

in the way that they have cohered according to kinship ties and age groups rather than by

\.
..u. .. r-
_. ~<
.. ..t .-

, o 0‘ r ‘

, .

-n .u u— a

. 'i
\ ‘x\ s

 

q
..‘ a ‘
' s .
|~O .
“‘-\».~..,_
,- .—1 1
"~ b: [\E.
‘.. , ‘
u
-.. _~\ .

 

 

‘5
“ as.
\ .‘lu
'-‘.
a
a
__“\
“x. ‘
- ‘ "u
0
V‘ Q
' II
. ‘F
L
A ’ -
.n
- u.
‘2 1.4.
.‘
1‘!
'-
\
., u‘ _ h
a.“
‘l
x.
.
\ ,-
'\_ ‘n .
.‘ 1
, "_\
.

 

 

. \
‘\
. .
g .4
m
\ u
.‘~\
- . h
n
. n
. H

 

12
bonds of impersonal interest. While many precolonial cultures in Africa included a plethora
of informal associations. their ascriptive nature set them apart from their counterparts in
Europe and the United States. C élestin Monga argues that African civil society cannot be
easily assessed in European terms, not least because "in the African case the leadership.
membership, and functioning of such structures are often shrouded in mystery" (Monga
1995, p. 362; 1996, p. 148). Consequently. one cannot assume that organizational

participation means the same thing there that it does in Europe and the United States. As the
late Ernest Gellner put it. "a modern conception of freedom includes the requirement that
identities be chosen rather than ascribed" (1994. p. 9). Some argue that the emergence of
class-based societies in the West engendered "horizontal social stratification based on non-
ascriptive characteristics" (Lewis 1992, p. 38). The process of class formation was thus an
important precursor to the emergence of what came to be known as civil society. Fukuyama
suggests that "a layer of voluntary associations" can temper ascriptive identities (F ukuyama
1995, p. 361). Bratton notes that in "scaling-up" from the community-level to become
national-level NGOs, one of the fundamental problems facing civil society organizations is
in managing the "internal social conflicts that arise when boundaries are expanded beyond
l(irrship-based groups to encompass “strangers” " (Bratton 1990, p. 92). At the same time. one
Should not dismiss traditional structures out of hand, as existing groups often have the
cOnfidence of local populations (Esman and Uphoff 1984, p. 239-246). Aili Mari Tripp
Shows how some new neighborhood organizations in Tanzania have succeeded in part
because of their incorporation of characteristics of longstanding social institutions (Tripp

1992). In any case, any serious treatment of civil society in the African context must be

13
informed by a knowledge of the characteristics of these organizations. including their
leadership, membership. and functioning.

While civil society is defined in part by its relationship to the state. the state also

needs civil society. The economic catastrophe of the 19803 made abundantly clear that
already-weak states could not be all things to all people. even in the unlikely event that they
actually wanted to be. This created the opportunity for the voluntary sector to take
ressponsibility for some service provision and development work. As a result. besides
serving to represent the interests of society. many NGOs and associations have ended up
actually implementing development projects. This highlights the variety of types of
organizations that are included under the rubric ofthe voluntary sector -- This heterogeneity
Often muddies the distinctions between it and the space occupied by the private sector and
the state (Anheier 1987; Esman and Uphoff 1984).

Given this heterogeneity. organizations in the voluntary sector are only of interest to
democratization theorists inasmuch as they contribute to the increased competition and
participation that are characteristic of democratizing regimes (Dahl 1971. 1989).5 Clearly,
the potential exists for them to enlarge public space, represent interests. hold governments
aecountable, and give voice to young people, women, and other previously marginalized
groups, but this is not a given. Experience shows that many organizations, NGOs in
particular, are vehicles for the ambition or economic gain of their founders, and are not

particularly representative of society at large. often consisting of well-educated men with

5 . . . . . .
I am Indebted to Nicolas van de Walle for this observation.

- I
1

‘~ .s..d

u

" a
.4
.' ’htha‘
m...” ‘rv

>- ...s .‘~_

O D
l
" —.5 n-..
f C
‘ Us.
“ .
" ’v-~| ‘
\ I
-

 

 

 

‘3 rm...
\
L. g
, ..
~o
"k -'l .
I“\ .'AI
.‘J" l-
.“l v‘
s. M‘x'
l
I \
d
. ‘5‘“
.‘I § ‘
I
\T' ‘1 (’
. 'K.
I
‘9“ ‘
‘\
"-i‘\ S
I - ~v1
I“
.i- I
..‘\u,."
-.,_“‘
c

 

.-
. \
-“
1.-
7n.
K. h
‘C. 4""
'l
‘ v
.' v,‘>
.

14
close ties to State elites. The actual composition ofa given voluntary sector is thus a useful
empirical question which can shed light upon its democratic potential.

The democratizing role ofcivic organizations. human rights groups. and the like is
clear. Less obvious are the liberalizing effects of a hunter’s federation or Robert Putnam‘s
bowling club. Nonetheless. if an organization represents the interests of its members it can
play a pluralist role in democratization. In other words. the sector of activity matters less

than the identity of a group‘s interlocutors. Furthermore. as we shall see below with regard
to the socializing effects of civil society organizations. sector of activity matters much less
than the nature of procedures.

Under authoritarian regimes, the associational sphere was dominated by community
groups that did not engage the state and national-level organizations that were created by the
State. Tenuous links existed between the two -- the intermediary level was almost vacant.
The role of either local particularistic groups or state creations in democratization is not
evident. Local organizations that ignore the state cannot be considered members of civil
Society because. Chazan argues, their very disengagement from the state undermines its
aUthority (Chazan 1992. p. 287).

Chazan’s strict criteria for inclusion in civil society include specific and well-defined
Objectives, distinct constituencies, participatory governance, autonomous resources, activities
beyond simply serving the needs of their members, and the ability to create links with other
Organizations with dissimilar objectives. The relative lack of common interests and unity
Of action among social groups in Africa has been judged a critical hindrance to the

Consolidation of civil society and the taming of post-colonial authoritarian states (Bayart

‘am,.

‘1.

\

.._.l
o

s.

.4.
b~ 7

 

‘-\

7“

 

.
.3\
.' y.
v.1 \
1 .~ ’
-
.
v.
a I ~
g.
.
.‘ u
T‘

 

15

1986. p. 1 19). Indeed. some would argue that civil society only exists where there "is a self-
consciousness of its existence and its opposition to the state" (Bayart 1986. p. l 17).

This concern with civil society as a counterweight to the state is excessively narrow.
but the concern with autonomy is important. The delicate balance between engagement and
autonomy is key to an understanding of civil society and its potential democratic effects.
Tripp identifies autonomy and interest representation as the defining characteristics ofcivil

society (Tripp 1994b). Indeed. these two characteristics are the yin and yang that ensure
equilibrium in the civil society cosmos. In a sense. autonomy and measured disengagement
define the space while engagement through interest representation give immediate political
relevance to the concept.

Some contend that existing definitions of civil society tend to favor large-scale.
urban, and elite organizations over smaller, rural. non-elite entities. For example, David
Hirschmann and Aili Mari Tripp argue for a broader definition of civil society which
ineludes non-elite, resource-poor organizations such as those of rural women. Such
OI‘ganizations have often been excluded from formal politics but perform vital social
functions, often those neglected by the State (Hirschmann 1994, Tripp 1994a).

While normative treatments see it as holding great promise. some do not see civil
SOciety as in and of itself a panacea for anti-democratic tendencies. Fatton suggests that civil
SOciety is a necessary but not sufficient condition for the existence of "civic community"
( 1 995, p. 72), which is characterized by "an active. public-spirited citizenry, by egalitarian
p01itical relations, and by a social fabric of trust and cooperation" (Putnam 1993, p. 15). In

the language of the present analysis, Fatton’s point is consistent with the argument that civil

l6

society is but a subset ofthe larger voluntary sector. While F atton suggests that treatments
ofcivil society are incomplete ifthey do not include class considerations. Callaghy suggests
a more general political economy analysis which considers issues of neo-patrimonialism in
state-society relations (Callaghy 1994). Clearly. understanding the balance between identity.
group interests, public space. and private behavior is critical to understanding African
politics.

The distinction between the positive and the normative provides a clearer set of

lenses through which to view these issues. "Society" is populated by an array of
organizations between the household and the state which constitute the voluntary sector.
This associational life is a necessary but not sufficient condition for the existence of "civil
society," which connotes a constructive force. a counterweight to the state but also a partner
and force for the democratization of public life. Thus. for the sake of empirical analysis. a
Civil society organization is defined in this dissertation as an autonomous. voluntary
0r‘ganization that is inclusive in leadership selection and decisionmaking, and engages the
State for purposes which are not exclusively personal and private. That is, it represents the
interests of its members. Further. profit-making activities are not its principal reason for
existence.

This definition does not exclusively privilege civic or development-oriented groups
0r the explicitly political. Instead, it considers the groups“ procedural characteristics and the
fact that they engage the state but do not seek political power to be determinant in whether
or not they meet the criteria for inclusion in civil society. Thus, while recognizing the

rlormative content of discussions surrounding civil society, we are left with a definition that

17

is empirically useful. Given the above—mentioned criteria. one can evaluate whether or not
an organization merits inclusion. This is the focus of Chapter 2 of this dissertation. Chapter
3 addresses the level of engagement ofnascent local civil society institutions with the state.
1.2.3 Political Socialization
While serving an interest aggregation function is a necessary condition for inclusion
in our emerging normative definition of civil society. it is not a sufficient condition --
illiberal entities in the voluntary sector are quite capable of agitating on behalf of common
interests but are not presumed to be as likely to contribute to the emergence and
consolidation of democracy. To have a complete picture of the role of civil society in
democratization, we must examine its potential socialization function. One should also note
that many organizations possess some but not all of the desirable characteristics sometimes
ascribed to civil society. The interest aggregation function of civil society plays a role in
Strengthening an accountable state. Organizations with little contact with the state could
nevertheless contribute to a socialization function of civil society.

While the first role presumed by those advocating the desirable effects of civil society
iS organizational in nature, the second takes as its unit of analysis the individual —- what
effect can one expect organizational participation to have on its members? Those who count
on the democratic socialization effects of the associations that constitute civil society make
two implicit assumptions. F irst. they assume that there is something inherently democratic
about associations. Second, they assume that exposure to these positive characteristics will

irlduce individual members to have more democratic values. To test empirically the potential

. .lh

 

‘ \
‘7 -—., .
" I n j .
' ~- 5 .
l.'-

"“~ - u
n ., I
t. ’i

' . 1|
- I.‘ g ‘
n, i .'. “

'~‘

..

_ .
' J'N‘
‘, I .
v..‘_ ‘
-'~.' '
"x... . H.‘
. ,~
~ 'H‘
.

“F .
. "li' -|
A“... ¢
.._.
t
. 7.
.._ p .
\. $.49.
h.
‘-

 

‘ 7 .4« ~
‘.u f
, 5‘,
,‘a
' ‘\. I.
\ _‘
.1 . p
.‘n.‘

 

N
I.
, '1‘.
L‘ \
w, .1.
“\l
c
‘-. .
.D
5- ‘.<
e. . .

18

of local organizations to contribute to democratic consolidation in this manner. one must
evaluate both of these assumptions.

The study of political socialization has long been a preoccupation of in Western

political science. Some have long suggested that the exposure of citizens to democratic
institutions from a young age performs the function of inculcating values presumed to be
favorable to democracy. such as efficacy. compromise. trust. and tolerance (Hyman 1959.
Dennis et al. 1968, Jennings and Niemi 1981. Niemi and Hepburn 1995). While most ofthis
body of work has focused on childhood political socialization. certain scholars have
acknowledged that beliefs and attitudes are not established immutably in childhood (Almond
and Verba 1980, pp. 399-400; Diamond 1994, p.234; Sigel 1995). Research on "political
learning" has attempted to identify the sources of changes in political beliefs among adults
(Bermeo 1992).

Most attention to socialization has examined the role of electoral institutions (e.g.,
I“iyman 1959, Campbell et a1. 1960, Jennings and Niemi 1981), though adult experiences in
r1Orr-political associations have also been suggested to be a source of the learning of political
Culture (Verba 1965, p. 551). Robert Putnam has argued that membership in horizontally-
Ol‘dered groups such as cooperatives. sports clubs. and mutual aid societies should be
positively correlated with good government and that institutional changes are reflected in
Corresponding, but gradual, changes in values (Putnam 1993, 1995). For example.
Organizational activism can foster solidarity which contributes to individual feelings of
Competence or efficacy with regard to the larger political system. Similarly, exposure to the

give and take of diverse groups within an organization can contribute to broader feelings of

l9

tolerance for different ethnicities or ideological points of view. Such participation can also
inculcate norms ofcompromise, conflict resolution. and respect for responsible authority and
the rule of law. On the eve of the current dramatic transitions. Muller. Seligson and Turan
echoed this point of view. suggesting that in nations with authoritarian pasts or where
electoral institutions have harbored clientelistic networks, "norms ofdemocratic governance
may be learned through participation in local organizations" (1987. pp. 29-30).

Of course, as discussed above. civil society cannot be assumed to embody the values
of anywhere but the milieu in which it evolves. Larry Diamond makes this point, saying that
democratic values will be acquired in civil society only if the organizations which make it
Up "function democratically in their internal processes of decisionmaking and leadership
selection" (Diamond 1994, p. 12). Michael Bratton makes a similar point, arguing that one
cannot expect organizations with liberal characteristics to emerge from a "neopatrimonial"

p01itical culture (Bratton 1994, p. 10). Bratton and Nicolas van de Walle define

neopatrimonialism as a regime type in which the defining characteristic is the maintenance
of authority through personal patronage rather than ideology or law -- “The distinction
between private and public interests is purposely blurred... in return for material rewards,
clients mobilize political support and refer all decisions upward as a mark of deference to
patrons” (Bratton and van de Walle 1994, p. 458). Arbitrary personal rule has characterized
African polities during the pre—colonial, colonial and post-colonial periods (Jackson and
I{osberg 1982, Callaghy 1984, Sandbrook 1986, 1993, Bayart 1993). It is thus unsurprising
that the African voluntary sectors have many of the weaknesses of the states with which they

Often interact (Lewis 1992).

 

-.,§- - .5 .
'- “'24"

-.~ >. s. 45

 

”J".

‘s... .

I
.‘u
0.. .
\,.\
v
-5 j .
. ‘ as
T...\
, . .
-1
L8
" s i 3‘
. K
~
" \
‘V w
_ O .
Vn‘ -
s ‘ k
‘ n
a.
_ ._
., a”.
“a.

 

 

 

\
I.-\.
I“ ‘t
‘. —' ‘
'Q v —1
b
" H
I- A n
c n‘.
. .
5‘

20

The discussion regarding the socializing effects oforganizational participation is a
venerable one. Tocqueville spoke ofcivic organizations as "classrooms of democracy" in
the American context a century and a half ago (Tocqueville 1945). More recently. in their
classic tome, The Civic Culture. Almond and Verba spoke of an emerging world political
culture in which "the participation explosion" would be key -- democracy was held to entail
more than just formal institutions (Almond and Verba 1963). While political culture
arguments had all but disappeared from the theoretical radar screen by the 1980's. the

emergence of nascent democracies in cultural contexts very different from those studied by
Almond and Verba and their contemporaries has led to a renewal of interest in these critical
issues. The socializing effects of local organizations will be the focus of Chapter 4.
1.2.4 Research Questions

Much of the empirical literature on civil society concerns the interest aggregation
function." However, these treatments tend to focus on the urban and national levels to the
neglect of the rural and local. Compared to the plethora of theoretical articles on civil
Society, as well as articles describing interest articulation. there is a dearth of empirical
testing of the socialization function. The current study describes the types of associations
most familiar to rural citizens in one African country. To evaluate the external
Characteristics of organizations, assess their fit with the Western notion of civil society, and
gauge internal levels of democratic procedure. local-level rural organizations in Mali are
eValuated according to observations on four dimensions considered important in the

literatures discussed above:

¥

" In the African context, see. for example. Lucas 1994; Holm. Molutsi. and Somolekae 1996; Degnbol 1996.
Burgess 1997. Smith 1997b, and Bingen 1994 and 1998.

1, I s membership ascriptive or voluntary?

2, Is the organization autonomous from control by the state and traditional authorities? In
other words. do external authorities have any significant control over organizational
activities?

3 . Is leadership selection conducted in an inclusive fashion? Is there any evidence of any
kind of systematic participation by the rank and file in this process?

4. Are decisions made inclusively? Again. do procedures exist which ensure systematic
participation by the general membership?

Once their relationship to Western notions ofcivil society is determined. data on the
pluralist function will be examined. Specifically, to what extent are local organizations in
Mali engaged with the state? What factors explain differences in civic action behavior?
Further, what factors determine civic action success? Finally. in light of the characteristics
of the organizations studied. hypotheses on socialization effects will be tested.

Before continuing. however. important contextual information on the Malian
democratic transition, decentralization reforms. and cooperative law reforms is presented.
1 -3 The Malian Democratic Transition

Democratic advances in Mali over the past eight years have been widely lauded as
emblematic of the rosy prospects for democratic consolidation on the continent (New York
Ti mes, 29 April 1996, 16 October 1996). In spite of more cautious recent assessments (New
f'Ork Times, 7 September 1997; Jeune A frique. 1-7 October 1997), the political openness of

I\’Iali, coupled with its rich associational life, make it a propitious venue in which to examine

cltizen participation in the process of democratic consolidation. Moreover, Mali is one of

the few countries in West Africa where political research raises neither hackles nor
eyebrows.

Before pro-democracy mass demonstrations in March 1991 led to the toppling of

General Moussa Traore's venal dictatorship. Mali had been ruled since its 1960 independence
from French colonial rule by two authoritarian regimes. Modibo Ke'ita's single-party Union
So udanaise-Rassemblemem Democratique Afi‘it‘ain (US-RDA) regime ( 1960-68) was widely
perceived as having a coherent developmental vision but its socialist ideology led to
excessive rigidity and over-centralization. Rural people complained that the statist regime
interfered excessively with agricultural production, even that of subsistence crops such as
millet. Traoré's military oligarchy (1968-91) spawned frequent charges of corruption but
prOSpered in power until a brutal massacre of at least two hundred pro-democracy
demonstrators led to public outrage and his ouster by sympathetic military officers. Colonel
Amadou Toumany Touré served as interim head of state from March 1991 until he
Shepherded in the country's first democratically-elected president. Alpha Oumar Konaré
(AOK), in June 1992.

The young life of Mali’s democracy has been marked by serious crises between
ADEMA and the opposition parties. In spite of mistrust and political turbulence, after the
fall of Moussa Traore' and continuing through most of AOK’s first term, there was agreement
atnong the political elite on the rules of the game. Political crises such as the recurrent
Strikes led by the Association des Etudiants et Eleves du Mali (AEEM) were resolved in a

Consensus-oriented fashion by the government, which usually included members of multiple

political parties, and the opposition. The April 1997 legislative elections marked a turning
point in this state of affairs.

These elections were marked by serious criticism. Opposition parties accused the
Commission Electorale Nationale Independante (C ENI) of. at the very least. incompetence.
if not outright corruption and collaboration with the ruling party. Once the serious flaws in
the electoral lists became clear, AOK annulled the results of the elections and scheduled a
new vote for July.7 The opposition implored him to cancel the scheduled May presidential
elections until all parties could agree on the integrity ofthe voter lists. but the scheduled May

1 1 first round of the presidential elections went forward anyway. All of the opposition
eXCept a wealthy crocodile skin merchant. Maribatrou Diaby, boycotted the election and
I(onaré won with over 85 percent of the vote. In spite of the impassioned protests of the
Opposition parties, none of their candidates were removed from the ballot, and the
government reported the meager vote percentages for all of the boycotting parties in an
Obvious effort to discredit them.8 New legislative elections were held in July and the
Opposition, which had spent its meager resources for the annulled first campaign, did not
Stand a chance against ADEMA, which took 130 of the 147 National Assembly seats. Other
pa-rties’ candidates who allied with ADEMA’s lists accounted for several more seats, leaving

the true opposition with fewer than ten.

7 Certain elements held that it was the Constitutional Court and not Konare who had the right to annul the

Elections. but some observers gave him credit for the quickness with which he attempted to restore credibility to the
rOcess.

_ 8 Of course, there was no easy solution to this -- other vote counts reported Konaré’s proportion of the vote

-: :81 as apcrcentage of the sum ofhis and Diaby‘s votes. putting him well over 90 percent and giving an inaccurate
usron of unanimity. One should note that American pundits routinely do this. excluding third party totals from
emocrat and Republican percentages. giving the incorrect impression that victors received a majority of the vote.

24

Local communal elections to put in place 682 new. primarily rural communes in

addition to the previously existing 19 urban communes have been postponed several times

since 1995. Municipal elections in the 19 were finally held on 21 June 1998. with the

remaining communes scheduled to vote on November 4. though repeated postponements

have made many skeptical of the govemment‘s commitment to meaningful decentralization.
These skeptics were validated in August when the government again postponed communal
elections to April 1999.” The June 21 round was again boycotted by the opposition. They
felt that dates were set in a cynical fashion, coming just two days after the departure of Carter
Center mediators. Preliminary reports suggest that voter participation was at an all-time low
(L 'Autre Afiique, 24-30 June 1998. pp. 24-25).

Since April 1997. there is a palpable decrease in civility among the various factions
of the political elite. For example. as stated by Mountaga Tall. president of the Congres
Nationale d’Initiative Democratique (CNID). "Konaré has established himself as absolute
1‘uler and it seems like a return to the days of a single party." In fact, because of the
govemment’s failure to address their accusations of fraud and other irregularities in the 1997
elections, Tall and other opposition members refuse to accord AOK with the respect of being
addressed as "Monsieur 1e president" (L 'Autre Afrique. 17-23 June 1998, p. 26). Sporadic
Violence between the opposition and the government has been both a symptom and a cause

of this decrease in civility.

Added to the political breakdown is the return of "la crise scolaire." High school

Students have been periodically on strike and university students only studied for a total of

9 Once the communal elections were finally held in April 1999. ADEMA won handily.

25
two months during the 1997-98 school year. Given the dramatic increase in the price of
millet, rice, and other commodities. unions have demanded across—the-board pay increases
of 20 percent, to which the govemment responded with only a 5 percent raise.

In spite ofall ofthis. the country is still viewed as hosting a favorable climate for the
survival and consolidation of democracy. Indeed, President Konaré has stated that without
an effort to strengthen opposition political parties Mali's "level of democratic culture will not
evolve" (Jeune Afiiqzre. 3-9 September 1997).10 It remains to be seen whether the

government and the opposition can bridge their significant differences in order to move
forward with the process of both institution-building and broader democratic consolidation.

Student strikes and union demands are symptomatic of the vivacity of Mali's

Voluntary sector. There has been an explosion of associational life in the years since
Traoré’s fall. As of June 1998, over 4000 associations and 650 NGOs were registered with
the government. The importance of the inclusion of representatives of civil society in
decisions which affect Mali’s population is on the lips of the highest officials in the
government. For example. the Minister of Finance, Souma‘r‘la Cisse’, recently affirmed the
need to include such elements in donor decisions on how to allocate development assistance
(L ’Autre Afrique, 17-23 June 1998, p. 50).

The dramatic increase in associational life in Mali is just one symptom of the
dynamic social and political situation. To fully understand the current environment facing
Malian civil society, one must be familiar with current reforms aimed at decentralization and
c()operative law. These are the subjects to which we turn in the next two sections.

\

'" Author's translation.

lil Plillllt

\

QJW "

in use»
. _ 0
..~. .
. . .

Q n— “K

_ .. .
' 1.... . ,
f» . '

5”

. V
~~ .wi.

“p

A‘-

\ “‘ n
v... \ .
.' m

I

é . 7"
u._ a v

”A
"“I.

"up-
— ‘ T l
"r5;“\ 1';-
_, ‘
,
.
i‘.

 

 

 

 

5.:
‘ i
\ _~ .
‘- .A
5
5.
\‘ 1'5
81
A
l
o. ‘
". ‘F'nh.
AT’L ‘
.“
\ : .
’ b—HI .
‘.\A‘ ‘I
U.
)-
~\ 'i'k -‘
-§
\g‘. i I
- ~\‘
.
s
\
..h A '
q

 

26
1.4 Politico-Administrative Context: Decentralization
Context is extremely important to the viability of civil society as a vehicle of
democracy. Hadenius and Uggla describe a five-stage continuum from a "hostile state."
which tolerates no independent civil activity. to a "benevolent state.” which actively
promotes autonomous organization (Hadenius and Uggla 1996). A turning point in the
evolution from totalitarianism to benevolence is the promotion of favorable institutional
structures, such as arenas for constructive interaction and a favorable legal and administrative
framework. This section describes Mali’s decentralization reforms and the next section
describes evolving cooperative law in Mali.
Many observers suggest that the most congenial institutional structure for civil
Sooiety is a decentralized one (Ostrom 1990). The disadvantages of central administration
deconcentration without devolution are well-documented, however. These include operating
in excessive isolation of the local environment, being insensitive to local needs and demands,
being an alien body making critical decisions without much real local participation, and
promoting centrally-designed integrated development programs of excessive complexity
(Hyden 1983, pp. 91-94). A truly-decentralized form of government can allow greater
participation by citizens and can facilitate meaningful participation in terms of policy voice
and development collaboration by truly local organizations that would not have the resources
to contribute on a larger scale. Indeed, in a devolved system where resources are generated
and spent at the local level, lack of funds would induce local officials to draw upon local
capacity (Curtis 1991; Brautigam 1992; Brown and Martz 1998). A delicate balance must

be struck between ensuring the participation of local citizens in decision making which

27
affects their lives while ensuring that local govemment structures are not captured by
longstanding clientelist networks.

Mali is currently in the midst of a transition from a modified prefectoral system of
local government inherited from the French colonial administration to a system that promises
to be quite devolved. While this transition has dragged on for several years and is
temporarily stalled once again, a significant investment has been made to create local
councils of elected leaders. albeit with substantial state oversight.

Mali’s current territorial admininstration is organized as follows. Below the national
level, the next level is the region. Mali has eight regions. with regional capitals in the cities
0f Kayes (lst Region), Koulikoro (2nd), Sikasso (3rd), Se’gou (4th), Mopti (5th), Timbuktu
(6th), Gao (7th), and Kidal (8th). The regions are divided into a total of 49 cercles, and these
are in turn divided into a total of over two hundred arrondissements. The District of Bamako
iS administratively separate from the regions, though it is surrounded by the Region of
I(oulikoro.

Under the new decentralized system, regions and cercles will still exist. However.
the arrondissements cease to exist, being replaced by a network of 701 new communes.
Municipal elections for the communes have been planned and postponed at least three times
Since 1996. As previously noted. elections finally took place in 19 urban communes in June
1 998 and in 682 rural communes in April 1999.

These elections will install communal councils (for a five year term) which are

e l ected by citizens and which will choose, in turn, a mayor and several deputies from among
t1leir numbers. The numbers of communal councilors depend upon each commune’s

DOpulation. Communal councils will range in size from 11 (in communes with fewer than

28

1 0-000 inhabitants. to 45 (in communes with over 200.000 inhabitants). Each mayor will
have three to five deputies. also according to population. These elected officials will receive
technical support from a secrézariut générul made up of state functionaries who will
supervise personnel. provide advice. and ensure the day-to-day functioning ofthe communal
administration. Support will also be provided by deconcentrated personnel of the state
development services. such as those under the purview of the Ministry of Health. Minisrry
of Primary Education. Ministry of Rural Development, Ministry of Public Works and
parastatal personnel. Local citizens will be given the right and responsibility of being
informed ofthe activities oftheir elected officials (Law No. 95-034. Articles 4-13. 18-38).
A variety of explanations have been advanced for the fact that communal elections
Were continually postponed until they finally took place in April 1999. All seem at least
Partially correct. For example, some believe that there had been significant donor pressure
tO decentralize but that entrenched interests in the government and administration lacked the
Will to implement a new system which would devolve away many rent-seeking opportunities.
Another explanation is that the national government and its “independent” electoral
Commission lacked the resources to organize local elections, especially considering the
GXpense of legislative and presidential elections in 1997. Political parties, especially those
of the opposition, are always short of resources and had difficulty running campaigns
throughout the country. This is one explanation for ADEMA‘S overwhelming victory when

the elections finally took place.
Mali’s decentralization program is accompanied by compelling rhetoric of
participation, with local citizens described as “the soul and motor of the new commune”

(Republic of Mali, Mission of Decentralization [undated], p. 9). It is intended to give

29
communes legal standing as well as the financial autonomy to provide essential services to
local citizens (Law No. 95-034. Article I). In so doing. the commune will be directly
responsible for its own development. gaining responsibility for formal education. literacy.
primary and maternal health. and management of local infrastructure and the environment
(Law No. 95-034, Article 14). While the communes are supposed to receive their just
portion of State budgets (including funds generated by the Regional and Local Development
Tax, the TDRL) for education. primary health. and other services, this will be insufficient
for communes to function effectively. Communal councils will have to be creative in finding
other sources of revenue generation. While they cannot waive any existing taxes. they are
allowed to institute supplemental local ones.
However, the new. soon-to-be-devolved responsibilities are subject to the oversight
(tutelle) of the national administration (Law No. 95-034, Articles 15-16, 229-235). The
Malian government insists that this is genuinely a system of oversight of legality but not an
imposition of hierarchical authority (Mission of Decentralization [undated], p. 29).
Nevertheless, for actions that are judged illegal, the state has the power to suspend or fire the
ITlayor or his assistants or dissolve the entire communal council. The state also reserves for
itself the nebulous role of “regulator of relations” between local officials. the administration.
Cleconcentrated state services, and civil society (Mission of Decentralization [undated]. p.
3 S ).
On the other hand, for certain decisions, the communal councils are required to seek
ihput from village councils or, in urban areas, heads of quartiers. Matters subject to such
i11pm include issues of land valuation (cadastre), public transport, private use of public lands.

access to water, organization of agricultural, pastoral, hunting, and fishing activities, and

30
environmental protection. among others (Law No. 95-034. Article 17). The Code dcs
Collectivités Territoriales describes these village chiefs and councils as the primary
representatives of the people of Mali (Law No. 95-034).

This raises interesting issues concerning how the new system differs from old systems
of oversight and accountability. One issue of particular importance is what Mahmood
Mamdani calls the “bifurcated State." where there are different forms of administration for
rural and urban people (Mamdani 1996, p. 8). Building upon this idea. Jesse Ribot wams
of pseudo-participatory approaches to development which divide West African citizens into
“citizens and subjects. the prior with rights and the latter privileges“ (Ribot 1998b, p. 28).
He is referring to an upwardly accountable system where urban dwellers are treated as
individual citizens and rural dwellers as communities of subj ects presided over by chiefs and
village councils. The point is taken that the representativeness of village ruling families is
often questionable, based solely upon their status as village founders or upon responsibilities
conferred by the French during colonialism.ll A decentralized system which employs the
system of government oversight in existence since colonial times and based on the
representation of rural people on hereditary grounds rather than democratic selection risks
being little different from past systems:

Decentralization with participation requires a switch from the decentralized

despotism of administrative apartheid to autonomous forms of rural

representative governance -- locally accountable governance bodies with real
powers of decision over financial and natural resources (Ribot 1998b. p. 5).

 

H While village chiefs are treated as the putative representatives of the people by the national government.
their role has been more localized ever since larger-scale chef: de canton were abolished at independence. For
example, in Niger. the traditional Che/3' de canton still serve as local representatives of the national administration.
having authority roughly equivalent to the chefs d ‘arrondissement in Mali. who are state functionaries and not
hereditary leaders. In Niger’s evolving decentralization system, a certain number of posts on elected local councils will
be reserved for chefs de canton. who will have ex officio status and not be subject to election.

A‘

31

Nevertheless. the promised system. once implemented. would represent a major
divergence from past Malian policy. as well as the systems in place in neighboring states.
At least as it concerns forest use. Ribot finds that the legal framework for Mali‘s
decentralization program devolves more power to communal councils than the analogous
programs in the neighboring countries of Senegal, Burkina Faso. and Niger (Ribot 1998b,
p. 20).

Given the creation ofa local territorial entity above the village but smaller than the
soon-to-be-defunct arrondissement, this should create greater proximity of the government
to the people and, consequently, greater opportunities for individuals and local organizations
to engage in civic action. In fact, Malian government documents make frequent reference
to the need for communal officials to consult with NGOs. associations. and the like in the

conception of communal development plans.

Crook and Manor suggest that three conditions are necessary for decentralization to
avoid capture by neopatrimonial elements and contribute to democratic development, in so
doing energizing civil society. First, local units must be governed in a genuinely democratic
manner. Second, local units must have the competence and technical capacity to make
decisions in a variety of domains. Finally, they must have sufficient resources to have a
certain degree of fiscal autonomy and not be subject to nefarious external control (Crook and

Manor 1994). On the first and third of these, the system as designed offers reason for
optimism, though external tutelle and the meager resources of the TDRL will constrain the
new localities. On the second, it is not yet clear how local governments will harness existing
government services to work for them. Without that, many localities will have serious

capacity problems grappling with a range of technical areas.

1.5 Legal Context: Cooperative Law

In facilitating the meaningful participation ofcivil society in emerging democracies.
policymakers must ensure the existence of a legal-rational framework which recognizes
freedom of association. facilitates participation. and allows the political space for
independent activity. One of the most important potential contributions to vibrant local
organizations is legal and administrative reform. A favorable legal environment can
facilitate associational life and lower the transaction costs of organizing (Ostrom 1990).
Mali’s evolving cooperative law is just the sort of reform which, if done properly. could
reverse a century of top-down, patronizing control of the rural sector.

Current efforts to reform Mali’s cooperative law raise important practical issues
concerning pursuit of income generating activities and access to credit. Perhaps even more
important, they raise issues concerning the right to free association, organization and
autonomy of the rural sector and who speaks for rural people. as well as how these leaders
are chosen. The current effort at reform is rooted in a history of central, top-down control of
rural interests. To understand the current situation, some history is necessary.

Under French colonial rule, a number of pseudo-cooperative farmer organizations
were established by experts working for the central administration of rural development in
Mali (Jones 1976, Jacquemot 1981, Belloncle 1982, Dembe’lé 1987). Beginning at the turn

of the 20th century, these organizations included the Greniers Villageois at the village level.
as well as, beginning in 1910, the Societe's Indigéne de Prévoyance (SIPs), which had a
variety of purposes, including input purchasing, introduction of new farming techniques and

the development of infrastructure. These organizations served as much the purpose of income

33
generation for local colonial authorities as for village development. After the Second World
War. there was a move toward increasing management responsibilities at the local level.
This was meant to be accomplished by replacing the SIPs with Societe's Mutuelles de
Production Rurale (SMPR) and then. in 1956. with Societe's Mutuelles dc Développement
Rurale (SMDR). These cercle-level institutions were meant to assist and guide local
cooperatives through the provision ofinputs and training. In 1960. with the SMDRs still in
place, an ambitious hierarchy of rural cooperatives was established. This system included
Groupements Ruraux dc Production et de Secours Mutuel (GRPSM) at the village level.
Groupements Ruraux Associés (GRA) at the arrondissement level, and Fe'dérations des
Groupements Ruraux (F GR) at the cercle level. While all of these organizations had the
official objective of promoting the capacity of farmers to collectively defend their interests,
these organizations did not emerge as a result of farmer initiative, and membership and dues
were compulsory (Dione 1989, p. 334; Ribot 1998a). They were all subject to repeated
official interference, even to the point of the government choosing their leaders and dictating
the manner in which they operated. They routinely served such purposes as tax collection,
compulsory cultivation of collective fields by villagers. and forced marketing of food grains
and other primary agricultural products at official prices. These structures also supplied
revenues to the colonial government through producer and export taxes.
After independence, the pattern of State control of cooperatives did not change much.
During the Modibo Kéita era (1960-68), the Malian village was idealized as a model of
socialist organization and a principal building block of the ideal socialist society. Beginning

in 1963, cooperatives were organized as a branch of the state and local branches of the

34

single-party State apparatus appointed cooperative leaders and ran cooperatives essentially
as local party units. Cooperatives had an absolute monopoly on commercialization of
agricultural products, but rural populations began to distrust them because of their close ties
to a heavy-handed state. A May 1968 Seminaire National Sllr la Cooperation Rural led to
the elimination of the SMDRs and greater support of GRPSMs and FGRs as "pre-
cooperative" structures striving for the designation of "Cooperative A gricole
Multifonctionnelle. "
General Moussa Traoré‘s November 1968 coup rendered the proposed reforms and
new laws irrelevant. though they officially remained in effect (Ribot 1998a, p. 9).
Cooperative policy foundered from 1968 to 1974. at which time Traoré‘s party, the Union
De'mocratique du Peuple Malien (UDPM) began promoting tons villageois. modern
cooperative structures based upon traditional, village-level solidarity groups. Beginning
about 1974, the Malian government began promoting Operations de Développement Rural
(ODRs) -- large, multi-sectoral development projects based within a clearly-defined
geographical area. Primary among these were the Office du Niger. the Operation [later
‘Office"] de la Haute Valle’e du Niger, and the cotton-producing parastatal, the Compagnie
Malienne de De'veloppement des Textiles (CMDT). These ODRs employed village
associations as vehicles of encadrement and tutelle, essentially training and oversight, which
was ostensibly developmental but also kept a tight leash on rural organizations, lest they
stray too far from official designs. During this time, ODR extension agents built local
capacity and eventually village associations took on some of the quotidien tasks of

agricultural commercialization. For example, in the CMDT zone, certain tasks, such as

'JI

3
weighing and valuing output. had always been performed by C MDT agents. Farmers in a
few villages began demanding that these tasks be assigned to literate villagers. This proposal
was well-received by CMDT officials because they saw the practical benefits of using it as
a means for collecting delinquent loans without coercion (Dione 1989. p. 335). This
permitted the CMDT to cut its administrative costs by substituting unpaid agents for paid
employees. and it also gave the govemment organ a means of coopting and maintaining
control over villagers (Bingen 1994. pp. 59-60).

A 1988 cooperative law paid lip service to greater freedom for cooperatives. For
example, it replaced state organization of cooperatives with voluntary membership. Rural
cooperatives would still be called tons villageois, and "pre-cooperative" village organizations
would be called associations villageoises (AVs). Through encadrement and tutelle of state
development services such as the Direction Nationale de l 'A ction Cooperative (DNACOOP)
and those provided by the ODRs, AVs could aspire to the legal standing and privileges that
came with cooperative status.

The 1988 law was very specific about mandatory fees to be levied on cooperative and
pre-cooperative structures. as well as the specific sectors in which such structures could be
created. Registration was channeled through the technical services of the national

bureaucracy, and a fairly arduous and complicated application and registration process was
instituted. In practice, organizations were required to have an official advisor from the state
apparatus, and it was required that this individual be invited to all general assemblies. where

he had the right to speak (Ribot 1998a, p. 12).

36

Though the law explicitly stated that these were "voluntary" organizations. it also
stated that membership was mandatory for all persons living in the geographical area covered
by a ton. Each village was only allowed to have a single ton. In contrast. villages were
allowed to have more than one AV. Clearly. in spite of the encouraging language. the new
law did not represent much ofa departure from the previous top-down situation.

Since the tumultuous events of March 1991 which marked the begining of Mali‘s
democratic transition, dramatic changes have occurred in Mali's rural sector. Emblematic
of this is the disdain with which State forestry agents are held in rural areas. This signals an
unwillingness to tolerate the imposed controlling structures which have been all Mali's rural
people have known since colonial times. At the same time, tentative steps taken by the state
suggest that the possibility of the creation of a more liberalized rural sector exists. For
example, the 1991 constitution permits freedom of all types of association. The December
1991 Etats Généraux du Monde Rural. though only consisting of "invited" representatives
of the rural sector, found that the 1988 cooperative law was not well-adapted to the needs of
rural people. Further, they recommended "promotion and support of cooperatives. tons. and
AVs" (RdM 1991, pp. 8. 29, cited by Ribot 1998a, p.14).

A move has been underway since August 1995 to re-write cooperative laws. The
Ministry of Rural Development and the Environment (MDRE) convened a national seminar
on cooperatives in December 1995, and one proposal was to group cooperatives, tons, and
AVs under the new appellation "societés cooperatives." With significant support from the
United Nations Development Program (UNDP) and the lntemational Labor Organization

( I LO), there have been at least eight regional meetings as well as international meetings to

37
develop the new texts. I’lowever. a L'SAID consultant recently commented that there is
reason for skepticism as to the representativeness of the process. as it only included rural
"representatives" chosen by the Ministry (Ribot 1998a).

Though temporarily blocked in May 1998 by the Ministry of Health's concerns about
the impact of the new law on the Associations de Sante' Communautaire (ASACOS) that run
rural Centres de Sante' ('ommunautaire (C SCOMs). no fundamental objections have been
voiced against the proposed law. The legislation is expected to be approved soon by the
Council of Ministers for debate in the National Assembly.

Though there is reason to fear that over-centralized state structures see the proposed
legislation as another opportunity for control of rural actors, there is sufficient cause for
optimism as well. For example, the proposed legislation states that the new societe’s
cooperatives are part of the private sector and not "instruments of the state" (RdM 1996. p.
9, cited in Ribot 1998. p. 16). If the spirit as well as the letter of such provisions are
respected, the emerging law will constitute a marked divergence from all rural-government
relations since the advent of colonialism in the 1890's.

The new groups will have the freedom to split or combine and will be allowed to
form unions, federations. and confederations. There will be a National Council of
Cooperation. There is some concern that this may be established by the government. For

example, some language in the proposed law suggests that there will eventually be a
Confederation of Malian Cooperatives, but until that time, the National Council of
Cooperation, created by the MDRE, will represent cooperatives at the national level. The

very fact that the government is deciding a priori who will be the new national-level

38

representative of the new. revitalized cooperative movement is worrisome. This same
concern has been voiced by a number of rural actors. including members of the MDRE
committee responsible for promoting the law (Ribot 1998. p. 25). This. indeed. smacks of
old wine in clumsily-recycled bottles -- The system of parent-teacher associations (APEs)
as well as "community” health committtees (ASACOs) have an elaborate hierarchy of
federations at the cercle. regional and national levels which have been imposed by the
powers that be with little input from below.

The danger, of course. in this would be that this could end up becoming another top-
down structure controlling the cooperative sector. At the very least, there is the risk that it
would become like the Chamber of Agriculture, a nominally-representative body which is
in fact composed largely of current and former state functionaries. l have found that many
state functionaries in such positions end up having a somewhat condescending attitude
toward rural people.

Ribot sees it as promising that the Rochdale Principles ofCooperative Organization
are stated in the proposed 1998 law, "creating an enabling environment for rural
organization, rather than an imposed state-run system of production and marketing in the
name of cooperation" (Ribot 1998, p. 19; see also Painter 1990). Clearly, the impact of the

proposed law will turn on the strength of the break it creates with the colonial and post-
colonial legacy of creating, training, and managing rural organizations for State purposes.
Front and center, then, is the degree of autonomy with which these organizations will be

entrusted. To what degree will state authorities seek to reinstitute encadrement and tutelle

39
under the guise of assisting weak organizations'? Or will they delegate these powers to the
ODRS, as was often the case in the past?

This is a critical period for the future of state-rural association relations in Mali. 1n

the past. much of the engagement of rural associations with the state was initiated by the
State. Indeed. the very inclusion of many Malian associations under the rubric of "civil
society" is problematic. For example. the Associations des Parents d 'Eleves (APEs) have
been frequently politicized in the past and were integrated into patron-client networks during
Moussa Traoré's regime. Creation of an ASACO. a community health association. is
required by the Ministry ofHealth before it will permit the creation of a community health
center (CSCOM). In one case seen first-hand by the author. an ASACO president in Koro
Cercle, Mopti Region, was frustrated because the other officers of the ASACO board were
ignoring requests to attend meetings. The president consequently requested the chef
d ’arrondissement, the local representative of the national administration. to issue a formal
Convocation requiring officers to attend board meetings!

For civil society to have meaning in representing the interests of its members and
their communities, engagement and civic action must be initiated by organizations
themselves, and those organizations must have the autonomy to determine their training
n(feds and other priorities on their own. The long-standing paternalistic attitude that rural
Deeple are incapable of determining their needs and managing their affairs will only lead to
COntinued interference in the affairs of associations. The degree to which local organizations

are initiating demands on the state is an important empirical question.

40

A significant component of the state role in overseeing the cooperative "movement"
in the past was the requirement that local state development agents evaluate AVs and
promote them from "weak" to "medium" to "strong" AV status. with a strong AV being
accorded the status of ton and the legal standing that went with it. In principle. the state

\vould have no such role with the new cooperative groups.
Interesting questions are raised by the movement toward meaningful decentralization.
Most important among these are those concerning evolving systems of representation. In the
past, AVs and tons were the principal interlocutors in local development. Will the new
System of elected commune-level councils change this? In fact. this should not have to be
a zero-sum game. Possibilities abound for mutually beneficial interaction between civil
society and local governments. Local organizations can serve as training grounds for local
Officials, both in practical issues of development and administration. but also in cultivating
experience with procedures and values which are critical to local democratic development.

1 .6 Data and Methodology

Clearly, these are heady times in the political development of Mali. The dramatic
Changes underway there make it both an interesting and a hospitable setting in which to
eXamine the role of local organizations in a nascent democracy. Earlier in this chapter. a
Variety of important empirical questions having implications for theoretical perspectives on
Civil society in emerging democracies were raised. These included the internal dynamics of
1Wired associations, the civic action behavior of such groups, and the socialization effects of
participation in these organizations. These issues will be addreesed in Chapters 2, 3. and 4,

respectively.

41

In order to bring a variety oftools to bear to answer complex questions on the nature
of civil society in rural Africa. a variety ofdata sources are used. These include a 2 l -village
political attitudes survey of individuals. semi-structured interviews with local organizational
leaders. village notables. and key informants in the agricultural sector. and USAID/Mali
performance monitoring data on community organizations (COS) and intermediary NGOs.

While the dissertation examines multifarious questions on the effects of
Organizations. an important part ofthe effects in question concern individuals. A 21-village
attitudinal survey was undertaken in October and November of 1995. This survey examined
the political attitudes of 252 adult rural Malians. Besides explicit questions comparing the
performance of the former UDPM regime and the current ADEMA regime, there were
numerous questions aimed at measuring "civic" values such as efficacy, tolerance, trust. and

Willingness to compromise. Also posed were numerous demographic questions and
questions about various sorts of participation. including voting, political party membership,
and organizational participation.

In order to gain an in-depth understanding of village-level dynamics, particularly the
procedures of rural associations, 52 semi-structured interviews were conducted. These
interviews. completed in November and December 1995, explored the nature of rural
associations and their links with local authorities, particularly village chiefs. the local
representatives of the national administration. Of special interest were issues of
organizational autonomy, the nature of membership, particularly whether they were
asCriptive or voluntary, and the internal procedures of these groups, specifically, were they

lrlClusive in decision making and leadership selection? Some of the interviews were also

42
conducted with key informants in the agricultural sector. For instance. various officials in
the CMDT, the cotton-producing parastatal. SYCOV. the cotton producers‘ union. and the
national Chamber ongriculture were interviewed.

During the course ofthe evolution of the dissertation. it became clear that the study
incompletely addressed issues ofinterest articulation. Specifically. what is the relevance of
participation in local associations to the national system? What is the relevance of peasant
organizations far from the corridors of power to a study aimed at studying issues of
democratic stability and consolidation? Fortunately. I was granted access to USAID/Mali

Democratic Govemance Strategic Objective (DGSO) performance-monitoring data on civil
society organizations. USAID. in its desire to help build and strengthen democratic
institutions in Mali, focuses almost exclusively on civil society capacity-building.
Consequently, the DGSO Team is cultivating a hierarchical network of organizational
support. where USAID finances programs of US. private voluntary organizations (PVOs)
Such as CARE and Save the Children-USA and these PV Os in turn form partnerships with
Malian non-governmental organizations (N G05) and community organizations (COs). This
democratic governance programming has evolved since 1995 but has only begun to be
Operational since mid-1997. In March and April 1997. USAID commissioned a baseline
Study of the performance of its partner Malian NGOs and C Os (Davis 1997b). A follow-up
Stlldy conducted in August-September of 1998 interviewed the leaders of 173 community
Organizations and 40 NGOS and federations, as well as about 40 local officials in
arrondissements and communes in which USAID had partners (Davis 1998). These data

Permit the examination of how multi-level civil society organizations fit into the national

43
political system. particularly what state and non-state entities they collaborate with and make
demands upon. Furthermore. they provide supplemental information on the nature and
internal procedures of such organizations.

All of the data sources used in this dissertation were collected over the course ofa
three year period. More importantly. however. each of the individual data bases were
compiled in a period of five weeks or less. Consequently. comparability ofdata within each
of the studies is unimpeachable. Each ofthe three chapters depends primarily upon one data
Source and thus constitutes a free-standing study. However, even across data sources.
Comparability is not expected to be a problem. For example. Chapter Two paints an
ethnographic landscape of organizational life in southern Mali in late 1995. Then. survey
data from 1998 is used to demonstrate two things. First. two different methodologies
discovered remarkably similar organizational characteristics. Second. not surprisingly. not
much changed in the characteristics of local organizations in Mali in that three year span.

1 .7 Dissertation Structure
Evidently, the theoretical problems raised at the outset raise a variety ofquestions.
These questions motivate the structure of the dissertation. The remaining chapters are
at‘I‘anged as follows.

In Chapter 2, the unit of analysis is the organization. Primarily through semi-
StI'uctured interviews but also using USAID survey data, I seek to uncover the nature of the
Village-level associations. How much do they have in common with the organizations seen
as constituting civil society in the industrial West? Are their characteristics and internal

proCedures of such a nature as would lead us to be sanguine about their contribution to

44
den‘iocratic stability. particularly in fostering civic attitudes? This study will permit the
generation ofa series of hypotheses to be addressed later.

Chapter 3 continues at the organizational level of analysis. exploring how local

organizations fit into the national political system. broadly construed. Are rural associations
even relevant to any discussion of Malian national politics? Or are such organizations so
disengaged from the state and its various tentacles that they bear no relevance to macro-level
political analysis? Specifically. what is the nature of local organization interactions with
local. regional. and national administrative officials. and what is the nature of their
relationships with other organizations in civil society. particularly federations and
intermediary NGOs? More precisely. are Malian civil society organizations involved to any
significant degree in civic action? Ifso. is this primarily a phenomenon of isolated. atomized
organizations. or is such action taken in concert with other elements in civil society? Finally.
regression analysis is used to discover the determinants of civic action behavior and civic
action success.

The analysis moves to the individual level in Chapter 4. Using the hypotheses
Presented at the end of Chapter 2, individual political attitudes and behavior are examined.
Chapter 2 asked the question "Is there any reason to expect Malian rural organizations to
Socialize their members into holding democratic values?" Based on the somewhat

pessimistic prognosis developed there. the analysis attempts to explain individual attitudes

on the basis of organizational participation and a variety of demographic variables.

aft”

4

Ur

Thednsenafionconchuksin(flunuer5\vuhasunnnnuyuptuthernahuthufingsof
the dissertation and an examination ofaltemative hypotheses and fruitful avenues for future
research. Cautious statements are made about the implications ofthese findings for donor

policy.

Chapter 2
Characteristics of Local Organizations in Southern Mali
There is little doubt that African civil society cannot be fully comprehended.
let alone assessed. by the classic instruments of analysis. Most institutions
that make up African society cannot be compared to those we see working in
Europe. where local, regional. and national structures meet regularly to
decide policy and choose leaders. In Africa. the leadership. membership, and
functioning of such structures are often shrouded in mystery... We cannot

define a specific African civil society without referring to its particularities
or the context in which it has emerged. -- Celestin Monga (1996. pp. 147-

148).

2.1 Introduction

Implicitly normative conceptions of civil society often assume that organizations will
have a salubrious impact wherever they arise. In contrast. I contend that organizations are
a product of the social and political milieux in which they evolve and cannot be expected to
behave otherwise. The present chapter examines the history, characteristics. and procedures
of four types of local organizations commonly found in villages throughout southern Mali.
IVIost previous treatments of civil society in Africa by political scientists have addressed
Organizations above the local level (e.g.. Bratton 1987. 1990. 1994; Bingen 1994, 1998;

Widner 1994; Holm. Molutsi, and Somolekae 1996). Where local organizations have been

treated, they have often been urban associations (e.g., Smith 1997, Woods 1994, Lucas
1 994). There are good reasons for this. In order to engage the state on behalf of constituents.

C lVil society organizations must have a critical mass of resources. Transaction costs of such

46

47
activity are much lower for urban groups (Bates 1981. 1990). and only larger-scale rural
groups are able to overcome barriers to engagement (Bratton 1990. 1994).

Nevertheless. the associational milieu in which the majority of Africa‘s people lead
their lives should not be overlooked for two reasons. First. in order to understand how far
the grasp of a weak state extends. one must understand the behavior and perceptions of rural
people with respect to that state. In other words, do state activities penetrate into the venues
W'here rural people engage in social and economic activity? If so, how does the local

interaction between state actors and rural people affect the behavior of the various parties?

Second, the associational contexts which structure social interaction for the

multitudes of people who rarely or never broach the cacaphonous quartiers of national
capitals are interesting in their own right, apart from the state. What are the types of
Organizations to which non-urban. non-elite Africans are most likely to belong? How are
they structured? How accountable are they to their members? Are they inclusive in their
Procedures or do traditional elites dominate their activities? Do oft-excluded groups such
aS women, youth, and the poorest of the poor find a voice in local organizations? Democracy
iS not only relevant to those who tread the corridors ofpower but also to those whose day-to-
day lives promise to be made better by inclusive and accountable procedures, regardless of

Scale.
Two sources of data are used to permit a systematic examination of organizational

Characteristics. First, I draw on fifty-two semi-structured interviews with rural association
leaders, village elders, technical experts and government and parastatal functionaries active
In the rural sector, particularly in southern Mali’s cotton zone. Then, USAID/Mali data on

Internal democracy in 173 community organizations are scrutinized to see how they

48
correspond to the initial conclusions. These data will enable us to refine hypotheses on the
role of rural organizations in the socialization of democratic values.

Procedural characteristics also have a bearing on how well organizations fulfill their
pluralist function. For example. in their study of over 150 organizations worldwide. Esman
and Uphoff found that organizations that are egalitarian in their approach to participation.
particularly those that conduct decision making in general assemblies rather than an
executive board. were more effective in their control over bureaucracies and in their own
claim-making (Esman and Uphoff 1984). Transparent decision making in general assemblies
can serve as a buffer against atavistic neo-patrimonial tendencies.

As discussed in Chapter 1. certain characteristics are considered important in order
for organizations to meet a W estem. normative conception ofcivil society with the potential
to contribute to democratization. For example. organizations must be autonomous from the
state, in order to avoid cooptation and to serve as a bulwark against tyranny (Tripp 1994b;
Hadenius and Uggla 1996). Voluntary membership is considered important because it
indicates freedom of association (Gellner 1994). Inclusive leadership selection and decision
making ensures participation by the rank and file and makes organizational leaders
accountable downward to their members rather than upward to patron-client networks
(Bratton 1994). Such procedures also contribute to the potential to inculcate enduring
democratic values in organizational participants (Diamond 1994).

These considerations are the most important in distinguishing between organizations

which are autonomous or coopted, accountable or unaccountable, and whether identities are
ascribed and rigid or freely chosen, flexible, and cross-cutting. Procedural aspects are

important to allow one to discern between organizations that may foster a commitment to

vlth

w

Ht...

0

,4‘.

Ni‘.

.

I

.u.

l
I
I

t .
'1"?
-““l-o..
a..-”

. '7
.~
~-..L~

I
.
2-.

J‘.‘

chi ‘4

I
".
~ t-A'h.
.g‘ a
u.“
I
. .1-
...“.
A, ‘ .
‘- A"!
.“‘
&
-\,
i ‘ A

\_'-‘_«..
. '~ 3.
.“
.‘x
\ u.
VHF»-
‘l
.s
‘

 

 

49

democratic values such as trust. tolerance. and compromise. These issues provide a set of
criteria by which to evaluate the democratic potential oflocal organizations. To evaluate the
external characteristics of organizations. assess their fit with the Westem notion of civil
society. and gauge internal levels of democratic procedure. local-level rural organizations
in Mali are evaluated according to observations on these four dimensions:

1. Is membership ascriptive or voluntary?

2. Is the organization autonomous from control by the state and traditional
authorities? In other words, do external authorities have any significant control over
organizational activities?

3. Is leadership selection conducted in an inclusive fashion? In other words. is there
any evidence of any kind of systematic participation by the rank and file in this process?

4. Are decisions made inclusively? Again, do procedures exist which ensure
systematic participation by the general membership?

Following the ethnographic evaluation of local organizations, these data will be
compared to the USAID data where possible to see ifthe characteristics uncovered hold for
a broader sample of local Malian organizations. First. some background is provided on the
region of Mali where most of these data were collected.

2.2 Background on the Cotton Zone

Mali’s cotton-producing zone is the rural area with the greatest degree of
associational activity. It extends from beyond San and Tominian, 450 kilometers northeast
of Bamako, near the Burkina Faso border. down past Kadiolo, along the Cote d‘lvoire
border, and over past Yanfolila and Kangaba, along the border of the Republic of Guinea

(Conakry). Due to the constraints of gradually-depleted soils in these areas and desires for

 

.‘..\

n‘.‘

 

 

 

50
ever-increasing production yields. commercial production has recently (in 1995-96)
expanded to include the area surrounding Kita, in Mali‘s westernmost First Region (Kayes).

Mali is Africa‘s number two cotton producer. after Egypt. This ambitious sector is
run by the parastatal C ompagnie Malienne pour le Der/teloppement des Textiles (CMDT).
The CMDT is currently a joint enterprise in which the Malian government owns 60 percent
of the capital and the French cotton giant C ompagnie Francoise pour le De'veloppement des
Fibres Textiles (CFDT) owns 40 percent. Ninety-Eight percent of Mali‘s commercial cotton
is produced in the CMDT zone (Bingen 1998, p. 270). with district offices in San, Fana,
Koutiala, Sikasso, Bougouni, and Kita. The remainder is produced in the adjacent zone in
which rural development activity is coordinated by the Operation de la Haute Valle'e du
Niger, which sells its cotton to the CMDT for processing and export. The combined CMDT
and OHVN cotton zone amounts to nearly fifteen percent of Mali’s surface area.

Since the decade preceding independence, the C FDT has been involved in Malian
cotton production.1 In 1974, the Malian government and the CF DT created the CMDT,
which currently has monopsony power in the Malian cotton market and has a monopoly on
all commercialization of Malian cotton. Cotton accounts for nearly half of Mali’s export
earnings and the cotton zone has household incomes significantly higher than the national
average.

Smallholder cotton production is fundamental to the success of the CMDT and the
other CF DT African partners in the international market. Their comparative advantage,

especially in light of the over-valuation of the CF A franc. is the fact that most Malian cotton

 

I But the French government first began exploiting cotton in the sub-region in the 19205. in order to palliate
French reliance on American cotton (Roberts 1996).

 

(L

.JF

(I

51

is cultivated by unpaid family labor (Bingen 1998. p. 273). In order to coordinate and train
this labor supply, the C MDT has promoted the creation ofAssociations Villageoises (AVs).
which are described in greater detail below.

CMDT field agents live closer to peasant farmers than most state officials. living in
villages rather than arrondissement clieleieux (Degnbol 1996a).2 Consequently. CMDT
agents are probably the representatives of the State with whom residents of that zone are
most familiar. Degnbol judges that interactions between rural populations and the CMDT
have been more positive than those with other government agencies. notably (and
notoriously) the National Forestry Service (,Eaux et F oréts). He attributes this to the fact that
the benefits of cotton production are easier to detect than those of other agencies. He also
suggests that higher salaries have generally made CMDT agents less corrupt than other
government officials (Degnbol 1996a. p. -3).

Nevertheless, relations between paysans and the C MDT have frayed in recent years,
primarily due to declining world market prices for cotton which resulted in stagnating
producer prices, stagnating CMDT wages, and a consequent increase in corruption (Degnbol
l996a, p. 24). Moreover. strong feelings developed that cotton producers were not being
given enough say in negotiations between the government, the CMDT, and the CFDT
(Bingen 1994, 1998; Coulibaly 1994; Degnbol 1996a, l996b). Further, discontent arose
about the CMDT’s monopsony power and producers felt that they should have the right to

market their cotton to whomever they chose. All these factors played a role in the creation

of the Association des A Vet T ons in 1991 and, subsequently, the national cotton producers‘

 

In spite of the parastatal distinction. the CMDT merits being treated as a state agency because it is
delegated primary responsibility for rural development in the areas in which it operates. Moreover. rural producers

make little distinction between the CMDT and the State (Degnbol l996a. p. 28).

52

union, the Syndicat des l’roclucteurs de Coton et Vivriers (SYCOV) in 1992 (Bader I992).
SYCOV has been hailed as a herald “of the grassroots potential for democracy that lies
beyond governmental institutions. “ though one of its biggest challenges will be ensuring
internal accountability of leaders to followers and avoiding co-optation by the CMDT and
the government (Bingen 1998, p. 284).3 The fear of co-optation is especially justified in light
of the asymmetry in institutional resources and the consequent reliance of SYCOV on the
C MDT for help with transportation and communication. Nevertheless. the existence of an
organization which can legitimately claim to represent the interests of a significant
percentage of rural Malians on the national stage is an important development in Mali’s
democratic transition.4

2.3 Organizational Characteristics in the Cotton Zone
2.3.1 Methodology

The data presented in the present chapter are aimed at evaluating the lay of the
organizational landscape in the cotton zone. We are interested in discovering the
characteristics most representative of organizations in southern Mali. In light of this, I chose
to sample villages with different levels of organizational activity. This permits getting a
broad sense of the range of dynamism and procedural approaches at play.

Because independent censuses or data on organizational activity across the cotton
zone do not exist, I was forced to rely on the CMDT's classification system. Given its
interest in the production of cotton, the CMDT promotes the creation of Associations

Villageoises (AVs) in order to provide a structure for the promotion of progressive growing

 

3 Studies by Bingen (1994. 1998) and Degnbol (l996a. l996b) provide excellent overviews ofthe issues at
stake in interactions between the CMDT and SYCOV.

4 . . . , . .
In addition to being the country s wealthiest zone. the cotton zone is also the most densely populated.

i-r.

1.1.1

lbl

45.4

the.

‘~b

H‘.

"‘uan

‘\
.
\\“
..H
t- -
l
“K
-
‘H.
'T

.‘E:

 

53

techniques as well as the commercialization of cotton. CMDT technical assistants classify
the level of organizational activity in a village as follows: A village classique is a village in
which there exists no "modern" AV (described below); A village A 1" simple is a village in
which a formal AV exists. but where no formal links have been established with neighboring
villages in order to promote development projects which are beyond the scale possible for
a single village; A village A V/ZAER (Zone d'Animation et Expansion Rurale) is a village
which has formed formal links of cooperation with nearby villages. Given the difficulty in
obtaining a data base on organizational strength, I used the C MDT classifications as a rough
proxy thereof. Accordingly. while recognizing the drawbacks of relying on the C MDT‘s
classifications, I provisionally classified the three categories as proxies for low. medium. and
high organizational activity, respectively.

In order to conduct the broader data collection program, I used CMDT records and
supplemented them with village lists from 1987 Mali census data to develop a sample frame
of over 4,000 villages and hamlets in the cotton zone. I then conducted a proportionate.
stratified random sample to select a total of twenty-one villages across the three different
categories. I selected eight A V/ZAER, 7 A V simple, and 6 villages classiques according to
their relative incidence among the 4000-plus villages and hamlets in the sample frame. The
political attitudes survey described in Chapter 4 was conducted in these 21 villages.

This sampling was consciously done in the spirit of heeding Robert Chambers'
admonition to avoid biases which obscure rural inequality (Chambers 1983. pp. 13-1 8). This
was accomplished by ensuring that the least organized villages were included, and by

ensuring that the sample was not biased in favor of easily accessible villages near paved

J—
lei»

 

 

.—

a)

I

54

roads. Less remote villages receive more regular contact with the national capital. are
generally more affluent. and consequently camiot be assumed to equivalent to more remote
villages. Thus. all types of villages must be represented in the proper proportions in order
to ensure representativeness.

The villages used for semi-structured contextual interviews were chosen in order to
ensure variation in terms ofdistance and remoteness from Bamako. organizational type. and
to some degree, ethnicity. I made a purposive selection of five of the twenty-one villages
surveyed and then added a Bambara cotton-producing village from the OHVN zone in which
I had done background research in 1994.

The following is a description of the characteristics of the five CMDT villages and
the single OHVN village:

"Fanadugu" is a Bambara village with a well-established A V simple in the F ana
region, about 100 kilometers from Bamako, 17 kilometers down a badly-rutted track from
the nearest paved highway.S Fanadugu has long been an important village. Its chief controls
land tenure in over forty neighboring hamlets with a total population of over 8,000, though
Fanadugu itself has fewer than 500 inhabitants. Because of the plentiful. fertile land in
Fanadugu, settlers from elsewhere in Mali (and often different ethnic groups) have formed
nearby hamlets in order to grow cotton.

"Koutialadugu" is a Miniarika village with an active A V/ZAER in the Koutiala region,
over 350 kilometers from Bamako, 8 kilometers down mediocre roads from the nearest

pavement. All of the association leaders and village elders interviewed spoke Bambara,

To preserve the anonymity of respondents. l have changed the names of the Villages studied. “Dugu is a Bambara
suffix meaning “town." “F010" means “first." and “filanan” means "second.“

55
including the women's group leader. Koutialadugu is the largest of the six villages studied
here. with a population of over 1500 inhabitants.

"Sikassodugu" is a Senufo village with a newA Vsimple ten kilometers from Sikasso.
its regional capital. though nearly 370 kilometers from Bamako. Proximity to Sikasso. as
well as the national highway leading to Bamako. are leading to the superseding of the Senufo
language by Bambara. the national linguafrancab At the present. very few villagers under
age 40 speak Senufo. Sikassodugu has a population of approximately 430 inhabitants.

"Bugunifolo" is a W asulu Peul village with an A Vsimple in Bougouni region. nearly
300 kilometers from Bamako and 121 kilometers down a decent dirt road from the nearest
paved road; The Wasulu Peul dominated an area near the Guinea border militarily one
hundred years ago at the time of Samory Touré’s conquests, but have themselves become
assimilated culturally and linguistically by their Bambara/Malinké neighbors. All the people
that we interviewed, therefore, spoke Bambara as their first language. Bugunifolo has nearly
800 inhabitants.

"Bugunifilanan" is a Bambara village classique in the Bougouni region, 210
kilometers from Bamako and almost 50 kilometers from the nearest paved road. It is
virtually inaccessible to four—wheeled vehicles.7 Bugunifilanan is the smallest of the villages
examined here, with about 200 inhabitants, according to a census carried out in 1995 by

visiting community health specialists.

 

6 Bambara, a Mande language closely related to Malinke and Dioula. is more correctly called Bamanankan
(and the people Bamanan), but I will employ the more familiar name.

7 .
The enumerators and I reached it on mopeds.

56

"Otwalidugu" is the single OHVN village included in the study. It is 55 kilometers
from Bamako and less than 20 kilometers from the nearest paved road. Otwalidugu is
inhabited almost exclusively by members ofthe Bambara ethnic group and has a population
of about 1 100. A very dynamic AV is active in Otwalidugu. comparable to an A I'D/ZAER.
though the OHVN development agency doesn't follow the same classification scheme as the
C MDT.

A note is warranted here on the relationship between village size and level of
organizational activity. Certainly, the existence of a direct relationship between these two
factors makes intuitive sense. Cotton producer organizations initially came into being at the
behest of the cotton parastatal. Consequently, larger, more influential villages were targeted
sooner and consequently often have a longer history of formal organizations. Moreover.
even without outside influence, larger villages tend to be influential in the affairs of their
area. Furthermore, many Malian villages lose able-bodied young men to exodus in search
of work for much of the year. As a result, extremely small villages may not retain a critical
mass of the energetic, literate elements most likely to animate organizational life.
Nevertheless, one must keep in mind that all of this is relative - all of the villages examined
here are small even by Malian standards, with not one being a chef-lieu of an arrondissement.
On most factors other than those of direct interest to us, these villages tend to be very similar
and one can be confident in maintaining the ceteris paribus assumption.

The village characteristics described above are summarized in Table 1.

 

 

 

 

 

 

 

 

£2st as 5382 :2 m :2 EM @6885
e853 8: as: E 2 .5. mm swsaeaé
sees: 82 as: as. w as. am sausages

who—Sum paw
885% com 8282 :2 t :2 a: sewage
:3 2355 8x 5282 as. .2 :2 8m 2&2st
«spasm com 33 as. 8 as SN aaeeswsm
ects—sac.— b_>=u< .53— .525 .5330 3:332
sate—:5 flue—$9.93 .aeeeafieaweO EPG 853m:— Eoc 353m:— oEaz own—.5

 

 

 

 

 

 

8:»...53230 ems—.5 - _ «Ea...

 

58

I visited each of the six villages for several days in late 1995. accompanied by a
research assistant/interpreter. Following protocol. we addressed ourselves to the village
chief, proffered cola nuts. and explained the purpose ofour mission. In every case. this was
the second or third visit by IMRAD personnel. and we were never refused in our request to
interview village elders and associational leaders.8 When possible and as a minimum. we
interviewed the village chief or his principal adviser. the president or secretary of the AV
(where applicable), the women's group president, the youth group president, and the hunters'
association president (where applicable). Occasionally. when we had more time and access.
or if initial informants were not very forthcoming, we interviewed additional people. We
interviewed six or seven leaders in every village.

Rural elites in Africa are not highly differentiated from the rural masses when
compared with rural elites in Latin America or Asia. A fact which demonstrates this is that
even village activists generally have very little formal education. Among the activists and
elders with whom we conducted the forty-one village-level interviews, only eight spoke
enough French to be even partially interviewed in French. and this figure is biased by the fact
that two of our rural leaders were also national office-holders in SYCOV, the national cotton
producers' union, and the nature of their positions required that they be functionally literate
in French. Consequently, thirty-three interviews were conducted entirely in Bambara and

the remainder were conducted at least partially in Bambara.

\

8 . .
. ’ The research was conducted while I was a research associate at Bamako’s Insrtut Malien de Recherches
APPIKI“¢¢$ au Developpement (IMRAD). IMRAD provided very important logistical support for this research.

i
l?“ A

.1...

than»-..

'. ”y
"ii-u}
N

HQ!) ~.‘
.....5 h

 

.g

19‘- ‘
Afi'fie.

"'sN “'
~ “4 ‘

3‘ -, '
"\i""i
M '
5.4‘
I\ 1
u; I
A‘. [Hi-o
‘K‘.

 

l.
1' ~ ‘
m .4
MR 1
~ I
‘A
\ i
‘

59

The twenty-one village political attitudes survey was conducted between October 5
and November 2. 1995. The semi-structured follow-up interviews were conducted between
November 21 and December 1 I. 1995. The bulk of the village visits were done in the weeks
immediately following the end of the rainy season. a period when travel to some remote
villages remains quite difficult. Because there is much agricultural work to be done at this
time of year, we were often required to conduct our interviews in the evening when
informants had finished their day's work.

The next section presents a detailed synthesis of data on rural organizations in
southern Mali.

2.3.2 Organizational Characteristics

Since independence, Mali's leaders have promoted the village and village
organizations as the locus of the country's economic development policy (Bingen 1994, p.
58) This concern with the village has manifested itself in different ways over time.
Nevertheless, "the main party that ultimately determines rural development. farmers. remain
the least represented at national debates on agricultural development planning," asserted
Dioné two years before the fall of Moussa Traoré. He claims that, since colonial times,
farmers have been excluded from policy decisions that affect them (Dione 1989, pp. 333~
334)

In fact, informal associations rooted in indigenous cultural institutions have existed

in W931 African villages since long before French colonial penetration (Lachenmann 1986,
Buijsrogge 1989, and Ouédraogo 1990). Buijsrogge points to traditional age class

assocrations among both males and females and cultural associations organized around

60

traditional religion and indigenous medicine as groups which have long had important

economic and social roles (1989. pp] 18-119).

Taking issue with treatments which she believes idealize the past. Danielle Jonckers
argues that an egalitarian society has never existed in Mali (.Ionckers 1994). Indeed. though
the subsistence economy in the past limited individualist behavior. lineage-based societies
rested upon a rigid sex- and age-based hierarchy. Affluent villagers would not allow their
neighbors to starve. but true solidarity was primarily based on direct biological relationships.
Nevertheless, Jonckers argues that existing social divisions are being exacerbated by new
forms of organization. For example. AVs, which focus on productive youth, can serve to
marginalize older villagers. AVs have an economic importance that the "traditional"
organizations did not have, in part because traditional groups were, and often still are,
focused on social purposes and any revenue earned went back into annual fétes. With the
AVs, technical training, literacy, production, and such matters have become more important
than the social aspect.

During French colonial rule. as well as after independence. a number of pseudo-
cooperative farmer organizations were established by experts working for the central
administration of rural development in Mali (Jones 1976, Jacquemot 1981, Belloncle 1982,

Dembélé 1987)? All of these organizations had the official objective of promoting the

capacity of farmers to collectively defend their interests. In fact, however, these
organizations were imposed from above and membership was compulsory (Dione 1989, p.

334). They were all subject to repeated official interference, even to the point of the

 

9 . . . . ., . .
See also the discussmn of Mali 5 cooperative law in Chapter 1.

'9‘ JP“ H

’5. leis“.
.

'J" .J‘ 91
.‘ti‘\\. 5.

[J"“l
h‘..|i\n\

 

[4‘5 HH
‘ “*‘s b s
n i

i i" >1

i.‘ “.k“
.\ ‘«p-\

I .
. t
AI'"'~n9.
.‘K. ‘
. lulu

 

 

 

61

govemment or parastatals choosing their leaders and giving them orders. They routinely
served the whims of the State in such matters as tax collection. compulsory cultivation of
collective fields by villagers. and forced marketing of food grains at official prices.

During the Kéita era (19(50-68), the Malian village was idealized as a model of
socialist organization and a principal building block of the ideal socialist society (Diarrah
1986, pp. 71-80). During this period. with assistance from French parastatals, Mali realized
notable gains in the production of cotton. which replaced peanuts as Mali's most important
cash crop. Nevertheless. villagers do not have many fond memories of the Kéita regime, as
it impeded capital accumulation by village cooperatives and was notorious for having a
labyrinthine bureaucracy. Furthermore, Kéita's promises of a radical regime based upon
grassroots democratic participation were not realized. The government's Milices Populaires
forcibly confiscated village grain stores and forestalled political liberties (Bingen 1994. p.
59). The Malian govemment's own Direction Nationale de l'Action Cooperative
(DNACOOP) has acknowledged that the collective fields and political excesses of the Kéita
era led rural people to be mistrustful of the government cooperative movement (DNACOOP
1989).10

In this section, I review the types of organizations and summarize internal dynamics
that I found most commonly in southern Malian villages: formal associations Villageoises
(AVs), women's groups, youth groups and hunters' associations. These illustrate the
diversity of organizations one currently finds at the local level in Mali -- The first two

organizational types described have generally come into existence since independence, while

 

“’ While noting this introspection. however. one must remark that it was expedient for DNACOOP to blame the
excesses of the previous regime.

 

 

62

the latter two represent the contemporary version of informal groups which have existed
since precolonial times. F urther. as we shall see. there is notable variation in organizational
characteristics as well. Needless to say. however. "It has become increasingly difficult to
differentiate the traditional from the modem, and the indigenous from the imported in
Africa" (Anheier 1987. pp. 426-427).

Of particular interest in the discussion which follows are degrees of organizational
autonomy, the nature of membership, and internal processes of decisionmaking and
leadership selection. Following the organizational descriptions, the organizational types will
be classified according to these four criteria. Distinguishing between organizations which
are voluntary in membership and those which are ascriptive is straightforward. Autonomy
is judged as follows: If there is evidence of regular participation of village chiefs or councils
of elders, the local representatives of the national administration, in an organization‘s internal
affairs autonomy is judged to be "weak." If, on the other hand, an organization usually
makes decisions that are not obviously constrained by these local authorities, autonomy is
judged to be "strong." Decisionmaking and leadership selection are both judged according
to how much they obviously take into account the preferences of rank-and-file members.
Where an organizational board of officers makes decisions affecting matters of significant
concern to the whole organization (including selecting leaders) without formal input from
the organizational membership or where the board‘s decision is a fait accompli
rubberstarnped by the rank-and-file, such a procedure is labeled "weakly inclusive."
Procedures where the opinion of the membership was regularly solicited before a provisional

decision was made would be labeled "strongly inclusive."

1'“ ..

l.l¢§

'\.
»

 

 

A. Associations Villageoises

More autonomous forms of farmer organization than had existed previously began
to emerge in the CMDT zone in the mid-19705. under General Moussa Traoré (Belloncle
1982. Dembéle’ 1987). Farmers were frustrated that some tasks critical to the marketing of
cotton, such as weighing and valuing output. were performed by C MDT agents rather than
locals. Farmers in a few villages began demanding that these tasks be assigned to literate
villagers (mostly school drop-outs). This proposal was well-received by CMDT officials
because they saw in it a mechanism for collecting delinquent loans (incurred during the
drought period of the early 19703). In exchange for being granted greater responsibility in
the marketing of their output, farmers' associations agreed to collect individual loans owed
to CMDT, sparing CMDT the trouble of coercively seeking loan repayment (Dione 1989,
p. 335). This permitted the CMDT to cut its administrative costs by substituting unpaid
villagers for paid officials, and it also gave the government organ a pretext for coopting
villagers (Bingen 1994. pp. 59-60). This is consistent with Bratton’s observation that
government technical staff often make use of local organizations to ensure service provision,
thus opening "opportunities for clients to engage administrators in on-the-spot bargaining
on programme implementation" (Bratton 1990, p. 95).

The evolution of the role of AVs paralleled a government recognition that "imported
Western cooperativism" was ill-suited to the Malian context. In 1982, hoping to promote
"voluntary collective participation," the Traore’ regime instituted the concept of Ton

Villageois, a village-level grouping based on the solidarity groupings which had existed since

- VJ» '—

We.”

'{J
.5 \
-
\i
ll.»
re?
.s.‘
iv r
ul ,

.-[‘

r- ?

5»

wt
u“.

1

.:‘[

 

o4

pre-colonial times (DNACOOP 1989. p. 11).” Under the new administrative system. AVs
were designated as pre-cooperative structures. Depending on their progress measured in
terms of literacy, bookkeeping skills. and diversity ofeconomic activities undertaken. among
other criteria, AVs were henceforth designated as being "weak," "medium." and "strong."
"Strong" AVs would be called "tons" and. under the purview of DNACOOP, would have
legal status and rights, while AVs would remain under the oversight of government regional
development operations such as the CMDT and the Operation Haute Valle'e du Niger
(OHVN) and would have no legal rights. In practice, it appears that promotion from AV to
ton status has often been made on the basis of technical assistants’ political and professional
expedience.12 In fact, Bingen has found that very few AVs request promotion to ton status
because their affiliation with the CMDT or the OHVN has provided them with administrative
support and financial guarantees necessary to request bank loans and engage in other formal
procedures. Consequently, they have no incentive to subject themselves to the administrative
hurdles involved in requesting a change of status (Bingen 1994, p. 65).

Currently, the appellation ton villageois is no longer being used in the CMDT zone.
More than one CMDT official informed me that this term was a vestige of the "politicized"
UDPM days, in contrast with CMDT organizational distinctions, which they claim constitute
a purely technical artifice through which economic development is pursued. AVs are the

fundamental link between rural producers and the CMDT and are used as a primary contact

 

” "Tan" is the generic Bambara word for “organization.“

‘2 For example, I've been told that whenever President Traoré was about to travel in the countryside. government
officrals would promote AVs to ton status, whether they merited it or not. in order to demonstrate that villages under
their supervision were making progress.

65
(after the village chief) in development-oriented overtures of outsiders coming into the
village.
AVs serve as versatile vehicles of agricultural development policy but are primarily

devoted to the needs of cotton producers rather than non-cash-crop producers. In spite ofthe

CMDT’s purported holistic concern with all aspects of cotton-producing villages’
development, poor villagers end up excluded from extension efforts, resulting in a widening
gap between affluent and poor villagers (Jonckers 1994). Amselle and Benhamou, agree.
arguing that AVs favor the emergence of rich peasants. in spite of the fact that they are
defended under the guise of "community development" (Amselle and Benhamou 1985).
These associations were created to facilitate the exploitation of cotton production, and this
remains their primary purpose.

Activities include collectively purchasing agricultural inputs on credit, facilitating
village-wide commercialization of cotton at harvest time, stocking of food grains for sale
during periods of food insecurity, securing of credit to permit the AV or individual farmers
to make capital investments in livestock or farm equipment, renting out equipment to other
villages, performing collective labor (1' 6k ebaara) for private individuals and in other villages.
and farming the association's collective field (foroba).

AVs have formal bank accounts, the contents of which they use for a variety of
purposes. AV expenditures include the purchase and maintenance of agricultural equipment,
advance money for the purchase of agricultural inputs for the AV, loans to members in
financial crisis (for which, depending on the village. they may or may not be required to pay

interest), and contributions to local development projects, including health centers, road and

66
school construction. and grain warehouses. Rebates (ristourne) on cotton revenue are used
for public goods such as these. The Otwalidugu AV treasurer told me that the village chief
and elders occasionally ask for AV money for village-wide purposes and they generally
accede to the elders' wishes. Thus. though primarily serving cotton producers. AV benefits
do often impact an entire village.

As discussed above. AVs were originally created in a top-down manner. beginning
in the late 19705 in the CMDT zone. Some previously unorganized villages now perceive
benefits to be gained from beginning or increasing cotton production and ask to create their
own AV. This is the case with Bugunifilanan. the smallest and probably the poorest of the
villages studied. Though still a village classique at the time that this research was conducted.
they were organizing an AV in cooperation with the C MDT during 1996. Before the CMDT
recognizes a new AV, it requires at least two villagers to become literate in Bambara and
come back and train other villagers. in order to ensure that a minimum threshold of
administrative competence exists in the association.

Following traditional age group associations. AVs are created as ascriptive
associations. That is, all adult residents of a village are members. However, over time some
AVs evolve, for all intents and purposes, into voluntary associations, in that they tend to
focus on issues of interest to cotton producers. so other villagers choose not to participate.”
Although the CMDT says that all villagers are members of their village's AV. in practice

most AV officers say that only cotton producers are members. A contrasting case is that of

 

'3 Because they are more accustomed to ascriptive groups than voluntary associations. many villagers seem not to
fully understand the concept ofa voluntary association. For example. the secretaries of the Bugunifolo and
Koutialadugu AVs both incorrectly told me that all villagers were automatically members of SYCOV. the national
cotton producers' union. even though they were aware that union cards were being sold in their villages.

. ‘ ‘
Hui-oh

{10: l.

V: ,v‘
|\|li‘ ll

‘.
'Mm'.

rah...

.lt

c
rue.—

;\
V ‘.L‘

l!
D

V: ‘
‘ . '7.
“A H.
‘\\
s

i‘ ~
o‘ ”I?“ 1
u“.

 

67

Otwalidugu. where all villagers are required to do collective AV work. regardless of whether
or not they grow cotton. The rationale for this is that there are some benefits from the
functioning of an AV -- such as school and road projects and access to credit -- that all
residents of a village enjoy. Otwalidugu discourages free riders by levying fines on villagers
who do not show up on designated community work days.

Largely because of strict guidelines promulgated by the C MDT, AVs have large
boards of officers with. at least on paper. a clear division oflabor. AVs almost always have
ten or more officers, but the most important officers are always the president and secretary,
as these are the only two members authorized to withdraw money from bank accounts, get
loans, and engage in most other formal economic transactions on behalf of the association.
One reason there are so many officers is that the villages have to balance loyalty, trust, and
competence in whom they choose as officers. Consequently, many villages have two officers
for every post, often an older man who is respected in the village and a younger man who
is literate and numerate.

Leaders are selected based on capability. particularly literacy in Bambara, trust, and
loyalty. Prominent families are disproportionately represented on AV boards. though some
villages have innovative representational procedures. In Otwalidugu, for instance, each
extended family (gwa) must have at least one member among the ten principal officers. In
Bugunifolo, in contrast, the association gets considerable input from the CMDT in choosing
officers. One village studied enjoys complete autonomy from CMDT agents because the
current chief‘s predecessor and a previous CMDT agent both died under mysterious

circumstances after having quarreled with him about land tenure issues. Villagers fervently

68

believe that the chiefkilled them with black magic. Ofcourse. his scary reputation threatens
AV autonomy because no villager dares controvert his will.

In most villages. however. elders exercise considerable control over who officers are.
For instance. in Sikassodugu. the feeble. elderly chief approves decisions. but his principal
adviser and the other elders make decisions about officers and the village tacitly accepts.
Officers are generally. though not exclusively, middle-aged men. I came across just one
village (Sikassodugu) where representatives of both the village women's group and the
village youth group were on the AV board and one other (Koutialadugu) where women but
not youth were on the board. This means that the majority of village adults, women and
young men, are under-represented in debates pertaining to a developmental vision for the
village.” One should further note that there is little turnover among these officers -- In most
villages, it appears that officers remain in their positions until they choose to resign or do
something wrong and are asked to step down.

Poor families play relatively little role in AVs. In the past. chiefs could call on the
ton to come to the assistance of poorer families. These days, Jonckers reports, tons have
become work groups for more affluent producers. and often are monopolized for the personal
benefit of chiefs. Such changes further serve to widen the gap between affluent and poor
villagers (Jonckers 1994. p. 128). Though I did not see direct evidence of Jonckers”
allegation at work, it is consistent with the dominance of AV boards by prominent families.

Many AV decisions are consensus-based among officers, with non-officers having

little direct say. For example, in Bugunifolo, the president and the AV board meet to make

 

M Jonckers agrees that AVs do not represent women, but she argues that youth gain power in AVs (Jonckers
1994). In comparison with pre-AV structures. she is probably correct. though in only a relative sense. Village youth
are still subordinate in the decision making process.

69

decisions and then inform the village. Among the board members. the majority rules and the
president has no veto. and the members of the association at large are obliged to accept the
board's decision. In Koutialadugu. where a very domineering. dynamic chief is in power. he
actually convenes AV meetings and the AV officers make no decisions without his approval.
In fact, the chief has the title tonfa. ”father of the organization." and Koutialadugu AV
officers are essentially conduits of the chiefs will.

AVs often wrestle with problems concerning the competence or honesty of their
officers. A big problem is asymmetry in numeracy. which leads numerate officers. most
often the secretary, to attempt to profit illegitimately from their position. Three of the five
AV or AV/ZAER villages that we visited reported this type of problem. Jonckers also
reports that corruption is commonplace in AVs (Jonckers 1994. p. 129). Clearly, the lessons
of large-scale state venality have been well-learned even at the lowest levels of Malian
society. This phenomenon must be considered before making any facile assumptions about
local organizations serving as "classrooms for democracy."

While chiefs have considerable input into most AV affairs, the increased prominence
of AVs at the village level could be perceived as a challenge to the authority of the chiefship
(cheflerie). AVs handle much larger amounts of money than chiefs and, whether for noble
or nefarious ends, it is in the chiefs interest to have some control over those funds. This
issue must be discussed in any account of village-level political dynamics. Bingen agrees
that AVs represent a dramatic change in the balance of power in CMDT villages, stating that
"[w]hile approved by village customary authorities, the AVs... have tended to marginalize

the older population and create a new class of ‘leaders’ whose power derives from their

70

association with the CMDT and not the community" (Bingen 1998. p. 28]). Given that
many chiefs gained power through association with the colonial French. the argument could
be made that the new ‘leaders’ are no less representative than many of their predecessors. but
AVs do, indisputably. represent a significant change in power relations in villages.

AVs also have an interest in amicable relations with village elders. In Otwalidugu.
the AV secretary frankly acknowledges that he and other founders of the AV had proposed
a self-effacing son of the chief as their president in order to pursue their developmental
objectives with minimal interference from the chiefand elders. Nevertheless, one should not
lose sight of the fact that this action is also perfectly consistent with a culture that values
consensus and social cohesion over conflict.

In Koutialadugu. a battle over succession to the chefferie led to government
authorities at the local arrondissement level resolving the dispute in favor of the man who
is now chief. This led to bad blood and lowered the legitimacy of the chief in the eyes of the
home quartier of his rival. an older relative. According to his opponents, the chief controlled
all the workings of the AV and embezzled funds from it. Complaints about his behavior led
him to force out AV officers unsupportive of him in 1993. This led to the virtual secession
of the rival’s quartier from the village and the creation of a second AV by ousted officers
of the original AV. Cases of this sort of village fragmentation, or éclatement, may reduce
village chief and elder control of AVs and thereby increase AV autonomy, but at the cost of
social cohesion.

AVs vary widely in the degree to which they work cooperatively with other villages.

Activities of short duration, such as seasonally teaming up to share the expense of renting

'r

 

¥

 

71

a truck to bring cotton or millet to market. are common. Formal project-oriented links to
other villages are very inconsistent. A cursory examination of both C MDT and OHVN
administrative divisions reveals frequent cases of villages not being grouped with other
villages with which they have long histories of social. cultural. economic. and political links
but with villages with which they have little historical affinity. For example. Otwalidugu is
in a separate OHVN secteur from a village six kilometers away with which they share many
familial ties and have historically engaged in social and economic activities. One would
think that villages could be grouped in such a way as to exploit natural affinities.
Interestingly, anecdotal evidence exists that the C MDT's creation of AVs has in some cases
been detrimental to horizontal ties between villages that had existed for generations. The
secretary of the Sikassodugu AV says that their work links with other villages had
diminished in the years since they had begun the process of establishing an AV because the
CMDT "said that each village must do its own work." This is a dramatic break from the past.
when Sikassodugu and its neighbors helped each other with labor-intensive agricultural tasks
for nominal or no compensation.

A charitable interpretation of the ostensibly irrational CMDT and OHVN
administrative divisions is that they have deliberately created cross-cutting cleavages to
avoid exacerbating regional affinities and fomenting inter-regional animosity. A neutral
interpretation is that they blindly drew lines on a map without considering any criteria other
than geography. A negative interpretation is that they deliberately sought to isolate
cooperative village clusters in order to forestall any solid fronts among the villages against

them. Without evidence for or against any of the three possibilities, the neutral interpretation

m. u i
the Iiil

Il'filv' ”1‘
s

- u, sand.

0., u

“it“.

t
"(w-q“ ll“
mum] ,g

’4" hIl-‘T‘: '1'

\u
'4

' i
‘T ”13‘;
u. (“‘5‘

ism-Mari I
’ l

su'“

F'viir'rv r n
V“

‘i
\Nt1‘L

.. ' > 1
A Ii“. h“,
.\ “1‘

 

 

.r; ‘
Q .
~ A.

'u. .
\tl ,“
x ‘5. ~
\

1

 

72

seems most plausible. Nevertheless. this sort of political (and economic) geography of
parastatal quasi-official administrative divisions would be a fascinating avenue for further
study.

B. Women's Groups

A recent report surveying rural organizations in Mali notes that. though male-
dominated organizations sometimes undertake initiatives that are directly or indirectly aimed
at supporting women‘s activities, women rarely have any significant role in decision making
in these organizations. In pursuit of managing their own economic and social activities.
women form village-level groups. Nevertheless, due to women lagging behind men in
becoming literate and numerate, young men often perform supporting roles in these groups,
primarily helping with bookkeeping and associated recordkeeping (IMRAD 1994).

In the cotton zone, ascriptive women's groups, formal or informal, have become
standard features of the rural landscape. As we found in Fanadugu, it is often unclear
whether these groups have been initiated by local representatives of the government’s
territorial administration or by village women themselves. In Otwalidugu, the only OHVN
village studied, a village-wide group had existed for decades before two Malian development
professionals affiliated with non-governmental organizations convinced village women to
start smaller groups to raise chickens for profit. Some groups, such as that in Bugunifolo.
are formally recognized by Malian authorities, while others are entirely informal. The
groups that we encountered had generally existed for ten to thirty or more years, by the
women's own admittedly hazy recollection. Often, women have long worked together

informally in their mutual economic interest. For example, in Fanadugu. before the

'\"

.
1,“.

1
fi‘r

t‘.

i“

73
formation of its present multivalent group. women had long cooperated to process and
commercialize shea nut butter.

Women's groups serve a variety of purposes. They are the locus of women's
collective activity in a village and as such mobilize women to serve village needs. They
permit income generation for the perpetuation of the group and allow members to
collectively battle financial constraints in their individual households. In order to generate
income, groups often cultivate a common field (foroba) or vegetable garden. do collective
work (iekebaara) in private individuals' fields, commercialize millet and rice to provide
village food security during the lean pre-harvest months. commercialize shea nut butter.
provide interest-accruing loans to members (and. occasionally, village men), and charge fines
to members who are late for collective work endeavors. More rarely, as in Otwalidugu.
women's groups engage in a wide array of small commercial activities. including dyeing
fabric, making soap, and raising chickens.

Expenses incurred by women's groups include paying for food and transport for
women going to classes to become literacy trainers. upkeep of local maternity clinics
([iginiso) and fiinds for infant nutrition, contributions to the chief for expenses incurred
hosting guests. and maintenance of common grain mills or water pumps. In most Malian
ethnic groups, women are entirely responsible for finding or buying meal condiments as well
as for clothing themselves and their children. Women's groups perform a vital role in serving
as a means for women to help each other fulfill these responsibilities. In an especially
elaborate example of this. the women of Bugunifilanan use collectively-generated funds to

buy clothing fabric and sandals for all village women and their children.

it" r'r-w.»
Lul. othu“,
5

l I

View tw,

““IH~ bll\
‘

‘.V I )r
“‘1 Ptiilt

. ., ‘1‘
)ihdjN it‘d

l

.i‘t'hai _
"Wilson

m; we

‘Ntu.’ t“

\ I

“ IVY; .
Ill.d“.‘v'l
I;

 

N‘P‘ti.

al
l
[/3
' 3
$-
‘7
H

74

The structure of women's groups varies by village. In some large villages. such as
Koutialadugu. one umbrella group unites all village women. while smaller groups that meet
with greater frequency exist in each quartier. In this case. the village-wide group only meets
during the less demanding months of the dry season. while the smaller quartier groups
cooperate on work projects throughout the year. Another arrangement is that seen in
Sikassodugu and Bugunifilanan. where sub—groups for young women and older women, all
married, exist within a single village-wide group. Sometimes sub-groups raise their own
funds while in other cases women's communal funds are centralized at the level of a large,
village-wide group.

Only married women take part in the activities of the women's associations in the six
villages studied. Unmarried women are normally members of a village's youth group until
they marry, at which time they cease to be members of the youth group and become members
of the women's group.'5 In some villages, elderly women do not participate in the main
village women's group but have their own informal group led by the first married among
them.

Leadership selection in women's groups is often heavily influenced by village chiefs
and elders. The selection of a president is based more on a subjective perception of
trustworthiness (danaya) and loyalty than on capability, though other posts are often filled
based on competence. Often, either the president or the vice president of a women's group

is the chief's wife. In the most extreme case of the chief and elders controlling a women's

group, Sikassodugu's president is the chief‘s wife and her vice president is the chiefs

 

IS . . . . . . .
This is in contrast to men. who remain youth group members even it they marry. only ceasrng to be
members if they pass the ceiling age (usually at about age 40).

n'ir ' nv.
c..l.\ . u
I O

i; ‘1 '
4311.01:

1 I i
r. . i r-
i‘nl'u “15‘

it r 2t“ 17‘-

Ini .,

””217

lflh.si r"

|\

0 "
.,.
h\ri

'71:: ' .
ul§fnbtx

I
We "ii

.( .
\"I-a 4

.1
-~'

75
principal adviser's wife. Women's group presidents are usually aged at least in their forties.
ifnot older, but I witnessed notable exceptions. In Bugunifilanan. for example. the president
had taken office at the age of about thirty. In F anadugu. a young. very bright. dynamic
woman was asked by the elders to head the woman's group but declined. stating that she had
elders and that they should be in charge. This wish was respected. but she nevertheless plays
a very prominent role in the affairs of the group.

Women's group leaders are generally not chosen for a specific term of office but
rather remain in office until they become too old. move away, or lose the community's
confidence. When a president dies or steps down, a village chief and elders normally have
a direct or indirect hand in the selection ofa successor. In Bugunifolo, for example. the chief
and elders appoint all nine members of the board of the village women's group. In
Koutialadugu, a powerful village chief appoints the president as well as all eleven other
members of the board, two from each of the village's six quartiers. In Fanadugu, women
have little say in village affairs in general but are left to their own devices in leadership
selection, with men technically having final say over women's leaders but in practice always
tacitly approving the women's own choices. In Bugunifilanan, the chief appoints the
women's president but women then assist her in selecting the other five officers. In practice,
they concede that they choose officers who they know will not elicit objections from the
chief.

Among the women's groups examined in detail, presidents held enormous sway over

the decisions made. In all cases, either the president alone or the president in concert with

other officers made decisions and then relayed them to the general membership. normally

T'fim
- u l».

‘TJYJY‘
“xvi:

“.0.
finch“;

Vi FPS“
“5“

s“ -“
\l
g 5

Ll‘ifiv,
i,”

5" 9.,

r

'-. ..
“a

Fin.

“‘1'
\ .
n

\

‘\,'.-.‘

‘N.‘

76

with no discussion by the group at large. In Bugunifilanan. the women‘s president
acknowledged the value of discussion by the board -- "A single person cannot solve all
problems: several people's ideas can result in a better idea" -- and board members are
allowed to challenge the president’s proposals. but the membership at large always accepts
a decision by the board. Otwalidugu was the most inclusive in women's decisionmaking.
There, the four most important officers (president, vice president. and two treasurers).
present their views on an issue to the membership at large and the members are allowed to
make alternative proposals if they are not happy with a proposal from the officers. In
practice, the membership usually acquiesces with little discussion.

Jonckers says that women have a certain autonomy in their own groups that they lose
when they join male-dominated groups. When in men's groups they lose their voice and
become dominated (Jonckers 1994.9pp. 131-132). This is encouraging, but in fact we saw
little voice on the part of women's organizations. Within women’s groups there seemed to
be great deference to leaders who were intimately connected to the village power structure.
External to the groups there seemed to be little voice in village affairs. This does not
contradict those who claim that apparent disengagement by African women often masks
strategies of seeking more private venues in which to be heard (Beck 1996. Patterson 1997).
Neither, however, does it in any way provide support for such a contention. While it is
plausible that female elites have the ear of male leaders, the present study cannot measure
such influence. In any case. the lack of inclusiveness in women’s group decision making

lends credence to the impression that the average woman has little voice in village affairs.

77

The degree to which women have influence over village-wide decisions varies
somewhat but is generally quite low. An extreme example of this is Fanadugu. The
matronne of the village matemity clinic informed me that women participate less in village
affairs in Fanadugu than in all five other villages in which she had worked -- "Here. women

N

do what their husbands tell them to do. end of story. As a case in point. the "husbands"
manage the most important women's project in the village -- the maternity clinic
management committee consists of the midwife and eight men. No members of the women's
group are on that committee. In a slight diversion from this extreme, the women of
Koutialadugu have two representatives. their president and vice president. on the AV board.
Their role is to inform women of decisions made but they don't voice their own opinions or
have a direct role in village decisions. The main village's women demonstrate this with their
unwillingness to comment on the fact that the seceding quartier's women no longer
participate in the village-wide women's group -- "This is a matter for men. We women don't
know anything about it." One village elder neatly summed up women's involvement by
saying "of course women participate in village decisions -- we make a decision and then
inform them." Though we saw little evidence of women’s political voice in the public
sphere, this is consistent with political theorists such as Carole Pateman who argue that
women’s political participation is best understood if no line is drawn between the public and
private spheres (Pateman 1991). Some have argued that African women seek more private,
less confrontational means of making themselves heard (Beck 1996, Patterson 1997). While

this is certainly plausible. my limited interactions with women’s groups, coupled with my

status as a male outsider. make it impossible for me to evaluate such arguments.

78

The secretary of the Otwalidugu AV. a well-respected middle-aged man who has
contributed a lot to the capacity-building of women's associations. told me that it was not
unusual for village elders to undermine women's organizations through the withholding of
permission for loans for women's economic activities. He attributes this to at least three
factors. First, men fear that women will be unable to meet the terms of the loan and the
village AV will be held liable for reimbursing it, as it is the AV that co-signs women's loans.
Second, men fear the specter of their wives becoming financially independent. Third, they
are simply averse to the idea of women's success drawing attention to their own indolence.
This is similar to the broader concept of fadenya (literally "father's children-ness”), where
one's half-siblings and villagers in general are resented for any successes or good fortune
they enjoy.

Women's groups do not appear to have many horizontal links with groups in other
villages. For example, in neither Koutialadugu or Bugunifolo did women conduct activities
with neighboring villages. In Bugunifilanan, an unorganized village by the CMDT's
estimation, women work with at least two neighboring hamlets. Though Koutialadugu's
insularity could result from its large size and complex quartier structure, no such factor
explains Bugunifolo's lack of significant ties to neighboring villages. This issue of
horizontal collaboration is important to a broad definition of civic community, where there
is a dense network of associations that, when necessary, work together to represent the
interests of their members. In general, the informal associations encountered in rural Mali

tend to act in a fairly insular, atomistic fashion, in sharp contrast to the ideal just described.

79

C. Youth Groups

One type of association with very deep origins in the cultural context of southern
Mali is the youth group. These ascriptive associations have roots in solidarity age groupings
and communal work groups such as the ciketon that have existed since long before colonial

rule. In the past. youth tons played a significant role in socializing desirable moral values

as well as in advising chiefs. When wars ravaged the country during the nineteenth century.
tons were called on to physically defend their villages against intruders. During the rule of
Modibo Kéita, he tried to create national youth groups that would provide ideological
support for his socialist regime. Similar local affiliates of the Union National de Jeunesse
Malienne (UNJM) existed during the Traore’ period. These formal youth groups officially
ceased to exist with the overthrow of Traore’, but informal organizations have continued to
this day.

Any effort to understand the role of local-level associations in Africa must be
informed by an awareness ofthe fluid nature of organizational distinctions. Asking villagers

where one informal group leaves off and another begins is often confusing and imprecise.

Based on extensive fieldwork among the Minianka. the ethnicity that prevails in
Koutialadugu, Jonckers argues that. though one is tempted to emphasize the continuity
between "traditional" associations and those that are of more recent origin, they are based in
fundamentally different forms of economic and social logic (1994. p. 121). This argument
clarifies the roots of social discord generated by new associations. For example, the

"éclatement" of many AVs, including the one in Koutialadugu, can be traced to conflict over

80

economic and power resources that have only developed in the years since the advent of
colonial rule.

It appears. however. that most villages have merged their ciketon with a more formal
grouping which has emerged since independence. although in at least one village
(Sikassodugu) the two still existed. with a division oflabor between the two groups. Among
the Wasulu Peul in Bugunifolo. the gwaminna is a social group similar to the Bambara
ciketon, and it exists side-by-side with the village youth group. These groups provide a locus
for social activity among younger villagers and a structure through which village leaders can
mobilize a robust labor pool for arduous collective endeavors. In some villages, however.
the increase in importance of AVs has resulted in a corresponding decrease in the importance
of traditionally-rooted youth groups. Nevertheless, in an age of economic precariousness,
youth groups can serve the vital role of stemming the rural exodus to Bamako and the Ivory
Coast by maintaining community solidarity among young adults.

This solidarity has clearly not disappeared entirely as is illustrated by the case of
Sikassodugu. Their youth group became virtually defunct a generation ago and young
people simply put together fEIes based on individual cash contributions. Then, about five
years ago, they re-activated the youth group. Why, I asked, if the social functions of the
group were still being performed on an ad hoc basis. was the group necessary? The president
told me that they saw active youth groups existing in other villages and wanted to
demonstrate that Sikassodugu, too, worked well together.

In five of the six villages studied, all unmarried women and all men aged 15 to 35 or

40, depending on the village, are members. but the women serve primarily as sources of

81

encouragement for the men. bringing water and food to work parties. and singing as the men
work. None of the six youth associations studied had female officers. Only Sikassodugu's
youth group had no female members.

Youth groups serve as an organizing structure to make young men available to
perform tasks to benefit the community at the behest of the chief. They make repairs on
village buildings, fix roads. clear fields. load trucks. and stock agricultural input warehouses.
They perform both unpaid collective work for their village as well as paid collective work
for their own budget. They levy fines against members who are disrespectful of others in the
group or show up late for collective work endeavors. With revenues from collective work.
youth groups often invest in technology such as generators, microphones, and speaker
systems to facilitate more exciting fetes (nyenaje). Other expenses include helping sick
villagers, assisting in funeral costs, helping with members' marriages or baptisms, and
providing no-interest loans to members in financial difficulty. The biggest expenses are
those related to throwing parties for village youth, a practice that has existed for generations.

It is instructive to emphasize that villagers engage in group activity for reasons that
go beyond simple economic incentives. For example, in Fanadugu the youth group had been
cultivating their own common field as well as performing remunerative collective work in
private fields, but found they did not have time to continue both. Though the collective field
was more profitable, they gave it up in order to do collective work. Their rationale was that
in doing collective work they were helping individuals who lacked adequate labor resources

and were thus providing an indispensable service to the village.

82

In four of the six villages. the president of the youth group was the oldest member
of the group and had succeeded someone who had stepped down because he had passed the
age limit. In Otwalidugu. the president of the youth group was not the oldest member but.
according to him, people had faith in him. According to an elder in the same influential
family, "he's stupid but well-known." Youth group presidents were noticeably less dynamic
than young men with key AV positions. but were often assisted by more dynamic. educated
younger men."’ Exode rurale. the mass departure of young. able-bodied men to the capital
and the coast in search ofwork, probably dilutes the talent pool available in the village to fill
such positions.

Other officer positions besides that of president are often chosen based on
competence. Often the president chooses his fellow officers with substantial oversight from
elders while his organization is passive in the process. But there are cases where the
organization actively nominates officers. In Bugunifilanan, for example. the organization
at large nominates officer candidates but the president has the final decision. In any case,
the boards of officers of these youth groups are generally smaller and less formal than the
boards of AVs.

Decision making procedures vary widely by youth group. In Koutialadugu, the youth
group is firmly controlled by the powerful chief. No important decision is finalized without
his approval. In other villages, the youth group president, sometimes aided by other board

members, makes firm decisions with little or no input from members. In Fanadugu, major

 

’6 Jonckers makes the point that. in contrast to women. young men gain power in AVs (Jonckers 1994). It
certainly makes sense that more promising young leaders would be drawn to AVs. which have more power and more
resources than youth groups. On the other hand. as with women. some young men may be drawn to groups where their
voice is not drowned out by older males.

I ..
uCkA

”9"?”
hhnbi
»

w. ..
1.. u

fl

.5 s.

0
'rl‘

«i\ if

I \
“’6qu
B ““

 

83

decisions are made by the association's board with input from the membership at large in a
general assembly. with the chief only intervening in cases of serious conflict.

Village elders have a strong hand in the affairs of youth groups. Because everyone
in a village knows everyone else and will have to live with them side-by-side for their entire
lives, as well as the fact that youth group officers are often the offspring of village elders.
youth groups virtually never challenge the wishes oftheir elders.

Cooperative efforts with other villages by youth groups appear to be diminishing. as
is the case with inter-AV collaboration. Nevertheless. such links still exist. For example.
the youth of Bugunifilanan still perform unremunerative collective labor for the next village
when they are asked.

D. Hunters' Groups

Hunter's associations (donsoton) have roots deep in the pre-Islamic religious
traditions of Malian ethnic groups.17 In former times. hunters often were given the task of
defending their village if it was threatened by outsiders or neighboring villages. In fact,
according to Jonckers. hunters' groups actually offered their services to the national
government during the periodic border conflicts with Burkina Faso. occurring most recently
in 1985 (Jonckers 1994, p. 128). To this day, Mali's hunters elicit a warm emotional response
even from urban dwellers. This deep cultural and religious significance makes hunters wary
of discussing their organizations with strangers for fear of inadvertently divulging secret

rites. Elders in only four of the six villages studied acknowledged the existence of hunters'

 

'7 ln Fanadugu. l was informed that the hunters' association had existed "since long before Samory." This is a
reference to Samory Touré. the great Malinké empire builder who led resistance to the French in the 18805 and 18905
and converted many of his subjects to Islam. At various times in his conquests during the period 1870-1898 he
controlled parts of what are now Guinea. Mali. Ivory Coast, and Sierra Leone. Samory is often used by older Malians
as a reference point for how recently a past event took place.

84

groups in their village. even though circumstantial evidence suggested to us that these groups
existed in the other two villages. ‘3 Reasons for denying the existence ofsuch an association
could be to preserve the secrecy ofmystical rites or for fear of having illegal hunting activity
reported to the authorities. In cases where hunters were interviewed. however, they provided
a rich window into associational patterns that have existed in Mali since long before colonial
rule.

Hunters' groups tend to be most active during the dry season. when game is easier to
spot and agricultural work is minimal. freeing them up to pursue their avocation. In some
villages, such as Bugunifilanan, all hunters are given a quota of game to kill during the dry
season. Those hunters who fail to meet the quota are required to provide millet beer
(chiapalo) and chickens for the hunters' end-of-season ceremonies. In Koutialadugu.
violations of association rules are penalized by a fine. in the form of cola nuts. beer or a goat.
depending on the severity of the offense. If activities are undertaken during the rainy season.
they may consist of hunters working the fields of their president, as occurs in Koutialadugu,
or performing collective work or working a collective field in order to raise money for
association expenses.

The hunters' associations encountered in the present study usually had an informal
fund (as opposed to a formal bank account) entrusted to their president. The principal

expenses of hunters' associations include giving or loaning money to members in financial

 

“‘ In Sikassodugu. for example. we asked elders in an initial discussion what sorts of organizations existed in their
village. They mentioned the nascent AV. the women's group. the youth group, and a potato producers' group. When
we mentioned that we had encountered hunters' groups in other villages. several elders became noticeably ill-at-ease.
conferred among themselves. and then announced confidently that they had never heard of such a thing in their village.

85
difficulty, paying for expenses of members who have to travel to meetings. hosting hunters
from other villages, and throwing big ceremonial fetes.

Hunters' associations constitute the most clearly voluntary associations that I came
across. in that membership was by choice rather than by ascriptive membership in a social
category. To become a hunter and to join a hunters' association one must be placed by one's
father in the care of a skilled hunter. The Otwalidugu chief pointed out that there are many
villagers who own rifles but are not in the association. He said that one had to love hunting
to join the association. Interestingly, hunters' groups seem to operate more independently
of village elders and internal village politics than do other groups. Protocol dictates that
hunters inform the village chief of their activities but he has little say in their affairs. For
example, in Koutialadugu, where personal animosity has resulted in the virtual secession of
one quartier from the village, as well as the creation of a second AV, the hunters from both
sides of the divide continue to share a single association.

Leadership selection is very straightforward among the hunters, and diverges from
other rural associations in that age or family is not the most important criterion. The
president of a hunters' association is usually the most experienced hunter, the member of the
village who has been hunting the longest. In other words, a man who began hunting at a
young age can become leader over an older man who began hunting at a later date. The chief
of Otwalidugu, the former president of the hunters' group who voluntarily gave up the post
when he became chief, likened leadership selection to a polygamous marriage. He said that
if a man takes a second wife who is older than the first she is nevertheless always the junior

wife. Such seniority is the custom in leadership selection among hunters.

86

In the Bugunifilanan and Otwalidugu hunters' groups. all decisions ofany importance
are made by the president. who is the only officer. In Koutialadugu. the president and his
vice president are the two most experienced hunters in the village. but they select two much
younger men to lead their peers on long hunting expeditions. These four officers make
decisions and then simply inform the other members without further discussion. In
Fanadugu. a frail. old. once-valiant hunter is the tonfa. literally, the father of the association.
while a younger. more physically-able hunter is president and runs the day-to-day affairs of
the association. One F anadugu hunter described the fond deference with which the tonfa is
treated -- "If he calls you. you go to him. It's like your father has called you. We respect him
and cannot question or refuse his decisions."

There is a national hunters' federation. the F e'dération Nationale de Chasseurs du
Mali, that engages in capacity-building among its constituent groups. About three years ago.
they encouraged the Fanadugu hunters to form a twelve-man board of officers, but it has yet
to become very functional. Fanadugu's hunters' association is still very much like a family,
with a benignly authoritarian father guiding his sons as he sees fit in spite of a modern board
of officers superimposed on the deeply-rooted, time-tested formula already in place. This
is the paradox of hunters' groups -- they constitute a genuinely voluntary form of association.
bringing together individuals in pursuit of common interests, but they govern themselves in
a more top-down manner than the ascriptive associations of their fellow villagers. Moreover,
in spite of the existence of the national federation, individual groups show virtually no

engagement with the State.

87

Historically. hunters and paysans in general have had a delicate relationship with
national authorities. particularly the field staff ofthe national forest service (Eater et F o‘rets).
During the corrupt days of the Traore regime. local officials often charged high taxes for
firewood collection and gave no receipt. allegedly set forest fires in order to arrest chiefs and
extort large "fines" from the villages. and "taxed" hunters heavily for killing game. This
relationship greatly compromised the legitimacy of the national government in the
countryside because the government officials with whom they had the most frequent contact
had a reputation for rampant venality. An important role of hunters' associations during this
period was to express solidarity when trumped-up charges were made against a member.
Interestingly, in the days immediately after the fall of Traoré, word of mouth carried many
stories of forest service agents being threatened with violence and run out of town. The rural
corruption of the previous regime has apparently largely disappeared since the transition
largely because, according to one Sikassodugu farmer. they fear the wrath of angry villagers.

E. Summary

Above are described four types of organizations typically seen in southern Malian

villages. The findings regarding each of these types are summarized in Table 2, below.

88

 

 

 

 

 

 

2,6305 >.§::.o> mzo_._.<_oomm<

mimic:— >2mo3 >28 3 weoem >_w:o:m .mzmhzoz
aims—o:—

o>62o£ >23 3 >28 3 23 3 o>cacom< 950:6 9sz03
aims—o:—

onEUE >28 3 >23 3 2mm 3 o>zatom< 950sz 5.39%

36205 beEEo> mzoFSUOmm<

032:2; >233 388352 Babes—2 >_8€oeo2 .mmmoanozm 20:00
ZOFUQJHW DEV—<2

Ezmzmnafix— -205..qu >ZOZOPD< 2:952sz ma>.—. ZO—h<N.Z<U~—O

 

 

 

 

 

=22 Eofizem E meeefioemm< .964...“qu he macaw—899.20 . N 033—.

 

 

89

The principal findings are the following. Hunters' associations have. primafacie.
external characteristics most similar to those of Western-style civil society. in that they are
a voluntary association which has strong autonomy both from the village chief. the local
representative of the national administration. and from state organs in general. However.
local hunters do not express themselves and advance their interests through engagement with
the state, leading one to question the role they can play in a nascent civil society. Their lack

of inclusiveness makes one doubtful regarding the potential for democratic socialization

effects. This failure to demonstrate inclusiveness suggests that members are not exposed to
procedures by which they could become accustomed to the ways of democracy.

Cotton producers' associations (AVs) most clearly meet the broader external and
internal criteria for civil society. in that they are moderately voluntary, have moderate
autonomy from the state and its vassals, and actively engage in the pursuit of development
objectives. This brings organizational activists into frequent interaction with state officials.
The fact that they actually engage in interest aggregation, have resources, and have
moderately inclusive procedures makes one moderately optimistic regarding their democratic
effects.

Women's groups and youth groups are both ascriptive associations and both
demonstrated very little autonomy and thus bear little resemblance to constituent
organizations in civil society as commonly conceived. For the same reasons involving a lack
of first-hand experience with democratic procedures, one is led to believe that they are

unlikely to have democratic socializing effects.

F"?~ .

i .
1thdtl‘. .

'VV’
1':

..‘ .
s

‘59.
it uh.
.L‘LVII,

”A

‘t
MI 4?
. 'M ‘

‘ci
3.

 

90

While AV activists have frequent contact with state and parastatal officials and are
loosely united under the auspices ofSYCOV. the national cotton producers' union. the other
three types of organizations studied have only tenuous links to the state as well as to other
groups in the voluntary sector. Both women's and youth groups had corporatist ties to
national organizations under the previous regime. and a national hunters' federation is trying
to unite hunters' associations to pursue their common interests. but links between these local-
level associations and the nation's capital seem only tenuous at best.

These differences among local associations are key to understanding the emerging
organizational landscape in Mali. Others have noted the diverging interests of men and
women, cash crop producers and not. Jonckers observes the evolving situation at the village
level and predicts inevitable confrontation (Jonckers 1994). Come what may, the mix of
“modern” AVs and “traditional” cikatons, women’s groups and mixed groups, engaged and
disengaged group is fundamental to understanding interest representation and socialization
potential at the local level.

2.4 Characteristics of Community Organizations Studied by USAID

The above discussion is very useful in its in-depth coverage of organizational
dynamics in six southern Mali villages. However. in spite of the careful methodology used
for selecting the villages, the generalizability of these results could be challenged.
Fortunately, for the purposes of evaluating these results, USAID organizational data exist on
similar attributes for a large set of local organizations which cover a similar geographic area.

In 1997 and 1998, the USAID/Mali Democratic Governance Strategic Objective

(DGSO) Team commissioned performance measurement surveys of their present and future

91

community organization (CO) and intermediary NGO partners.IQ These studies have been
critical to the evaluation and evolution of their program. which is grounded exclusively in
civil society capacity-building. The following discussion refers primarily to the 1998 study.
Further discussion and analysis of the USAID data is provided in Chapter 3.

The study administered two principal survey instruments -- one to all forty of their
partner intermediary NGOs and federations. and one to a sample of 173 community
organizations. 137 of which collaborated with USAID partners and 36 of which were a
control group of non-U SAID-partner organizations. Consequently. the USAID data provide
a representative sample of their CO partners in the regions of Koulikoro. Mopti. Segou.
Sikasso, and the District of Bamako. While one cannot be certain of how well the sample
represents the universe of all local organizations in this geographic area. it is instructive to
note that there were not notable differences between the control group and the target group
(Davis 1998). This leads one to believe that the USAID data provide a fairly representative
portrayal of associational life in central and southern Mali.20

These survey instruments were designed to measure performance according to a wide
variety of criteria, for which previous consultants had devised indicators. While USAID’s
indicators do not perfectly correspond to the categories evaluated in the first part of this

chapter, they are close enough to permit comparison. In order to measure the level of

 

'9 The two surveys were carried out by consultants from the Washington. DC. based Management Systems
lntemational. including myself. and Bamako-based Info-Stat (Davis 1997b. 1998).

20 As will be described in greater detail in Chapter 3. the original sample was drawn disproportionately in
order to ensure adequate sub-samples along the strata of partner PVOs and gender. The analysis for USAID reporting
purposes was then weighted to account for vast divergences in CO numbers by PVC, as well as the very low number of
women’s groups working with USAID. Because the present analysis is not trying to represent the universe of USAID
partners but to give an idea of the general population of local organizations (for which there exists no reliable sampling
frame). the analyses are not weighted.

92

internal democracy in COS. previous consultants developed a four-point index according to
commonly-accepted criteria. Based on the 1997 data collection experience. these criteria
were modified slightly (Davis 1998). resulting in the following:

1. Is the CO voluntary in membership or is it ascriptive? That is. do members
automatically belong according to their membership in a social category such as gender or
an age group?

2. Is leadership elected for fixed terms? If not. what type of leadership selection
procedure does the organization employ?

3. Does the organization possess written by-laws?

4. Did at least 60 percent of the rank-and-file members attend the most recent general
assembly? This is intended as a rough measure of rank-and-file membership participation
in organizational affairs.

These standards correspond roughly to those used earlier in the chapter. The first one
is identical to the voluntary/ascriptive criterion used above. The second corresponds to the
concern with inclusive leadership selection. broadly construed. The fourth constitutes a more
institutional interpretation of the concern with inclusiveness in decisionmaking. Thus, the
only topic from the earlier study not included by USAID was the concern with organizational
autonomy, and this is indirectly addressed by the leadership selection variable, as it
distinguishes between cases where village leadership selects CO leaders and those where
they do not.

The third criterion was included for the following reason -- The 1997 index made two
ofthe four criteria, leadership selection and convening of meetings, contingent upon whether

or not they conformed to procedures prescribed in organizational by-laws. In order to avoid

93
penalizing COs twice ifthey did not possess by-laws. existence ofby-laws was made a stand-
alone criterion. The reader should recognize that this is as much a measure of
“formalization” as of “democratic governance.” Nevertheless. it is included in this internal
governance index because it is also an indication of having clearly-stated. rather than
arbitrary, procedures.

As this index is a compound indicator for which the individual items are of interest
in their own right, cross-tabulations for each of the four components are first presented. At
the conclusion of this discussion. a table is presented calculating a distribution of scores for
the internal democracy index.

For precision in understanding the nature of organizational membership, USAID
originally wanted to determine whether their partners and potential partners were voluntary
in both founding and membership. The frequencies for "voluntary in founding" and
"voluntary in membership" were virtually identical in the 1997 study. Consequently. for
simplification. as well as to hold them accountable only for that which they currently have
the power to change, only the "voluntary in membership" variable is included in the index.
This variable elicits compelling information on the nature of membership in Malian local

organizations. Table 3 displays the findings.

94

Table 3 -USAID Internal Democracy Criterion 1: Membership Type

 

 

 

 

Women's Groups Mixed Groups Total
Voluntary Membership 32 66 98
76.19% 50.38% 56.65%
Ascriptive membership 10 65 75
23.81% 49.82% 43.35%
Total 42 131 173
100.00% 100.00% 100.00%

 

 

 

 

 

 

Pearson chi2(l') = 8.6261 Pr = 0.003

Community organizations in the sample were slightly more likely to be voluntary in
membership than ascriptive (56.65 percent to 43.35 percent). Perhaps more interesting,
women’s organizations were far more likely to be voluntary than were mixed organizations.
The explanation for this is relatively simple. The composition of the women’s group sample
is less varied than the mixed sample, with 26 of the 42 women’s organizations being credit
institutions, 85 percent of which described themselves as voluntary in membership. In the
case of mixed groups, 84 were AVs or parent-teacher associations (APEs). the majority of
which described themselves as ascriptive. The high percentage of ascriptive associations.
even among organizations such as APEs which do not have their origins in “traditional”
organizations, confirms the earlier observation that Malians often conceive of membership
differently than Westerners.

The 1995 ethnographic data reported above, combined with the fact that at least
eighty percent of the USAID sample consisted of organizations at least indirectly engaged
with an international partner, lead one to believe that the actual proportion of ascriptive
organizations in the general population of Malian local, rural associations is even higher.

The fact that such a large portion of organizations are based on ascribed rather than chosen

95

identities suggests a lower likelihood of multiple memberships and the cross-cutting
cleavages of affiliation which have been long held to promote trust. tolerance. and
compromise (Komhauser 1959).

Another variable inextricably linked to the cultural context from which it springs is
leadership selection among the local. primarily-rural organizations in the sample. The next
table shows the distribution of leadership selection procedures among the groups.

Table 4 - USAID Internal Democracy Criterion 2: Leadership Selection Procedures

 

 

 

 

 

 

 

 

 

 

Women's Groups Mixed Groups Total

Leadership elected 9 26 35
for specific time period 21.43% 19.85% 20.23%
Leadership elected 2 16 18
for open period 4.76% 12.21% 10.40%
Consensus of members 26 62 88

61.90% 47.33% 50.87%
Selection by village notables 5 27 32

11.90% 20.61% 18.50%
Total 42 131 173

100.00% 100.00% 100.00%

 

Pearson chi2(3) = 4.3683 Pr = 0.224

 

Here, one-fifth of the C05 surveyed elected their leaders for fixed terms. There was
virtually no difference between women’s groups and mixed groups reporting this approach.
On the other hand, twice as many mixed groups as women’s groups (20.61 percent to l 1.90
percent) reported that their leaders were chosen by village notables. Again, I believe that this
result is more a consequence of the types of groups in each gender sub-sample than anything
systematic across all COs -- Only two of the 26 women’s credit institutions had their leaders

chosen by village notables, while 18 of 47 APEs reported that village notables selected their

96

officers.“ Also worth noting is the fact that half ofall groups surveyed reported choosing
their leaders by consensus. While this does not meet USAID’s standard of democratic
leadership selection. it is consistent with arguments of the existence of "democratic a
l’africain." in which the points of view of all members of the community are heard before
a decision is made. Unfortunately. however. the data are too general in their categories to
give an idea as to how inclusive the consensus process was in each case. In other words. it
is difficult to know how much of a voice females. younger males, and relative latecomers to
a community have in decisions?2 If these frequently-marginalized groups have little say. the
term "democratic" may be an overstatement.

Though USAID did not explicitly consider issues of organizational autonomy. the
fact that 18.50 percent of the groups surveyed report that village notables select their leaders
suggests considerable infringement on the autonomy of a substantial portion of local

organizations. This seems to be a reliable ballpark estimate, as the 1997 study came up with

the estimate that 15.70 percent of USAID partner community organizations had their leaders
selected by village notables. The 1995 data reported above. combined with the fact that at
most of the USAID sample consisted of organizations at least indirectly engaged with an
international partner, lead one to believe that the proportion of organizations having leaders

chosen by village notables is even higher in the general population.

 

2’ Again. the fact that “modern" APEs are so linked to traditional structures and procedures is a notable
finding. One should note that village authorities have long been tied to the regime in power. combined with the fact
that APEs had their origins in the educational policies of the Traore’ regime. In light of this. involvement of the
chef/erie in the internal affairs of APEs is not surprising. Current analyses of the evolving Malian political context
point to the reform of APEs as being critical to eliminating local corruption and making APEs effective and
representative, as well as to restore their credibility (Charlick et al. 1998. p. 40).

22 As is illustrated by Henri Konan Bédie’s cynical maneuver to rule Alassane Ouatarra ineligible to be

president of C6te d‘lvoire because his parents were born in what is now Burkina Faso, the term “latecomer” is
culturally relative.

 

97

Table 5. below displays the distribution ofCOs according to whether or not they were
able to show the existence of formal by-laws.

Table 5 - USAID Internal Democracy Criterion 3: Existence of By-Laws

 

 

 

 

 

 

 

 

 

 

Women's Groups Mixed Groups Total

Existence of by-laws 10 37 47
confirmed 23.81% 28.24% 27.17%
Existence of by-laws 18 46 64
claimed but not confirmed 42.66% 35.1 1% 36.99%
No by-laws 14 48 62

33.33% 36.64% 35.84%
Total 42 131 173

100.00% 100.00% 100.00%

 

Pearson chi2(2) = 0.8427 Pr = 0.656

Interestingly, the results shown here are virtually identical for mixed and women’s
groups. Women’s groups were slightly more likely to claim the existence of organizational
by-laws but were slightly less likely to produce those by-laws upon demand. For the
purposes of the internal governance index, COS were only given credit for meeting the
criterion if they were able to actually show their written by-laws to the enumerator. The
reason for this is that there is significant risk of response distortion -- Organizations would
want to impress a data collector reporting back to a present or potential donor -- so one
cannot take claimed-but-unconfirmed by-laws at face value. Community health associations
(ASACOs) and one of the two types of women’s credit unions were the most likely to show
by-laws (both over 50 percent).

Table 6, below, shows the distribution of organizations according to percentage

attending their most recent general assembly.

98

Table 6 - USAID Internal Democracy Criterion 4: General Assembly Attendance

 

 

 

 

 

Women's Groups Mixed Groups Total
Attendance at least 60% 2 6 8
4.76% 4.62% 4.65%
Attendance 0-59°/o 1 7 8
2.38% 5.38% 4.65%
No Attendance Records 39 1 17 156
92.86% 90.00% 90.70%
Total 42 130 172
100.00% 100.00% 100.00%

 

 

 

 

 

 

The usefulness of this indicator is severely impeded by the fact that over 90 percent
of the respondents were unable to present any kind of attendance records for their
organization. This lack of recordkeeping attests to the lack of formalization of many of the
organizations in the sample, but the issue is more complicated than just that. As noted
above, nearly half of the organization leaders surveyed reported that their CO was ascriptive
in membership. When organizational leaders report an organization such as an APE or AV
to be ascriptive. this usually signifies that the entire village automatically belongs. This
raises at least three problems with measuring attendance. First. if everyone residing in a
circumscription or geographic space belongs to an organization the membership/non-
membership dichotomy is negated and the idea of measuring membership attendance
rendered less meaningful. Second, even if an entire village’s residents are considered to be
members, there is usually no expectation that every adult resident attend meetings. In fact,
the most common approach seems to be to ensure either that each household (du) or
extended family (gwa) is represented, but this is not always the case. Consequently, the

methodological implications of calculating general assembly attendance become unwieldy,

r'Q JP.)
bkbtb

l 6‘ r
03.016;

1.7% Ni;

9‘. ..'
0. LCM:

rm . - -
111.6135“

99

as the researcher would have to clarify a variety ofdefinitions on a village-by-village basis
before any meaningful calculation could be done. Third. even where villagers are clear on
who is expected to attend a general assembly. the logistics of recording hundreds of attendees
and then knowing the precise number of adults in a village in order to calculate percentages
are extremely dificult. Again, this highlights the difficulty of applying Western conceptions
of certain political terms to very different cultural contexts.

Of the 16 CO leaders presenting usable infomiation. exactly half met the standard of
60 percent participation in their most recent general assembly. This is somewhat
encouraging, though potentially misleading. This indicator is intended to reflect rank-and-
file participation in an organization. One cannot automatically assume that these results are
applicable to the entire universe of C03 (or even to the rest of the sample) for the following
reason -- It would be reasonable to assume that organizational leaders who go to the effort
to keep attendance records tend to value such participation, ceteris paribus. more than
leaders who do not. Consequently, any effort at extrapolating non-missing data to the entire
CO population would vastly overstate the proportion meeting the attendance criterion.

For the reasons stated above, the very absence of usable data for this indicator is an
interesting result. This. coupled with the relative paucity of organizations demonstrating the
existence of by-laws demonstrates that the majority of C03 working with this particular
major bilateral donor are somewhat informal in nature. This level of informality in
recordkeeping and procedures may come as a surprise to those accustomed to studying
regional or urban organizations. Indeed, the intermediary NGOS interviewed as part of this

same study are notably different from the C03. Fifteen of twenty-five NGOs surveyed were

100

able to demonstrate with written documents and records that they convened meetings in a
manner consistent with their by-lavvs.23

The internal democracy index aggregates CO scores based on the four criteria
presented above. The distribution of scores by women’s and mixed-gender groups are

presented in Table 7.

Table 7 - USAID Internal Democracy Index

 

 

 

 

 

 

 

 

[Indexz voluntariness + leadership alternation + by-laws + general assembly
attendance;
one point for each item; a score of 4 = "democratic;" 0 = "undemocratic" ]
Score Women's Groups Mixed Groups Total
Four 0 0 0
0.00% 0.00% 0.00%
Three 7 11 18
16.67% 8.46% 10.47%
Two 5 22 27
1 1.90% 16.92% 15.70%
One 22 55 77
52.38% 42.31% 44.77%
Zero 8 42 50
19.05% 32.31% 29.07%
Total 42 130 172
100.00% 100.00% 100.00%

 

 

 

 

 

 

The most notable result here is that only about one-quarter (26.17 percent) of all the
community organizations surveyed received at least two of a possible four points on the

internal democracy scale. and none received a perfect score.

 

23 The reader will note that the NGOs were held to a stricter standard and a higher percentage nevertheless
met it. In fact, though internal democracy and sound management indices were calculated for both C05 and NGOs. the
criteria were different and uniformly more rigorous for the NGOs. This was done for the simple reason that USAID
and a series of consultants felt it reasonable to hold them to a higher standard. and this approach was supported by the

data.

101

While there are slight differences between the distribution of the women’s groups and
that ofthe mixed groups. cell expected frequencies are too low to generate a valid chi-square
statistic. This can be remedied by combining the top three categories. Once this is done. the
difference between the two groups is insignificant. Nevertheless, the irregular shape of the
distribution of women‘s groups. with a small peak at “three." a slight trough at “two" and the
mode at "one." merits examination. This can be explained as follows --The 27 mixed
organizations in the non-target control group were distributed similarly among the different
scores to the mixed-gender USAID target groups. In contrast, all of the women’s groups in
the control group only scored zero or one on the internal democracy index. This partially
explains the unusual distribution of the women’s groups in the sample. The remainder of the
difference is explained by sector -- Only one of the 49 APEs and local education committees
(the largest sectoral group in the mixed sub-sample) in the sample scored three out of four
on the index. In comparison, six of the 26 women’s credit institutions (the largest sectoral
group in the women’s sub-sample) scored three out of four. Most of the high scorers among
the mixed groups were ASACOs, eight of which scored three out of four.

2.5 Conclusion

The USAID data confirm earlier observations in several respects. First, a large
portion of local organizations in Mali are ascriptive in membership. The different cultural
conception of organizational membership has important implications regarding the
democratic effects of organizational participation. In particular, if an individual belongs to
a group simply because of an immutable characteristic, he or she may not necessarily assume
that the group’s leaders have to do anything like interest representation. Moreover, that

individual may not have any assumption that his or her point of view must necessarily be

102

taken into account by leaders. In other words. there may be no pressure for inclusiveness or
accountability from below.

A second earlier result confirmed by the USAID data is that a significant minority
of organizations report that traditional village elites choose their leaders. In fact. the
percentage reported here presumably understates the interference of village notables in CO
affairs because it only notes cases of total control over leadership selection. It does not
include cases where, as noted above. for example. AV leaders choose members of a village
founding family as president in order to reduce conflict. There is also no measure for the
portion of CO boards composed, for whatever reason. of village elites such as founding
families.

A third way in which the USAID data confirm previous observations is in the
conclusions about general assembly attendance and what that says about rank-and-file
participation in CO affairs and decision making. Even if we were to take the best case
scenario, extrapolating from those organizations for which data were available, this means
that fully half of all organizations do not meet the general assembly attendance standard. In
fact, as previously stated, we suspect that actual numbers are significantly lower than this,
calling into question the degree of accountability of CO leaders.

Also of interest in the data is the fact that the degree of democracy among women‘s
groups and mixed groups is not significantly different. It is important to note that the
distribution of C05 in the USAID sample was quite different from that in the 1995 in-depth
interview data. Most notably, AVs were not remarkably different from other organizations

in the sample. which were generally more formal entities than in the previous study. That

103
said. it is nevertheless striking how similar results from the two studies were in many
respects.

Even though the 1995 organizational types are somewhat different from the USAID
organizations. the distinction between women's groups and other groups is worth tracking.
The observations of Pateman. Beck. Patterson and others that women‘s participation differs
from men‘s (Pateman 1991. Beck 1996. Patterson 1997). coupled with Jonckers‘ pessimistic
predictions ofinevitable confl ict (Jonckers 1994). make one curious about differences among
women’s groups and others in civic action activity. That will be just one of the hypotheses
tested in Chapter 3. Chapter 4 will permit the examination of the democratic socialization
effects of local organizations in light of the present findings regarding organizational
autonomy. membership. and procedures.

These data demonstrate that local organizations in rural Africa are often very different
from the idealized associations that W estem democratic theorists were describing when they
spoke of the role of civic organizations as classrooms of democracy. Rural associations in
Mali perform vital social, economic. and cultural functions that one must comprehend in
order to understand the context upon which liberal democratic institutions have been
superimposed. However. researchers and policymakers must be extremely cautious about

assuming the similarity of entities which have evolved in radically different contexts.

31c
dim-

Chapter 3

The Interest Aggregation Function of Local Organizations

Bolokoloni den kelen; a te se ka bele ta.
(Bambara Proverb: "a single finger cannot lift a stone")

3.1 Introduction

This chapter continues at the organizational level of analysis undertaken in the
previous chapter. but here the scope of the study is expanded. Beyond examining how local
organizations correspond to Western conceptions of civil society, we explore how local
organizations fit into the national political system, broadly construed. In Chapter 1. the two
principal functions of civil society were discussed. The first is an interest aggregation
function -- to aggregate the interests of individuals and groups. The second is an educational
function -- to socialize values that are conducive to democracy. The former is discussed in
the present chapter. while the latter is presented in Chapter 4.

The internal dynamics of rural organizations are intrinsically important because they
affect the quality of life and social justice in the areas where most Africans live. Internally
democratic, representative organizations have the potential of giving voice to non-
autochthones, women, youth, and other traditionally less-privileged groups. They can also
serve as counterweights to village founding families. village elders, and other local elites.
However, issues of organizational self-governance and local democracy become even more
important if links can be established between local politics and the national system. Ideally,

local democratically-structured organizations are linked into the regional and national levels.

104

OT,

A3

5.
r1.

Slim

v1“

105

both to represent the interests of members as well as to ensure democratic accountability at
all levels. The present study has increased relevance if links can be demonstrated among
various components of civil society. as well as with the various levels of government and
administration. Specifically. what is the level of engagement or disengagement between
these local groups and the various levels of state administration? More specifically, what
is the nature of local organization interactions with local, regional, and national
administrative officials. and what is the nature of their relationships with other organizations
in civil society. particularly federations and intermediary NGOs? More precisely. are Malian
civil society organizations involved to any significant degree in civic action? If so, is this
primarily a phenomenon of isolated. atomized organizations, or is such action taken in
concert with other elements in civil society? Answers to such questions can inform debates
on a constructive role for civil society participation and rural politics in democratic
transitions. Further, it can inform interactions between states and donors and civil society.

A variety of plausible explanations exist for why individuals combine into groups to
pursue collective interests. but these explanations were largely developed outside of Africa
and do not fully respond to the unique constellation of factors at play in African settings.
Rather than engaging in procrustean attempts to make Africa fit into rigid, pre-conceived
theoretical constructs, we must evolve theories which account for the cultural and historical
specificity of Africa. But any informed discussion of the African context can draw lessons
from frameworks developed elsewhere.

Purely pluralist approaches assume that individuals unite freely in a variety of groups
according to diverse interests with cross-cutting cleavages which preserve social and political

stability (Komhauser 1959). Different organizations respond to the needs of different

~Ii-t

at».

we.

106

interests and. taken together. they balance out each other as well as the State. The State. for
its part. serves as a relatively neutral arbiter in policy debates. These organizations must be
autonomous from the State but engage it on some level. seeking to influence policy (Hodgkin
1956. Wallerstein 1964). Esman and Uphoff find that too close ties to government agencies
are negatively correlated with claim-making and control over bureaucracy. By the same
token absolute autonomy is not effective, because there must be a certain amount of
engagement for effective claim-making to take place (Esman and Uphoff 1984).
“Engagement with autonomy” permits enough contact to place demands while allowing
enough distance to not be coopted.

Many have pointed out that pluralist arguments are unrealistic in the African setting,
where, as already discussed. groups are often formed based on lineages, ethnic, or other ties,
and these groups are as likely to be ascriptive as voluntary in membership (Gellner 1994;
Monga 1995, 1996). Other criticisms of a purely pluralistic approach have been leveled. For
example, African states are held to be far from neutral in mediating the various. often
conflicting demands made by societal actors (Bratton and Bingen 1994, p. 21).

Neo-Marxist approaches posit that individuals form groups based on class interests
and that the state is a tool for realizing the hegemonic program of the dominant classes (Leys
1975, Sklar 1979). These arguments are less convincing in largely rural societies with tiny
middle classes and no significant industry. Nevertheless. in a rural economy such as the
planter-dominated coffee and cocoa sectors in C6te d’lvoire, plausible neo-Marxist
arguments could be made.

Both pluralist and neo-Marxist approaches make more sense at the urban level,

particularly in national capitals. than at the local level in Africa. At the local level. lineage

107

often dominates. Nevertheless. individual arguments can be made of organizations serving
class interests. For instance. as discussed in Chapter 2. at their founding most AVs included
entire village populations in their membership. In fact. however, they have usually become
the exclusive reserve of cash crop producers. For example. cotton farmers in the CMDT
zone have become a source to be reckoned with. Encompassing a range of levels of wealth
and technology use, they do not constitute a homogeneous group. but they represent a
significant swath of rural Mali above the poorest of the poor.1

Pluralism is also more plausible among urban groups because of the relative ease of
organizing. For example, “urban consumers are geographically concentrated and
strategically located” (Bates 1990. p. 156). Rural dwellers. in contrast, are widely scattered
and are at the mercy of poor transport and communications infrastructure. This greatly
increases the transaction costs involved in collective action. Moreover, urban settings bring
together people of diverse ethnic and religious origins in a way that is seen less frequently
in more-homogeneous rural settings.

Corporatist approaches suggest that social stability is maintained by all important
interests being pulled into the ruling apparatus (Schmitter 1974. Shaw 1982, Remmer 1989.
Nyang’oro 1989, Bianchi I989).2 Nyang’oro points to formal organizations serving as the
official representatives of mass constituencies as “quasi-corporatist” structures (Nyang’oro

1986/ 1987). In the Malian case, the creation of health (ASACOs) and education (APEs)

 

l In fact. while civil society can level the playing field. the poorest ofthe poor are rarely well-represented in
civil society organizations (Esman and Uphoff 1984). This lack ofpresence can be attributed to insufficient leisure
time, education differences. or a lack of influence.

2 Corporatism is an institutional framework by which structures for interest representation are either created
or co-opted by the state. State-approved interest groups gain favor and predominance over non-approved groups. This

mutually-beneficial arrangement shares power and influence with moderate groups and is aimed at ensuring stability
for relatively-uninclusive regimes.

108

associations. as well as cooperatives (Ions villageois) and related federative structures under
state ititiation have a quasi-corporatist flavor. Without making a holistically corporatist
argument. Bingen points out that. since colonial times. Malian peasants have participated in
politics in an essentially corporatist fashion. through state-controlled cooperatives or party
organizations. He describes the current risk of SYCOV. the cotton producers’ union. being
coopted as essentially a corporatist project of the C MDT and state elites (Bingen 1998).

Nevertheless. corporatist arguments fall short in Africa for a variety ofreasons. Most
fundamentally, as exemplified by the examples just cited. “corporatist” projects are
piecemeal and not systematic. Where corporatist behavior has appeared, it has failed, for a
variety of reasons. First, African states do not have the capacity to dominate all political
space in their societies, room exists for grassroots organizations to maneuver. Second, class
formation which would permit coherent occupational blocs is impeded in Africa by regional
and ethnic identities. Third. the leaders of African states have rarely ruled in a consciously
corporatist way, constructing networks based on blocs of influence. In fact, some argue that
these factors, combined with Africa’s status straddling patrimonial and rational-legal
authority systems, make neo-patrimonialism a more convincing framework for viewing
African politics (Bratton and Bingen 1994, pp. 21-22).

Accordingly, Callaghy, Bayart, Bratton and van de Walle and others have analyzed
the African State as a nexus of neo-patrimonial relations (Jackson and Rosberg 1982;
Sandbrook 1986, 1993; Bayart 1993; Callaghy 1984, 1994, Bratton and van de Walle 1994,

1997).3 The logic of African politics has usually been driven by personalities and patron-

 

Under neopatrimonialism. politics 15 not based on ideology btit on the distribution of resources. power. and

prestige. It is "a form of governance based chiefly on personal loyalties. patron-client networks. and coercion"
(Sandbrook 1993. pp. 27-28).

109

client networks rather than by rational-legal frameworks. Bratton takes this a step further.
arguing that this lens can refine our perception of associational life and interest
representation in Africa as well (Bratton 1994). This point is well-taken -- It stands to reason
that organizations would be a product of the milieu in which they evolve. Interests are not
advanced on relatively even footing before a more-or-less neutral arbiter. Whole networks
are built around shared rent-seeking and divided spoils. Iforganizational leaders are drawn
into these webs, they become accountable upwards to their patrons and not downward to
their members. For instance, the existence of a very affluent “clan CMDT” at the highest
bastions of Mali’s ruling ADEMA party raises the risk of SYCOV leaders being drawn into
patron-client networks. Where in corporatism the rewards for co-optation are power and
influence for one’s ideological agenda (though there also exists the risk that one’s agenda
will become watered-down), in neo-patrimonialism the rewards are primarily material. The
consequences for SYCOV’s constituents are, therefore, arguably worse if their leaders are
drawn into neo-patrimonial networks, because downward accountability has little role in neo-
patrimonialism, except inasmuch as rewards are distributed to loyal clients.

As discussed above, African kinship-based organizations can pose problems when
analyzed in terms of normative Western constructs such as civil society. Noted
practitioner/academics have advised that new institutions should be made compatible with
“traditional norms” (Esman and Uphoff 1984; Korten 1990). Bingen makes the point that.
unlike the historical experience of farmer organizations in the West, “there is a tendency in
Sub-Saharan Africa for community-based organizations to exist, and in many instances to
continue, as a viable constituent unit within interest-based organizations” (Bingen 1998, p.

281). But Hadenius and Uggla caution that

) r
illumil

‘ . ‘»?
1.0“h

P
v

”FA. “
‘u‘tJNJ‘

to;

1'1

Ak‘l'

'rI

11's" ,4“,
' 4
““4
L ihl'g .
6'. “h
\LL

l 10

traditional norms. rituals, and patterns of authority are part of the reason why

a strong and viable civil society is absent in many Third World Countries...

While grassroots organizations based on traditional structures do not have to

be dominated by elites. new organizations emerging from below are less

likely to be so (Hadenius and Uggla 1996. p. 1625. citing Fisher 1994).

While it is the fashion in many of the rosier portrayals ofcivil society to paint them
as representative bodies. one should note that the local associations described here do not
normally represent the interests of the entire community. For example, women’s groups and
youth groups can in the best case represent the ensemble of these under-represented
constituencies. While pluralist explanations do not require that an organization be broadly
representative, they suggest that organizations represent the interests of their members. In
fact, however, they can also be controlled by founding families or other traditional village
elites and serve neo-patrimonial ends. Neo-patrimonial behavior is difficult to measure, but
two factors lead one to believe that it is fairly common in the villages studied. First.
founding families tend to have inordinate influence in the organizations studied. Second, in
most of the villages described in Chapter 2 there was anecdotal influence of fiduciary
malfeasance by organizational leaders. This combination of elite domination and use of
common resources for private ends leads one to believe that neo-patrimonial behavior is
fairly common among the sorts of groups studied here.

Each of these approaches -- pluralism, neo-Marxism, corporatism, and neo-
patrimonialism -- is plausible and offers stimulating insights to debates on the functioning
of African politics. The two which intuitively have the most to offer to analyses of the role
of civil society are pluralism and neo-patrimonialism. These are useful lenses through which

to evaluate the behavior of organizations and their leaders. There is an obvious tension

between these two points of view and those who see pluralism at work must be vigilant, for

l 11
today’s accountable leader could be coopted into tomorrow‘s link in the neo-patrimonial
chain.

Other. more directly policy—oriented. treatments of local organizations examine how
they fit into development planning and policy. Organizations in civil society have been
posited to have a wide range of developmental functions. In their broad study of local
development-oriented organizations with primary accountability to their membership. Esman
and Uphoff identified a variety of “types and tasks” of local organizations (Esman and
Uphoff 1984). Eschewing the complexity of a typology developed by David Leonard and
his colleagues (1982). they break down the range of local organizations into three categories:
local development associations. cooperatives. and interest associations.4 They acknowledge
that these categories are not mutually exclusive. For example, the AVs described in the
present study meet most of the criteria for both the local development association and
cooperative categories. The Leonard typology, consisting of 24 types of “local participatory
organizations” would certainly eliminate such ambiguity but with no more analytical
leverage, Esman and Uphoff insist (Esman and Uphoff 1984, p. 68). They break local
organization tasks into four general categories: 1. Intra—organizational tasks such as
planning, goal-setting, and conflict management; 2. Resource mobilization and management;
3. Service provision; and 4. Extra-organizational tasks such as control of bureaucracy and
what they call “claim-making” on government. Of course, Esman and Uphoff view such

tasks as outputs, which are not valuable in their own right but only in as much as they

 

4 Local development associations are area-based. bringing together most of an area’s residents for varied
development purposes. The defining characteristic of cooperatives is the pooling of member resources. Interest
associations can unite for functional purposes. such as education or health. or along social category lines such as sex or
religion (Esman and Uphoff 1984. pp. 61-64)

112

contribute to the outcome of local organization performance (Esman and Uphoff 1984. p.
82).

Michael Bratton expands on the concept of “claim-making.” drawing on Hirsclunan‘s
concept of “voice” (Hirschman 1970; Bratton 1990). Bratton sees claim-making as being
limited to attracting resources from an existing program into other localities. He sees policy
voice as being somewhat more ambitious. such as seeking the initiation of altogether new
programs or the modification of existing ones (Bratton 1990. p. 95). He specifies the
conditions under which organizations representing the rural poor can gain policy influence
and alter the allocation of public resources. These include: a homogeneous membership, a
balanced, federated structure with strong links between levels, a focused program,
constructive informal ties with political leaders, and a solid domestic funding base.

The action of civil society organizations seeking policy voice is an example of
“society-sponsored engagement,” where citizens engage in collective action in order to
influence the state or even seek control of state power (Bratton 1994). It is not a given that
all organizations make this choice. as they also have the option of avoiding the state through
such manifestations of disengagement as apathy, exodus, the informal sector. or other means
(Azarya 1988, 1994; Bratton 1994). Bratton links the choice of strategy to access to power,
which he defines as a combination of political autonomy and political capacity (Bratton
1994, pp. 235-239).

Independent resources are quite important to organizational autonomy. Organizations
must be able to generate resources which are sustainable and leave them beholden (and
perceived as beholden) to no one. Membership dues are a good place to start, but small, rural

organizations cannot rely on meager dues alone. Organizational revenue sources must

113

expand beyond membership dues without relying excessively on grants from external donors
or the state. Outside support can reduce initiative. a necessary entrepreneurial spirit. and also
cultivate a “psychology of dependency” (Uphoff 1991. p. 499). Moreover. it can contribute
to the sort of neo-patrimonial relations described above. making organizational leaders more
accountable to external leaders than to their membership. Support from international donors
can also reduce legitimacy or credibility in the eyes of an organization‘s interlocutors or can
permit the state to discredit them as externally controlled.

Organizational political capacity depends on the ability to articulate shared interests
within an organization or across organizations. a tradition of collective action. a favorable
economic environment and a somewhat predisposed state. Hadenius and Uggla’s five stages
of state treatment of civil society, ranging from a hostile, totalizing state through a non-
encouraging but non-interfering state to a benevolent state that actively promotes
autonomous organization. constitute a useful framework through which to view prospects

for civic action (Hadenius and Uggla 1996, p. 1629).
Like society, the state also has the options of engaging or disengaging (Azarya 1988,
l 994; Bratton 1994). Examples of state-sponsored engagement include economic planning,
6f forts to regulate social behavior and interaction, control market supply and demand,
particularly with regard to foodstuffs, and the creation of public enterprises. Examples of
State-sponsored disengagement include economic privatization, as well as personalistic
behavior such as corruption which does not extend state authority but rather compromises
it- Constructive actions on the part of both the state and civil society can lead to mutual
engagement, where civil society voices important demands and the state considers these

Clernands in its policy making.

114

A collaborative research and action project called the Research Group on Agricultural
Politics in Africa (RGAPA) sheds light on the role of farmer organizations in agricultural
policy in Africa. This research. based on case studies from Kenya. Zambia. Zimbabwe.
Senegal. and Mali. has been synthesized and summarized by Michael Bratton and Jim
Bingen (Bratton and Bingen 1994). The project participants made interesting findings
regarding practical lessons which improve prospects for successful policy advocacy by
farmer organizations. For example, they found that organizations require a strong popular
base and independent resources in order to have credibility and stake out independent policy
positions. Moreover, successful policy advocacy results not from organizational “loners” but
from coordinated efforts by a multitude of groups. They also found that the case studies
pointed toward the need for leadership accountability and rule of law within farmer
organizations. This implies the existence of organizational by-laws which provide for
regular elections, fixed terms, and leadership alternation. A final important lesson Bratton
and Bingen point to is the importance of capacity in fundamental areas such as literacy,
numeracy, accounting, and sound management skills in general (Bratton and Bingen 1994.
pp. 23-25).

Links between organizations in civil society are particularly important. Renshaw
makes the point that civil society is a relational concept. It takes on its clearest definition
when juxtaposed with other autonomous organizations as well as the state (Renshaw 1994).
“L’union fait la force” -- It is not sufficient to have numerous, atomistic organizations in
operation. These groups must have horizontal and vertical ties which permit them to share
capacity and coordinate civic action activity (Clark 1991). Vertical ties are important for

linking local organizations into the national system. Though many have expressed wariness

115

of vertical ties in the context of patron-client networks. it is not vertical ties in and of
themselves which pose a problem, but rather the direction of accountability. Where leaders
and organizations are not forced to answer to their base. they risk being coopted by the state
or drawn into patron-client networks. Just as leaders must be accountable to rank-and-file
members within organizations. intermediary NGOs and federations must be accountable to
local organizations in inter-organizational hierarchies.

The scale issue raises a fundamental paradox for civil society policy voice in Africa.
In spite of the fact that neo-patrimonial pressures pervade both state and society at all levels,
accountability is much easier when scale is small -- people know their collaborators and
interlocutors. and a reservoir of time-eamed information and trust pervade a transaction or
interaction. As one gets further away from the local level, uncertainty increases, reducing
trust. Because information reduces uncertainty and increases trust, accountability is more
easily ensured when scale is small. However, the probability of successful civic action
increases as scale increases. There is thus a tension between increases in the probability of
successful policy advocacy and the ability of rank-and-file members to hold leaders
accountable.

Along with differences of scale and the urban-rural dichotomy, important differences
exist in civil society along gender lines. In light of social, political, and economic contexts
which privilege men over women, political theorists such as Carole Pateman challenge the
liberal distinction separating the public and private spheres, a distinction that is fundamental
to definitions of civil society (Pateman 1991). Building on this idea, and suggesting that
personal and family life are deeply affected by state policies, Aili Mari Tripp holds that

women’s exclusion from formal political institutions has resulted in their seeking of

116

alternative strategies to gain control over their lives. Among these strategies are
disengagement from disadvantageous interactions with the state. Similarly. women are prone
to disengage from male-dominated associations which fail to take their interests into account
(Tripp 1994b). This disengagement is often also induced by cultural and religious
conceptions of appropriate roles for women (Folbre 1986a. 1986b). Women’s own
associations often prioritize service provision over interest representation (Tripp 1994b).
This does not mean that women do not voice their opinions. only that they make seek more
private, less confrontational, means of expressing themselves (Beck 1996, Patterson 1997).
The marginalization of women in both electoral politics and traditional power structures. as
well as exclusion from cash crop production and low rates of female school attendance rates
pose special challenges for women‘s political participation at all levels. Even in local
venues, one expects these factors to lead to lower feelings of efficacy, impeding civic action
activity.

The above discussion brings to light certain issues that are fundamental to an
understanding of interest aggregation behavior by organizations. F irst. many organizations
do not engage in such activity. having decided for whatever reason to not engage the State.
Second, there are a wide array of levels at which such State engagement could take place,
from local representatives of deconcentrated ministries, to legislators, to development
parastatals. Third, such action can be taken individually by atomistic organizations or in
horizontal or vertical collaboration with other voluntary sector entities. Fourth, the simple
decision to engage in civic action does not ensure its success. One could reasonably expect
levels of organizational capacity and inclusiveness to have an impact upon prospects for civic

action as well as a successful outcome to it.

117

In the interest of examining these issues. and understanding the extent of rural
organization claim-making upon or policy voice to public officials, the following questions
are raised:

1. Do local Malian organizations engage in civic action? This is civic action broadly
construed, and includes both Esman and Uphoffs limited notion ofclaim-making as well
as Bratton’s more ambitious policy voice.

2. If they do engage in civic action of some kind, what is the scale of that activity? More
specifically, with what levels of public officials. administrators, or parastatal enterprises do
they initiate contact?

3. Is this civic action undertaken in a solitary fashion or in collaboration with other
community organizations or regional or national NGOs or federations?

4. Are these organizations successful in civic action? That is. by their own report. do they
manage to convince officials to change unpopular decisions or in some other way result in
a tangible outcome desired by the organization?

5. Do they seem representative of their membership base? Without being prodded. do they
speak of the concerns of rank-and-file members? This is important, as base participation
increases representativeness and forces accountability.

6. What are the determinants of engaging in civic action? Specifically, what organizational,
political and demographic factors influence such behavior? For example. are women’s
groups less likely to engage in various claim-making activites than mixed groups?

7. What are the determinants of successfiil civic action?

The next section explains the data and methodology used to examine these issues.

118

3.2 Data and Methodology

The following analysis employs USAID/Mali Democratic Governance Strategic
Objective (DGSO) performance measurement data on their partner civil society
organizations. USAID. in its desire to help build and strengthen democratic institutions in
Mali, decided to focus exclusively on civil society capacity-building. Consequently. their
data on the preformance of local organizations represents the most comprehensive data
available on civil society and civic action in Mali.

To understand the questions that were asked and the resulting data, one must be
aware of the general contours of USAID/Mali’s program. The Mission aspires to the
following broad goal: That "Mali achieve a level of sustainable economic, social, and
political development that eliminates the need for concessional foreign assistance"
(USAID/Mali Results Plan). The 1996 - 2002 strategic plan aims at achieving the program
goal by focusing on three highly interrelated "strategic objectives" within the social,
economic, and political realms.5 Among these strategic objectives contributing to USAID
Mali's program goal is a Democratic Governance Strategic Objective (DGSO). The DGSO
is exclusively civil society based and foresees the day when: "Community organizations in
target communes are effective partners in democratic govemance, including development
decisionmaking and planning" (USAID/Mali Results Plan).

The DGSO is cultivating a hierarchical network of mutual support between USAID
and a range of partners. This network is organized as follows: USAID finances programs

of US. private voluntary organizations (PVOs). including CARE, Save the Children/USA,

 

5 In addition. there are two “special objectives.” one aimed at the improvement of information dissemination
and communications. and the other aimed at helping Mali’s northern three regions emerge from the Tuareg rebellion of
the early 19903 and move toward sustainable development.

119

World Education. and the Cooperative League of the United States of America (CLUSA).°
These PVOs in turn usually form partnerships with Malian non-governmental organizations
(NGOs) and federations at the national and regional level. as well as with community
organizations (C Os) at the local level.7 Some of these intermediary NGOs and federations
assist the PVOs in the implementation of their programs while others simply receive
assistance, particularly capacity-building training in such areas as conflict resolution,
management, project design, etc.8

This democratic governance pro gramming has evolved since 1995 but has only begun
to be operational since mid-1997. In March and April 1997, USAID commissioned a
baseline study of the performance of its partner Malian NGOs and C03. The product of this
was a baseline sample survey of 168 C05 that partner PVOs expected would take part in
future democratic governance programming. A census of all 30 USAID-partner Malian
intermediary NGOs was also conducted. A second round of DGSO performance
measurement was conducted in August and September of 1998, with a sample of 173 C05
(137 USAID partners and 36 non-targeted organizations), as well as a census of all 25 Malian
NGOs and 15 federations collaborating in democratic governance programming. The

subsequent analysis relies primarily on the 1998 study, with occasional references to the

1997 study. These data permit the examination of how civil society organizations at various

 

6 CLUSA operates under the name National Cooperative Business Association (NCBA) in the United States.
CLUSA prefers to be referred to as a cooperative. not a PVO. For simplicity’s sake. however. and with no offense
intended, I will use the common parlance.

An exception is CARE. which intervenes directly wrth community organizations, With no intermediary.

8 In my experience. some Malian NGO leaders resent serving as implementers of the programs of
international NGOs. Many would prefer to be more equal partners in development. having more autonomy to do their
own programming.

120
levels fit into the national political system. as well as how extensively they collaborate with
each other.

These USAID data were collected for the very applied purposes of their monitoring.
evaluation and reporting purposes. The 1997 baseline indicators survey and the 1998 follow-
up study constitute a fundamental component of their Performance Monitoring Plan
approved by USAID/ Washington. While the C O samples cannot be claimed to be perfectly
representative of the southern Mali local organization universe, they do exhibit substantial
variability and provide a vehicle through which to understand civic action activity and its
determinants. Moreover. these studies provide the most systematic data available on a range
of associations operating in southern Mali, as well as how they are interconnected. The
present section provides information on the nature of the data and the sampling methodology,
in order to set the proper context in which to view the results.

The two performance measurement studies described here were conducted in order
to meet USAID/Mali’s reporting requirements vis-a-vis USAID/Washington, as well as the

Mission and PVC partners” needs to adapt and improve programming. Legitimate questions

could be asked about the generalizability of the USAID CO sample. Because the different
PVOs have different programs and different criteria for inclusion in these programs. one
could reasonably surmise that their respective sets of CO partners vary in how representative
they are of the larger CO population in the regions under study. Because anecdotal evidence
suggests that some US PVOs have minimum criteria in terms of the capacity of potential
partners, it would be reasonable to suspect that the present sample displays greater average
organizational capacity than the typical CO in those regions. The principal purpose in

employing the USAID data is to take advantage of the best available data set on local

121
organization civic action behavior. all the while recognizing that the data set may not be
perfectly representative ofthe universe of Malian C Os. By the same token. the fact that there

were not large differences between the target C O and control group results suggests that the

sample does not deviate greatly from the universe."

3.3 The Sampling Frame and Design10

In order to define a sampling frame. the list of every single member of a population.
the population must first be defined. Because of the early stage at which DGSO
programming found itself at the time of the data collection. it was difficult to define the
population exclusively in terms of C Os receiving DGSO assistance. In the context of the
new integrated cooperative agreements between USAID and the PVOs, it is often difficult
to dis-integrate sub-components of a holistic program to isolate cases of DG support. For
this reason. the decision was made to define the population of the C 0 study as follows: “All
community organizations collaborating with USAID DGSO partners under USAID financing
as of 1 August 1998.” This means that. as long as a CO collaborates with one of the PVC
partners, even if they have not engaged in specific DG activities, they are included in the
sampling frame.” Of the 137 USAID-partner C05 in the sample. most have or will be

collaborating on issues of health. primary education. or sustainable economic growth, in

 

9 The USAID sample was drawn in a larger geographic area than the 1995 data reported on in Chapters 2 and
4. The 1995 cotton zone data were drawn from an area falling entirely within the boundaries covered by the USAID
study. Seventy-two of the 173 C05 (41.62 Percent) in the sample are in the cotton zone. F orty-three more COS (24.86
percent) come from circumscriptions in Segou and Koulikoro regions that are immediately contiguous to the cotton
zone.

to The sampling information that follows describes the 1998 study. For more detailed information on
sampling frames and sampling methodology. see Davis I997b and Davis 1998.

I Because different PVOs were at different stages in their DG programming. the minimum threshold level
of collaboration in order to be included in the sampling frame was quite low. In fact. anecdotal evidence suggests that
some C05 in the sample had not had more than superficial contact with USAID partners.

122
addition to being a part ofdemocratic governance programming. The control group C Os are
from the same sectors as the target (‘05. Consequently. all of the C05 in the sample are
engaged in social and economic development activities.

Given the concerns of the DGSO Team concerning the demographic factors of
greatest interest. USAID‘s consultants drew a stratified random sample which ensured the
inclusion of women‘s groups. mixed gender groups. and each of the PVOs' partner C05 in
sufficient numbers to do sub-group analysis. This was a decision that was taken after careful
reflection, and the consultants encountered numerous challenges in implementing it.

The most obvious pitfall with stratifying by PVO is that these partners vary widely
in the number of C03 with which they work. Totals of CO partners range from 37 for CARE
to well over 600 for Save the Children. Consequently, primafacie, any aggregate statistics
for a sample which was weighted equally by PVO would overstate the importance of CARE
and underestimate the other partners. When considering that the individual CO is the unit
of analysis in the USAID studies, it made sense that each international NGO’s partner COs
should have an importance in the data analysis equal to their proportion of the total number
of USAID partner COs. To draw a representative sample while ensuring proportionality
would have required a sample of well over 400 community organizations spread over almost
as many different villages in the four regions being studied, which cover an area half the size

of the state of Texas.‘2 USAID resources did not permit a study this large, not least because

of the fact that it was conducted under arduous rainy season conditions. Consequently, the

 

'2 As it was. enumerators visited well over 150 different villages distributed throughout the four regions. All
of this was accomplished during the worst weeks of the rainy season.

123
sample was drawn with a priority on ensuring minimum sample sizes of thirty in each sub-
group of interest.

Given the disproportionate nature of the sampling methodology, any analysis
attempting to present statistics aimed at drawing a portrait of the population of USAID-
partner C 05 would have to be weighted. as was done in the USAID reports (Davis 1997b and
Davis 1998). ‘3 However. the purpose of the present analysis is simply to use the USAID data
to examine trends in civic action among a variety of local organizations is four of Mali’s
eight regions. Because the analysis is not interested in the USAID-partner CO universe but
in the universe of C05 in general, weighting is unnecessary.M For this reason. I will describe
the sample in some detail before presenting unweighted results and analysis.

A comment on gender issues in sampling is called for here. Ensuring adequate sub-
group numbers of women’s organizations was important to USAID, as they target women’s
organizations and the indicators on which they are evaluated by Washington require separate
results for women’s groups and for C08 in general. An initial effort was made to make a
three-way distinction between women’s. men’s, and mixed COs, but it became clear that the
most clear and pertinent distinction was between women’s C05 and all others. Though the
difference between women's organizations and all others is clear, the distinction between

men's and mixed organizations is far less so. For example, associations Villageoises (AVs)

 

13 . . . . . .

In a perfectly representative probability sample. every subgroup appears in exact proportion to its
incidence in the population at large. Thus. every stratum is weighted equally in that its incidence in the sample equals
its incidence in the population. so it has a weight of one (say, .25 percent of population/.25 percent of sample = 1.0).
In a disproportionate sample. sub-groups must be weighted in order to correct for over- or under-sampling. For
example, if a sub-group constitutes twenty percent of the population but only ten percent of the sample. its weight
would be .20/. 10. that is. two. Half as many of the sub-group were sampled as should have been, so the weighting
system effectively doubles the stratum in order to correct for undersampling.

14 . . . . .- .
In any case. weighting is necessarily artttrcral and should only be conducted when absolutely necessary to
simulate a representative sample along stratifications of interest.

124

in the cotton zone were formed in most cases as all-village organizations but cotton is a
"man's crop," so in practice only men participate in most AVs. Similarly. as was described
in Chapter 2, young men dominate most youth groups and females often have only a
superficial subservient role. with men holding all leadership positions. It is therefore
plausible that some would call these men‘s organizations while others would call them
mixed. Consequently. there is little reliability in distinguishing between men's and mixed
organizations. because two different equally-informed CO leaders could often describe the
same organization as either. Therefore. the decision was made to include men's and mixed
groups as a single category.
The sample that was ultimately drawn was distributed as shown in Table 8.

Table 8 - The CO Sample, By Gender and PVO Partner

 

 

 

 

 

 

 

Partner Women’s Groups Mixed Groups Total
CARE - 35 35
CLUSA 14 22 36
Save-USA 9 24 33
World Education 10 23 33
None (Control) 9 27 36
Total 42 131 173

 

 

 

 

 

 

The over-riding priority in the sample design was to ensure that adequate sample
sizes existed for women’s groups. mixed groups, and for each of the four PVOs in order to
allow cautious analysis to be done for these sub-groups. There was also a concern that the
sample be geographically representative, but no stratification was done by region. In any

case, because the various PVOs have different geographic areas of focus, the consultants

125
were confident that any random sample drawn would yield a reasonable distribution by
region. The distribution of C05 by region and PVO is displayed in Table 9. With the
exception ofthe District of Bamako. the regional totals all exceed thirty. making cautious

statistical comparisons possible.

Table 9 - Distribution of Target CO Sample by Region and PVO

 

 

 

 

 

 

 

Region CARE CLUSA Save World Ed None Total
Bamako - - - 16 0 16
Koulikoro - 14 - 17 5 36
Mopti 26 - - - 7 33
Ségou 9 l 8 - - 9 36
Sikasso - 4 33 - 15 52
Total 35 36 33 33 36 173

 

 

 

 

 

 

 

 

 

The sub-group totals therefore permit a range of interesting comparisons. As
previously stated, the choice of sample size entailed reconciling statistical power with
resource constraints. Given the well-known trade-offs between sampling error and data
collection error (Henry 1990, pp. 33-34), the sampling methodology carefully balances these
important considerations. A few additional comments on sampling methodology are in
order: Efficient sample size estimates assume a simple random sample without multiple
stages or deliberate stratifications. Sampling variability increases when cluster sampling is
used and decreases when stratification is used (Henry 1990). Gains from stratification occur
when the variability within a stratum is small (homogeneity) and the differences between
strata are large. Preliminary analysis in the present study showed consistent differences by

PVC and by gender. The decisions to carefully stratify and to do a census of all CARE COs

126
contribute to a significant reduction in error.15 Nonetheless. the relatively small size of the
CO sample makes it more difficult to discover statistically significant differences between
groups, because fine-grained differences require larger samples to detect. Consequently. a
relatively small sample allows us to trace general contours in the data while being unable to

make definitive statements about finer nuances.

Non-response per so was not a problem in the study. However. given the fact that
the study was conducted during the rainy season, village inaccessibility did pose a problem.
Anticipating this. the Malian data collection firm was provided a back-up list of C03 to
contact in case of becoming “mired“ in such a difficulty. Of course, being forced to
substitute accessible COs for inaccessible ones could bias the results -- for reasons of low
access to information and a lack of infrastructure and economic vibrancy, among other
reasons, one would expect remote villages to differ from accessible villages. Ultimately. 18
substitutions were made. 13 of them in the cercle of Kolondiéba (Sikasso Region), because
of flooding which made certain areas inaccessible to motor vehicles. One cannot ignore the
failure to meet with ten percent of the organizations in the sample. but every effort was made
to replace organizations in unattainable villages with organizations resembling them in every
other characteristic possible. Intuitively, one would expect any resulting bias to be in an
upward direction, but I do not believe it to be large.

In order to permit USAID to begin to work toward meaningful impact assessment of
its DGSO program, a control group of non-USAID-partner COs was included in the study.

Nothing resembling a reliable sampling frame of such organizations exists. so a purposeful

 

'5 All CARE COs were included in the sample because they only totalled 37 and it made sense to interview a
couple of more organizations to eliminate sampling error for that segment of the sample. Unfortunately. two C05 in

the Mopti region were inaccessible. bringing the final CARE total to 35. All CARE COs are of mixed gender.

127

quota selection of non-target organizations was drawn with approximately the same
distribution of characteristics as in the target sample. The quota sampling method for

selecting a control group was conducted as follows. Eight arrondissements were chosen with
the intention of approximating the geographic distribution of USAID‘s PVO partners‘ zones
of operation, while avoiding arrondissements where there would be a high probability of
stumbling upon COS already collaborating with USAID Partners.”

In each arrondissement. data collection supervisors were instructed to go to the chef
d ’arrondissement and ask for five C Os meeting very specific criteria in terms of sector of
operation and distance from paved roads or major non-paved roads.‘7 These criteria were
developed in order to approximate the same distribution of sectors and remoteness in the
target sample, both in general and by region. For example, because among the DGSO
partner PVOs only CARE operates in the Mopti Region, we instructed data collectors to
identify and interview parent-teacher associations (APEs) and primary health associations
(ASACOs) in Mopti Region because those are the predominant types of CO CARE works
with.

To demonstrate the similarity of the control group to the target group sample, Table

10 displays their respective distributions by sector of operation.

 

'6 Slight geographical deviations exist between the target sample and the control group. as follows: Because
the Koulikoro Region is saturated with World Education and CLUSA partner COS. only five control group COs were
drawn in that region. in the central arrondissement of Kangaba. To avoid a high probability of inadvertently selecting
target groups, we chose five more control group C05 in the arrondissement of Keleya. in the Sikasso Region but
bordering Koulikoro. For this reason. the control group over samples Sikasso and under samples Koulikoro, but I do
not believe it to be a source of significant bias. The other notable difference is the fact that the control group includes
no organizations based in Bamako. Because Bamako is the smallest portion of the sample. and because the control
group over samples Sikasso, the wealthiest region of the country (and thus. though not comparable to Bamako. it is the
region closest in wealth to Bamako). we believe that serious threats to its usefulness are mitigated.

17 Again. this was done because of the belief that villages easily accessed by good infrastructure differ

systematically. ceteris paribus. from more remote villages. See Robert Chambers‘ comments on “spatial biases: urban.
tarmac. and roadside" (Chambers 1983. pp. 13-16).

128

Table 10 - Comparison of Target C08 and Control Group COs, by Sector

 

 

 

 

 

Sector Target COs Control Group COs Total
Health 28 6 34
20.44% 16.67% 19.65%
Education 39 10 49
28.47% 27.78% 28.32%
Economic 70 20 90
51.10% 55.55% 52.02%
Total 137 36 173
100% 100% 100%

 

 

 

 

 

 

All three of the major sectors in which DGSO partner COs work match up very
closely in terms of proportions of the target and non-target samples. All three sectors are
within 4.45 percent in variation between the two groups. Consequently, while USAID’s
impact assessment methodology is only beginning to be developed. the control group
provides a useful point of reference.

The predominant types of community organizations represented in the sample are as
follows. Health COs include both village-level health committees (CVSs) as well as
Community Health Associations (ASACOs). These ASACOs are charged with running a
Community Health Center (CSCOM), which normally serves the populations of ten to
twenty different nearby villages. Representatives of each of these villages make up the
general assembly of the ASACO.

A digression on units of analysis is warranted here. Ideally, the present study would
only include roughly comparable units. It was judged problematic to include CVSs and
ASACOs in the same study, as they are at different levels in a hierarchy of health

organizations ranging up through cercle-level and regional health association federations, as

129
well as a national federation.” Nonetheless. in the grand scheme of things. ASACOs and
C VSs are both local organizations. as well as important types of USAID-partner COS. so they
were included.

The education COs included both state-sponsored parent-teacher associations (APEs)
as well as management committees of community-initiated schools. C03 in the economic
sector were the most heterogeneous. including producer cooperatives. associations
Villageoises (AVs), women‘s credit unions, and the like. As the health and education COS
were all at least nominally mixed-gender organizations, all of the women’s groups in the
sample are in the economic sector. Twenty-six of the 42 women’s groups in the sample were
women’s credit institutions.

3.4 Results

The present section provides data on the determinants of civic action in southern
Mali. First, cross—tabulations by organizational gender type are presented on a number of
important indicators of civic action activity and inclusiveness. These cross-tabulations are
useful because they show the distributions of observations on a certain type of behavior more
explicitly than would regression results. Cross-tabulations on variables of interest are
followed by regression analysis of the determinants of civic action and civic action success.

Gender type is included because the gender distinction is arguably the most clear
deviation from homogeneity in rural associations. Moreover. it is a stratification along

which there are often significant differences, as suggested by Jonckers in the Mali context

 

'8 F elascom. Ferascom. and F enascom. respectively. Associated with both ASACOs and APEs. there is a

whole hierarchy of federations that originated as quasi-State creations. The fact that these organizations originated as

state creations is problematic but. as was seen in the case of AVs (Chapter 2), which were generally originally initiated
by the CMDT or the state. organizations evolve.

130
and Beck and Patterson in research on rural organizations in Senegal (Jonckers 1994. Beck
1996. Patterson 1997). This is particularly important because women tend to be relatively
powerless in village decision making. so it is essential to compare their organizational
behavior with that of male-dominated groups. Failure to do so would mask important
tendencies in the data.

A number ofbehaviors are of interest. First. are rank-and-file members consulted in
the internal decision making process? This is important in that it provides insight into the
level of inclusiveneness and representativeness of an organization. Second, does the
organization engage in civic action at the local level or beyond? Third, is this civic action
successful? That is. does it result in a changed decision or some other concrete result which
benefits the organization. its members, or the community at large? Related questions
concern whether or not these organizations have legal recognition, arguably the most
fundamental level of engagement with the state, and whether they speak of an intermediary
NGO or federation which represents their interests.

3.4.1 Description of Civic Action Behaviors and Their Prevalence”

Perhaps the most basic indicator of engagement with the state is whether an
organization is legally recognized by the state. With this in mind, CO leaders were asked
whether they were "officially recognized by the Malian Administration." While State
recognition is formally granted by the Ministry of Territorial Administration in Bamako, in

practice many organizations have de facto recognition by the state but only possess

 

19 A number of terms could have been used to describe the behaviors examined here. For example. we could
have used Esman and Uphoffs term “claim making,” or Bratton’s term “policy advocacy,” though Bratton‘s term
implies action more ambitious than most of the actions described here. Though it is quite general. I opted to keep the
term "civic action.” which USAID used in the study (Davis 1998).

131
documents originating at the cercle-level representatives of Malian government technical
services. For example. this is the case with ASACOs. Accordingly. enumerators were told
to first ask about official recognition and then to ask for a réce’pissé or another official
document. Enumerators then recorded whether the document demonstrated "local" (usually
cercle-level) or national-level recognition. The results of these queries are displayed in
Table 11.

Table 11 - COs Volunteering Evidence of Legal Recognition

 

 

 

 

 

 

 

 

 

 

 

Women's Groups Mixed Groups Total

National-level recognition 6 1 7

14.29% 00.76% 4.05%
Local-level recognition 5 27 32

11.90% 20.61% 18.50%
Recognition claimed, 22 78 100
no evidence shown 52.3 8% 59.54% 57.80%
No legal recognition 9 25 34
(or don’t know) 21.43% 19.08% 19.65%
Total 42 131 173

100.00% 100.00% 100.00%

 

The most immediately striking result here is that far more women’s groups have
national—level legal recognition. This can be explained by a particularin in the sample -- All
six of the women’s groups reporting national-level recognition were women’s credit union
committees, and there are no equivalent mixed associations in the sample. Once both forms
of legal recognition are combined, however, the results for women’s groups and mixed
groups are virtually identical. No chi-square statistic is reported here because the smallest

expected cell frequencies are too low. However. once both forms of recognition are

I32
combined into a single category. the resulting differences are far from being statistically
significant.

More generally. while fully 80.35 percent ofCO leaders interviewed claimed that
their organizations had legal recognition. only 22.55 percent were able to display a document
demonstrating proof of recognition. This suggests several things. First of all. some CO
leaders may not understand the legal implications of the phrase "official recognition."
Second, as previously mentioned. the fact that the survey was conducted during farmers’
busiest period of the year meant that many respondents did not have documents close at hand
when they were interviewed. Finally. in instances where this was not the case. one questions
how systematic record keeping is for many of these organizations. While donors may
consider formal organizations superior to informal organizations and take record keeping to
be a manifestation of a certain level of formalization, local organizations in Mali have
functioned for years on a basis of trust, memory, and oral contracts (Bingen 1998, p. 281).
Thus, where such documents exist they often have to be found and dusted off for the benefit
of nosy outsiders and otherwise may not be often referred to.20 In light of this. one would
prefer to rely on the respondents’ own report of being officially recognized as an indication
of their engagement with the state. Unfortunately, this is a case where there is a clear risk
of response distortion, as CO leaders could fear the consequences of admitting that they are

not legally recognized.2|

 

20 One sees this frequently in the case of AVs. APEs. and ASACOs. where external bodies have often
initiated their creation. Cookie—cutter by-Iaws are put in place and the necessary official documents are issued. and
then they are put away for safekeeping. 1 met with ASACO leaders in Mopti region who proudly showed me the by-
laws of their national federation (Fenascom) when asked for their associational by-Iaws.

2’ This would not be entirely without founding. An ongoing discussion in the donor community in Mali
concerns whether or not they should only target organizations with legal recognition.

133

Given the concern about (‘05 being potentially integrated into patron-client networks
and not representing the interests of all of their members. a series ofquestions was asked to
determine how systematically leaders responded to a problem. Specifically. was the problem
analyzed and was an action plan developed? Respondents were asked about their civic
action process only in the most general terms. and every detail of their responses was
recorded painstakingly. When C O leaders responded to a question concerning who in the
organization had taken part in these civic action activities. enumerators recorded whether or
not there was any mention of rank-and-file members. While this question is inadequate to
establish conclusively their level of inclusiveness. it nevertheless provides some idea
thereof.” These results are displayed in Table 12.

Table 12 - COs Showing Evidence of Rank-and-File Participation in Civic Action

 

 

 

 

 

Women's Groups Mixed Groups Total

Civic Action with 6 36 42
Rank-and-File Participation 14.29% 27.48% 24.28%
Civic Action without 8 45 53
Rank-and-File Participation 19.05% 34.45% 30.64%
No Civic Action 28 50 78

66.67% 38.17% 45.09%
Total 42 131 173

100.00% 100.00% 100.00%

 

 

 

 

 

 

Pearson chi2(2) = 10.4411 Pr = 0.005
In order to give a sense of the full range of behavior, those not engaging in civic
action at all are included in this cross-tabulation. Of the 95 organizations giving examples

of civic action behavior, slightly less than half (42/95 = 44.21 percent) gave any indication

 

7 . . . . .
2’ The reader is referred to the internal democracy drscussron in Chapter 2.

134

that rank-and-file members participated in planning their C 05’ civic action activities. The
strong significance of this result is interesting. and should be interpreted carefully. In fact,
the reason that there is such a strong difference between women’s and mixed groups is that
women’s groups are far less likely to engage in civic action. Fully two-thirds of women’s
C O leaders reported not having engaged in any civic action at all in the previous twelve
months. There is no significant difference in rank-and-file participation between mixed-
gender and women’s groups once those not engaging in civic action are excluded from the
analysis. Put another way. among C Os who actually engage in civic action activities, there
is no statistical difference between the rates of inclusiveness of the two groups.23

While apostles of pluralism would wish to see higher rates of inclusiveness. the fact
that nearly half of COS engaging in various civic action activities are judged to be at least
somewhat inclusive is an important finding. One could say that rank-and-file input in the
civic action planning process is a common occurrence but not one to be taken for granted.
This provides some information filling the void left by the lack of organizational attendance
records, as described in Chapter 2.

This established, the frequency with which civic action is initiated is now examined
in greater detail. The first level of civic action about which respondents were asked was that
pertaining to their most local official representatives of the national administration. For the
city of Bamako and regional capitals such as Koulikoro, Mopti, Ségou, and Sikasso,
communes have become operational. For all rural areas, municipal elections to choose

communal councils (and, from them, mayors) have yet to take place and the old system of

 

23 Of women’s groups engaging in civic action. 6 of 14 (42.86 percent) were judged inclusive. Among
mixed groups engaging in civic action. 36 of 81 (44.44 percent) met this standard.

135
arrondissements is still in place. Table 13 presents C O leaders’ responses to whether or not
they had initiated contact with local officials regarding a pressing deveIOpment problem.

broadly defined.”

Table 13 - Participation in Civic Action at the Arrondissement or Commune Level

 

 

 

 

Women's Groups Mixed Groups Total
One or More Occasions 5 59 64
1 1.90% 45.04% 36.99%
No Contact 37 72 109
88.10% 54.96% 63.01%
Total 42 I31 173
100.00% 100.00% 100.00%

 

 

 

 

 

 

Pearson chi2(1) = 14.9793 Pr = 0.000

When this study was first undertaken in 1997, there was some discussion of including
a variable operationalizing civic action undertaken at the village level. However. during the
pre-test of the CO questionnaire, it was discovered that all organizations engaged in such
activity, and with such frequency that they were unable to count it. Indeed, it was considered
a matter of protocol to keep the village chief and council informed of an organization’s
activities, and it was impossible to draw a line defining where protocol ended and civic
action began. Consequently, the decision was made to count civic action activities at the
arrondissement or commune level or above.

It is striking that even at this most local level of the national administration, women’ 5
groups almost never initiated contact with local officials “concerning a development

problem,” as the question was worded. In contrast. almost half of mixed groups initiated

 

24 During the pre-test of the CO questionnaire in 1997. responses to a broad question on civic action
activities revealed some confusion. Once clarifications were offered. virtually all examples of civic action given
concerned development-related problems, broadly defined. Consequently, civic action questions were worded to ask
precisely about action taken to resolve development problems.

136

such contact on one or more occasions in the preceding 12 months. Because municipal
elections have yet to occur to make decentralization a reality in 682 rural communes. the
soon-to-be-obsolete arrmrdr‘sscmcnt represents the frontline of the national administration.
Births and deaths are recorded there. new organizations are expected to register there. and
protocol dictates that all development-oriented missions visiting or intitating activity within
the circumscription pay a visit to the chef d'arrondissement. Arrondissement officials
complain that they no longer get the respect they feel they deserve. Conversations with these
local administrators often reveal a nostalgia for the halcyon days of the Traoré regime, when
they were treated with great deference by local people and visitors alike. Consequently, it
is pithy that only a bit over one-third of local organizations reported consulting these
officials. However, as we shall see. one should not read too much into this statistic, as COs
went to these local officials more frequently than any others for development problems.

Respondents were queried about several other types of civic action, including contact
with cercle-level officials. local development agents. development parastatals such as the
CMDT. as well as contact with one’s député, his or her representative in the national
assembly.25 The distribution of organizations contacting their national assembly

representative is displayed in Table 14, below.

25 When the DGSO baseline study was conducted in 1997. questions were also asked about regional and
national-level civic action. Because the results were extremely low for these. the same questions were not asked in the
1998 study. The reader will note that even in confining the analysis of civic action to relatively-local levels. relatively
few COs exhibit any civic action at all. so 1 do not feel that much information is lost with this simplification.

137

Table 14 - CO-Initiated Contact with Their National Assembly Representative

 

 

 

 

Women's Groups Mixed Groups Total
One or More Occasions 2 20 22
4.76% 15.27% 12.72%
No Contact 40 11 1 151
95.24% 84.73% 87.28%
Total 42 131 173
100.00% 100.00% 100.00%

 

 

 

 

 

 

Pearson chi2(1)= 3.1621 Pr = 0.075

These results are striking in how rare community organization leaders report
contacting their national assembly representatives. Given that. prior to the full
implementation of the new system of decentralization, one’s de'pute' is the elected official
most immediately accountable to him, the lack of contact is shocking. Again. there is a
nearly statistically-significant difference between women’s groups and mixed groups.26 In
any case, while mixed groups are slightly more likely to contact their representatives, the
results are unifome low. Results are particularly surprising considering that these results
represent responses by organizational leaders. This leads one to wonder how much actual
contact national official representatives actually have with their constituents. In the United
States, one would typically contact his or her mayor or city councilor or other state or local
officials to make demands concerning local problems, but one could also contact one’s
congressperson or senator to seek express burning local needs that he or she is capable of

resolving.27 Furthermore, the absence of anything resembling a state or provincial

26 At first glance at Table 14, one should be worried that there may be expected cell frequencies that are too
low to generate a valid chi-square statistic (Alreck and Settle 1995. pp. 289-90). In fact. the upper left cell has an
expected frequency of 5.34. just barely meeting the generally-accepted cut-off of five.

The sobriquet of New York 5 erstwhile Senator Pothole. Alphonse D Amato. IS revealing in this regard.

be

Cl‘

on
113‘.
for

3.111

Cit
Par
Poli

in}

138
legislature in Mali would lead one to suspect that a Malian might be. all else being equal.
more likely to contact his National Assembly representative than an American would his
congressperson.

The absence ofthis type of contact reveals. at the very least. an important difference
between the two political cultures. One wonders what underlying phenomenon the lack of
civic action activity with national assembly representatives reflects. Is it a lack of political
efficacy that impedes local organization leaders from initiating such contact. or does the
National Assembly lack legitimacy with the primarily-rural respondents to the CO survey?
In the RGAPA study cited above. Bratton and Bingen note that national level farmer
organizations in Africa rarely approach legislators. They attribute this to the fact that
national assemblies are correctly perceived by these organizations as rarely playing a
formative policy role. Consequently. they speculate that democratic consolidation in Mali
and elsewhere may lead farmers to consider legislative channels to be an effective means of
being represented (Bratton and Bingen 1998, p. 18).

It will be interesting to see what impact the new decentralized system will have on
civic action activity. Some believe that bringing politics closer to citizens may facilitate
participation and interest in public affairs. For example. Jonathan Fox holds that “pluralist
politics must be learned. and subnational governments make a good school” (Fox 1994, cited
in Hadenius and Uggla 1996. p. 1631). When evaluating the arrondissement-level results
as well as the national assembly representative results, above, it is clear that neither
proximity (the arrondissement) nor the status of being an elected representative of the peOple
(the députe') is enough to elicit the attention of community organization leaders. Of course,

the big change with the communal elections will be to combine closer structures (the

1 9

DJ

commune) with elected status (the mayor and communal councils). Time will tell whether
this is a winning combination.

Because the Malian State has ceded much of its rural development and extension
responsibilities to development parastatals such as the Office du Niger and the Compagnie
Malienne de De'veloppement des Textiles (CMDT). these entities have become the closest
representatives ofthe state for many rural people. In fact. agents of these enterprises engage
in a fair amount of dispensing of general developmental advice, and they are perceived as
having more developmental resources than the state itself.28 In this context, the following
table is revealing.

Table 15 - CO-Initiated Contact with a Development Parastatal

 

 

 

 

 

 

 

 

Women's Groups Mixed Groups Total
One or more occasions 9 31 40
21 .43% 23.66% 23.12%
No contact 33 100 133
78.57% 76.34% 76.88%
Total 42 131 173
100.00% 100.00% 100.00%

 

 

Pearson chi2(1) = 0.0894 Pr = 0.765

Most interesting about this result is that the percentages of women’s and mixed groups
contacting a development parastatal for a development problem are almost identical. While
the percentages engaging in this sort of claim-making are still low, they are considerably
higher than the percentage contacting their representative. Again, because the C03 in the

sample can all be loosely described as development-oriented organizations. it is not

 

28 In fact. as previously mentioned. CMDT officials often take themselves for a state in and of themselves.
When doing research in the cotton zone in 1995, l was told that my research permission granted by the Malian State
was inadequate and that I had to obtain special permission from the CMDT to talk to “their" villagers!

140
surprising that they would be more likely to contact development agencies than local
administrators or bureaucrats. Bratton and Bingen concur, noting that much policy advocacy
by African farmers is directed at line ministries of the executive branch or at parastatals
(Bratton and Bingen 1994. p. 18).

In fact, these results are low, if anything. for the following reason -- Only half of the
organizations in the sample fall into the broad “economic” sector. with the rest falling into
the education and health sectors. Those in the economic sector, primarily AVs and
cooperatives, would be more apt to contact a development parastatal than would those in the
other sectors. One wonders if health and education COs have initiated contact with entities
not evoked in the questionnaire. This is doubtful. however, as a question asked about civic
action at the cercle level, including with local representatives of State technical services (e.g.,
linked to the ministries of health or primary education). and this received very few
affirmative responses.

Above, the observation was made that the new rural communes may elicit greater
civic action in that they combine greater proximity with the status of being elected. We have
now seen that bodies with explicitly developmental responsibilities received more attention
from the C05 in the sample. The fact that developmental tasks and administration of the
Local and Regional Development Tax (TDRL) will be devolved to the communal councils
gives one more reason to be optimistic about the prospects for genuine participation in the
new rural communes.

The next table aggregates all CO-initiated contact with arrondissement, cercle, or

parastatal officials, as well as contact with one’s elected representative.

141

Table 16 - COs Reporting Civic Action Activity below the Regional Level (all kinds)

 

 

 

 

 

Women's Groups Mixed Groups Total

At least two forms of 3 38 41
civic action 7.14% 29.01% 23.70%
One form of civic action 1 1 43 54

26.1990 32.820/0 3l.210/o
No civic action activity 28 50 78
at all 66.67% 38.17% 45.09%
Total 42 131 173

100.00% 100.00% 100.00%

 

 

 

 

 

 

Pearson chi2(2) = 12.5928 Pr = 0.002

Here again. mixed groups are shown to be twice as likely as women’s groups to
engage in any civic action at all, and four times as likely to have engaged in two or more
different kinds of civic action. Nearly 62 percent of all mixed groups report at least one type
of civic action activity in the preceding twelve months, while only one-third of women’s
groups report the same level of engagement. This raises truly puzzling issues. The C05 in
this sample have had a range of developmental experiences, but the vast majority (137/ 1 73
= 79.19 percent) have had at least superficial contact with PVOs or national NGOs
collaborating with USAID. The fact that almost 40 percent of mixed groups and two-thirds
of women’s groups report no civic action at all of any kind in the previous twelve months
is startling. This begs a whole array of questions. Why is this? Have all of their
development needs been met? Are they content with the links they may have with the
USAID system? Do they lack initiative in seeking other sources of funding? Do they lack
efficacy in engaging the Malian State?

To this point, the discussion has addressed civic action. without considering the

result. Next we consider the reports made by the C03 in the sample themselves regarding

142
their effectiveness in the civic action activity they have engaged in. This survey question
was phrased as follows: "[in the cases of civic action of which you’ve just spoken] How
many times in the last twelve months has your action convinced authorities to change a
decision or otherwise resulted in something concrete?" This formulation represents a change
from the previous year. The USAID DGSO indicators require measurement of
"development decisions affected." In the 1997 baseline survey. CO leaders were only asked
whether they had been able to convince authorities to change decisions.29 Later. it was
decided that this formulation privileged COs having adversarial relationships with the
authorities, in that COs with positive but non-confrontational relationships with local
officials would be scored as having had no success. In an effort to remedy this. the
somewhat nebulous language "otherwise result in a concrete result" was added in order to
give credit where COs made claims and their action bore fruit. These results are displayed

in Table 17.

Table 17 - COs Reporting Civic Action Success

 

 

 

 

 

 

 

 

 

Women's Groups Mixed Groups Total

At least two civic action 2 17 19
successes 14.29% 20.99% 20.00%
One civic action success 4 35 39

28.57% 43.21% 41.05%
No civic action success 8 29 37

57.14% 35.80% 38.95%
Total 14 81 95

100.00% 100.00% 100.00%

 

 

29 . . . . . . . ,
Without knowrng the counterfactual. that is. what the oftrcrals would have done in the absence of the LO 5
contact, this is difficult to verify. Raising a similar measurement issue. Bratton and Bingen observe that "it is easier to

have apparent policy impact when the policy preferences of farmer organizations and governments are roughly

consonant" (Bratton and Bingen 1994. p. 20).

 

143

No chi-square statistic is reported here because the expected cell frequencies are too
low to for such a statistic to be valid. Even after combining the top two rows and re-running
the analysis. I did not find a statistically significant difference between the behavior of
women's groups and mixed groups. Nevertheless. this problem with statistical significance
can be seen as a consequence ofthe very small number of women’s groups having engaged
in civic action. Comparing numbers ofwomen’s groups and mixed groups reporting no civic
action successes. one sees a large divergence. Just over a third of mixed groups reported not
a single instance of civic action success. while nearly sixty percent of women’s groups
reported the same total lack of success. While the lack of statistical significance should not
be dismissed lightly, a clear pattern is evident here. Not only are organizations made up
exclusively of women far less likely to engage in civic action, but when they do. they are
also less likely to report success in those endeavors.

As a validity check. CO leaders were asked to clearly explain the circumstances
surrounding at least one case of civic action success. Enumerators were then instructed to
code the response as having met the standard of proof only if a significant amount of detail
was provided. The high percentage of respondents able to meet this standard of proof is
reassuring.30 Of the 58 CO leaders who claimed to have had an impact at least once. 51
(87.93 percent) were able to present a precise description of at least one case of successful

civic action. Methodologically, this is an encouraging result. Survey research is fraught

30 As a further validity check. respondents were asked to provide written documentation of their civic action
success. and only 14 of 58 (24.14 percent) were able to do so. This is consistent with the relative lack of formal
recordkeeping in many of these organizations and should not be automatically taken as evidence of response distortion.
Moreover. one should note that a serious problem encountered in conducting data collection during the rainy season
Was that rural CO leaders were often interviewed in their fields and had difficulty producing the numerous continuation
documents that the enumerators requested.

lllll

do '

desi

Res

flirt

plil’l

ilt’l'l

C00

C011

144

with threats to the validity of the data collected. Two of the most significant threats have to
do with ability to recall events occurring in the past and the incentive to give socially
desirable answers (Fowler 1995). Both of these are at play in this line of questioning.
Respondents have to first understand the questions. then identify cases of civic action in
which they have engaged. and finally pass judgment on the success of these activities.
Further. their interlocutor in the interview is reporting back to potential development
partners. and respondents may thus feel a need to impress the enumerator.

Another type of engagement with the state is collaboration in the provision of
development services. This is somewhat ambiguous in its potential effects, as it can court
cooptation or dependency. infringing upon organizational autonomy. Nevertheless, USAID
considers such collaboration to be primarily positive, as it embodies the spirit of cooperation
between civil society and local governments that they wish to cultivate in the emerging rural
communes. USAID consultants were wary of broaching the subject too directly, for fear of
some sort of response distortion, perhaps because of negative or positive perceptions of the
state or a desire to impress the enumerator in some way. With this in mind, enumerators
proceeded as follows. First. they introduced the t0pic by saying

Sometimes community organizations work in partnership with projects.

NGOs, the administration, parastatals. or other entities in order to provide

services to the community. For example, community organizations may

receive aid in the form of financial assistance, materials, or personnel which

permit them to assist their villages or communities. In return, these

organizations contribute physical labor, a portion of the financial cost, or
something else.

Once the enumerator provided this contextual information, he or she asked the following
question -- “In the last twelve months. have you collaborated with any sort of organizational

entity in this sort of service provision?” If the respondent answered affirmatively. he or she

145

was asked to list the activities they had undertaken which met these criteria. Enumerators
then coded the response in terms of whether or not they mentioned any cases of partnership
with state officials or technical services. Results for this line of questioning are presented
in Table 18.

Table 18 - COS Reporting Any Kind of Developmental Partnership

 

 

 

 

 

Women's Groups Mixed Groups Total
Partnership with outside 12 62 74
entities (self-reported) 28.57% 47.33% 42.77%
No partnership 30 69 99
71.43% 52.67% 57.23%
Total 42 131 173
100.00% 100.00% 100.00%

 

 

 

 

 

Pearson chi2(1) = 4.5711 Pr = 0.033

First, in terms of reporting any kind of partnership at all, mixed group leaders were
far more likely to report being engaged in such activity than women’s group leaders (47.33
percent, as compared with 28.57 percent), and the chi-square statistic for this result is
significant at the .05 level. Fully 71.53 percent of women’s groups and slightly over half of
the mixed groups (52.67 percent) report no partnership whatsoever. This is a somewhat
startling result, as eighty percent of the sample are indirect partners with USAID. The highly
negative response could be partially related to the very precise question wording. For
example, the USAID assistance may consist exclusively of training support, or it may be
more directive than collaborative, or it may be only in its very early stages. Nevertheless,
these results lead one to question the level of presence USAID’s partner PVOs and NGOs

have at the local level.

The next table breaks down the above numbers to show the frequency of partnership

with state entities.

Table 19 - COs Reporting Developmental Partnership with State Entities

146

 

 

 

 

 

 

Women's Groups Mixed Groups Total

Partnership with State 4 23 27
entities (self-reported) 9.52% 17.56% 15.61%
Partnership, 8 39 47
but not with State 19.05% 29.77% 27.17%
No partnership 30 69 99

71.43% 52.67% 57.23%
Total 42 131 173

100.00% 100.00% 100.00%

 

 

 

 

 

Pearson chi2(2) = 4.6165 Pr = 0.099

Mixed groups are about twice as likely as women’s groups to engage in partnership
with the state but, again. the most pertinent observation is that such partnership is uniformly
low -- engaged in by 17.56 percent of mixed groups and only 9.52 percent of women’s
groups. The chi-square statistic is nearly significant, but this is somewhat misleading --
once the 99 C05 reporting no partnership at all are excluded, the proportion of women’s
groups and mixed groups engaged in partnerships with the state as a pr0portion of all groups
engaging in partnership (in their respective categories) is almost exactly equal (women’s:
4/12 = 33.33 percent; mixed: 23/62 = 37.10 percent).

In order to examine vertical links in civil society between local organizations and
regional or national NGOs or federations, the following question was asked: “Sometimes
local organizations collaborate with intermediary organizations such as NGOs or federations.
Does at least one such intermediary organization represent your interests to the authorites?”

If they responded in the affirmative, they were asked to name the intermediary

147

organization(s). after which they were asked the open question “How would you evaluate
their representation of your interests?” Enumerators were trained to evaluate the answer for
evidence that the respondent felt that at least one partner represented his or her C O’s interests
effectively. and then to code accordingly. The results for this line of questioning are
displayed in Table 20.

Table 20 - COS Reporting Collaboration with Intermediary Organizations

 

 

 

 

 

 

 

 

 

 

Women's Groups Mixed Groups Total
Effective collaboration 10 32 42
25.00% 24.43% 24.56%
Collaboration, but not 5 9 14
effective (or “don’t know) 12.50% 6.87% 8.19%
No collaboration 25 90 115
62.50% 68.70% 67.25%
Total 40 131 171
100.00% 100.00% 100.00%

 

While exactly three-quarters of those reporting collaboration said that they were
effectively represented (42/56 = 75.00 percent). it is important to note that over two-thirds
of all C03 in the sample reported no such collaboration at all. This suggests that there is
relatively little vertical linkage between different types of organizations at various levels in
Malian civil society.

Horizontal links are as important as vertical links for the health of civil society.
While we do not have systematic evidence on these ties. we can make reasoned conjectures.
As mentioned in Chapter 2, AVs often collaborate together in groups of approximately ten

called A V/ZAERs. Further, APEs and ASACOs both participate in extensive networks of

148
federations. Other than these cases. which the reader should note are generally state- or
parastatal initiated. systematic horizontal collaboration is relatively rare.
3.4.2 Determinants of Civic Action Behavior

In the present section. we will be looking both at civic action behavior as well as
successfitl civic action behavior. Civic action behavior is an additive index ranging in value
from 0 to 4. As described above. respondents were queried about four levels of civic action.
including: 1. arrondissement officials, 2. cercle-level officials or local state development
agents, 3. development parastatals such as the CMDT, and 4. national assembly depute's.
Respondents received a score from 0 to 4 based on the number of levels at which they
engaged in civic action.

Successful civic action was a three point scale, scored as follows: Zero points for no
civic action, one point for civic action with no reported success, two points for at least one
self-reported success. and three points for two or more self-reported successes.

From the above discussions, it appears that a number of factors can potentially
influence civic action behavior:

The level of internal democracy in an organization is expected to be related to civic
action because an intemally democratic organization is more likely to make claims on behalf
of its members. Furthermore, the leaders of such an organization are more likely to be hald
accountable by members and therefore to try to produce tangible results for members. The
phenomenon of internal democracy is measured using the four-item index described in
Chapter 2. Reiterating. the components of the index (“demgov”) relate to membership,
leadership selection, formal rules, and rank-and-file participation. COs are given one point

each for voluntary membership, leadership alternation with fixed terms, possession of formal

I49

by-laws. and at least sixty percent participation at their most recent general assembly.
Consequently. the index may take on values ranging from a low of zero to a high of four
points.

Similarly. organizations which systematically consult their members are believed to
be more likely to engage in advocacy on their behalf. The variable “rankfile.” as measured
in Table 12. above. is included to operationalize this phenomenon.

Because of the belief that an organization must reach a certain level of formalization
and organizational capacity to engage in civic action (Bratton and Bingen 1994). the effects
of management systems are also examined. This is also measured on a four-point index
(“soundmgt”), with points awarded for the existence of formal financial systems, evidence
of strategic planning, at least fifty percent of a CO’s officers being literate in French or a
national language such as Bambara (bamanankan). and the systematic collection of dues.

It stands to reason that organizations which have formed a partnership with the
government or its development agencies in the execution of development activities would
be more likely to engage the state by making claims upon it. The dichotomous variable
“partnership with government (“partgov”), as measured in Table 19. above, operationalizes
this concept.

Institutionalized links with government are not the only form of networking which
can result in more engaged local organizations. Effective representation by NGOs or other
intermediary organizations such as federations can increase levels of civic action by
increasing feelings of efficacy (as opposed to isolation), as well as in setting an example and
in building local organization capacity. As displayed in Table 20, above. The variable

“ngorep” has three possible values: 0 if a CO reports that no intermediary organizations

150
represent their interests. 1 if there is representation but the CO is not satisfied with its level
of effectiveness. and 2 ifthe C 0 judges itselfeffectively represented by at least one NGO or
federation.

Two variables are included to measure diversity of financial resources and, indirectly.
the resulting autonomy. The variable “revenu” is dichotomous. measuring whether
organizations report revenue sources other than member dues or a USAID-financed NGO.
The variable “dues” is also dichotomous, operationalizing CO leaders’ answers to the
question “are your members required to pay regular dues?”3‘ “Revenu” and “dues” are
expected to be positively associated with civic action.

A number of demographic variables are included to investigate their effects on civic
action activity and effective civic action. These include the age of an organization (“age”),
the number of villages it represents (“size”), gender status (“gender”: all women or mixed).
the region in which it is located, and the primary sector of activities in which it engages.

“Age” is not expected to have a significant impact on civic action for the following
reason. Organizations founded long before the advent of the democratic transition in 1991
are expected to have been influenced by the neo-patrimonial context in which they evolved.
One can be a bit more sanguine about the prospects of younger organizations. However, one
cannot assume that the democratic ferment of the capital city and major regional capitals
quickly filtered down to the remote rural settings in which most of the C05 in the sample are
based. Moreover, even given such exposure, one would expect it to only have a measurable

effect after some passage of time.

 

3i . . . . . . . . .
This IS not precrsely the same question asked in the sound management index. which required evidence of

a system for recording the payment of dues. Nevertheless. this variable is expected to be strongly correlated with that
index.

151

For the same reason that one expects "scaling up" of C Os and NGOs to increase civic
action activity. one expects that as the number of villages represented by a CO increases the
civic action activity ofthe C 0 would also increase . cetcri's paribus. This expectation seems
logical in light of the fact that a larger constituency would result in increased demands for
representation and accountability.

As noted earlier. because of their subordinate position in society and the alternative
strategies that women employ in African politics (Hirschmann 1994, Beck 1996. Patterson
1997), women’s organizations are expected to engage less in civic action than mixed
organizations. We have even more reason to expect this given the many significant
differences between women‘s groups and mixed groups described above.

Region is operationalized as a series of dummy variables for Bamako (“dbamako”),
Koulikoro (“dkoulikor”), Ségou (“dsegou”), Sikasso (“dsikasso”). and Mopti (“dmopti”).
Normally. one would expect region to influence civic action activity in the following manner.
Organizations in the District of Bamako would. for reasons of proximity and capacity (e.g.,
higher literacy) engage in more civic action than, say, organizations in Mopti, which is
remote, economically poor, and where people are, on average. poorly educated. One would
expect Sikasso, as the wealthiest region of the country. to rank relatively high while Se'gou
and Koulikoro, which are quite heterogeneous in terms of wealth and remoteness, to be
somewhere in the middle. Unfortunately. in this sample. PVO partner is heavily correlated

with region. as PVOs tend to focus on particular regions, rather than being spread all over

152
the country (see Table 9. above).33 PVOs. in turn. tend to work predominantly in particular
sectors. resulting in region and sector being highly correlated.33

Primary sector ofintervention is not expected to have a bearing on civic action in and
of itself. but several intervening factors lead to the hypothesis that there will be some
relationships. For example. the 26 credit associations in the sample are exclusively made up
of women. For this reason. one expects the same negative relationship between credit groups
and civic action as one sees between women’s groups in general and such activity. Similarly.
because APEs are clientelistic relics of the Traoré era. one would expect these education
groups to be less engaged in civic action than other groups. Sector is operationalized through
a series of dummy variables for credit groups (“dcredit”). other economic groups such as
AVs, cooperatives and the like (“decon”). parent-teacher groups such as APEs (“dape”). and
ASACOS (“dasaco”).

The correlations between the variables described above are shown in Table 21,
below. In each cell in the correlation matrix, the first proportion provides the direction and
strength of relationship found in the USAID sample. The second proportion is the
probability of the observed relationship occurring randomly if no such correlation existed in
the larger population. In other words. the lower the probability, the more confident we are
that there is a systematic relationship between two variables. As per convention, we are most

interested in relationships where the probability is less than .05.

 

32 Save the Children-USA operates exclusively in two cercles of the Sikasso Region. CARE is only in Mopti
and Se’gou. World Education is only in Koulikoro and Bamako. and CLUSA primarily works in Koulikoro but is
initiating activities in the other three regions and Bamako.

33 CARE works primarily in the health and education sectors. CLUSA in the economic sector. and World

Education in the education sector. Save the Children-USA is the only PVO studied here that is fairly evenly active in
all sectors.

153

Table 21 - Correlation Matrix for USAID Democratic Governance Data Set

 

 

soundmgt

demgov

size

gender

revenu

dues

dcredit

decon

dape

dasaco

dbamako

dkoulikor

dsegou

dsikasso

dmopti

rankfile

- -- 2__. + _.

C) C)

CDC)

.2718
.0003

.0012
.9876

$11:

civic soundmgt

(j. (77

C)
._J
\O in

I‘x.‘ C)
C) KO

LPC‘H
0 J

l\) O
l\) to
J) P.)
43 O J) (I)

\J (.11

.4647

.0264
.7376

.2522
.0008

.0716
.3489

.1488
.0507

.1355
.0754

.1176
.1235

.0694
.3640

.0636
.4061

.0017
.9818

.0646
.3984

.1859
.0143

.1707
.0247

.0584
.4455

.2267
.0272

QC)

CC)

.0487

.5334

.1279
.1060

.1227
.1108

.0135
.8609

.2391
.0017

.0974
.2063

.0808
.2951

.2302
.0025

.1694
.0272

.2800
.0002

.1505
.0500

.1147
.1362

.0557
.4710

.1078
.1617

.0026
.9806

demgov

.1

partgov

._-—.———--————-—-.—-——————_-——.———-————-——.———_n—————.—--——--—-——-

.0000

L») C)
U: Ox
Ox

(D
t o l—J

.1320
.0871

.1374
.0795

.0949
.2142

.0205

.7891

.1080
.1574

.0917
.2301

.0656
.3912

.0229
.7650

.2648
.0004

.0823
.2817

.0243
.7513

.0150
.8450

.0307
.6882

.0344
.6528

.0742
.4749

F4

00

.0000

.0014

.9861

.0358
.6498

.0383
.6181

.1173
.1253

.0268
.7274

.1659
.0296

.1101
.1504

.0872
.2552

.1131
.1396

.2757
.0003

.0547
.4756

.2291
.0025

.0350
.6484

.0174
.8213

.0192
.8536

 

 

Table 21 (continued)

 

 

 

I size gender revenu dues dcredit decon dape
.......... +_____-..-______--__-____.___._-__.......__.__--_....____..___.....___.._._.____...__
size I 1.000:

I
gender . — 260* 1 '0!

0 000-
revenu I —;.124r 3.1533 1.0300

.l :1 4.1;“ 0 09514

I
3”@S I —0.u3:: e 3985 C 3268 l i 70

0 04“ C 0506 W 7264
dcredlt I -0.1699 0.7427 1029 0 1, - 1 0100
0.030" 0.0000 C .780 0 0097
decon I -0.0574 0.0129 0.2871 —0 0475 ~0.3223 1.0000

I 0.4655 0.8661 0.0001 0.5345 0.0000

I
dape I -0.1017 —0.3559 —0.2329 —0.0526 -0.2644 -0.4817 1.0000

I 0.1952 0.0000 0.0020 0.4917 0.0004 0.0000

I
dasaco I 0.3965 -0.2381 —0.1565 —0.0045 —0.1769 -0.3223 -0.2644

I 0.0000 0.0016 0.0398 0.9534 0.0199 0.0000 0.0004

I
dbamako I -0.1046 0.2846 0.0196 —0.0073 0.4241 -0.2446 0.0650

I 0.1827 0.0001 0.7975 0.9245 0.0000 0.0012 0.3954

I
dkoulikor I -0.2347 —0.1574 0.0815 0.0690 —0.1359 0.1086 0.2150

I 0.0025 0.0386 0.2865 0.3672 0.0746 0.1550 0.0045

I
dsegou I 0.0506 0.2079 —C.0612 0.2481 —0.0164 0.0791 -0.1642

1 0.5195 0.0061 0.4238 0.0010 0.8309 0.3009 0.0308

I
dsikasso I 0.0411 0.0110 0.3372 -0.1021 0.0418 0.1766 -0.3561

I 0.6017 0.8853 0.0000 0.1811 0.5850 0.0201 0.0000

I
dmopti I 0.2174 —0.2749 —0.4290 -0.2031 -0.2042 —0.2197 0.3152

| 0.0052 0.0003 0.0000 0.0074 0.0070 0.0037 0.0000

I
rankfile I ~0.0397 ~0.0113 0.2134 0.0485 —0.0291 0.2018 -0.2790

I 0.7074 0.9132 0.0379 0.6409 0.7797 0.0499 0.0062

1 dasaco dbamako dkoulikor dseqou dsikasso dmopti rankfile
__________ +_____.._______-_..________.__-_____--___--_-__..._______-_....___-____
dasaco l 1 0000

I
dbamako I -0.1343 1.0000

I 0 0782

I
dkoulikor I -0.2156 -0.1636 1.0000

I 0.0044 0.0315

I
dsegou I 0.1829 —0.1636 -0.2628 1.0000

I 0.0160 0.0315 0.0005

1
dsikasso I -0.0288 -0.2093 -0.3360 —0 3360 1.0000

I 0.7073 0.0057 0.0000 0.0000

1
dmopti I 0.1663 -0.1550 -0.2489 -0.2489 —0.3183 1.0000

I 0.0287 0.0417 0.0010 0.0010 0.0000
rankfile I -0.0285 -0.2363 0.0848 -0.0948 0.4654 -0.3557 1.0000

I 0.7838 0.0212 0.4140 0.3606 0.0000 0.0004

 

 

 

As predicted. there is a strong positive relationship between the sound management
index and civic action behavior (p = 0118).“ There is a similarly strong relationship
between the internal democracy index (“demgov”) and civic action (p = .0171).
Interestingly. this relationship holds for sound management with successful civic action
(“succivic”). but internal democracy and successful civic action are not significantly
correlated.

Partnership with govemment (“partgov”) is also significantly correlated with civic
action (p = .0130) and then has almost exactly the same non-significant relationship with
successful civic action. It is revealing to note that partnership with government is
significantly correlated with internal democratic governance, leading one to surmise that one
of these two variables will fall out of significance when running regressions which control
for the influence of multiple variables. The NGO representation variable (“ngorep”) is
modestly correlated with civic action but falls just short of significance (p = .1072).
Interestingly, ngorep and demgov are correlated (p = .0318) which begs the question of the
causal direction of this relationship. In other words, do NGOs, which the USAID study
found to be on average much more internally democratic than COs (Davis 1998), foster such
procedures in their local partners. or are more intemally democratic COs simply more likely
to “scale up” than their less democratic peers? Or is there another important intervening
variable? Also interesting. NGO representation explanatory variable is even closer to

significance in its relationship with successful civic action (p = .0774).

 

3" The reader will note that the actual correlation coefficients are almost uniformly low. rarely exceeding .30.
Reynolds points out that “[gJenerally speaking. the higher the variation. the larger a measure of association can be.
other things being equal” (Reynolds 1984, p. 64). Because most of the variables are ordinal or nominal in nature and
consequently exhibit less variation than interval-level data. the discussion of correlation concentrates primarily on

Significance level.

156

Consultation of rank and file members has absolutely no relationship with civic
action behavior but is strongly related to successful civic action. This is probably due to the
high number of C05 in the Sikasso region exhibiting rank and file participation. as fully 52
of the C Os studied were located in Sikasso. Both rankfile and dsikasso are strongly related
to succivic but not civic. It is appealing to believe that organizations in the wealthiest region
of the country are more inclusive. However. the lack of correlation between rankfile or
dsikasso and demgov leaves one puzzled as to the cause for the strong link between a single
region and organizational inclusiveness.

Organizations with alternative revenue sources had no particular relationship with
civic and succivic. but organizations charging regular dues have a borderline significant
relationship with both of them in a negative direction. However. this surprising relationship
is explained by the fact that the women’s organizations in the sample were significantly more
likely to charge dues than were their mixed counterparts.

As expected. women’s credit groups have a negative relationship with civic action
behavior. as well as with successful civic action (though the latter relationship falls just short
of significance -- p = .0754). None of the other sectoral dummy variables have significant
relationships with the two civic action variables.

Most of the regional dummies have no relationship with the civic action variables.
Two exceptions are worth noting. The dummy variable for Ségou Region is inversely related
to civic action behavior and just short of being significant (p = .0634). The Segou dummy
is also inversely related to civic action success. and this correlation is significant (p = .0143).
Also, as noted above, the Sikasso dummy is totally unrelated to civic action behavior in

general but is significantly correlated with successful civic action (p = .0247) in a positive

157
direction. In Mopti. ASACOs are significantly positively correlated with sound management
and internal democracy. which is strongly correlated with civic action. For this reason. while
its relative remotenesss and poverty would lead us to expect less civic action in Mopti. the
strong correlation between sector and region prevents us from examining this issue.

Having examined bivariate relationships. more complex multivariate relationships
can now be examined using multiple regression analysis. Both of the dependent variables
of interest, civic and succivic, are ordinal in nature. Ordered logistic regression (“ordered
logit”) is the appropriate technique for this circumstance (Kennedy 1992).

Table 22. below. displays the results of an ordered logit regression with simple civic
action behavior as the dependent variable and the sound management index, internal
democracy index, partnership with government. NGO representation. size. and gender as
explanatory variables.

Table 22 - Ordered Logit Estimates: Determinants of Civic Action

 

 

 

 

 

Iteration 0: Log Likelihood =-204.76007
Iteration 1: Log Likelihood =-189.76509
Iteration 2: Log Likelihood =—189.42745
Iteration 3: Log Likelihood =-189 42653
Number of obs = 161
chi2(6) = 30.67
Prob > chi2 = 0.0000
Log Likelihood = -189 42653 Pseudo R2 = 0.0749
civic I Coef Std. Err z P>Izl [95% Conf. Interval]
......... +-__-___-______-___-_________---___-_______-_____-____-__________-__
soundmgt I .36281 .1832975 1.979 0.048 .0035534 .7220665
demgov I .2920736 .1808662 1.615 0.106 -.O624176 .6465648
partgov I .7470157 .4120745 1.813 0.070 -.0606355 1.554667
ngorep I .3146024 .174199 1.806 0.071 -.0268215 .6560262
size I .0053802 .0153784 0.350 0.726 -.0247609 .0355213
gender I -1.512002 .4233198 3.572 0 000 -2.341694 —.6823106
_cut1 I —.8347267 .5721223 (Ancillary parameters)
_cut2 I .7890386 .578152
_cut3 I 2.022042 .6203572
cut4 I 3.799193 .814476

 

Ordered logit’s iterative estimation procedure aims to maximize the logarithm of the

likelihood function. The pseudo R2 statistic is not interpreted in the straightforward

 

158

explained-variance fashion that true R2 is in ordinary least squares (OLS) regression.
Consequently. the chi-square statistic calculated for the entire model provides a more useful.
if rough. evaluation of model specification. This chi-square statistic tests the null hypothesis
that all of the coefficients in the model equal zero. In this case. the null hypothesis is
rejected. as p = .0000.

In the regression results displayed in Table 22. only soundmgt (p = .048) and gender
(p = .000) are statistically significant at the .05 level, but partgov and ngorep are very close
to significance (p = .070 and .071. respectively). and demgov is not far behind (p = .106).
Because demgov is heavily correlated with soundmgt, partgov, ngorep, and size, the
regression should be re-run without demgov. The results of this modified specification are
shown in Table 23.

Table 23 - Ordered Logit Estimates: Determinants of Civic Action

 

 

 

 

 

Iteration 0: Log Likelihood =—204.76007
Iteration 1: Log Likelihood =-191.00505
Iteration 2: Log Likelihood =-190.73738
Iteration 3: Log Likelihood =-190.73683
Number of obs = 161
chi2(5) = 28.05
Prob > ch12 = 0.0000
Log Likelihood = -190 73683 Pseudo R2 = 0.0685
civic | Coef. Std. Err. z P>Izl [95% Conf. Interval]
......... +_--_____--____--_-_-___--__---_--__-____-_---_____-__-----_-__----_
soundmgt | .4012128 .1821949 2.202 0.028 .0441175 .7583082
partgov I .8416668 .4081867 2.062 0.039 .0416356 1.641698
ngorep I .3573199 .1721446 2.076 0.038 .0199227 .6947171
size I .012345 .0149011 0.828 0.407 -.0168606 .0415506
gender I -l.370412 .4097493 —3.345 0.001 -2.l73506 -.5673183
_cutl I -.798367 .5716003 (Ancillary parameters)
_cut2 I .8037447 .5767886
_cut3 I 2.013401 .6180856
cut4 | 3.77684 .8119788

 

 

In the new specification, all of the explanatory variables except size are shown to be
significant. In fact, size is heavily correlated with gender. Other specifications of the model

were tried in which gender was dropped, but size never became significant. One of the most

159
robust results is the negative correlation between women‘s groups and civic action activity.
This relationship held in all specifications ofthe model and across sectors and regions.
As was discovered in Table 21. the correlates of civic action and successful civic
action are not the same. The next table displays results when the same independent variables
are regressed with successful civic action as the dependent variable.

Table 24- Ordered Logit Estimates: Determinants of Successful Civic Action

 

 

 

 

 

 

Iteration 0: Log Likelihood =—206.01127
Iteration 1: Log Likelihood =-195.07396
Iteration 2: og LliéilhOQd =-194.96183
Iteration 3: Log LifieilhOOd = —l94.9617
Number of obs = 161
chiZIS) = 22.10
Prob > ch12 = 0.0005
Log Likelihood = -194.9617 Pseudo R2 = 0.0536
succivic I Coef. Std. Err. z P>lzl [95% Conf. Interval]
_________ +__-_-______-_----_______-_-_--_-__-________-_-_-___-_-_-__-____----
soundmgt I .4161205 .1782714 2 334 0.020 .0667149 .765526
partgov I .4 528:9 .3914033 1.087 0.277 —.3418505 1.192422
ngorep I .2890352 .1700645 1.700 0.089 -.0442851 .6223555
size I -.0118023 .0137401 -0 859 0.390 -.0387324 .0151278
gender I -1.413815 .4075869 —3.469 0.001 —2.21267 -.6l4959
_cutl I -1.066343 .5738904 (Anoillary parameters)
_cut2 I ~.0943715 .5680333
cut3 I 1.370612 .5904803

 

Among the determinants of successful civic action. government partnership, NGO
representation, and internal democracy no longer have the effect they were shown to have
on simple civic action behavior. though ngorep is close to significance. The fact that internal
democracy has a positive effect on actually engaging in civic action but is not significantly
related to the outcome of that action is an important finding. The inverse significant
relationship between gender and successful civic action is also worth noting -- not only are
women’s groups less likely to engage in civic action, but they are less likely to be successful

in the civic action that they undertake. These results are discussed in greater depth below.

160
3.5 Findings

The findings of the above analysis can be summarized as follows. First, sound
management practices such as systematic financial systems and dues collection, strategic
planning, and capacity in literacy and numeracy are strongly correlated with both the simple
act of engaging in civil action. as well as having success in such endeavors. This robust
result corresponds with previous findings on the importance of technical and managerial
competence in civic action (Bratton 1990, Bratton and Bingen 1994).

Second, internal democratic govemance is strongly correlated with the decision to
engage repeatedly in civic action but has no measurable relationship with civic action
success. This is an intuitively appealing result. It stands to reason that a democratic
organization would be more constrained by its members to engage in action aimed at
representing their interests. but it is not immediately evident that this would make them more
likely to succeed in such endeavors. Such success is more likely to follow from increased
capacity that is measured in the sound management variable.

Third, partnership with state entities in development activities or services is at least
somewhat related to civic action activity, but not civic action success. This is a less
important result, as it simply suggests that familiarity breeds continued contact but does not
ensure success.

Fourth, effective representation by an NGO or federation is significantly related to
civic action activity and nearly-significant in its relationship with civic action success. This
is an intuitively appealing result but is not particularly robust. The effects of this variable
are no doubt muted by the fact that there is some correlation between sound management and

NGO representation, though it is not quite statistically significant. Another measure of

161

scaling up, number of Villages represented. did not turn out to be significantly related to
either dependent variable.

Finally, perhaps the most robust result of all was that women’s groups are
significantly less likely than mixed—gender groups to engage in civic action. Moreover, even
when women’s groups do make the leap to engage in claim making, they are significantly
less likely than mixed groups to report success. This is consistent with previous observations
about women's behavior with respect to civic action. Under conventional approaches to
claim making and political participation, women’s lack of civic action is particularly
puzzling for the following reason -- due to the particular type of women‘s organizations in
the sample, women’s groups as a whole actually scored slightly higher than men’s groups on
the sound management index. Consequently, we can at least cautiously rule out lower
capacity as a reason for lower civic action. We are left wondering whether more deep-seated
factors such as feelings of low political efficacy are responsible for this result.

Given male domination of the public sphere, the low efficacy explanation is
compelling but incomplete. Moreover, the present data are informative as to what women
do not do but less informative about activity in which they do engage. For example, women
may use informal avenues to express their demands (Beck 1996, Patterson 1997), or they
may express their needs to village elders or male association leaders who in turn express
them to arrondissement or other state officials.

It is true that if we seek to evaluate women’s political participation and voice on the
same terms as that of men, the outlook is “somber in the extreme” (Hirschmann 1991, p.
1679). By the same token, some argue that elite and non—elite groups cannot be evaluated

on the same terms and that the failure to engage the state should not be seen as equivalent

162

to impotence — "There are many local women‘s groups that do not engage the state. yet are
vital to the welfare of their communities precisely because the state has so little impact”
(Tripp 1994a. p. 125). In providing health. education and other social services. women‘s
groups fill a void of state service provision and provide for their members and their
communities. According to this point of view, women‘s groups that do not engage the state
are not passive objects of patriarchy and their lack of formal influence but rather seek
avenues of participation where they can be more effective. Seen in this light, the women’s
groups in the present study can be recognized as providing essential services to their
members, such as ensuring that women and children are properly clothed or providing
routine and emergency loans to those in need.

The next chapter explores the effect of organizational participation on political

values.

Chapter 4

The Educational Function of Local Organizations

K0 to nyOgOn ta la. 0 de b8 kartabugu men si la.
[Bambara Proverb: "If we accept each other's beliefs. the straw house will endure
Ionger."]

4.1 Introduction

The relationship between culture and political and economic institutions is a
venerable topic. It is frequently alluded to, but less often addressed systematically, and even
more rarely measured. The present chapter takes up this subject in the interest of delving
deeply into how individuals are tied into a social and political system. In particular, I
evaluate the effect of organizational participation upon individual political attitudes. Some
have hypothesized that in countries with authoritarian histories, civil society can play a
socializing role which is played by electoral institutions in Western democracies.‘ As was
shown in Chapter 2, we have little reason to be sanguine about democratic effects of local
organizations in southern Mali. Here I use individual-level political attitudes data to evaluate
hypotheses along those lines.

The chapter is arranged as follows. First, I outline macro-level debates on the
relationship between culture and institutions. Second, the key issues in the political
socialization and political learning literature are summarized. Then, I examine the

relationship between organizational participation and political attitudes. From the literature

 

I See. for example, Muller. Seligson, and Turan 1987.

163

164

review coupled with the Chapter 2 findings on internal organizational characteristics.
hypotheses on these relationships are generated. Fifth. a methodology for sampling and
investigating these questions in the context of individual political attitudes is presented.
Next. I present an overview of the survey sample and data set. A description of the
demographics of the sample and a description of results on key political value indicators are
then provided. F inally. regression analysis of political behavior and attitudes is presented
and discussed.
4.2 Theoretical Issues
4.2.1 Culture and Institutions

Political theorists have long been intrigued by the relationship between culture and
political behavior and institutions. In spite of the fact that culture is a nebulous concept, the
intuitive appeal of the relationship has resulted in references to it throughout recorded
history.2 For example, Aristotle noted that ethical virtue was a product of habit and held that

our moral dispositions are formed as a result of the corresponding activities...

It is therefore not of small moment whether we are trained from childhood in

one set of habits or another; on the contrary it is of very great, or rather of

supreme importance (Aristotle, as quoted by Fukuyama 1995, p. 36).
John Stuart Mill placed similar importance on the importance of citizens’ values for the

survival of democracy. For Mill, democracy could not be preserved without the support of

a public that valued it intrinsically:

 

2 Culture is a nebulous term that has been defined in many different ways. Geertz cites 11 different definitions of
culture (Geertz 1973, pp. 4-5). Jamieson provides and overview of 160 definitions of culture used across the social
sciences (Jamieson 1980). F ukuyama defines it succinctly as “inherited ethical habit” that can encompass ideas and
values as well as actual social relationships (Fukuyama 1995. p. 34).

In political science. Almond and Verba implicitly define culture as “a set of orientations toward a special set
of social objects and processes." and they explicitly define political culture as the “specifically political orientations --
attitudes toward the political system and its various parts. and attitudes toward the role of the self in the system.“ while
recognizing that the boundary between political culture and other aspects of culture is not as clear as the name might
suggest (Almond and Verba 1965. p. 12). The idea of political culture aggregates the political psychology of
individuals into a distribution of orientations in an entire body politic.

165

Representative institutions necessarily depend for permanence upon the
readiness of the people to fight for them in case of their being endangered.
If too little valued for this. they seldom obtain a footing at all, and if they do.
are almost sure to be overthrown as soon as the head of the government, or
any party leader who can muster force for a coup de main, is willing to run
some small risk for absolute power (John Stuart Mill 1991, quoted by
Hadenius and Uggla 1996. p. 1636).

Speaking in less cataclysmic terms. Tocqueville spoke of the role of mores (mwurs) to
democratic stability. He said that “habits of the heart” as well as more conscious mores were
"one of the great general causes responsible for the maintenance of a democratic republic in
the United States” (Tocqueville 1969. p. 287).3 At the same time, Tocqueville recognized
that the causal direction between culture and institutions was not clear cut. Writing in the
18403, he noted that the east coast of the young United States had experienced democratic
government for longer than the western frontier had, permitting Easterners to have “formed
the habits and conceived the ideas most favorable to its maintenance. Their customs,
opinions, and forms of behavior have been gradually penetrated by democracy...”
(Tocqueville 1969, p. 308).

Beginning over a century after Tocqueville discussed these issues in Democracy in
America, much ink was again spilled over the causal direction of the relationship between
culture and political institutions. In their classic tome, The Civic C ulture, Almond and Verba
postulated that attitudes such as feelings of political efficacy, positive affect for the political
system, and trust for fellow citizens were extremely important to the establishment,

consolidation, and stability of a democratic form of government (Almond and Verba 1963).

 

3 Speaking recently and addressing a much-changed world. Hadenius and Uggla echoed Tocqueville:
[Dlemocracy’s prospects do not just depend on what rules of procedure apply or what
organizational channels stand open to the citizens. For it is the citizens themselves who furnish the
critical foundation Stone. The growth and preservation of democracy depend ultimately on the
support this form of government has in the hearts and minds of the people (Hadenius and Uggla
l996,p.1622)

166

They brought survey research methods to bear on these issues. using data from five nations
to argue that “unless the political culture is able to support a democratic system. the chances
for the success ofthat system are slim" (Almond and Verba 1965. p. 366).4

But their ideas did not go unchallenged. Others agreed with Tocqueville’s anecdotal
evidence that those with a longer and more profound experience with democracy and its
institutions were more likely to behave democratically. Almond and Verba’s premise that
civic culture was a cause of democracy was turned on its head by Brian Barry, who argued
that democracy usually takes hold for reasons much more complex than values, but that a
civic culture would likely emerge from prolonged experience with democracy and the
concomitant institutionalization of accountability and the reliance upon a competent,
informed citizenry (Barry 1978, pp. 50-52). More recently, Schmitter and Karl made a
similar argument (Schmitter and Karl 1991, pp. 82-83). Gabriel Almond responded to Barry
by arguing that the key hypothesis of The Civic Culture was more complex and entailed
reciprocal causation between structure and political culture (Almond and Verba 1980, p. 29;
cited in Muller and Seligson 1994, p. 650). Nevertheless. their original work focused much
more on one causal direction than the other.

Ronald Inglehart took up the “culture begets institutions” torch for Almond and
Verba with his sophisticated argument postulating the move toward “postmaterialism” with
generational replacement and a consequential increase in support for democracy in more

affluent societies (Inglehart 1977, 1990; Abramson and Inglehart 1995).5 The greater

 

4 In a flight of fancy and hubris, Almond and Verba speculated that were Tocqueville still alive. perhaps he would
have employed survey research methods similar to their own (Preface. Almond and Verba 1963).

5 “Materialist” concerns include economic and physical security. while “postmaterialist” values highlight freedom.
self-expression, and quality of life issues (Inglehart 1977).

I67

economic and physical security experienced by children in nations with steady economic
growth leads to the value change which makes mass publics more supportive of democratic
institutions.

While Inglehart and his intellectual comrade-in-arms, Paul R. Abramson, have
mustered considerable empirical support for their claims (Inglehart 1977. Inglehart 1990.
Abramson and Inglehart 1995), they have not gone unchallenged. Muller and Seligson use
data which overlap with Inglehart’s to test an elegant causal model examining relationships
between (primarily economic) structural properties of states, civic culture attitudes of the
general public, and change in levels of democracy (Muller and Seligson 1994). Challenging
the results supporting the effects of attitudes on years of continuous democracy (Inglehart
1990), they point out that. among other problems, the attitudinal data used by Inglehart came
from a period after the years of continuous effect. In other words, Inglehart’s purported
cause occurred after its effect (Muller and Seligson 1994, p. 636)! They also find
specification error in using the dubious variable “percent of labor force in service industries”
as one of only two macro societal determinants of democracy.6 Muller and Seligson examine
a broader range of macro societal variables, including income inequality (operationalized as
income share of the most affluent quintile), wealth, and subcultural pluralism. The
attitudinal variables they include are interpersonal trust and support for gradual reform.
While examining a range of alternative hypotheses, including Inglehart’s, they ultimately
show that the strongest single determinant of democracy was low income inequality and that

civic attitudes are not a significant factor in long-term democratic stability. Further. they

 

6 “Extremism in the pursuit of parsimony is no virtue. and moderation in the pursuit of parsimony is no vice“
would be an appropriate comment on this type of mis-specification (apologies to Barry Goldwater).

168

show that interpersonal trust is an mm. and not a cause. ofdemocratic stability (Muller and
Seligson 1994. pp. 646-647).7

Interest in the salubrious effects of a hospitable cultural environment has not been
confined to political scientists. Francis Fukuyama has recently produced an ambitious.
erudite work on the contributions of trust to economic life. Unfortunately, however. be
virtually punts the entire issue of causal direction. saying “...I will not make use of this
distinction between culture and social structure because it is often difficult to distinguish
between the two; values and ideas shape concrete social relationships, and vice versa”
(Fukuyama 1995, p. 34).8 While the reciprocal causation argument is intuitively appealing.
Fukuyama is not a very careful exponent, murky as his analysis is about precise causes and
effects. He does suggest that trust and its byproduct, “spontaneous sociability,” are crucial
to organizational innovation and, consequently, the creation of wealth (Fukuyama 1995, p.
37). He also suggests that both stable democracy and a successful market economy have the
“prior factor of social capital” as a determinant (Fukuyama 1995, p. 356). Thus, though his
analysis is oblivious to the political culture literature, his statements are most consistent with
Inglehart’s general position.

Commenting on political culture in the African context, Naomi Chazan suggests that
a positive experience with democracy prior to independence would have socialized certain

norms and improved prospects for success of post-independence democracies (Chazan 1994).

 

7 Inglehart and Abramson have responded with their own methodological critiques of Muller and Seligson
(Abramson and Inglehart 1995. pp. 3. I61). Principal among them is that Muller and Seligson do not properly
operationalize long-term stability of democracy, focusing as they do on changes from the 19705 to the 19805.

3 . . ,, . . . . ,. . .

Investigating the soc1al Virtues and the creation of prosperity, Fukuyama correctly pomts out the neglect of
economists to consider the importance of culture to economic behavior. However, he overlooks most of the broad
political science literature on political culture, including Inglehart‘s work on post-materialism.

169
Her point is consistent with Barry and the more moderate findings of Muller and Seligson.
An earlier paper by Muller and Seligson and Ilter Turan postulated that, in countries with an
authoritarian past. voluntary associations may serve as a more effective agent of socialization
of democratic values than electoral institutions (Muller. Seligson. and Turan 1987).

As the above discussion demonstrates. the debates on culture and institutions are an
important link between the micro level of individual attitudes and behavior and the macro
level of structure and process (Rokkan and Campbell 1960; Almond and Verba 1963).
Culture is the aggregation of individual orientations toward a system. As such, it constitutes
a vertical scale link between the individual and the system and the local and the national. In
many African countries in which there is one dominant metropolis, culture also constitutes
an important link between the urban center and the rural periphery. Where state institutions
are weak, one must consider the interaction between institutions and the broader political
culture. In such cases, it is also important to examine social institutions extending beyond
areas of direct state control.

Cultural attributes and individual attitudes have been linked to how human beings
behave in collective settings. The ability to associate in relative harmony and with some
measure of effectiveness has been held to depend on a certain common heritage of shared
values. Whether called “social capital” (Coleman 1990, Putnam 1993), “social energy”
(Hirschman 1984), or “Spontaneous sociability” (Fukuyama 1995), there is a groundswell
of interest in the cultural prerequisites for effective collective action. Hadenius and Uggla
define social capital succinctly as “the proclivity to join in common efforts” (Hadenius and
Uggla 1996, p. 1624). Fukuyama defines it as “a capability that arises from the prevalence

of trust in a society... [which] is usually created and transmitted through cultural mechanisms

170

like religion. tradition. or historical habit” (Fukuyama 1995. p. 26). He coins the expression
“spontaneous sociability." to describe a subset of social capital concerning the specific
"capacity to form new associations and to cooperate within the terms of reference they
establish” (Fukuyama 1995. p. 27).9 Putnam defines social capital as “features of social
organization, such as trust. norms. and networks, that can improve the efficiency of society
by facilitating coordinated actions” (Putnam 1993, p. 167).

The literature on social capital points to the complexity and recursive nature of the
links between culture and institutions. Cultural norms generate social capital which fosters
effective collective action. At the same time, there is much discussion of the importance of
institutions in the socialization of individual political values. So, values are important to the
initiation of collective action and its institutionalization as civil society but prolonged
experience with civil society can also cultivate values. As discussed in Chapter 2, the nature
of the associations in civil society has much to do with the types of values fostered by
organizational participation.

4.2.2 Political Socialization and Political Learning

The literature on political socialization provides a more specialized debate on the
actual processes by which democratic or non-democratic citizens are created. Early efforts
at research into political socialization marked an important bridging of work in educational

psychology and political behavior (Hyman 1959).'0 Early work in political socialization

 

9 Although one fails to see what Fukuyama adds to Putnam’s observation that "spontaneous cooperation is
facilitated by social capital” (Putnam 1993. p. 167).

'0 Political socialization studies have been fraught with a lack of conceptual clarity (Greenstein 1970). Sigel
points out four frequently encountered definitions: 1. “[L]earning pure and simple... the state of a person’s political
knowledge and comprehension;" 2. The developmental sequence by which political knowledge and comprehension are
acquired and the continuity over time of such knowledge and attitudes; 3. The process by which a given society’s
norms and behavior are acquired; 4. Formal civic education (Sigel 1995. pp. 17-18).

I71

focused primarily on childhood socialization (Greenstein I965. Hess and Tomey I967.
Easton and Dennis 1969). These studies were later largely discredited for falling victim to
two questionable assumptions. First. researchers in political socialization appeared to
assume that what was learned prior to adulthood remained virtually unchanged throughout
one’s life. Second. these early learning experiences were assumed to have a significant effect
later on. And what was learned earliest was assumed to be most important (N iemi and
Hepburn 1995).

While culture certainly changes very slowly, it does not automatically follow that
individual values are established immutably in childhood. This nuanced view began to gain
currency as a result of the rapid change that occurred in Western societies during the 19605.
For example, levels of political trust declined far too rapidly to be explained away by
generational replacement (Miller 1974). Such anomalies, combined with the difficulty of
assessing early learning and its significance, have led some researchers to advocate a
redirection to studying political socialization later in life. For example, Niemi and Hepburn
have recently argued for political scientists to redirect their energies toward studying “the
period of maximum change” in psychological and social orientations, which they place
approximately in the range from age fourteen to twenty-five (Niemi and Hepburn 1995).
Morduchowicz and her colleagues have argued that in countries with lower proportions of
educated citizens than the United States it may still be worthwhile to study slightly younger
children, as they begin to have adult responsibilities at a younger age (Morduchowicz et a1.
1994). The key concern in such assertions seems to be that socialization depends on social

factors outside of just the home and school. At the same time, most political socialization

172

advocates do not seem to be advocating a study of lifelong learning (Niemi and Hepburn
1995.p.9i

The focus on political socialization among just children and young adults has not
gone uncontested. however. For example. in a study of West Germany. the United States.
and the Netherlands. Kent Jennings found that young peoples’ orientations did not change
more than that of cohorts aged thirty or more. Indeed, he found that “many orientations
never do harden” (Jennings 1990, p. 346).H Accordingly. research has been done on adult
political socialization. but such studies have been few and far between (Sigel and Hoskin
1977; Si gel 1989, 1995,). Such work advocates a "life span perspective” of macro and micro-
level influences on individuals. Furthermore, they counsel more attention to indirect
socialization agents such as participation in social movements (Sigel 1995. pp. 19. 21).
J ustifying the fact that his edited volume on political culture in developing countries ignored
childhood socialization altogether, Larry Diamond argues that new values and beliefs can be
learned well into adulthood and adds that the seeds of cultural change are not to be found in
early socialization but rather in “distinctive carriers of new political ideas and norms,”
including autonomous groups in civil society (Diamond 1994, p. 234).

Addressing the case of Eastern Europe, Sadowski appears to agree, arguing that
members of autonomous groups that she studied “got practice in the art of self-organization
and in the strategy of confronting political elites” (Sadowski 1994, p. 166). She notes

citizens’ lack of “democratic skills and traits,” including tolerance, a willingness to

 

H This sentiment is echoed by Sidney Verba in a re-evaluation of his earlier work. where he notes that past studies
had not taken into account “the degree to which political attitudes are labile and could be affected by political events”
(Verba. in the conclusion to Almond and Verba I980. pp. 399-400).

I73

compromise and trust in public officials and implies that these skills and traits can be learned
through organizational participation (Sadowski 1994. p. 183).

With the recent increase in American proselytizing about the virtues of democracy.
studies have been commissioned to discover how best to foster democratic stability through
socialization. For example. recent studies commissioned by USAID sought to assess the
impact of the large sums ofmoney they spend annually on civic education programs (Bowser
et al. 1997, Sabatini et al. 1998).12 Examining a variety of adult, informal programs in
Poland and the Dominican Republic, Sabatini and his colleagues found that “the immediate
impact of civic education on democratic values was inconsistent and generally small in
magnitude” (Sabatini er al. 1998, p. 28). On the other hand, they found that such programs
had the greatest effect on levels of political participation but did not find participation to be
directly associated with democratic values or even civic competence (Sabatini er al. 1998,
p. 28).

Concerns with adult political socialization and civic education effects dovetail nicely
with Nancy Bermeo’s concept of “political learning” (Bermeo 1992).” Building on Robert
Axelrod’s “evolutionary approach to norms,” Bermeo defines political Ieaming as “the
process through which people modify their political beliefs and tactics as a result of severe
crises, frustrations, and dramatic changes in environment” (Bermeo 1992, p. 274; Axelrod

1986). Her key contribution is to distinguish between tactics and values, saying that “even

 

‘2 In 1994, USAID spent over $23 million on civic education programs. This wasjust one component of their
global push to support democracy (Sabatini er al. 1998. p. 6).

13 Useful though Bermeo’s approach is. her choice of terminology is regrettable, as devotees of political
socialization have long used “political learning” as a synonym for their enterprise. In fact, Berrneo’s approach differs

considerably from other research on socialization. Indeed. she only makes one passing reference to previous work in
that field (Bermeo 1992, p. 274).

174

when political learning takes a prodemocratic form. it does not necessarily involve the
emergence of a deep normative commitment to democracy per se” (Bermeo 1992. p. 275).”
In this. she echoes Dankwart Rustow. who made the point that “democracies can be built by
people who are not truly democrats” (Bermeo 1992. p. 275. citing Rustow 1970). Larry
Diamond makes a similar point with respect to Africa. suggesting that a culture of resistance
to authoritarian abuses does not automatically mean that people have adopted the positive
attributes of democratic citizenship. such as allegiance to state authority and a willingness
to compromise. Moreover. political beliefs can often be motivated by instrumental factors
(Diamond 1994c. pp. 232-233, 235). Sadowski implies support of this point of view in her
suggestion that citizens can learn political strategies through participation in civil society
organizations (Sadowski 1994. p. 166).

Another important point made by Bermeo is that a critical mass of learners, who are
for whatever reason amenable to democracy. must emerge before democracy can begin to
take hold. She feels that the Ieaming experience of elites is particularly important in this
regard (Bermeo 1992, pp. 275-276). Diamond agrees with this focus on elites. While
political culture is not a precondition for the initiation of democracy, he says that this process
is facilitated by attitude shifts among key elites, who assist in the dissemination of these
values to a wider circle of elites and eventually the citizenry at large (Diamond 1994c. p.
239). While lacking data to test it, Muller and Seligson offer the hypothesis that mass
political attitudes are caused by elite attitudes (Muller and Seligson 1994, p. 648). Political

scientists have long thought that elite attitudes may be more important than mass attitudes.

 

i4 , . . . . . . . . . . , . . .

Bermeo s pornt that motivations are not ObVIOUS from behavror 15 Similar to Amartya Sen 5 critique of the notion
of “revealed preference” in economics (Sen 1973). The common thread is that one cannot assume that a given
behavior reveals a particular preference or value.

175

For example. over four decades ago. Stouffer found that community leaders were more
tolerant and had more democratic values than non-leaders ( Stouffer 1955). This is consistent
with the above finding that in transitional settings. elite attitudes may evolve before mass
attitudes. Almond and Verba. on the other hand. saw no reason to believe that elite attitudes
would differ significantly from non-elites (p. 352).

4.2.3 Organizational Participation and Political Attitudes

African states have differed from the European and South American cases treated by
Sadowski and Bermeo because of the weakness of the typical African state -- As much as
African authoritarians may have liked to have total control, they could not eliminate
independent space for civil society because they lacked the wherewithal and capacity to
penetrate deeply into society. Nonetheless, as the recent organizational explosion occurring
in nascent democracies such as Mali suggests, there were far fewer venues for citizen
participation prior to the openings created by democratization. Moreover. the opportunities
that did exist were often created or closely supervised by the state or its agents. Just as those
in Eastern Europe had to learn self-initiated organizational behavior after generations of
communist rule, Africans have had to wrestle with the legacies of authoritarian pre-colonial
and post-colonial rule and neo-patrimonialism, as well as the very different organizational
heritage of ascriptive kinship and lineage based groups.

An underlying theme in the above discussion is found in the frequent references to
the relationship between autonomous groups in civil society and the evolution of political
values. As Fukuyama says, “a thriving civil society depends on a people’s habits. customs.
and ethics” (Fukuyama 1995, p. 5). It is intuitively appealing that values such as trust and

spontaneous sociability increase the likelihood of participation in civil society associations.

176

Indeed. the civic education programs promoted and evaluated by USAID "promote the
knowledge. skills. and values necessary for democratic participation” (Sabatini et al. 1998.
p. 10. emphasis added). At the same time. though civic education programs cannot
realistically be expected to impart democratic values in the short run. one can expect that
practical civic education can encourage participation for instrumental reasons, and this
participation can contribute to the institutionalization of democratic norms.

Verba. Nie. and Kim define political participation as “acts by private citizens that are
more or less directly aimed at influencing the selection of governmental personnel and/or the
actions they take” (Verba et a1. 1978, p. 1). Much of the attention devoted to political
participation has been devoted to electoral participation. However, under liberal regimes
much quasi-political participation occurs in other venues. Weak state penetration of society
or the existence of a passively tolerant state or one that actively promotes autonomous
organization leads to alternative venues for such participation. Citizens can organize and
communicate with each other or can attempt to influence the policy process by
communicating directly with officials in the sort of claim making or policy voice described
in Chapter 3.

Group activity has long been posited to be a major determinant of political
participation (Verba,Nie and Kim 1978.Putnam 1993, Sabatinietal. 1998).l5 Organizations
can serve to increase feelings of political efficacy and consequently have an impact upon
electoral participation. Apart from that, however, organizational participation can be

important in its own right. For example, as mentioned above, organizational participation

 

IS - . . . . . . . . .
For a careful study of the correlates of political partrcrpation in the African context. see Simon 1997.

177

has been hypothesized to play a role in socializing attitudes which are favorable to the
support of democracy. thereby creating good democratic citizens (Muller et al. 1987).”
Gibson and his colleagues describe a democratic citizen as a person who
believes in individual liberty and who is politically tolerant. who holds a
certain amount ofdistrust of political authority but at the same time is trustful
of his fellow citizens, who is obedient but nonetheless willing to assert rights
against the state. who views the state as constrained by legality and who

supports basic democratic institutions and processes (Gibson et al. 1992, p.
332).

The qualities described by Gibson et al. are each the subject of a vast and rich literature in
political science. Below. I briefly discuss the individual qualities that, taken in the aggregate,
are believed to be most essential to the survival of democracy.

The importance of trust to collective action and thus to political and economic life
has recently been highlighted in a number of studies (e.g., Putnam 1993, Fukuyama 1995).
It has long held to be the basis of social bonds extending beyond the family and is thus
implicated in studies of politics. sociology, and economics. According to Hadenius and
Uggla, “Trust is the foundation of every form of cooperation, and thus the basis for viable
institutions as well” (Hadenius and Uggla, p. 1626). It is difficult to separate the components
of trust from its manifestations and effects. Among these are a willingness to cooperate, to
tolerate opposition, and to live with the uncertainty inherent in democracy.

Closely linked to trust is the concept of tolerance. Tolerance involves the degree to
which citizens are willing to tolerate viewpoints with which they disagree. This

disagreement is absolutely fundamental to a definition of tolerance. American public debates

 

‘6 Of course, supposed socializing effects are only likely to occur under certain conditions. as discussed in Chapter
2. For example, Hadenius and U ggla stress the importance of “longstanding participation in democratic structures”
(Hadenius and Uggla 1996, p. 1622. emphasis added). Vertically-structured. clientelistic organizations cannot be
expected to have the same effect.

178

over the past decade over "political correctness” and "multiculturalism” often suggest that
our society needs more "tolerance” ofthose who are different. but exponents of this point of
view seem to be asking for something more. They seem to be expecting “acceptance” of
people of different lifestyles and backgrounds. The democratic standard of tolerance is less
exigent. Democratic citizens do not have to accept everything. just tolerate. Tolerating the
expression of unpopular viewpoints is fundamental to the definition of liberal democracy.

At the same time. there are boundaries beyond which one could argue that democratic
citizens need not tolerate. One of the most thorny issues in a democracy is whether
individuals and society as a whole should tolerate those who advocate the oppression or
extermination of others, or those who advocate the establishment of authoritarian secular or
religious rule (Stouffer 1955, Sullivan, Piereson, and Marcus 1982).'7 Because there are
limits to tolerance, even among those who would normally be considered to be of a
democratic spirit, recent studies question whether intolerance should be judged in terms of
the most extreme groups (e. g., Nazis, Stalinists, skinheads) or by using less incendiary but
still controversial cases (e.g., draft resisters. socialists, conservative Christians)(Gibson
1993)

Support for political rights is a more specific manifestation of tolerance. Should any
or all groups in society be accorded the right to criticize the government, the right to
organize, and the right to petition? To what extent do citizens endorse the extension of such

democratic liberties unconditionally? Or do citizens believe that rights should only apply in

 

l7 . . . . . . . . ,,
A case in pomt is the current controversy in the United States over whether or not extremist “right to life
groups are responsible for individuals who reSpond to their incendiary rhetoric with bombings or assassinations.

I79

certain circumstances or only be extended to certain groups? There is a fundamental tension
between support for democracy versus social order (Gibson. Duch and Tedin 1992).

Another important characteristic of a democratic citizen is that she feel capable of
influencing the system and not just be a passive subject. This is often described as entailing
a sense of political eflicaev. that is. a feeling of empowerment and confidence that one’s own
opinions are important and that one can personally have an impact on the system. According
to Campbell, Gurin, and Miller, political efficacy is “the feeling that individual action does
have. or can have. an impact on the political process. i.e., that it is worthwhile to perform
one’s civic duties” (Campbell, Gurin. and Miller 1954. p. 190).

Sabatini and his colleagues assert that efficacy does not exist in a vacuum and posit
the importance of a broader feeling of civic competence. Civic competence combines the
more abstract feeling of efficacy with more concrete political knowledge and skills which
facilitate individual participation. Political knowledge consists of a reservoir of information
on basic rights, how democracy works, the structure of the system, and role of the individual.
as well as practical knowledge about parties. candidates, and public officials (Sabatini et al.
1998). Important skills include critical thinking, planning, resource generation, problem
solving, and civic action techniques. Efficacy is the link between knowledge and skills and
effective activism. While knowledge and skills can be taught directly, cultivating efficacy
is a more nebulous, elusive, and long-term proposition. F inkel has done extensive research
on the reciprocal effects between participation and political efficacy (F inkel 1985, 1987).

Nevertheless, participation can be key to building the political skills and cultivating
the efficacy that are key to civic competence. It fosters the transition from passive subject

to active citizen (Milbrath and Goel 1977, Goel 1980). Since before Lipset, social scientists

180

have spoken of the contribution oforganizational participation in this regard. Organizations
have socialization effects because individuals voluntarily adhere to a range of organizations
and interact with others of varied backgrounds in terms of reli gion. ethnicity, socio-economic
status. ideological persuasion. and other characteristics (Lipset 1960). Because these
socialization effects involve general values such as openness. trust and tolerance. the positive
effects can result from participation in a broad range oforganizations and not just those that
are explicitly political.

System support is also considered important because it is a direct manifestation of the
how ingrained support for democracy has become in the hearts and minds of citizens. More
specifically. do citizens support all democratic institutions, including the judiciary and the
legislature? Moreover, do they support the extension of broad rights to all other citizens
(Muller 1979, Gibson et al. 1992)?

4.3 Hypotheses

A commonly held but rarely tested assumption is that civil society associations foster
the socialization of democratic values. In Chapter 2, I made the argument that this premise
could only be plausible if the associations under consideration were democratic in their
internal procedures. In such a case, it would be logical to argue that prolonged participation
and, hence, exposure to democratic values would contribute to the adult socialization of such
values. Examining the internal workings of four different types of village-level
organizations found in Southern Mali, I found that these organizations were all weakly-
inclusive in leadership selection and most were also weakly-inclusive in their decision
making procedures. Consequently, most village-level groupings seemed to have little

potential to contribute to the socialization of democratic values among their members and

181

leaders. Local-level cotton producers' associations (associations villageoises. or AVs) were
found to have more potential than women's. youth. and hunters' associations. but only modest
optimism was expressed in this regard.

In more internally democratic organizations. more profound levels of participation
would be thought to increase the likelihood that participants would be inculcated with civic
values. For example, organizational officers would be more likely to be socialized to hold
democratic beliefs than more passive rank-and-file members. In the present case. however,
the non-inclusive procedures cause us to be less than sanguine in this regard.

In the event that an organization was deemed to have procedures which nurtured
democratic values, one would expect more profound levels of participation, such as
leadership, to lead to even more favorable socialization effects. Given the weakly-inclusive
nature of procedures of the organizations under study here, however, one should not expect
any significant difference in attitudes between the general membership and the leadership.
Bearing the Chapter 2 results and the above discussion in mind, the following hypotheses
are offered.

Hypothesis 1 .' Simple members of local-level rural associations in Mali (other than AVs)
will not hold more democratic attitudes than non-members.

Hypothesis 2: Simple members of cotton producer associations (AVs) will hold slightly
more democratic values than non-members.

Hypothesis 3: Officers of local-level rural associations in Mali will not hold significantly
different attitudes than simple members.

The next section describes the sampling and survey methodologies employed.

Special attention is paid to how the hypotheses were operationalized.

4.4 Methodology
4.4.1 Methodology Selection

In order to sort out broad trends in Malians' political attitudes toward their political
system and in light of the dramatic changes taking place. the decision was made to design.
implement and analyze a survey of political attitudes and orientations in rural Mali. Survey
research methods have a long and rich history in the treatment of cultural and attitudinal

variables.

However. the drawbacks of survey research, especially in non-Westem settings, are
well-known (see, e.g., Chambers 1983, pp. 51-59). They include issues of superficiality,
simplistic quantification of complex phenomena, problems with translation and other
stimulus equivalence issues, among others. Attitudinal surveys raise even more thorny
issues. Such surveys ask questions aimed at eliciting underlying values. The problem with
this approach is that it “confuses opinions with habits” (F ukuyama 1995, p. 38). Important
challenges can be raised concerning the malleability or reliability of attitudes over time and
what they say about underlying values.

Consequently, quantitative methods could not (and should not) be attempted without
the groundwork laid by rich ethnographic studies. Nevertheless, informed survey research

allows cautious generalizations to be made to a broader array of cases. Recognizing the

drawbacks of the method and taking these as a caveat to proceed cautiously and not overstate
claims, we can benefit from the advantages of such an approach. In particular, careful survey
research permits controlling for variation in context and background through a series of
demographic variables in order to draw rigorous conclusions about the determinants of

values. With these issues in mind, I conducted a political attitudes survey among elders,

I83

rural organization leaders. and heads ofhousehold in twenty-one villages of southern Mali's

cotton zone in October and November of 1995.

A caveat is necessary concerning the nature of the research question and the data.
Implicitly. there is a time element in the research question - what impact does prolonged
exposure to the procedures of particular types of organizations have on the behavior of
participants? Ideally. we would have access to time series data on a panel of respondents and
be able to measure how their values changed over time. Given the paucity of political
attitudes data in rural Africa. we are forced to simulate this time series element using cross-
sectional data — where possible, we correlate length of membership and values. In other
cases we just correlate dichotomous measures of membership with values. Caveat lector!
The reader must recognize that we are inferring the existence of a long, slow process from
cross-sectional data.

4.4.2 Sampling

The feasibility of sample frame development played a role in the delineation of the
area under study. Because of the importance of organizational activity to the study, an effort
was made to include villages representing a range of levels of organizational activity.
Unfortunately, little central data exist on the density of organizational activity at the local
level in Mali. Moreover. such activity varies markedly from region to region. The one
institution possessing organizational information to this degree of detail is the Compagnie
Malienne pour le Développement des Textiles (CMDT), Mali's cotton-producing parastatal.
In order to ensure comparability of data, I restricted the study to the operating area of the

CMDT.

184

The sample frame was developed using CMDT lists. as well as the 1987 Malian
census. A multi-stage. proportionately-stratified random sample was drawn from the
population of over four thousand villages and hamlets in southern Mali's cotton-producing
zone. The villages were classified in three categories. according to level of organizational
activity. as designated by the CMDT. as described in Chapter 2. Reiterating. villages
without cotton producers' organizations are classified as villages classiques. Villages with
a simple cotton-producers' organization (associations villageoises, or AVs, in CMDT
parlance) constitute their middle level of organizational activity. Finally, villages having
complex arrays of organizational networks and collaborative activity with other AVs are
classified as A V/ZAER (Zone d’Animation et Expansion Rurale) Villages.

Twenty-one villages were randomly selected according to the proportions of different
village types in the population. Consequently, eight AV/ZAER villages, seven simple AV
villages, and six villages classiques were chosen. These villages were widely dispersed over
the CMDT's working area and included villages situated from 100 to 450 kilometers from
the capital, adjacent to and remote from paved roads. mono-ethnic and multi-ethnic, and
ranging in population from barely 100 inhabitants to about two thousand inhabitants. In
short, the surveyed villages constituted a reasonable cross section of rural southern Mali. A
sense of the diversity of these villages is provided by Table 1 in Chapter 2.

In each village. enumerators were instructed to compile lists of organizational
officers, village elders, and heads of household not fitting into either of the first two groups.
From these lists, they randomly selected four from each category for a total of twelve

interviews per village. The total sample size was 252.

4.4.3 Operationalization

The purpose of the present study is to operationalize indicators of organizational
participation and political attitudes. as well as several demographic control variables, in order
to assess the effectiveness of rural organizations in performing their so-called educational
function.

The survey instrument included questions on political participation and associational
life, political authority. political accountability. economic and political knowledge. policy
preferences. attitudes and orientations toward trust, efficacy, tolerance, materialism/post-
materialism, and a broad array of demographic questions. All of the surveys were
administered using Bambara (also known as Bamanankan), the linguafranca of southern
Mali. The questionnaires were carefully translated and back-translated between English and
French and then French and Bambara before the pre-test and then painstakingly revised
based on the pre-test. Much of the questionnaire was based on the 1993 USAID/Zambia
survey directed by Michael Bratton and Beatrice Liatto-Katundu. The survey instrument can
be found in Appendix 2.

The dependent variable is political values as measured by respondents' expression of
political attitudes. I was interested in measuring a broad range of attitudes and looked at the
possibility of a composite civic values scale but was concerned that it included excessively
diverse factors and it would be very arbitrary to assume that such values and orientations as
trust, efficacy, and tolerance should be weighted equally. Consequently, I chose to compare
results on two different individual dimensions of political values, employing each in turn as

a dependent variable in the analysis. One option was to employ individual responses to

186

attitudinal questions as the dependent variable. The disadvantage ofthis approach is that a
single question is unlikely to capture the rich complexity of a concept being measured. For
this reason. the decision was made to create a dependent variable that was a composite of the
responses to several questions. each weighted according to factor loadings.

Factor analysis assumes that observed variables are linear combinations of fewer.
unobserved source variables. It permits the drawing of inferences about a given causal
structure from the correlations between a series of variables. Factor analysis permits one to
use high correlations between observed variables to represent data in terms of a smaller set
of unobserved, hypothetical source variables (factors) (Kim and Mueller 1978a, 1978b).
Underlying factors are identified statistically but not named. Consequently, one must have
a very good theoretical reason for hypothesizing the identity of a factor.

The two dimensions of political values examined here are political efficacy and
political tolerance. Political efficacy was chosen because it is arguably the dimension of
political values most closely related to organizational participation. That is, the concept of
political efficacy describes feelings of competence when engaging the political system. It
is therefore plausible that those engaged in organizational interactions would have increased
self-confidence when participating in the political system. Political tolerance was selected
as the second dimension of civic values used because it seemed to best embody the open-
mindedness required in a democratic polity. The concept of political tolerance is
operationalized by measuring the respondent's willingness to let social and political groups
with which he may have little in common participate in the political system (Sullivan et al.

1981, 1982). It can also be measured using questions which elicit one's willingness to

187

compromise with groups with which she disagrees. The questions employed in the two
scales are displayed in Table 25. These questions constitute well-established means of
operationalizing these values (Bratton and Liatto-Katundu 1994).

In the construction of the factor-analyzed scales. covariation among the first five
variables in Table 25 was assumed to spring from one principal factor. the fact that they all
measure dimensions of "political efficacy." In fact. a single factor was found to explain
68.82 percent ofthe variance in responses to these questions.

Similarly, covariance among the last five variables in Table 25 was assumed to stem
from the underlying commonality that all of these variables tap into dimensions of "political
tolerance." Interestingly. a single factor explained almost all of the variance in question
responses (95.40 percent).

The difference between the factor-analyzed scales and standard additive indices is as
follows. Survey questions are used to operationalize concepts. In Table 25, some questions
are better proxies than others in operationalizing dimensions of political efficacy or
tolerance. In a standard additive index, "democratic" answers to each question would be
weighed equally in their contribution to the value purportedly being measured. Because each
question is different in its contribution to the value. this is an imprecise means of
aggregation. Factor analysis permits the discovery of the importance of each question in the
measurement of the concept and weighs them accordingly, thus offering a more accurate

measurement of the level of tolerance or efficacy of each respondent.

188

Table 25 - Survey Questions Used in Value Scales

 

For each question. respondents were asked to choose between two statements, A and
B, indicating the one that most closely reflected their attitudes or behavior.

 

Political Efficacy Scale:

MAKEPPLNS
A. It is not wise to plan too far ahead. because many things may turn out to be a matter of luck.
B. I always try to plan ahead. because I feel I can make my plans work.

INFLUOTH
A. In discussions about politics with friends and neighbors. I can influence the opinions of
others.
B. As far as politics are concerned. friends and neighbors don't listen to me.

COMPGOV
A. Government sometimes seems so complicated I cannot really understand what is going on.
B. The way that government works is generally understandable to people like me.

DEPSLIST
A. As a community. we are generally able to make our depute'(s) listen to our problems.
B. We are usually unable to make our de'puté(s) listen to us.

GOVSOLVE
A. Government officials and politicians are capable of solving the problems that Mali presently
faces.
B. Government officials and politicians are not very effective at carrying out programs to solve
national problems.

 

Political Tolerance Scale:
EXPRVIEW
A. If people have different views than you do. they should be allowed to express them.
B. It is dangerous and confusing to allow the expression of too many points of view.
ORGAFFIL
A. If people want to form a community organization. they should affiliate with the ruling party.
B. If people want to form a community organization, they should feel free to do so
independently.
MUSLPARTY
A. Even though Mall is primarily Muslim, Christians should be allowed to form a Christian
political party.
B. Christians should not be allowed to form a Christian political party.
POLCOMPR
A. To compromise with one's opponents is dangerous because you betray your own side.
B. The only way we can all get along is to compromise with each other.

 

 

 

The principal explanatory variable employed was organizational participation.
measured in a variety of ways. Dummy variables were created to distinguish between
individuals who were only members of village-level cotton producers' associations
(associations villageoises. or AVs), other organizations but not AVs, or both AVs and other

organizations.

I89

Along with organizational participation. several independent control variables were
used. specifically age. wealth. and education. Age is important in African societies. where
elders are automatically accorded great deference. For this reason. one expects a positive
relationship between age and political efficacy. There is no cultural or theoretical reason to

expect any particular relationship between age and tolerance. on the other hand -- Abramson

and Inglehart predict more democratic values to be held by younger cohorts. but report that
“in societies with very little growth in per capita income. such as Nigeria. there is only a very
weak relationship between age and values” (Abramson and Inglehart 1995. p. 7).

Wealth or income is difficult to measure in parochial rural settings where people are
unfamiliar with social science research and the motives of outsiders are suspect. I therefore
employed several proxy measures for wealth. These included asking about ownership of
farm technology such as mechanical seeders and plows. asking about ownership of bicycles
or mopeds, and inquiring about cotton production and hectarage. These are necessarily
imperfect proxies for wealth. The Western political culture literature predicts a positive
association between wealth and democratic values.

I employed several measures of education. including questions on time spent in
formal French-language educational settings, time spent in functional literacy programs
employing national languages, and a question simply asking the respondent whether he or
she could read, defined as whatever that meant to the individual. Education is expected to

be positively-associated with democratic values such as efficacy and tolerance.

190

The next section provides a description of sample demographics. as well as
illustrative results for survey questions on the various values. The following section
provides regression results.

4.5 Descriptive Statistics on Organizational Participation and Political Attitudes
4.5.1 Sample Demographics

The sample consists of 230 men and only 22 women. Those belonging to the
Bambara ethnic group constituted by far the plurality of those surveyed. 38.1 percent. The
next most numerous group were the Minianka. 18.7 percent of those surveyed. followed by
the Peul and the Bobo. at 12.3 and 9.5 percent, respectively.‘8 Seventy-seven percent of
those surveyed were Muslim. Animists constituted 14.3 percent of the sample, while
Christians of all denominations comprised 7.1 percent. Interestingly, all thirty-one of the
Peuls surveyed reported that they were Muslim, as did 89.5 percent of the Bambara, while
none of the Bobo and only 51 percent of the Minianka were Muslim. Two-thirds of the Bobo
were animist, as were 27.6 percent of the Minianka.l9

According to the respondents' own descriptions of themselves, 91 of252 respondents
(36 percent) had a role more important than simple membership in their village-level cotton
producers' association (association villageoise, or AV). An even greater number, 109 (43.3
percent), claimed posts of responsibility in other organizations. The self-described

organizational activists. those who have leadership roles in either an AV or another type of

 

‘8 Malian law prohibits the publishing of data on the distribution of ethnic groups. so it is impossible to know how
closely the sample corresponds to the population in this regard.

19This closely mirrors the conventional wisdom. which suggests that Mali is about eighty percent Muslim (though
Muslims claim that this percentage is higher). Because the north of Mali has fewer animists and Christians than the
south. these numbers seem to reflect the population quite well. The greater probability of the Bobo and the Minianka
to not be Muslim also squares with rough estimates of the national demographic profile, though I cannot be sure how
closely the Christian/animist breakdown mirrors reality.

I91

village-level association. constitute 58.7 percent ofthe sample. Village council members
constitute 32.5 percent (82 respondents) ofthe sample. Sixty-two of the 252 respondents
(24.6 percent) held no organizational leadership or village council positions.

At first glance. sample literacy statistics closely mirror those of the United Nations
Development Program (UNDP) for Mali. The UNDP reports that 27.2 percent of Malians

are literate.BU

We asked respondents whether they could "read and write, in French or a
Malian language." to which 34 percent answered "yes" and 66 percent answered "no." At
closer examination. the question posed a more vague and less rigorous standard than that of
the UNDP. Indeed. an astounding 70.8 percent of the respondents reported one or fewer
years of any kind of education and a full 78.8 percent reported no more than three years of
education. Only 17.2 percent of those surveyed had ever had any formal, F tench-language
schooling and only 6.0 percent had six or more years. The maximum amount of F rench-
language schooling of any respondent was eleven years. In other words, not a single
respondent had completed secondary school. Non-formal education advocates would be
encouraged to note that a full 32.9 percent of the respondents reported having attended
functional literacy classes in a Malian language. though only 20.2 percent (i.e.. 61.4 percent
of those reporting functional literacy training) had taken more than one year of these
classes.2| The low level of literacy of the respondents is important because of its

implications for capacity-building among rural constituencies and particularly its impact

upon feelings of political efficacity. While literacy is low in the sample in general, it is

 

20 Literacy is defined as "The percentage ofpcrsons aged fifteen and over who can. with understanding. both read
and write a short simple statement on their everyday life" (UNDP 1995. p. 225).

21 . . . . .- . . . . .
Functional literacy IS difficult to measure. short of administering a test. as such programs vary wrdely in length
and intensiveness.

I92

important to note that only 20.9 percent of the non-activists claimed to be able to read. as
opposed to 37.8 percent of the activists. So while literacy is uniformly low throughout all
sub-groups in the sample. it is nearly twice as high among organizational activists.

Every respondent but one (99.6 percent) said that they farmed. Of these. village
activists and council members were more likely than non-activists to grow cotton, Mali's
principal cash crop. Among the first group, 84.7 percent reported cultivating cotton, while
only 69.4 percent of the non-activists did the same.

4.5.2 Orientation Toward Politics

When asked about the frequency with which they discuss politics with other people,
57.6 percent of those responding said that they discussed politics "less often than monthly,"
with many of these respondents insisting categorically that they never discussed politics.
While an even 30 percent of those surveyed reported that they discussed politics at least
weekly, only a single respondent acknowledged discussing politics every day. Men were
much more likely to discuss politics than women.22 Not surprisingly, council members and
activists in village-level associations were much more likely to discuss politics than non-
activists (44.7 percent versus 24.2 percent).

4.5.3 Political Knowledge

Activists and council members were far more likely to seek political information than

were non-activists. Among those who reported having listened to the radio news, non-

activists were over four times as likely to listen to the news less often than once a month

 

‘) t o a o n o u v . u v a .
2‘ The chi-square assocration for this relationship is Significant at the .002 level. so I report it in spite of the small
number of women responding to this question (twenty). Any relationship reported in the text reflects an association for
which the chi-square statistic is significant at the .05 level or better. unless otherwise reported.

193

(14.5 versus 3 .2 percent). Not surprisingly. the non-activists were also far less likely to have
read a newspaper recently -- 8.1 of the non-activists had read a newspaper. compared with
26.3 of the activists and council members.

Survey enumerators posed several open-ended questions related to the political and
economic knowledge of those surveyed. When respondents were asked to explain "what is
meant by the term ‘devaluation’?." referring to the January 1994 devaluation of the CF A
franc, 108 of the 136 respondents who ventured answers gave responses along the lines of

H H

"our money is worth less. prices are higher." or "we're more miserable now."23 One should
note. however, that fully forty-five percent of the respondents volunteered the response "I
don't know" to this question. Examination of these respondents is instructive. Two-thirds
(66.1 percent) of the non-activists volunteered this response, compared with only 38.4
percent of the activists and council members. Similarly, 77.2 percent of the women surveyed

said they didn't know, compared with only 42.1 percent of the men. While the difference

appears significant and resonates in light of the Chapter 3 results, the reader should note the
very small number of women in the sample (n = 22). On the other hand, all of the women
were organizational leaders. unlike the men, so the frequency of “don’t know” responses is
particularly striking.

When asked "who is responsible for devaluation?" fewer people were at a loss for an
answer. In this case, 61 respondents (24.2 percent) said they didn't know. Interestingly,
many respondents who didn't acknowledge knowing what it was were nonetheless willing

to attribute responsibility. I interpret this to be the result of people having heard

 

23 The CFA franc is directly pegged to the value of the French franc. The former rate was 50 CFA to 1 French
franc. but the rate has been 100 to 1 since January 1994.

194

"devaluation" expressed with a negative connotation. so they are happy to assign blame even
if they cannot clearly describe the phenomenon. Of the 190 who ventured a response, 55.2
percent blamed "the government" or the more nebulous "our leaders." while 17.3 percent laid
the blame directly on the doorstep of President Konare. Only ten percent said that "the
whites" or "the F rench" were responsible when, in fact. devaluation could not have occurred
without the assent of the F rench.” Again. analysis of the 61 "don't know" responses is useful
-- 43.6 percent of the non-activists in the sample gave this response. as opposed to only 17.9
percent of the activists and council members. When this response is compared in terms of
gender, over twice as many women as men (45.5 versus 22.2 percent) said that they didn't
know.

The answer "don't know" can be interpreted in more than one way. Primafacie, it
demonstrates a lack of knowledge, but the issue is more complicated than this. An incorrect
answer could reflect a lack of education, but the failure to even venture a guess suggests
something even more profound, a lack of empowerment. The two are certainly related.
Indeed, there is a very strong negative relationship between claiming to be literate and

likelihood of answering "don't know" -- Among those who say they can read in French or

in a Malian language, only 22.4 percent responded "don't know" to the question asking them
to say what devaluation was, while 56.4 percent of the self-identified illiterate respondents

gave that response. This corroborates the repeated finding in other contexts that the principal

correlate of "don’t know" responses is low education (Schuman and Presser 1996, p. 114).

 

24 . . . . . . . . .
Given the frequency With which educated elites infonnally suggest their disenchantment with continued close

ties between Mali and France. one suspects that this response would appear more often in a survey done in Bamako or
among more educated respondents.

195

Because most village-level literacy programs are administered in cooperation with
AVs or women's groups. this suggests that organizational membership is an intervening
variable which affects one's probability of being literate and thus being more capable of
informing oneself. As stated. I think that the failure to venture a guess to factual questions
is a function of political efficacy as well. and it is to this topic that we now turn.

4.5.4 Political Efficacy

In the domain of questions aimed at measuring one's personal sense of efficacy and
how that relates to political issues, rural Malians tended to give answers which expressed an
ethic of group solidarity over personal capacity and pursuits. For example. when asked
whether they worked better in a group or alone, 78.2 percent chose the former option. When
asked whether they put their main effort into improving their own lives or those of younger
family members, 92.5 percent chose the latter response. The vast majority of the respondents
clearly saw their personal success in a communitarian context. For neither of these questions
was there a significant difference between the responses of council members and
organizational activists and non-activists.

When asked to choose between "it is not wise to plan ahead, because many things
turn out to be a matter of luck" or "I always try to plan ahead because I feel I can make my
plans work," in terms of which best characterized them, over two-thirds (67.9 percent) chose
the first statement. This acquiescence to the role of luck is consistent with responses we saw
to other survey questions. For example, when asked an open-ended question on why they
were better or worse off than four years ago, 8.8 percent volunteered that their current status

was an affair that only God or destiny could determine. When asked an open-ended question

196

concerning how they had changed their lives because of crime, almost a quarter (23.0
percent) volunteered that they had done nothing because this matter was in the hands of God.
fate. or luck. Again. the chi-square association for the "plan ahead" question was not
significantly different for non-activists and activists.

When asked a more direct question concerning political efficacy. 71.8 percent said
that they were able to influence the political opinions of their friends and neighbors, while
25.8 percent didn't know or felt that those close to them didn't listen to them. In this case.
activists and council members were significantly more likely than non-activists to express
confidence in their persuasiveness (76.3 versus 58.1 percent). Intuitively, the analogous
relationship holds for efficacy with respect to influencing politics. While half (49.6 percent)
of all respondents agreed that "as a community, we are generally able to make our deputies
listen to our problems. activists and council members were considerably more likely to
express this efficacy than non-activists (53.7 percent, compared with 37.1 percent,
respectively).

4.5.5 Political Tolerance

Rural Malians displayed remarkable levels of political tolerance in response to
several questions. For example, 68.25 percent of those surveyed felt that people with
different views should be allowed to express them. Similarly, 57.2 percent approved of
community organizations operating independently, without being affiliated with the ruling
party. However, other results are mixed in what they imply about the political tolerance of
rural Malians. For example, while 65.5 percent of respondents acknowledged that "in Mali

today, we have a real choice among different political parties and candidates." 58 percent felt

197

that "the best fomi of government is one that gets things done." rather than "one elected by
its people." Moreover. a majority of those surveyed opposed the right of minority Christians
to form their own political party. and non-activists were even less tolerant than activists and
council members (69.4 percent opposed. compared to 57.3 percent). Interestingly. the more
active group overwhelmingly disapproved of the use of violence being used to further
political goals, with 82.3 percent expressing opposition. The non-activist group, on the other
hand, was much more evenly split. with only 56.4 percent expressing disapproval, and a full
43.6 percent saying that "in Mali. the use of violence is sometimes justified to achieve
political goals." This willingness of the non-activists to countenance the coercive use of
force marks a striking difference among different sub-groups of farmers.

4.5.6 Political Trust

Respondents were asked to state their level of trust in various social groups and
institutions. They were asked to use a scale of one to ten, with "1" signifying that "I do not
trust them at all" and "10" signifying that "I trust them completely."

Not surprisingly, rural Malians expressed the highest degree of trust for those groups
with whom they had close contact. Groups further removed from the villagers' personal
experience tended to be ranked lower or the respondent freely professed not to know how to
rank them. For the three groups closest to their day-to-day lives, family, relatives, and
neighbors, the mean ratings were 9.32, 8.52, and 8.02, respectively.

Malians from the three regions in which the survey was conducted, Koulikoro,
Ségou, and Sikasso, were rated 6.58, 6.75, and 7.56, respectively. Gao, Timbuktu, and

Kidal, the three northern regions of the country from which Tuareg rebels had recently been

198

waging a secessionist war. garnered trust levels of5.19. 5.02. and 4.74. Interestingly. Mopti.
the central region which separates the southern three regions from the northern three regions.
earned a mean rating of 5.80. which leaves us with an intuitively appealing negative
relationship between trust and geographical distance.

Trust levels for citizens of nearby countries tended to be near or below the bottom
end of how fellow Malians were rated. Ghanaians scored the lowest. with a mean rating of
4.09. while Ivoiriens scored the highest. at 5.45. Given that many Malians spend protracted
periods of time working in Cote d’Ivoire, as well as the fact that there are liguistic and
cultural links between the two neighbors. this result resonates as accurate. The Ghanaians'
score was not surprising either. as Malians commonly deride natives of Anglophone
countries as "drug dealers" and "bad barbers," and it is consistent with the mean rating of

S

4.56 received by Sierra Leoneans.2 I was, however. surprised by the low rating of 4.50

given to the Burkinabe, considering its proximity to the survey region and cultural and ethnic
similarities between Burkina F aso and Mali.

The scores given to remote northern Malians and those from nearby countries were
surprisingly high, close to the center of the scale, but are most useful when evaluated relative
to other scores. For example, the mean respondent's rating for his own ethnicity was 8.80.
higher than even the mean ratings for other family members. Therefore, when one compares
the rankings for people originating in the north of Mali or in Anglophone countries, for

example, one can detect a fair level of distrust.

 

25 When the Ghanaian junior national soccer team came to Bamako in 1995 and defeated their hosts. Malians
lamented falling at the hands of "les mauvais coiffeurs," a reference to the Ghanaians and Sierra Leoneans who operate
roadside barber shops in Mali's capital city.

199

Other relationships are also instructive. One expects the high rating given to Islamic
organizations (8.2) but the relatively high score given to Christian organizations (6.84) is
somewhat surprising?" On second glance. given the tolerant nature of Malian Islam and
Malians' general respect for practicing believers of any stripe. the score seems quite
plausible. Trust for SYCOV, the national cotton producers' union. is startlingly low (4.62).
but rings true when compared to responses to other survey questions. Only 25 percent of
those surveyed acknowledged being a member of the nascent advocacy group. and of the
relatively few responses (n=74) to the question "what does SYCOV do for farmers?," 45.9
percent responded "nothing" and 14.9 percent responded that they didn't know. Anecdotal
evidence from ancillary comments made by some of those surveyed suggest that many
farmers are disillusioned that some have bought membership cards but seen little advocacy
and no concrete results. Somewhat surprisingly, the CMDT, the cotton-producing parastatal,
received a 7.73 rating. This could be an artifact of mistrust of the research team and villagers
not wanting to "bite the hand that feeds them" (as the CMDT has a monopsony on cotton
purchasing) or it could be a genuine reflection of satisfaction over the contribution of cotton
to village-level economic fortunes.27

Trust for national-level institutions was also quite high. Local representatives of the
central government received a mean rating of 6.88. the National Assembly was rated at 7.16.

and the police garnered a score of 6.46. This could be the popular maxim, "familiarity

 

26 Though a Muslim Malian professional told me that he trusts Christians more than Muslims for the following
reason -- Muslims in a Muslim country are just conforming to social pressure, while Christians are violating
convention for reasons of faith. Therefore. he reasons. a Christian is more likely to be principled and treat someone
fairly in a business deal.

27 When asked whether the C MDT was justified in having exclusive rights to buying cotton in Mali. respondents
were evenly split, 1 18 to 1 12 in favor. with 22 saying that they did not know.

200

breeds contempt." turned on its head — In other words. local citizens have very little contact
with these institutions. Because crime was often mentioned as a concern. the high marks for
the police could be an affirmation oftheir importance. With respect to the other institutions
the high marks are somewhat puzzling. Interestingly. The National Assembly rating
provides the reverse relationship to what one finds in the United States. where people tend
to respect their congressperson more than the institution as a whole. Rural Malians gave
their own deputies a mean score of 6.5 1. compared with the 7.16 given to the institution as
a whole.

The raw results on the attitudinal questions suggest that village elders and
organizational leaders are slightly more predisposed to democracy than other heads of
household. These results are very preliminary, however. Given marked difference in wealth
and education between the two kinds of leaders on one hand and the simple heads of
household on the other. one must control for those differences in order to isolate the effect
of participation.

4.6 Results of Regression Analysis

As mentioned previously. the different dimensions of political values are highly
correlated, so two for employed here for the purposes of illustration. Table 26 displays the
results of a robust regression run with the political efficacy scale as the dependent variable

and organizational participation run as the principal explanatory variable.28

 

28 . . . . . . . . . . .
Robust estimation methods aim at achrevrng better efficrency than ordinary least squares (OLS) in Situations

where the standard OLS assumptions are violated. In cases where data are ideal. robust regression still achieves nearly
the efficiency of OLS.

201

Table 26 - Regression Results: Political Efficacy and Simple Membership

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Variable Coefficient Standard Error t p
AV member -.1197283 ...227230 -0.538 0.591
Other member .1478541 .2978075 0.496 0.620
Both AV & Other .0780954 .2141602 0.365 0.716
Age .1096353 .0444827 2.465 0.014
Produces cotton? -.9315810 .1970043 -4.729 0.000
Can read? .0721113 .1055558 0.683 0.495
Constant -1.5800350 .4066387 -3.886 0.000
n = 245
r2 = .188

Interestingly, there is no statistically-significant relationship between any
combination of simple participation (that is, membership without necessarily holding an
office) and political efficacy. Whether an individual belongs to a cotton producers'
organization (AV). any other type of organization, or belongs to an AV and least one other
organization, there is no relationship with the level of political efficacy felt.

In contrast, there is a strong positive relationship between age and efficacy. This is
not surprising -- in a gerontocratic society, where age is respected and a certain authority
comes with age, this is the relationship that we would expect.

On the other hand. the strong negative relationship between the rough proxy for
wealth, cotton production, and efficacy runs counter to the predictions made by the political
culture literature. Nevertheless, put in perspective, this surprising result may not be as
counter-intuitive as it initially appears. At the time of this research, Mali was only four years
removed from an authoritarian regime in which the overriding ethic was neo-patrimonialism.

Moussa Traore’ sat atop a pyramid in which followers who did his bidding were rewarded for

202

their loyalty. The persons interviewed for this study were predominantly village and
organizational leaders. In Mali's villages these posts are more often than not occupied by
members ofa village's founding family or its allies. Village chiefs are chosen from a line
of succession within the founding family and have long served as the village-level
representatives ofthe national administration. Consequently. they were long integrated into
vertical patron-client networks and their families felt the benefits of this influence. Thus.
given the resistance of pre-democratic elites to a system which could greatly reduce their
influence. this inverse relationship between wealth and efficacy in the new system is not
surprising.

Various specifications of the education variable had the expected sign but were
statistically insignificant. The variable used in Table 26 was a dichotomous variable where
respondents self-identified themselves as readers or non-readers. Only 34.29 percent of
those interviewed identified themselves as readers. Other specifications included asking
about participation in functional literacy programs or attendance of formal French-language
or Franco-Arab schools. Data on functional literacy are unreliable because of individual
difficulties in recalling duration of study and are complicated by the wide variation of
intensity of study. There is no universal standard by which to compare responses to this
question. Data on formal schooling are more helpful, but there was relatively little variation
on this variable. The vast majority of respondents had no formal schooling whatsoever, none
had more than eleven years. and few had more than six years. Stated differently, there was
not a single lycée graduate and only very few primary school graduates in the whole sample.

making for relative homogeneity in minimal formal educational experiences. Alternatively,

203
ifour rival hypothesis about wealth holds. the high correlation between education and wealth
could mute the effect of education.

The value of the square of the correlation coefficient. r3. is quite encouraging. In
attitudinal research. where it is rare to explain more than twenty—five percent ofthe variance
on the dependent variable. a value of.188 for r2 is substantial (Witte 1997. p. 170).

At first glance one may expect multicollinearity between the AV membership and
cotton producer variables. While this hypothesis is intuitively appealing. this is far from the
case. It is true that fully 160 of the 252 respondents produce cotton and are AV members.
Put differently. 78.43 percent (160/204) of cotton producers are AV members. Interestingly
however, 72.92 percent (35/48) of the non-cotton producers are also AV members. This is
not altogether surprising. as we noted in Chapter 2 that non-cotton producers often share the
benefits of pubic goods generated by these organizations. As the small degree of difference
between the above percentages would lead us to expect. the correlation between these two
variables is a meager .0518 and a chi-square test found them to be statistically independent
(p=0.41 1). We can therefore unambiguously rule out multicollinearity.

To test the robustness of this result under different specifications. logistic regressions
were run on individual dichotomous components of the efficacy scale with virtually identical
results. Specifically, age and cotton production were still the only explanatory variables for
which a significant relationship was found, and the direction of these relationships remained

consistent.

204
Table 27 displays the results ofa robust regression run with the political tolerance
scale as the dependent variable and organizational participation as the principal explanatory
variable.

Table 27 - Regression Results: Political Tolerance and Simple Membership

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Variable Coefficient Standard Error t p
AV member -.01 17589 .1972658 -0.060 .953
Other member .2089406 .2723448 0.767 .444
Both AV & Other .0286467 .1849139 0.155 .877
Age .0120893 .0349795 0.346 .730
Produces cotton? -.564 l 604 .2077696 -2.715 .007
Can read? -.1072418 .0800908 -l.339 .182
Constant -.7275134 .3587199 -2.028 .044
n = 245
r2 = .0838

The tolerance scale yielded very similar relationships to the efficacy scale. The one
notable difference is that age is no longer significant, as was predicted. There is no reason
to expect a relationship here between age and tolerance. Once again. the organizational
participation variables are not revealed to have any significant effect on attitudes. The
wealth proxy, cotton, continues to have a strong inverse relationship with democratic values.

As was done with the previous relationships, logistic regressions were run on
individual dichotomous components of the tolerance scale with results that were identical
to these in terms of direction and strength of relationships. This gives added confidence
regarding the robustness of the results.

Given the lack of evidence that simple organizational membership has any impact

upon political values, I now examine whether a more profound level of participation has any

205

effect. Similar regressions were run with the same dependent variables and demographic
independent variables while employing village and organizational leadership as independent
variables. I included a dichotomous independent variable representing organizational
leadership in order to measure the effects of that level of participation. For purposes of
comparison. another dichotomous variable for village council leadership was included.
Table 28 displays these results.

Table 28 - Regression Results: Political Tolerance and Organizational Leadership

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Variable Coefficient Standard Error t p
Village Council —.2932477 .1 108986 -2.644 .009
Local organization .245 7925 .1044061 2.354 .019
leader
Age .0257321 .0367604 0.700 .485
Produces cotton? -.5469955 .19601 15 -2.791 .006
Can read? -.1571 860 .0929284 -1.691 .092
Constant -.6239525 .3347904 - l .864 .064
n = 245
r2 = .1 189

While startling. the strong inverse relationship between village council membership
and political tolerance is consistent with the previously cited alternative hypothesis about the
inverse relationship between wealth and democratic values. Village ruling families and their
allies are justifiably nervous about the changes that democratization and decentralization
reforms could engender at the local level. For example, in one village an elder of the village
founding family told me that "democratization can only go so far -- a non-Coulibaly [his

family name] will never be chief. Jamais!"

206

Here. for the first time. one of the participation variables has a positive effect on
individual levels of political tolerance.29 The reader should note. however. that the positive
and significant organizational leadership result is the least robust ofthe results. Under other
specifications. it became insignificant. Nevertheless. given our pessimistic predictions of
this variable not having any effect. we must reflect on how it could be plausible for
organizational leaders to hold more democratic values than their rank-and-file members in
organizations without democratic internal procedures. For example. it is conceivable that
organizational leaders are more likely to travel outside the village on association or village
business. Regular exposure to the democratic ferment of regional capitals could plausibly
have a socializing effect that their association did not.

The previous relationships between tolerance and age and wealth remain. As in
Table 27, there is a counterintuitive negative relationship between the education variable and
tolerance, but that variable is insignificant. Interestingly, the findings on age, wealth.
education, and simple membership effects were consistent across a variety of specifications
of the model.

In the next chapter, the above results will be interpreted in the context of the
substantive findings from the other chapters. In addition. tentative policy recommendations

and avenues for further research will be proposed.

 

29 Even if we traced these views to participation. there would be an endogeneity problem -- with the present cross-
sectional data, one cannot easily sort out issues of self-selection. In other words. how do we know that individuals
with greater levels of democratic values at the outset are not for this reason impelled to join organizations?
Longitudinal data or reliable data on length of membership and leadership would be needed to help us to sort out this
issue.

Chapter 5
Conclusion
Une communaute’ de base ne saurait re'aliser des operations de de'veloppement
correctes que dans la mesure ou elle baigne dans une atmosphere de confiance et
de sécurité.1
Bernard Le'déa Ouedraogo (1992. p. 3)
5.1 Introduction
The past decades dramatic transitions in many African countries have provided
ample fodder for debates on many topics pertaining to democratic transitions and democratic
consolidation. Among the most important of these concern state-society relations, the nature
of civil society in Africa, and the relationship between political participation and political
culture. Besides raising important theoretical questions, these issues are important for their
policy implications. Donors spend hundreds of millions of dollars annually on promoting
democratic governance in Africa. and much of this money is devoted to support for civil
society. Civil society is in vogue among scholars and practitioners because of its purported
pluralist and educational functions in nascent democracies. That is. the wide array of
autonomous associations and non-governmental organizations that make up civil society are

held to play an important role in both interest representation and the socialization of values

which support democracy.

 

“A local community will be unable to carry out proper development activities
except insofar as it is suffused with an atmosphere of trust and security.”

207

208

Using data from West Africa‘s Republic of Mali. this dissertation contributes to
ongoing debates in three ways. First. I examine the notion of civil society as conceived in
the West and how it is used in academic debates concerning the applicability of the concept
to African politics. Second. I explore the level of engagement of local organizations in Mali
in civic action activities representing the interests of their members. Included in this analysis
is an investigation of the determinants of civic action behavior by local organizations. Third
and finally. I examine the relationship between organizational participation and political
attitudes.

In any research enterprise, the investigator should examine his reasons for
undertaking it -- What does it contribute to existing knowledge? Why is it important? In the
present project, for example, one could reasonably inquire as to the relevance of local, rural
organizations to African politics and to political science as a discipline. Given concerns
about societal disengagement and the huge disparities in economic and social indicators
between rural and urban areas, one could reasonably attribute little political importance to
remote areas where state influence barely penetrates. In fact, however, one can also posit
bridges from the micro level of individuals to the macro level of the national political system.
Such a bridge is potentially provided by local organizations, which can aggregate the
interests of individuals.

These organizations are of interest to political scientists for at least three reasons.
First, in aggregating the interests of individuals, they provide a vehicle for organized
collective action. As the title of this dissertation, drawn from a Bambara proverb, suggests,
Malians value coordinated effort toward a common goal. While African communitarianism

has often been excessively idealized, the fact remains that collaboration with friends and

~Ta L.

209

neighbors is an essential pattern in the social fabric. Individuals take part in a wide array of
associations at various levels. These associations often collaborate formally or informally
with regional or national NGOs or federations. which participate in turn in development
debates. represent the interests of local collaborating organizations. and seek technical
support and financial assistance from donors and the government. So. while local citizens
and their organizations may seem far removed from the corridors of power and thus
irrelevant to the larger political process, many clearly have links to the capital and its
intrigues. Moreover. with the long-delayed but promising decentralization reforms that are
underway in Mali, power will be more evenly distributed at various levels of the state
hierarchy. This will both make officials more accessible to direct entreaties by local
organizations as well as create a large number of new elected positions which need filling.
Civil society actors will be well-placed to answer this call.

A second reason for which local organizations merit the attention of political
scientists is that they provide a crucial training ground for the orientations of citizens toward
the larger political system and their resulting public behavior. For meaningful democracy
to come into being, democratic institutions must be built on a firm foundation of support in
the hearts and minds of citizens. This is directly related to a third reason for attention to local
organizations in Mali. Political scientists should not only be interested in institutions and
processes on the grand stage of international relations or national-level policy. Local
institutions and political behavior are important because they bear a direct relevance to issues
of equity and social justice in the venues with which the vast majority of Mali’s citizens have
the most contact. Indeed, given the debates on “uncaptured” peasants, and their “exit” or

“disengagement” from national politics, intensive investigation of local politics is the best

210

way to shed light on these issues and would be relevant even if one were unable to
demonstrate clear links between the local and the national levels.

This chapter provides a summary and synthesis of the key findings of the dissertation.
I also suggest tentative policy implications and survey fertile terrain for further study. The
remainder of this concluding chapter is arranged as follows. First, I discuss the fit between
local organizations in southem Mali and normative conceptions of civil society. Second, I
examine the extent of civic action among local organizations as well as the determinants of
such behavior. Third. I summarize results concerning the democratic socializing effects. or
lack thereof, on organizational participants. Next. I make tentative policy recommendations
and examine promising research directions raised by the present study, and then conclude
with a few final general observations on local organizations and democracy in Mali.
5.2 The Fit Between the Notion of Civil Society and the Cases Studied

In the first chapter, considerable space was devoted to exploring the various
definitions of civil society that scholars have offered. The point was made that the broad
definition often employed which defines civil society as the arena between household and
state is so vague as to render it a residual category without much explanatory power. Much
of the imprecision comes from the fact that many make no distinction between civil society
and society in general. while at the same time imbuing the concept with normative content
which may or may not exist in a given empirical reality. In other words, many observers see
associational life as an unmitigated good without demonstrating convincingly that this is the
case.

Another conceptual concern with civil society is that it evolved in the West under

very different historical circumstances than currently face nascent civil societies in Africa.

21 1

More specifically. while it preceded the emergence of democracy in Europe. the two are
evolving in tandem in African countries such as Mali. This political development
phenomenon is further complicated by the fact that until very recently, and to this day in
many remote areas. rural Africans have more experience with “traditional” kinship or lineage
based groups rather than those with the voluntary membership which is fundamental to most
definitions of civil society. Many rural African organizations straddle the distinction
between the household and local community, where ties are based on loyalty and personal
bonds, as opposed to voluntary organizations, where bonds are based on a more generalized
trust or social capital.

For the purposes of the present analysis, civil society has been defined as follows.
It is a subset of the voluntary sector, the array of organizations between the household and
the state. Associational life is a necessary but not sufficient condition for the existence of
civil society, which is characterized by a network of autonomous, voluntary organizations
that are inclusive in decision making and leadership selection. Civil society engages the state
in order to represent the interests of its members and may take adversarial or conciliatory
postures with respect to the state. In addition, civil society organizations may undertake a
wide range of activities, from religion to self-help to bowling, but profit-making activities
cannot be the principal reason for existence.

Though civil society is a theoretical construct, the above definition provides an
empirically useful set of criteria for evaluating how close a given case comes to meeting this
standard. Needless to say, within a given geographical space there is a particular
constellation of associational life. It may be densely inhabited or quite sparse, and the mix

of organizations present which meet and do not meet the criteria for inclusion can vary

212

widely. To have a favorable impact on democratic consolidation. one would want a critical
mass of organizations meeting the normative criteria for inclusion under the rubric of civil
society. Similarly. the chances for democratic consolidation are impeded if illiberal
organizations surpass a certain numerical threshold.

Chapter 2 draws on extensive data on organizational characteristics in southern Mali.
The analysis employed in-depth interviews with leaders of a variety of organizational types
in six villages, as well as USAID organizational-level survey data on governance and
management in community organizations. These data permit cautious generalizations about
the nature of associational life in the region. Clearly, local organizations perform a wide
variety of essential social. economic, and cultural functions. For example, youth groups and
associations villageoises (AVs) often perform collective labor to benefit their villages.
Women’s groups ensure that women and children are well-clothed and maternity centers are
staffed and run effectively. Various groups raise funds to give donations or loans to
members in need. Hunters’ groups serve as the guardians of sacred traditions.

In terms of local organizations’ contributions as “classrooms of democracy,” one is
less sanguine, however. Most organizations studied were ascriptive rather than voluntary.
With rare exceptions they showed little autonomy, depending inordinately on village elders
or the cotton parastatal for guidance. Almost without exception, they showed little
inclusiveness in leadership selection and decision making. Furthermore. with the exception
of AVs, local organizations showed little engagement with the state.

In the organizations studied in depth, a paradox emerged. The associations
villageoises in the cotton zone bore the characteristics most closely comparable to civil

society as conceived in the West but were often not entirely autonomous from the CMDT.

213

Other organizations were somewhat more autonomous from the state but tended to be
ascriptive and not voluntary, subject to meddling by local elites. and were not very inclusive
in membership or decision making. Moreover. many organizations demonstrated little
engagement at all with entities beyond their immediate surroundings. One gets the clear
impression that local organizations that meet all of the requirements for inclusion in the
exclusive club of civil society are almost as rare as a Malian rain in March. Again. however,
this is not to say that these groups do not perform other vital functions.

Clearly. even in ascriptive organizations, individuals have the choice of whether or
not to participate actively and many choose to do so. Other treatments of collective action
ask how atomized individuals can be induced into taking part in collective action. In rural
Africa, the problem is somewhat different, as individuals are much less mobile and more
tightly bound to their communities than are most individuals in the contemporary West.
People engage in collective action for deep cultural and social reasons that cannot be
caricatured as economic incentives. This is not meant to idealize or romanticize “traditional”
settings in which women. youth, outsiders, and ethnic minorities are often marginalized. The
point should simply be made that “community” is conceived of differently in different
cultures.

The upshot of this is that organizations meeting a normative standard of civil society
replete with democratic effects are rare. The rosy depictions of civil society serving as a
gilded vessel from which Malians can drink the cool water of democracy are clearly
overstated. Nevertheless, on a case by case basis, organizations can be evaluated regarding
their potential. Moreover. they can be evaluated on more than one dimension, perhaps

having greater potential in one domain than another. For this reason. the research project

214

looked separately at both the interest aggregation and socialization effects of local Malian
organizations.
5.3 Local Organizations and Interest Aggregation

While local organizations and their procedures are intrinsically interesting. they have
added relevance if a connection can be demonstrated between these groups and regional or
national politics and public policy. There is a delicate balance to be struck here. Many
consider civil society to have meaning only to the extent that it interacts with the state, but
excessive interaction can lead to cooptation or other interference and a consequent loss of
autonomy -- “civil society is subject to the same contradictory imperatives of autonomy and
determination that characterize all intermediate spheres,” as Ehrenberg puts it (Ehrenberg
1999,p.235)

I suspect that the relationship between serious claim-making or control over
bureaucracy and level of engagement with the state resembles an inverted U -- at both low
and high levels of engagement, policy voice is low. In the former case. it is low because of
lack of opportunity and possible intervening variables such as lack of efficacy or a belief that
one’s opinion will not be heard. In the latter case, organizational leaders who have been
corrupted or coopted by proximity to the state are less likely to challenge their benefactors.
Only when there is a healthy balance between autonomy and proximity will significant policy
voice occur. This perspective could be described as pluralism tempered with realism. While
recognizing the potential for organizations to represent the interests of their members, it also
acknowledges the distorting effects of neo-patrimonialism and the fact that states are
generally not the neutral arbiters of interests posited by theory. Engagement with autonomy

is a delicate balance to strike.

215

With such issues in mind. USAID data on Malian local organizations’ civic action
behavior were used to examine questions regarding the level of organizational engagement
with the state through claim making on development issues. Also of interest were the degree
of such engagement and with which representatives of the state. the level of collaboration
among non-state actors such as associations and NGOs, and the success and
representativeness of such actions. These data were used to uncover the most important
determinants of civic action behavior and successful civic action among the organizations
studied.

Almost half of the organizational leaders interviewed reported engaging in no civic
action activity at all in the previous year, including at the most local levels of the government
bureaucracy. Among the local organizations reporting civic action activity, nearly forty
percent claimed no success in this activity. When civic action was undertaken it was more
often than not without input from rank-and-file members. Only one third of all the
organizations in the sample. even among partners of USAID, reported civic action successes
in the previous year. Only about one-third of the organizations studied reported collaboration
with NGOs or federations.

In other words, in terms of volume, success, representativeness, and collaboration,
civic action among local organizations in rural Mali is generally insignificant. This is not
surprising given Mali’s long, authoritarian history and lack of incentives for local
organizations to participate in national politics. Before expecting people to internalize
democratic values, they must at least recognize the instrumental usefulness of the system.
The current explosion of associational life leads one to believe that this is beginning to

happen. Civic action is occurring, it just happens to be a small minority of organizations that

216

are doing it. These tend to be mixed gender organizations and especially organizations at a
higher level of aggregation. particularly regional or national NGOs.

The data are instructive and intuitively appealing regarding determinants of civic
action. Sound management practices and systematic financialsystems have a strong positive
effect on both engagement in civic action as well as success in such an endeavor. Internal
democratic governance has a strong positive relationship with engagement in civic action
activities but has no relationship with civic action success. Both of these relationships make
sense. One would expect increased capacity to give organizational leaders confidence to
engage in civic action as well as the competence to succeed. Intemal democracy, on the
other hand. would compel leaders to represent the interests of their members but would not
guarantee success.

Effective representation by an NGO or federation is positively related to civic action
activity and is nearly-significant in its relationship with civic action success. This is
intuitively appealing as well. Links with intermediary organizations are vertical links that
tie a local organization into the regional or national system. Such links increase the
likelihood that a local organization will come into contact with state agents. Furthermore,
the know-how and expertise of intermediary organizations, coupled with the knowledge that
“l’union fait la force,” could lead organizations to be more assertive. The nearly significant
relationship between intermediary organization representation and civic action success can
be explained as follows. The question about effective representation inquired about both
intermediary NGOs and federations. The USAID study found that NGOs exhibited
considerably greater capacity than community organizations (COs), while federations were

not notably more competent than the C05 which make them up. I therefore suspect that the

217
inclusion of both NGOs and federations in the question muted the effect of NGOs’ greater
capacity on civic action success.

Perhaps the most robust result of all is that women’s groups are less likely to engage
in civic action and are less likely to report success when they do engage in such behavior.
This result is particularly remarkable considering the fact that the types of women’s C05 in
the sample tended to be ofequal or greater capacity than the mixed organizations. Because
organizations with greater capacity tended to engage in more civic action and be more
successful in it, the discrepancy between women’s and mixed groups is all the more striking.

With the recent restructuring of the Chamber of Agriculture, the impending reform
of the cooperative law, and the government and the CMDT’s recognition of SYCOV as a
legitimate representative of cotton producers, there are increasing avenues for representation
of rural interests on the national stage. Because it is a quasi-statal institution the officials of
which have often been perceived as condescending to their rural constituents, the Chamber
of Agriculture is an imperfect representative of rural people. The jury is still out on whether
cooperative law reform will reverse nearly a century of top-down organization of farmers in
order to nurture the growth of autonomous organizations. The emergence of SYCOV and
other institutions in civil society (e. g., revitalization of parent-teacher associations, APEs)
bodes well for Malian democracy in that it increases pluralist outlets. However, SYCOV
possesses few independent resources, increasing the risk of cooptation by the CMDT. Thus,
while fraught with potential impediments, citizens of rural Mali face more opportunities than

ever to engage the state.

218
5.4 Local Organizations and Their Educational Effects

It is not surprising that associational life in Africa is often fraught with the same
illiberal characteristics as political society. Neo-patrimonial polities rife with patron-client
linkages can not be expected by some immaculate conception to yield democratic
organizations. These organizations. in turn. cannot be expected to cultivate pro-democratic
values if they are not internally democratic. This is. in fact, what the analysis showed.

Building on the Chapter 2 analysis of the internal functioning of rural Malian
associations, hypotheses regarding the expected socialization effects of these associations
were generated. These hypotheses predicted that the low levels of inclusiveness in decision
making and leadership selection would result in these associations having little effect in
fostering democratic values, though cotton producers' organizations (AVs) were
hypothesized to have more of an impact than other organizations. Because of the nature of
these organizations, not even participation at the more profound level of being an
organizational leader was expected to foster civic values.

Weighted scales measuring political efficacy and political tolerance across several
indicators were devised to serve as "civic values" dependent variables. These were in turn
regressed on measures of organizational participation and a series of demographic variables.
The results can be summarized as follows. Not a single specification of simple associational
membership was found to significantly contribute to individual levels of civic values. This
includes even membership in cotton producers' associations which, while not very inclusive
in their internal procedures, were expected to have the best potential of the local associations

studied here. In some specifications of the model, organizational leadership was found

219

statistically significant in its contribution to civic values. but the result was not robust -- It
was significant in some specifications and insignificant in others. In contrast. wealth and
village council membership were found to have a strong negative effect on civic values.
These were robust results that survived a wide variety of specifications. Village ruling
families and their allies are understandably worried that democratization and meaningful
decentralization could result in dramatic changes in the nature and distribution of local
power. These issues are particularly relevant as we begin to discover the effects of the rural
municipal elections that were held in April 1999.

The next two sections present policy recommendations and fruitful research avenues
stemming from the present work.
5.5 Recommendations

While the unbridled optimism of the most ardent advocates of the socialization
effects of civil society appears unwarranted in the rural Malian context, this does not mean
that development strategies targeting civil society are inappropriate. On the contrary, Malian
civil society is experiencing impressive growth, and the evolution of a contre-poids to the
state could strengthen the prospects for long-term democratic consolidation. On the other
hand, an explosion of growth among ascriptive, uninclusive organizations could portend
chaos as plausibly as democracy.

Mali’s democratic transition faces daunting challenges, most notably low government
revenues and a total lack of trust between the ruling ADEMA party and the political
opposition. These are linked, because low revenues result in challenges in organizing and

running elections, and missteps in this domain contribute to the downward spiral of

220

government-opposition relations. A recent manifestation of this was the repeated
postponement of the local municipal elections that finally took place in April 1999. At the
present writing (August 1999) there are reports of growing popular disenchantment
manifested by unrest in Bamako’s streets.

Nevertheless. with these challenges come notable opportunities. For example, when
the long-planned decentralization reforms finally take place. substantial devolution of state
services to the local level will occur. The Mission of Decentralization purveys appealing
descriptions of the participatory nature of the new decentralized communes. Substantial
resources will also be devolved to the commune level, though local officials will have to be
creative in finding other revenue sources. While considerable latitude for state intervention
remains. the opening created by devolution will permit a greater number of local citizens to
become public officials. while also creating opportunities for partnership between
autonomous local organizations and the state. While participants should be vigilant that non-
elected neo-patrimonial elements such as village chiefs not be ceded too much power, the
emerging system appears to devolve more power and resources to communal councils than
the analogous systems in neighboring countries.

Another promising development is the project currently under way to dramatically
reform Mali’s cooperative law. The emerging law could mark a dramatic paradigm shift
from the French-introduced system of top-down, state-controlled cooperative movements
which have existed throughout the present, waning century to a system of autonomous rural
organizations which interact with the state on their own terms. While the current state

rhetoric is appealing, advocates of autonomous associations must ensure that civil society

221

chooses its own representatives and does not passively allow itself to be represented by a
state-dictated National Council of Cooperation or any other such quasi-corporatist body.

This calls for continued emphasis on capacity-building among civil society
organizations at all levels. Not-explicitly political skills such as functional literacy,
numeracy, and financial bookkeeping should be taught in order to foster competence,
efficacy, broader participation, and greater transparency. The results of the present
dissertation highlight the importance of such capacity to the initiation and success of civic
action agendas.

On such a base could be added "applied" civic education programs emphasizing
participation, oversight. accountability, and partnership in the new decentralized
municipalities. That civic education programs do not often yield significant gains in
democratic values is unsurprising, given the short time frame usually devoted to them. The
fact that they often increase short term participation bodes well, however. If the government
follows through on promises of greater local control of the local and regional development
tax (TDRL), local capacity will be needed to manage and use these funds productively.
Civic education and a spirit of collaboration with societal actors will increase the chances of
success of such initiatives. It could also foster democratic procedures which hold leaders
accountable to their members and result in meaningful claim—making.

Reiterating, the results of this dissertation suggest that activities building both
management capacity and internal governance of local organizations will be a wise
investment, as it will increase constructive civic action, setting a positive example and

crowding out Callaghy’s "resurgent society." Pluralist politics requires the acquiescence of

222
the state but also requires capacity and confidence on the part of societal actors. Support for
programs aimed at such a result is key to avoiding the politics of confrontation and chaos.

In this vein. capacity-building must be undertaken in a spirit of striving to at least
indirectly remedy voids of inclusiveness and autonomy. For example, every effort should
be made to encourage and facilitate the participation of marginalized groups such as women.
youth. and non-producers ofcash crOps. Training programs which favor existing elite groups
may improve local capacity but will do little to rectify imbalance of power. Those designing
capacity-building programs must be explicit about their objectives, because these programs
cannot be assumed to have democratic effects without carefully designing them in such a
way as to ensure it.

Such capacity-building programs should be implemented by neutral parties bearing
credibility with rural people. To wit, state-linked agencies that have a history of
condescension and control should be kept at arm’s length. Indeed, local state agents, most
of whom were socialized into the bureaucracy in the authoritarian era, need education about
the evolving system as much as do rural farmers and civil society participants.

Finally, capacity-building needs exist at the most elementary level, literally -- Mali
has one of the lowest rates of primary school attendance in the world. Any program aimed
at grassroots empowerment must address the low levels of both formal and non-formal

literacy in rural areas, both as a needed skill in itself as well as a means of improving

efficacy. participation, and ownership in the new decentralized system.

 

223
5.6 Future Research Directions

This dissertation has yielded interesting but still preliminary results. These
compelling results call for further studies to expand upon and confirm their implications. For
example. ethnographic studies must be conducted to understand better the variety of civil
society types and environments in order to permit prediction of outcomes in diverse
circumstances. As the present work has shown, the cultural context in which groups emerge
has an important effect on how they behave. their effects on society. and their impact upon
their members.

Subsequent political attitudes surveys with ample questions on different types of
organizational affiliations and levels of participation would permit further testing of
hypotheses concerning how political socialization occurs in different settings. Much remains
to be learned about the interaction between social groupings and attitudes and how they

impact upon democratic consolidation.

The relationship between organizational autonomy, level of engagement with the
state, and level of policy voice merits further study. The USAID data give a good sense of
engagement levels, which are in general moderately low, but do not operationalize
organizational autonomy. Further study is needed to test more rigorously the relationship
between engagement and policy voice. As I suggested earlier in this chapter, it may be an
inverted U where policy voice is only high where a balance is struck between autonomy and
engagement.

The civic action data raised interesting questions regarding why women’s groups
were both less likely to engage in civic action as well as less likely to report success when

they did engage in such activity. In the present sample, lower capacity was ruled out as an

 

224

explanation for these phenomena. The suggestion was made that perhaps such factors as low
political efficacy explained the lower participation of women’s groups in such behavior.
Unfortunately, the political attitudes data set did not include a large enough sample size of
women’s group leaders to test this hypothesis. Consequently. it would be productive to
further examine such issues. Understanding the lack of engagement and policy voice of
women’s organizations will go a long way toward more effectively striving for greater gender
equality in African politics.

The present study only briefly touches on horizontal and vertical links among civil
society organizations. Preliminary indications suggest that regional and national NGOs have
greater capacity and internal democracy and are consequently more active and more effective
than federations at representing the interests of local organizations. This stems in part from
the greater level of education of many NGO leaders, while federation leaders. especially at
the sub-national level, are usually not much more educated than local organization leaders.
Moreover, health and APE federations, among others, have histories of heavy state control.
These issues of scale and concertation among civil society groups are fundamental to the
issue of interest aggregation.

5.7 Concluding Remarks

While no transition is without its challenges, these are auspicious times for Malian
democrats. Nearly a century of authoritarian rule by the French as well as post-colonial
regimes has resulted in a polity which is only now beginning to participate on a large scale.
Much work still needs to be done to ensure that popular energy is constructively channeled
in order to better serve the masses of the country’s population that live in rural areas. The

popular ferment will only be constructive if participants in associational life have the

225
resources and the skills to represent the needs of their sector. Attention to the needs of a
rapidly-growing voluntary sector will also work toward the long-term cultivation of values
which provide fertile ground for the flourishing of Malian democracy. but this process will
take many years.

In the more immediate future. Mali's decentralization reforms portend potentially
significant changes in the contours of local politics and offer an opportunity to marry the
local reality with more participatory. inclusive procedures. Longstanding hereditary power
could erode in the face of meaningful municipal elections and Mali's nascent civil society
will likely serve as a proving ground for future local officials. Then. more than ever, it will

be important to understand the forces that can foster a sense of civic community in the

milieux where the vast majority of the country's citizens live.

APPENDIX A

USAID Community Organization Questionnaire

226

 

L'Equipe Gouvernance Democratique de l'USAID
et Info-Stat
Questionnaire sur la C ollecte des Données de Base pour le Suivi de
la Performance des Organisations Communautaires
Aor’it-Septembre 1998

____________________________________.l
r————————————-—1—____________________[
Bonjour. Je m'appelle [Enquéteurz Dites votre nom]. Je travaille pour un bureau d'étude du nom Info-

Stat qui appuie l'Agence Americaine pour le Developpment lntemationale (USAID) qui assiste votre
ONG partenaire dans le financement de son programme. Dans le cadre de ce projet. je voudrais vous
demander certaines infomiations sur votre organisation.

 

Vous n'étes pas obliges de prendre part a cette e'tude et vous pouvez ne pas répondre a certaines
questions. Vos réponses serviront a aider l'USAID mais n'auront pas d'impact positif ou négatif sur
votre organisation. L'interview vous prendra environ une heure de temps mais nous aurions besoin
aussi d'informations complementaires sous forme de document. Acceptez-vous la procedure? [si oui,

 

remplissez les lignes ci-dessous]

l-———_———————_————-—l
l Code

Identification]

 

 

 

 

 

Enquéteur Date E =
De'but de l'interview Fin de l'interview

3 =
Superviseur

 

 

Nom de l'Organisation

 

 

 

Secteur: 21 = ASACO 22 = CVS

31 = APE publique 32 = APE privé 33 = CVE
41 = Cooperative 42 = AV 43 = Groupement
61 = CEC 62 = GGLS 71 = Groupe civique
O=Amm

 

 

Genre d'Organisation: l = Hommes ou mixte 2 = Uniquement des femmes

 

Poste occupé par enquété dans cette organisation

1 = President 2 = Vice-President 3 = Secretaire 4 = Trésorier
5 = Animateur

O = Autre

 

 

 

PVO Partenaire: l = CARE 2 = CLUSA 3 = Save-USA 4 = World Education

 

ONG Intermédiaire Partenaire

 

 

 

 

 

Depuis combien de temps collaborez-vous avec cette ONG? [Soyez aussi précis que
possible]

 

 

 

 

Rggion: 0 = Bamako 2 = Koulikoro 3 = Sikasso 4 = Ségou 5 = Mopti

 

227

228

 

Cercle: 0 = Bamako 22 = Kati 23 = Kolokani 24 = Nara
31 = Sikasso 32 =Kolondieba 41 = Ségou 42 = Macina
52 = Djenné 53 = Koro

 

 

 

 

 

 

Arrondissement [postcodification]
Future Commune [postcodification]
Ville (& Quartier) ou village [postcodification]

 

 

[Historique de l'Organizationl

 

Code

 

l.

Quelle est la mission de votre organisation? Plus clairement, dans son document de creation ou

dans la conception des adherents, elle a été crée’e pour accomplir quoi?

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2. Depuis combien de temps votre organisation existe-t-elle? [Si ne'cessaire, notez #
mois] ans
3. Depuis combien de temps votre responsable principal actuel detient-il ce poste?
ans
4. Votre bureau est constitué de combien de membres?
5. [Pour les associations mixtes] Combien de ces membres du bureau sont des femmes?
6. Combien de vos membres du bureau ont fréquenté l'école forrnelle (NB: francaise ou
franco-arabe) jusqu'a au moins la fin de la 4ieme année?
7. Combien de vos membres du bureau peuvent lire et écrire dans le bamanankan ou une
autre langue nationale? [Soulignez qu’on veut savoir 1e nombre qui arrivent a lire et
écrire]
8. Combien de vos membres du bureau n'ont atteint ni la 4ieme année de l'école francaise
ni la capacité de lire et écrire en langue nationale?
9. Gardez-vous les listes de presence de vos reunions du bureau ou des proces verbaux de
vos reunions qui donnent le nombre de presents? 0 = Non l = Oui
10. [Si oui] Puis-je avoir les listes de presence de vos reunions du bureau?
[Le nombre des membres du bureau présents a la demiére réunion]
1 1. Combien d'adhérents comptent votre organisation? [Total = bureau + adherents
simples]
12. Gardez—vous 1a liste de presence ou des proces verbaux de vos Assemblées Générales?

 

0=Non1=Oui

 

 

 

 

 

229

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

13. [Si oui] Puis-je avoir un document qui donne 1e nombre d'adhe’rents presents a la
demiere Assemblee Générale? [Nombre d'adhérents presents a la demiere AG]
14. V05 adhe’rents payent-ils des cotisations périodiques? 0 = Non 1 = Oui
15. [Si oui] A combien s'élevent les cotisations?
Montant = [par] Pe'riode = Cfa Pa!
an
16. Combien de vos adherents sont ajour dans le paiement de leur cotisations?
17. Veuillez nous montrer 1a or) vous enregistrez le paiement de cotisations.
0 = Non confirmé
l = Confirmé qu’ils enregistrent syste’matiquement le paiement de cotisations
[Auto-Gouvernance] Code
18. Actuellement, qui est considéré étre adherent de l'organisation?
O = Tout le monde dans le village ou spécifiquement un certain age. une certaine
classe ou selon le sexe.
1 = Seulement les personnes qui choissisent d'y adhérer.
19. Est-ce que vous avez un statut et reglement intérieur? 0 = Non l = Oui
20. [Si oui] Veuillez me les montrer.
0 = Non confirmé 1 = Confirmé
21. Comment constituez-vous les membres de votre bureau?
l = Nomination par les notables du village ou quartier
2 = Par consensus des membres
3 = Par vote par l'Assemblée Générale pour une pe'riode qui n'est pas précisee
4 = Par vote par l'Assemble’e Générale pour une pe'riode bien détenninée
0 = D'autre modalité [expliquezl
22. Quelle est la fréquence des reunions du bureau?
7 = Pas de fre’quence fixe (autant que ne'eessaire)
6 = Plus d'une fois par semaine
5 = Une fois par semaine
4 = Plus d’une fois par mois, mais pas de facon hebdomadaire
3 = Une fois par mois
2 = Au moins une fois par semestre
1 = Une fois par an
23. Quelle est la fréquence de vos assemblées générales?
7 = Pas de fre'quence fixe (autant que nécessaire)
5 = Une fois par semaine
4 = Plus d'une fois par mois, mais pas de facon hebdomadaire
3 = Une fois par mois
2 = Au moins une fois par semestre
1 = Une fois par an
0 = Jamais

 

 

 

 

 

[Pratiques du Management]

Code

 

 

 

 

 

 

 

 

 

 

 

 

 

 

24. Votre organisation est-elle officiellement reconnue par l'administration Malienne?
0 = Non l = Oui
25. [Si oui] Veuillez me montrer votre récépissé ou autre document officiel.
0 = Non confirrne
1 = Reconnaissance par l’administration locale confirmée
2 = Reconnaissance par l’administration nationale confirmée
26. Enregistrez-vous les e'critures comptables? 0 = Non 1 = Oui
27. [Si oui] Veuillez me montrer votre systeme d'enregistrement comptable.
0 = Non confirmé l = Confirrne’
28. Avez—vous fixé pour votre organisation des objectifs a atteindre dans les cinq (5) ans a
venir?
0 = Non 1 = Oui
29. [Si oui] Quels sont vos objectifs?
a.
b.
0 = Zero ou un objectif donne
1 = Deux objectifs concrets dont les membres de l'organisation ou de la
communauté en géne’ral bénéficieront
30. [Si l'organisation est hommes ou mixte] Quelle est l'impact de votre organisation sur la
communauté?
0 = Pas de preuve d'analyse d'impact sur lafemme
1 = Preuve des précédentes analyses d'impact sur lafemme
31. En plus de votre ONG interrnédiaire partenaire et/ou vos cotisations périodiques, avez-
vous d'autres sources de revenus?
0 = Non l = Oui
32. [Si oui] Veuillez me citer une ou deux importantes sources de revenus en plus d'une

 

aide quelle conque de IEnquéteur: Prononcez le nom de l'ONG partenaire nomme’e sur

la premiere pagel et de vos cotisations.

[Explez d’autres partenaires, activites lucratives, credit, etc.]

a.

 

b.

 

0 = Non confirmé l = Confinné
[Enquéteun C'est confirmé si au moins une autre source existe]

 

 

 

k)
b)

 

[Action Civique]

 

Code

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

activites?
O = Pas mention des adherents simples 1 = Les adherents simples ontparticipé

 

33. Vous avez ici une liste de certaines actions que les organisations font. Pour chacune d'elles.
veuillez me dire si votre organisation a exécuté cette activité durant les 12 demiers mois.
[Codifiez de a a d comme suit:

0 = Jamais
1 = Une fois ou plus]

a. Avez-vous contacté votre chef d'arrondissement ou votre mairie concemant un
probleme de développement? (Exple: Eau. education, credit, agriculture, sante’. etc.)

b. Avez-vous contacté votre commandant de cercle ou les services locaux des ministeres
concemant un probleme de développement?

c. Avez-vous contacte’ les agents d’une societé mixte comme l'OHVN, la CMDT, l'Office
du Niger, l'Opération Riz. ou ODEM pour un probleme de développement?

d. Avez-vous contacté votre député concemant un probleme de développement?

34. [S'il y a eu des cas de contacts dans les réponses a la Q33; si non, allez a la 0.40] Rappelez-vous
et identifiez le probleme le plus complexe pour lequel votre organisation a contacté
l'administration durant les 12 demiers mois. Expliquez.

35. [Par rapport a la 0.34] Comment avez-vous démarché et quel a été le resultat de votre action?
[Enquéteurz Ne guide: pas. Relevez simplement toutes les démarches entreprises par la
personne sondée]

a. "Nous avons analyse un Probleme ..." 0 = Non mentionné 1 = Mentionné

b. "L'organisation a proposé une solution ..." 0 = Non mentionné 1 = Mentionné

c. "...Nous avons forrnule' un plan d'action..." 0 = Non mentionné 1 = Mentionné

d. "Nous avons travaille avec des groupements ou entite’s en déhors de notre
organisation..."

0 = Non mentionné 1 = Mentionné
36. [Toujours par rapport a la Q34] Qui dans votre organisation ont pris part a ces

 

 

 

 

L)
DJ
IJ

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

37. [S'il y a eu des cas de contacts dans les re'ponses a la Q33] En vous référant sur votre
discussion de vos contacts avec les dirigeants de l'administration de l'arrondissement
ou commune. du cercle. des societes mixtes ou d’autres autorite’s: Combien de fois
dans les 12 demiers mois votre organisation 3 pu convaincre les autorités afin d'obtenir
un changement des decisions prises ou d'aboutir a un resultat concret?

0 = Jamais
1 = Une fois
2 = Deux ou plusieurs fois

38. [8'11 cite au moins un cas d'influence] Citez des exemples prouvant comment votre
action a pu influencer au moins une decision de de'veloppement.
a.

b.
0 = Pas d'exemple concret cite’
l = Au moins un exemplc concret cite

39. Veuillez me montrer n'importe quel document de'crivant clairement la nature de cette
activité. 0 = Non confirmé l = Confinné
[S'ils ont pu influencer au moins une decision]

40. Parfois des organisations de base collaborent avec des organisations intermédiaires

a. telles que les ONG et les federations. Est-ce qu'au moins une oganisation
interme'diaire re'pre'sente vos inte’réts aupres des autorités?

0 = Non l = Oui

40. [Si oui] Quelle(s) ONG ou féde’ration(s)? [Nommez-les]

b.

41. [Si Oui] Comment estimez-vous que cette (ces) organisation(s) réprésente vos intéréts

a. aupres des autorités?

0 = Aucun partenaire represente les inte'réts de l'OC avec efficacite’
1 = Au moins un partenaire réprésente les intéréts de I'OC avec efficacité

41. Votre organisation sert combien de villages? [Nombre de villages]

b.

42. Connaissez-vous des nouvelles federations cre'e'es pendant les 12 demiers mois pour

 

apporter les attentes de la population aupres de I'administration?
0 = Non 1= Oui

 

 

 

 

 

43.

[Si oui] Veuillez me les nommer.

 

 

 

 

 

 

[Droits et Devoirs]

 

Code

 

Pour qu'une organisation communautaire puisse faire son travail, elle a besoin de travailler avec les

services de l'administration locale et les élus locaux . En principe, vous avez certains droits et devoirs
qui facilitent cette interaction. Aux questions suivantes, repondez par rapport a votre connaissance du
systeme de la Décentralisation. 0 = Non 1 = Oui

 

 

 

 

44. Bien que mon enfant remplisse tous les criteres de scolarisation. l'Etat a-t-il 1e droit
de lui refuser l'entree a l'e'cole?

45. L'Etat a-t-il 1e droit de refuser la reconnaissance légale a true organisation?

46. L'Etat peut-t-il ge’rer la TDRL comme il veut?

47. Avez—vous le droit d'assister aux sessions ordinaires du conseil communal?

 

 

 

IDéveloppement Communautaire: Collaboration avec I'Administration]

Code

 

48.

Parfois les organisations communautaires travaillent en partenariat avec les projets,
les ONG, I’administration, les sociéte’s mixtes, ou d’autres organismes afin de donner
des biens publics a la communauté. Par exemple, des organisations communautaires
peuvent recevoir de l’aide en forme d’assistance financiere, matérielle ou du
personnel pour leur perrnettre d'aider leurs villages ou leurs communautés. De leur
part ces organisations contribuerrt en main d’truvre, des cotisations ou autres.

Pendant les 12 demiers mais, avez-vous collaboré avec un organisme quelquonque
dans la prestation de services publics?
0 = Non 1 = Oui

 

 

49.

 

[Si Oui] Donnez-nous des exemples sur comment vous étes engages dans ce genre de
collaboration et avec qui. [12 demiers mois]

 

 

 

0 = Pas d'exemple concret cité de partenariat avec les autorités ou les services de
1 'état
1 = Au moins un exemplc concret cité d’une telle collaboration avec l 'état

 

 

 

234

 

 

Veuillez me montrer des documents que vous avez qui peuvent clarifier la nature de
cette interaction.
0 = Non Confirme’ 1 = Confirmé

 

 

lDéveloppement Communautaire: Prestation de Services et Renforcement de
Capacité]

 

Code

 

Au de'but de cette interview. vous avez parle' de la mission de votre organisation. [Enquéteurz
Faites un rappel a l'enquéte de sa réponse a la 0.1] Maintenant. j'aimerais discuter les activite’s que
vous menez et leur rapport avec cette mission.

 

51.

Veuillez citer des activites ou services de développement que votre organisation a foumi pendant
les 12 demiers mois (Dans les domaines tels que: sante’. eau. education. crédit. agriculture.

se'curité. etc.).

 

 

 

 

C.

 

 

52.

[8'11 y a des activites ou services] Panni les activites ou services mentionne’s ci-dessus
y-a-t-il des activites qui ont démarre'es au cours des 12 demiers mois?
0 = Non 1 = Oui [Si non. allez a la 0.55]

 

53.

[Si oui] Quels sont les nouvelles activites ou nouveaux services de développement
qui sont clairement indiqués dans le cadre de la mission de votre organisation?

 

 

 

0 = Pas de nouveaux services dans le cadre de la mission
1 = nouveaux services dans le cadre de la mission.

 

 

54.

 

[Si oui a la 0.52] Quels sont les nouveaux services ou activites qui sont hors de la
mission de votre organisation?

 

 

 

0 = Pas de nouveaux services hors de la mission
I = Nouveaux services hors de la mission.

 

 

 

l J
OJ
L]!

 

 

 

 

 

 

 

55. Les membres de votre bureau ou les adherents simples de votre organisation ont-ils
suivi des formations durant les 12 demiers mois? 0 = Non l = Oui [Si non.
l’interview prend fin]

56. [Si oui] Specifiez ces formations, s'il vous plait? [0uoi. et par qui?]

57. Comment ces formations ont été utiles pour votre organisation?

 

 

 

0 = Aucune indication de changement au niveau de l'organisation ou d'utilisation
des connaissances.

1 = Evidences d'au moins un changement organisationnel ou d'utilisation des
connaissances apprises.

 

 

 

Merci beaucoup. Vos réponses ont été tres utiles.
[Enquéteurz Veuillez écrire sur la premiere page l'heure a laquelle l'interview a pris fin.]

[Enquéteurz Veuillez noter ci-dessous toutes les observations qui vous semblent interessantes ou les
commentaires foumis volontairement par l'enquété. Veuillez e'crire ces commentaires en indicant le

nume’ro de la question. Utilisez le verso de la feuille si nécessaire]

 

 

 

 

 

APPENDIX B

Political Attitudes Questionnaire

236

Numero du Village Numero de l'Enquete

 

Institut Malien de Recherches Appliquées au Développement (IMRAD)
Questionnaire sur les Attitudes Civiques

Bonjour. Je m'appelle [Enqueteun Dites votre nom]. Je suis de l'IMRAD. Je ne représente ni le gouvernement. ni
aucun parti politique. L'IMRAD est un institut privé qui veut recenser les points de vue du monde paysan a propos des
organisations rurales et du systerne de gouvernement dans notre pays. Etant agent de ce projet. j'aimerais vous poser
quelques questions.

II n'y a pas de réponse correcte ou fausse -- toute reponse sera la bienvenue.

La participation au sondage est volontaire et nul ne sera term a répondre obligatoirement a toutes les questions. Personne
ne sera inquiétée pour ses opinions car les noms seront tenus dans l‘anonymat. et les réponses strictement confidentielles.
Cela vous permettra de parler avec franchise. Cet interview prendra a peu pres une heure de temps. Voulez-vous

continuer?

[Si oui. remplissez le tableau ci-dessous.]

 

 

 

 

 

 

 

 

 

Date Nom de l'Enquéteur
Region Cercle Arrondissement
Village Grande Famille (Gwa) Assoc. Conseil Exploit. ‘

 

 

 

[Vie Associative]
1. Etes-vous membre d'une Association Villageoise (AV)?
Oui Non_

2. [S'il est membre] En quelle année étes-vous devenu membre de cette organisation? (année)

3. Lequel des postes. membre dc bureau ou animateur, tenez-vous au sein de cette organisation?
1. A partir de (année)
2. A partir de (année)

 

4. Quels sont les avantages de cette organisation sur vous?

 

 

 

 

5. Attendez-vous mieux d'elle?

 

 

6. [S'il est membre de l'AV] Dans les demiers 12 mois, a combien de reunions de votre AV avez-vous été present?
Toutes les réunions_ La plupart_ certaines_ aucune_

7. Etes-vous membre du SYCOV (1e Syndicat des Cotoniers et Vivriers)? Oui_ Non

8. [Si non] Pourquoi pas?

 

 

 

9. [S'il est membre on a été membre du SYCOV auparavant] Lequel poste, membre de bureau ou animateur, tenez-vous
actuellement ou teniez—vous auparavant au sein du SYCOV?

1 . Quand?_-_(années)

2. Quand?_o_(années)

237

238

10. Que fait le SYCOV pour les producteurs?

 

 

 

ll. Vous attendez-vous encore a quelque chose de mieux de lui?

 

 

 

l2. Etes-vous membre d'autres organisations communautaires tell que des groupes religieux ou sportifs ou un grin ou
un groupement d'intérét économique (GIE)?
Oui Non

13. [Si oui] Quelles organisations? Tenez-vous des postes dans ces organisations?

 

 

 

1. Organisation: a partir de (année) Poste
2. Organisation: a partir de (année) Poste
3. Organisation: a partir de (année) Poste

 

l4. Dans les demiers 12 mois. a combien de reunions de votre (vos) organisation(s) avez-vous été present?
Toutes les réunions__ La plupart__ certaines_ aucune_

[Autorité Politique]
Nous vous rappelons une fois de plus qu'il n' ya pas de réponses correctes ou fausses. Nous nous interessons tout
simplement a vos points de vue.

15. Peut-on affirmer aujourd'hui que l'autorité de l'état est respectée au Mali?
Oui. tout a fait_ Peut-étre_ Je ne sais pas_ Pas tellement_ Non, pas du tout_
16. Pensez-vous que les femmes doivent avoir 1e meme droit de voter que les hommes lors des elections?
Oui. tout a fait_ Peut-étre_ Je ne sais pas_ Pas tellement_ Non. pas du tout_
17. Les forces de l'ordre ont-elles trop de pouvoir dans ce pays?
Oui, tout a fait_ Peut—étre_ Je ne sais pas_ Pas tellement_ Non, pas du tout_
I8. Doit-on Iaisser voter quelqu'un qui ne comprend pas tous les enjeux?
Oui, tout a fait_ Peut—étre_ Je ne sais pas_ Pas tellement_ Non, pas du tout_
19. Les postes d'élus. doivent-ils étre le seul monopole des hommes?
Oui, tout a fait_ Peut-étre_ Je ne sais pas_ Pas tellement_ Non. pas du tout_
20. Doit-on permettre a un agent des forces de l’ordre de tirer sur quelqu'un qui fuit le lieu d'un crime d'une maniere
suspecte?
Oui. tout a fait_ Peut-étre_ Je ne sais pas_ Pas tellement_ Non, pas du tout_
21. Les commercants devraient—ils avoir 1e droit de bastonner un voleur attrappé en flagrant délit?
Oui. tout a fait_ Peut-étre_ Je ne sais pas_ Pas tellement_ Non, pas du tout_
22. Le gouvernement doit-il detenir un citoyen et surseoir a son proces indéfiniment?
Oui, tout a fait_ Peut-étre___ Je ne sais pas_ Pas tellement_ Non, pas du tout_
23. Pour mieux faire son travail. le gouvernement doit-il dormer plus de postes publics aux jeunes cadres?
Oui, tout a fait_ Peut-étre_ Je ne sais pas_ Pas tellement_ Non. pas du tout_

[Responsabilite’ Politique]

24. Les pots-de-vin sont-ils rares de nos jours chez les fonctionnaires et les politiciens maliens?
Oui, tout a fait_ Peut-etre_ Je ne sais pas_ Pas tellement_ Non, pas du tout_

25. L'appartenance ethnique a-t-elle un impact sur le succes de quelqu'un dans la politique ou le gouvernement?
Oui, tout a fait_ Peut-étre_ Je ne sais pas_ Pas tellement_ Non, pas du tout_

26. Est-i1 correct qu'un ministre aide son village natal avec des projets de développement?
Oui, tout a fait_ Peut-étre_ Je ne sais pas_ Pas tellement_ Non, pas du tout_

27. La preoccupation majeure de la plupart des fonctionnaires et des politiciens est-elle de s'enrichir?
Oui, tout a fait_ Peut-étre_ Je ne sais pas_ Pas tellement_ Non. pas du tout_

28. La region du president beneficie-t-elle de plus d’apports gouvernementaux que les autres regions?
Oui, tout a fait_ Peut~étre_ Je ne sais pas_ Pas tellement_ Non, pas du tout_

 

 

 

29.

239

La corruption était-elle pire sous l‘ancien gouvernement UDPM que maintenant?
Oui. tour a fait_ Peut-etre_1e ne sais pas_ Pas tellement_ Non. pas du tout_

[Savoir Economique]

30.

31

Vous débrouillez-vous mieux aujourd'hui qu'il y a cinq ans? Oui Non

Si oui. pourquoi“? Si non. pourquoi pas?

 

 

 

Que signifie la devaluation?

 

 

 

33

34

Economiquement. votre famille se porte-t-elle mieux, pire. ou de la meme maniere qu'avant la devaluation?
Mieux_ Pire_ La méme_

Pourquoi nous avons subi urie devaluation?

 

 

35

procurer de l'argent pour les financer. A votre avis. d'ou doit venir cet argent?

Si on veut améliorer des services d'intérét public (routes. santé. eau et assainissement). 1e gouvernement doit se

 

 

 

[Savoir Politique]

 

 

 

 

 

 

 

 

 

 

 

 

36. Selon vous. quel est le r61e de l'Assemblée Nationale?

37. Avez-vous déja entendu parler de la Mission de Décentralisation? Oui Non

38. [Si oui] Quelle est sa tache'.’

39. Au Mali, y a-t-il une difference entre un parti politique et un gouvernement ou sont-ils plutét les memes?
Différent_ Méme_ Je ne sais pas_

40. La chefferie traditionelle doit-elle jouer un réle dans la facon de gouverner le Mali d'aujourd'hui?
Oui Non Je ne sais pas

41. [Si oui] Quel rOIe?

42. Ecoutez-vous les inforrnations radio-diffusées? Oui Non_

43. [Si oui] Comment écoutez-vous ces informations?

Tous les jours_ Plusieurs fois par semaine_ Une fois par semaine_
Une fois par mois_ Moins qu'une fois par mois_

240

44. Quelle(s) station(s) de radio e'coutez-vous habituellement‘.’ [Dites a l'enquété de specifier]

 

 

 

45. Suivez-vous les informations ORTM televisees? Oui Non

46. [Si oui] Comment regardez-vous les infos télévise’es?
Tous les jours_ Plusieurs fois par semaine_ Une fois par semaine_
Une fois par mois_ Moins qu'une fois par mois_

47. Depuis 1e 26 mars 1991, avez-vous lu les journeaux nationaux? Oui Non

 

48. [Si oui] Lesquels? [Dites a l'enquété de specifier]
Combien de fois par mois lisez-vous les journaux? Fois

 

[Confiance Politique]
[Enquéteun Montrez l'échelle a dix points et expliquez-le]

 

18 NM PAS DU TOUT J'AI mm
CONFIANCE EN EUX (ammo: EN EIJX
ll 1 2 3 4 5 6 7 8 9 10 ll

 

 

 

 

 

 

 

 

 

 

49. Sur l'échelle de un a dix. quel degré de confiance avez-vous des groupes suivants?

[Enquéteurz Donnez un score numérique de un 21 dix. NSP = Ne sait pas]

Votre famille immediate _ Des Maliens de la Région de Gao _ Des Algeriens _
Vos parents hommes __ " de la Région de Kayes _ Des Burkinabé _
Vos parentes femmes _ " de la Région de Kidal _ Des Ghanéens _
Vos voisins _ ” de la Région de Koulikoro _ Des Guinéens _
De votre groupe ethnique _ " de la Région de Mopti _ Des Ivoiriens _
Votre député _ " de la Région de Ségou _ Des Mauritaniens _

" de la Région de Sikasso Des N igeriens

 

" de la Région de Tombouctou Des Sénégalais

Des Sierra
Leonais

Des Francais

241

50. Sur l'éclielle de un a dix. quel degre’ de confiance avez-vous des institutions suivantes?

[Enqueteurz Donnez un score nume’rique de un a dix. NSP = Ne sait pas]

La police _ La Banque Mondiale _ L’IMRAD _
L‘arrondissement _ Les syndicats urbains _ La Radio ORTM _
L'Assemblée Nationale _ La CMDT _ La Radio Kayira _
Les ministres du _ Le SYCOV _ RFI _
gouvernement

Les organisations _ Le Plan International _ La Voix de __
Islamiques l'Amérique

Les organisations _ Le Journal Essor _
Chrétiennes

[Efficacité Politique]
Voici quelques affirmations. Choisissez A ou B selon votre approbation.

51. A. D'habitude. je travaille mieux dans un groupe.
B. D'habitude. je travaille mieux seul. A_ B_
52. A. Je foumis plus d'effort pour améliorer ma propre vie.
B. Je foumis plus d'effort pour améliorer la vie de mes enfants et d'autres jeunes parents.

A B_

53. A. II n'est pas sage de planifier a tres long terme car beaucoup de choses dependent du destin.
B. Je planifre a long terme car j'ai confiance dans ma capacité a realiser des projets.
A_ B_
54. A. Lors des discussions politiques. je parviens a influencer mes amis et voisins.
B. Mes amis et voisins ne m'écoutent pas quand il s'agit de débats politiques.
A_ B_
55. A. Souvent la gestion gouvernementale est telle que je ne peux la bien comprendre.
B. La gestion gouvernementale est telle qu'elle est généralement comprehensible pour les gens comme moi.
A_ B__
56. A. Pour realiser des objectifs communautaires, il est mieux que l'un des membres de la communauté touche
individuellement des responsables appropriés.
B. Pour réaliser des objectifs communautaires, i1 vaut mieux créer un comité charge de poser publiquement
les problémes a qui de droit.
A_ B_
57. A. En tant que communauté, nous arrivons habituellement a nous faire écouter par nos députés.
B. En tant que communauté, nous n'arrivons pas d’habitude a nous faire écouter par nos députés.
A_ B_
58. A. Les fonctionnaires et politiciens sont a mesure de résoudre les problemes auxquels 1e Mali est confronté

actuellement.
B. Les fonctionnaires et politiciens ne sont pas tres efficaces dans la mise en oeuvre des programmes qui
peuvent résoudre des problémes nationaux.
A_ B_
[Préférences de I'Intérét Public]
59. A. Nous nous débrouillions mieux avant Ia devaluation, avec des bas prix aux producteurs, des bas prix
d'intrants, et des bas prix des biens de consommation.
B. Nous nous débrouillons mieux actuellement, avec des meilleurs prix aux producteurs, des prix d'intrants
plus élevés. et les prix des biens de consommation plus élevés.
A B

60. A. Le gouvernement doit embaucher beaucoup de jeunes diplomés sans emploi meme si cela revient cher au
pays. .
B. Les moyens limités du pays ne permettent pas d'embaucher et le poussent au licenciement. A__ B

 

 

61.

242

A. Le gouvernement devrait augmenter les bourses des eleves et étudiants.

B. Le gouvernement doit plutét financer l'éducation primaire. A_ B

A. La CMDT est justifie d'avoir 1e monopsone de l'achat du coton au Mali.

B. Les cultivateurs devraient avoir 1e droit de vendre leur coton oii ils veulent et a qui ils veulent.
A B

[Tolerance Politique]

63.

64.

66.

67.

68.

69.

70.

A. Si les gens ont des points de vue différents des miens. ils devraient avoir 1e droit de les exprimer.
B. 11 est dangereux d'avoir plusieurs points de vue car cela pre'te a confusron.
A_ B_
A. Si des gens veulent une organisation communautaire, ils doivent s'affilier au parti politique au pouvoir.
B. Si des gens veulent une organisation communautaire. ils doivent avoir 1e droit de le faire independamment.
A_ B_
A. Bien que le Mali soit un pays oir les musulmans sont majoritaires. les Chrétiens devraient étre permis de
creer un parti politique Chrétien.
B. Les Chrétiens ne doivent pas étre permis de creer un parti politique Chrétien.
A_ B_
A. Les groupes ethniques et religieux devraient avoir le droit de créer leur propre parti politique.
B. Les partis politiques a base ethnique ou religieux divisent 1e peuple et doivent étre interdits.
A_ B_
A. Les moyens violents sont parfois justifies au Mali pour realiser des fins politiques.
B. Les moyens violents ne sont jamais justifies pour les fins politiques au Mali.
A_ B_
A. La meilleure forme de gouvernement est celle élue par le peuple.
B. La meilleure forme de gouvernement est celle qui est efficace.
A_ B_
A. Au Mali actuel, nous avons un vrai choix des partis politiques et des candidats.
B. Au Mali actuel, nous sommes en voie de retourner dans un e'tat a parti unique.
A_ B_
A. Tendre la main aux adversaires est dangereux car cela peut amener a trahir ses alliés.
B. Pour la paix sociale. il est essentiel de s'ouvrir aux adversaires.

A_ B_

[Satisfaction de Vie/ (Post) Materielisme]
71. Comment étes-vous satisfait:

de votre travail“?

Tres satisfait_ Assez satisfait_ Pas tellement satisfait_ Pas du tout satisfait_
de votre situation financiere?

Tres satisfait_ Assez satisfait_ Pas tellement satisfait_ Pas du tout satisfait_
de votre état de santé?

Tres satisfait_ Assez satisfait_ Pas tellement satisfait_ Pas du tout satisfait_
en général, avec la vie que vous menez?

Tres satisfait_ Assez satisfait_ Pas tellement satisfait_ Pas du tout satisfait_

72. Quand vous regardez votre vie d'aujourd'hui. comment étes-vous satisfait:

par rapport a votre vie d'il y a un an?
Beaucoup plus satisfait_ Un peu plus satisfait_ Un peu moins satisfait_ Beaucoup moins satisfait_

par rapport a votre vie il ya cinq ans?
Beaucoup plus satisfait_ Un peu plus satisfait_ Un peu moins satisfait_ Beaucoup moins satisfait_

243

73. Quand vous regardez devant vous. face a votre avenir et celui de vos enfants. comment prévoyez-vous votre
satisfaction:

darts un an?
Beaucoup plus satisfait_ Un peu plus satisfait_ Uri peu moins satisfait_ Beaucoup moins satisfait_

dans cinq ans?
Beaucoup plus satisfait_ Un peu plus satisfait_ Un peu moins satisfait_ Beaucoup moins satisfait_

74. De la liste suivante des objectifs. lequel choisiriez-vous comme etant la priorité principale pour le pays dans les dix
ans qui viennent? Lequel chorsiriez-vous comme la deuxieme priorite du pays dans les dix ans qui viennent?
[Enquéteurz Ecrivez quel objectif l'enqueté choisit comme premiere priorité et lequel il choisit comme deuxieme]

A. Maintenir l’ordre dans le pays

B. Donner au peuple plus a dire dans les decisions du gouvernement

C. Combattre les rehaussements des prix

D

E

 

 

 

. Proteger la liberté d'expression
. Je ne sais pas [peut (are 1‘ choix, 2‘ choix. ou tous les deux]

 

 

75. Le crime est-i1 un grand probleme dans votre vie? Oui Non

76. [Si oui] Dans quelle maniére avez-vous change votre vie a cause du crime?

 

 

 

77. Quelles sont des choses que I'ADEMA a mieux fait que I'UDPM?

 

 

 

78. Quelles sont des choses que l'ADEMA a fait pire que I'UDPM?

 

 

 

79. Quelle est votre evaluation générale de la performance du gouvernement ADEMA ces demiers trois ans?
Tres bien_ bien_ passable__ mauvais_

[Participation Politique]
80. Etes-vous inscrit pour voter? Oui Non

81. [Si non] Pourquoi pas?

 

 

 

 

82. Avez-vous voté dans le deuxi‘eme tour de l'élection presidentielle en avril 1992? Oui Non

83. [Si non] Pourquoi pas?

 

 

 

 

84. Avez-vous l' intention de voter dans les elections municipales prévues par la Mission de De'centralisation pour l'année
prochaine?
Oui Non

 

 

85. [Si non] Pourquoi pas?

244

 

 

 

 

 

86. Comment souvent discutez-vous de la politique avec les gens?

87. Supportez-vous un parti politique? Oui_ Non

88. [Si oui] Lequel?

Tous les jours_ Plusieurs fois par semaine_ Une fois par semaine_
Une fois par mois_ Moms souvent qu'une fois par mois_

 

89. Avez-vous une carte de membre pour ce parti? Oui Non

90. Pouvez-vous dire. a partir de la liste suivante. l'importance que vous attachez a l'un ou l'autre groupe ou personne

cites?

Votre famille Trés important_ assez important_ pas important_
Votre groupe ethnique Tres important_ assez important_ pas important_
Votre conseil de village Tres important_ assez important_ pas important_
Votre chef d'arrondissement Tres important_ assez important_ pas important_
Votre (vos) député(s) Tres important_ assez important_ pas important_

91. Pendant les demiers cinq ans, avez-vous participé a une ou plusieurs des activités suivantes? Dans l'avenir. pourriez-
vous participer a une ou plusieurs de ces activites?

Assister a une reunion communautaire
Déja fait_ Pourrait faire_ Ne le ferait jamais_ Ne sait pas_

Assister a une reunion d'élection
Déja fait_ Pourrait faire_ Ne 1e ferait jamais_ Ne sait pas_

Travailler pour un candidat ou un parti
Déja fait_ Pourrait faire_ Ne le ferait jamais_ Ne sait pas_

Aller chez un chef coutumier pour demander son aide a resoudre un probleme
Déja fait_ Pourrait faire_ Ne le ferait jamais_ Ne sait pas_

Aller chez un Chef d'arrondissement pour demander son aide pour resoudre un probleme
Déja fait_ Pourrait faire_ Ne le ferait jamais_ Ne sait pas_

Aller chez un député pour demander son aide pour resoudre un probleme
Déja fait_ Pourrait faire_ Ne le ferait jamais_ Ne sait pas_

Signer une petition adressée au gouvernement on a ses agents
Déja fait_ Pourrait faire_ Ne le ferait jamais_ Ne sait pas_

Participer a une manifestation paisible
Déja fait_ Pourrait faire_ Ne le ferait jamais_ Ne sait pas_

Participer a true manifestation violente
Déja fait_ Pourrait faire_ Ne le ferait jamais_ Ne sait pas_

92. [Enquéteurz Identifiez l'action la plus engagée] Que vouliez-vous réclamer par cette activité?

 

 

245

["Background" Social ]
J'aimerais conclure cet interview avec quelques questions sur vous et votre menage. Rappelez-vous que nous n'allons
pas demander vorre nom. donc vous pouvez vous sentir libre a re’pondre franchement ct ouvertement.
93. [Enquéteurz Remplissez le sexe de l'enquéte] Hornme Femme
94. Etes-vous marie? Oui Non
95. Quel est votre groupe d'age‘.’

Moins de 20 20-moins de 30 30-moins de 40 40-moins de 50

50-moins de 60 6041101115 de 70 70-moins de 80 80+

96. Quelle est votre ethnic?

 

97. Quelle est votre religion?

 

98. Quel est le nom du village natal de votre grand-pére paternel? [Enquéteurz Si l'enquetée est une femme mariee, quel
est le village natal du grand-pere paternel de son mari?]

99. Combien d'années de d'ecole avez-vous faites (par exemple. medersa. l'école francaise. ou des cours
d'alphabetisation fonctionelle?

Medersa ans
L'école francaise ans
Alphabétisation fonctionnelle ans

100. Pouvez-vous lire et écrire dans une langue (le francais ou une langue nationale)?
Oui Non

101. [Si oui] Quelle(s) langue(s)?

 

 

 

102. Etes-vous membre de vmre conseil de village?
Oui_ Non_
103. Cultivez-vous?
Oui_ Non_
104. [Si oui] Cultivez-vous du coton?
Oui_ Non_
105. [Si oui] Combien d'hectares de coton votre famille produit—elle?
_ hectares

106. Parmi les choses suivantes. lesquelles appartiennent au membres de votre exploitation?

Un multiculteur? Our_ Non_
Une charrue? Oui_ Non_
Une mobylette? Oui_ Non_
Une télévision? Oui_ Non

107. Combien de personnes logent dans votre ménage? personnes

108. Dans les demiers douze mois, avez-vous eu un boulot rémuneré?
Oui_ Non_
109. [Si oui] Quel sorte de boulot? Ot‘i?

 

110. Avez-vous déja voyage hors de votre village? Oui_ Non

 

 

 

246

11 1. [Si oui] Oti avez-vous voyage? Dans ma region A Bamako
Ailleurs au Mali [Dites 00.]

Ailleurs en Afrique [Dites ou.]

Hors d'Afrique [Dites 00.]

1 12. Avez—vous quelque chose a ajouter?

 

 

 

 

 

 

 

MERCI BEAUCOUP DE VOTRE AIDE.

Bibliography

247

Bibliography

Abramson. Paul R. And Ronald Inglehart. 1995. Value Change in Global Perspective.
Ann Arbor: University of Michigan Press.

Allison, Graham T.. Jr.. and Robert P. Beschel, Jr. 1992. "Can the United States
Promote Democracy?" Political Science Quarterly. 107: 81-98.

Almond. Gabriel A. and Sidney Verba. 1963. The Civic Culture: Political Attitudes and
Democracy in Five Nations. Princeton. NJ: Princeton University Press.

Almond, Gabriel A. and Sidney Verba. 1965. The Civic Culture: Political Attitudes and
Democracy in Five Nations (Abridgement of 1963 edition). Boston: Little,
Brown.

Almond, Gabriel A. and Sidney Verba. 1980. The Civic Culture Revisited. Boston:
Little, Brown.

Alreck, Pamela L. And Robert B. Settle. 1995. The Survey Research Handbook, 2nd
Ed. Chicago: Irwin.

Amselle and Benhamou. J. Réactualisation de lafilie‘re coton au Mali. Paris: Minister of
Foreign Affairs. 1985.

Anheier, Helmut K. 1987. "Indigenous Voluntary Associations, Nonprofits, and
Development in Africa," In Walter Powell, ed., The Nonprofit Sector: A Research
Handbook. New Haven: Yale University Press, pp. 416-433.

Associates in Rural Development. Inc. (ARD). "Governance in Democratic Mali: An
Assessment of Transition and Consolidation and Guidelines for Near-Term

Action." Final Draft Report to USAID-Mali, July 1994.

L 'Autre Afiique. 24-30 June 1998. "Mali: L’Adema se Taille la Part du Lion aux
Municipales." pp. 24-25.

L ’Autre Afrique. 17-23 June 1998. "Mali: Dialogue des Sourds." pp. 26-27

L ’Autre Afrique. 17-23 June 1998. "Le Mali et ses Partenaires Revoient Leur Copie."
pp. 48-50.

248

249

Axelrod. Robert. 1986. "An Evolutionary Approach to Norms." American Political
Science Review. Vol. 80 (December).

Azarya. Victor. 1994. "Civil Society and Disengagement in Africa." In John W.
Harbeson. Donald Rothchild. and Naomi Chazan. eds.. Civil Society and the State
in Africa. Boulder. CO: Lynne Rienner.

Bader. C. "Au Mali. des paysans s’organisent." Le Monde Diplomatique. September
1992.

Banfield. Edward C. 1958. The Moral Basis ofa Backward Society. Glencoe. IL: The
Free Press.

Barry. Brian. 1978. Sociologists. Economists. and Democracy. Chicago: University of
Chicago.

Bates, Robert H. 1990. "The Political Framework for Agricultural Policy Decisions." In
Carl Eicher and John Staatz. eds. Agricultural Development in the Third World,
2nd ed. Baltimore, MD: Johns Hopkins Press, pp. 154-159.

Bates. Robert H. 1981. Markets and States in Tropical Africa: The Political Basis of
Agricultural Policies. Berkeley: University of California Press.

Bayart. Jean-Francois. "Civil Society in Africa." In Patrick Chabal, ed., Political
Domination In Africa: Reflections on the Limits of Power. Cambridge, UK:
Cambridge University Press, 1986. pp. 109-125.

Bayart, Jean- Francois. 1993. The State In Africa: The Politics of the Belly. London:
Longman.

Bazin, Jean. 1985. "A chacun son Bambara,‘ in Jean-Loup Amselle and Elikia M'Bokolo
(editors), Au Coeur de l'ethnie. 87-127.

Beck, Linda. 1996. "Democratization and the ‘Hidden" Public Sphere: Continuity and
Change in the Participation of Senegalese Women in Patronage Politics." Paper
Presented at the Annual Meeting of the American Political Science Association.

Belloncle, Guy. 1982. La question paysanne en Afrique noire.

Bermeo, Nancy. 1992. "Democracy and the Lessons of Dictatorship." Comparative
Politics. (April) pp. 273—291.

250

Bingen, R. James. 1994. "Agricultural Development Policy and Grassroots Democracy
in Mali: The Emergence of Mali's Farmer Movement." African Rural and Urban
Studies. 1:1, pp. 57-72.

Bingen. R. James. 1998. "Cotton. Democracy, and Development in Mali." The Journal
of Modern African Studies. 36:2, pp. 265-285.

Blair. Harry. "Donors, Democratisation and Civil Society: Relating Theory to Practice."
In NGOs, States, and Donors: Too Close for Comfort? London: MacMillan,
1997. pp. 23-42.

Bowser, Georgia, Joseph Temba, Philip Alderfer, and Michael Bratton. 1997. "The
Impact of Civic Education on the Knowledge, Values, and Actions of Zambian
Citizens." Report Prepared for USAID/Zambia. East Lansing: Michigan State
University (February).

Bratton, Michael. 1989. "Beyond the State: Civil Society and Associational Life in
Africa." World Politics. 41, 3: 407-430.

Bratton. Michael. 1994a. "Civil Society and Political Transition in Africa." IDR
Reports. Boston: Institute for Development Research, Vol. XI, No. V

Bratton, Michael. 1994b. "Micro-Democracy? The Merger of Farmer Unions in
Zimbabwe." African Studies Review. 37, 1: pp. 9-37.

Bratton, Michael, and R. James Bingen. 1994. "Farmer Organization and Agricultural
Policy in Africa -- Introduction." African Rural and Urban Studies. 1,1, pp. 7-30.

Bratton, Michael, and Beatrice Liatto-Katundu. 1994. "Political Culture in Zambia: A
Pilot Survey." Political Reform in Africa Working Paper Number 7, Michigan
State University Department of Political Science.

Bratton, Michael and Nicolas van de Walle. 1994. "Neopatrimonial Regimes and
Political Transitions in Africa." World Politics. 46.

Bratton. Michael and Nicolas van de Walle. 1997. Democratic Experiments in Africa:
Regime Transitions in Comparative Perspective. Cambridge, UK: Cambridge

University Press.

Brown, Michael and Stephanie Martz. 1998. "Decentralization and Local Initiative
Concepts." Presentation at Club du Sahel, June 22.

Buijsrogge. Piet. 1989. Initiatives paysannes en Afrique de l’Ouest. Paris: l'Harmattan.

251

Burgess. Stephen F. 1997. "Smallholder Voice and Rural Transformation: Zimbabwe
and Kenya Compared." Comparative Politics. 29: 127-150.

C allaghy, Thomas. 1984. The State-Society Struggle: Zaire in Comparative Perspective.
New York: Columbia University Press.

C allaghy, Thomas. 1994. "Civil Society, Democracy. and Economic Change in Africa:
Adissenting Opinion About Resurgent Societies." In John W. Harbeson, Donald
Rothchild. and Naomi Chazan, eds., Civil Society and the State in Africa.
Boulder, CO: Lynne Rienner. pp. 231-253.

Campbell, Angus, Philip E. Converse. Warren E. Miller, and Donald E. Stokes. 1960.
The American Voter.

Campbell, Angus, Gerald Gurin. and Warren E. Miller. 1954. The Voter Decides.
Evanston, IL: Peterson.

Chambers, Robert. 1983. Rural Development: Putting the Last First. Essex, UK:
Longman.

Chazan, Naomi. 1994. "Between Liberalism and Statism: African Political Cultures and
Democracy." In Larry Diamond, ed. Political Culture and Democracy in
Developing Countries. Boulder, CO: Lynne Rienner. pp. 59-97.

Chazan, Naomi. "Africa's Democratic Challenge: Strengthening Civil Society and the
State." World Policy Journal. Spring 1992, 9, 2, pp. 279-307.

Chazan. Naomi. 1982. "The New Politics of Participation in Tropical Africa."
Comparative Politics. 14: 169-189.

C lark, John. 1991. Democratizing Development: The Role of Voluntary Organizations.
West Hartford, CT: Kumarian Press.

CMDT. "Analyse des causes et des consequences de l'effritement de la cohe’sion sociale
au sein des associations villageoises." San, Mali: CMDT Direction Régionale, 8
June 1994.

CMDT. "Commission pedagogique concemant l'action ZAER." Koutiala, Mali: CMDT
Service F ormation, April 1987.

Coleman, James S. 1990. Foundations of Social Theory. Cambridge. UK: Cambridge
University Press.

 

 

252

Coulibaly, Chéibane. 1994. Mali: Les Ruraux et la Gouvernance Democratique.
Bamako: Insitut Malien de Recherches Appliquees au Développement (IMRAD).

Coulibaly. Chéibane et al. "Les organisations paysannes a l'Office du Niger, les
mecanismes de prise de decision et la problematique des structures villageoises
d'appui aux projets." Bamako: IMRAD. August 1988.

Crook. Richard, and James Manor. 1994. "Enhancing Participation and Institutional
Performance: Democratic Decentralization in South Asia and West Africa."
Paper for ESCOR and Overseas Development Administration. London.

Dahl, Robert A. 1971. Polyarchy: Participation and Opposition. New Haven: Yale
University Press

Dahl, Robert A. 1989. Democracy and Its Critics. New Haven: Yale University Press.

Davis, John Uniack. 1999. "Classrooms of Democracy? The Educational Prospects of
Malian Civil Society.” Chapter in Democracy and Development in Mali, R.
James Bingen. John Staatz. and David Robinson, eds. East Lansing: Michigan
State University Press (forthcoming).

Davis, John Uniack. 1998. "Democratic Governance Strategic Objective Performance

Measurement Survey 11: Data Analysis Report”. Report to USAID/Mali, Contract
number AEP-5468-I-00-6006-00, Washington. DC: Management Systems
lntemational. 15 October.

Davis, John Uniack. 1997a. "Civil Society and Civic Values: The Case of Rural
Organizations in Southern Mali." Paper presented at the African Studies
Association Annual Conference, Columbus, Ohio, 13 November 1997.

Davis. John Uniack. 1997b. “Democratic Governance Strategic Objective Baseline
Indicators Survey: Data Analysis Report.” Report to USAID/Mali, Contract
number AEP-5468-I-00-6012-00, Washington, DC, Management Systems
International, 23 June.

Davis, John Uniack. 19970. "The Intemal Dynamics of Local-Level Rural Associations
in Mali: The Prospects of a Nascent Civil Society." Paper presented at the annual
meeting of the Midwest Political Science Association, Chicago, April.

Davis, John Uniack. 1996. "Fonctionnement inteme des associations rurales au Mali:
Des ‘ateliers de la démocratie'?” Published in two parts by Cauris, Bamako. Mali,
8 November and 22 November.

1 r-flmi'fi

 

253

Decalo. Samuel. "The Process. Prospects, and Constraints of Democratization in Africa."
Afi‘ican Affairs. 91(1992). pp. 7-35.

Degnbol, Tove. 1996a. "Changing Patterns of State-Society Interactions in Mali."

Roskilde University International Development Studies Working Paper No.
5/1996.

Degnbol. Tove. 1996b. "The Interactive Approach to the Study of State and Society
Relations." Forum for Development Studies. 2: 327-351.

Dembe'le', K. 1987. "State Policies on Agriculture and Food Production in Mali." in

Thandika Mkandawire and Naceur Bourenane. The State and Agriculture in
Africa.

Dennis, Jack, Leon Lindberg, Donald McCrone, and Rodney Stiefbold, 1968. "Political
Socialization to Democratic Orientations in Four Western Systems."
Comparative Political Studies, 1: 71-101.

Diamond, Larry. 1996. "Toward Democratic Consolidation." Chapter 19 in Larry
Diamond and Marc E. Plattner, eds.. The Global Resurgence of Democracy (2"d
Ed.). Baltimore. MD: The Johns Hopkins University Press, 1996, pp. 227-240.

Diamond, Larry. 1994a. "Rethinking Civil Society: Toward Democratic Consolidation."
Journal of Democracy. 5.

Diamond, Larry, ed. 1994b. Political Culture and Democracy in Developing Countries.
Boulder, CO: Lynne Rienner.

Diamond, Larry. 1994c. "Causes and Effects." In Larry Diamond, ed. Political Culture

and Democracy in Developing Countries. Boulder, CO: Lynne Rienner. pp. 229-
249.

Diarrah, Cheick Oumar. 1986. Le Mali de Modibo Kéita. Paris: Harmattan.

Dioné, Josue'. 1989. "Informing Food Security Policy in Mali: Interactions between
Technology, Institutions and Market Reforms." Unpublished Ph.D. Dissertation,
Department of Agricultural Economics, Michigan State University.

Direction Nationale de l'Action Cooperative (DNACOOP). 1989. Receuil de T extes
Legislatifs et Reglementaires Regissant le Mouvement Coopératifen République
du Mali.

254

Easton. David, and Jack Dennis. 1969. Children in the Political System. New York:
McGraw-Hill.

Ehrenberg. John. 1999. Civil Society: The Critical History of an Idea. New York: New
York University Press.

Eicher. Carl K. And John M. Staatz. eds. 1990. Agricultural Development in the Third
World, 2nd ed. Baltimore. MD: Johns Hopkins Press.

Esman, Milton J. And Norman T. Uphoff. 1984. Local Organizations: Intermediaries in
Rural Development. Ithaca, NY: Comall University Press.

F atton. Robert, Jr. 1995. "Africa in the Age of Democratization: The Civic Limitations
of Civil Society." African Studies Review. 38: 67-99.

 

Finkel, Steven E. 1987. "The Effects of Participation on Political Efficacy and Political

Support: Evidence from a West German Panel." Journal of Politics. 49, pp. 441-
464.

F inkel, Steven E. 1985. "Reciprocal Effects of Participation and Political Efficacy: A
Panel Analysis." American Journal of Political Science. November, pp. 891-913.

Fisher, Julie. 1994. "Is the Iron Law of Oligarchy Rusting Away in the Third World?"
World Development. Vol. 22. No. 2, pp. 129-143.

Folbre, Nancy. 1986a. "Cleaning House: New Perspectives on Households and
Economic Development." Journal of Development Economics. 2225-40.

Folbre, Nancy. 1986b. "Hearts and Spades: Paradigms of Household Economics." .
World Development. 14, 2: 245-255. I

F owler, Floyd J. 1995. Improving Survey Questions: Design and Evaluation. Sage:
Thousand Oaks. CA.

Fox. Leslie M. USAID/Mali DGSO Performance Monitoring Plan (1996-2002). 16
April 1996.

Fukuyama, Francis. 1995. Trust: The Social Virtues and the Creation of Prosperity.
New York: The Free Press.

Geertz, Clifford. 1973. The Interpretation of Cultures. New York: Basic Books.

255

Gellner, Ernest. I994. Conditions of Liberty: Civil Society and its Rivals. London:
Penguin.

Gentil. Dominique. Les .110mements Coope'ratifs en Afrique de l'Ouest: Interventions de
l'Etat ou Organisations Paysannes? Paris: Harmattan. 1986.

Gibson, James L. 1993. "Alternative Measures of Political Tolerance: Must Tolerance
be 'Least-Liked’?" American Journal of Political Science. 36(May), pp. 560-
577.

Gibson, James L., Raymond M. Duch. and Kent L. Tedin. "Democratic Values and the
Transformation of the Soviet Union." Journal of Politics. 54, pp. 329-371.

Goel, M. Lal. 1980. "Conventional Political Participation." In David Horton Smith et
al., eds., Participation in Social and Political Activities. San Francisco: Jossey
Bass.

Greenstein, Fred 1. 1970. "A Note on the Ambiguity of ‘Political Socialization’:

Definitions. Criticisms, and Strategies of Inquiry." Journal of Politics. 32, pp.
969-978.

Greenstein, Fred 1. 1965. Children and Politics. New Haven. CT: Yale University
Press.

GREFCO/IMRAD. "Le mouvement paysan au Mali" April 1994.

Hadenius, Axel, and Fredrik Uggla. "Making Civil Society Work, Promoting Democratic
Development: What Can States and Donors Do?" World Development, Vol. 24,
No. 10 (October 1996). pp. 1621-1639.

Harbeson, John W., Donald Rothchild. and Naomi Chazan, eds. 1994. Civil Society and
the State in Afiica. Boulder, CO: Lynne Rienner.

Henry, Gary T. 1990. Practical Sampling. Newbury Park, CA: Sage.

Hess, Robert D. And JudithV. Tomey. 1967. The Development of Political Attitudes in
Children. Chicago: Aldine.

Hill, Polly. Development Economics on Trial: The Anthropological Case for a
Prosecution. New York: Cambridge University Press, 1986.

’31....

 

256

Hirschman, Albert 0. "Notes on Consolidating Democracy in Latin America." In Rival
Views of Market Society and Other Recent Essays. Cambridge. MA: Harvard
University Press. 1992.

Hirschrnan. Albert 0. 1984. Getting Ahead Collectively: Grassroots Experiences in
Latin America. Elmsford: Pergamon Press.

Hirschmann, David. 1994. "Urban Women, Civil Society. and Social Transition in the
Eastern Cape. South Africa." African Rural and Urban Studies], 2, pp. 31-48.

Hirschmann, David. 1991. "Women and Political Participation in Africa: Broadening
the Scope of Research." World Development. 19, 12. pp. 1679-1694.

Hodgkin, Thomas. 1956. Nationalism in Colonial Africa. London: Frederick Muller.

Holm. John D., Patrick P. Molutsi, and Gloria Somolekae. 1996. "The Development of
Civil Society in a Democratic State: The Botswana Model." Afiican Studies
Review.

Huntington, Samuel P. 1991. The Third Wave: Democratization in the Late Twentieth
Century. Norman: University of Oklahoma Press.

Hyden. Goran. 1996. "The Challenges of Analysing and Building Civil Society." Africa
Insight. 26, 2: pp. 92-106.

Hyden. Goran, and Michael Bratton. eds. Governance and Politics in Africa. Boulder,
CO: Lynne Reinner, 1992.

Hyman, Herbert. 1959. Political Socialization. New York: Free Press.

Ignatieff, Michael. 1995. "On Civil Society: Why Eastern Europe's Revolutions Could
Succeed." Foreign Aflairs. 74.

Inglehart, Ronald. 1990. Culture Shift in Advanced Industrial Society. Princeton, NJ:
Princeton University Press.

Inglehart, Ronald. 1988. "The Renaissance of Political Culture." American Political
Science Review. 82, 4, December.

Inglehart, Ronald. 1977. The Silent Revolution: Changing Values and Political Styles
Among Western Publics. Princeton, NJ: Princeton University Press.

, I"

 

 

257

Institut Malien de Recherches Appliquees au Développement (IMRAD). 1994. "Le
mouvement paysan au Mali." unpublished manuscript.

Isaacman. Allen. and Richard Roberts, eds. 1995. C otton, Colonialism, and Social
History in Sub-Saharan Africa. Portsmouth. NH: Heinemann.

Jackson, Robert H. and Carl G. Rosberg. 1982. Personal Rule in Black Africa.
Berkeley: University of California Press.

Jacquemot. Pierre, editor. 1981. Le Mali, le paysan, et l'e'tat.

Jamieson, Ian. 1980. Capitalism and Culture: A Comparative Analysis of British and
American Manufacturing Organizations. London: Gower.

Jennings, M. Kent. 1990. "The Crystallization of Orientations." In W. Kent Jennings.
ed. Continuities in Political Action. Berlin: de Gruyter. '

Jennings. M. Kent. and Richard G. Niemi. I981. Generations and Politics. Princeton:
Princeton University Press.

Jeune Afrique. 3-9 September 1997. "Mali: En attendant le proces." pp. 6-7.

.leune Afrique. 1-7 October 1997. "Mali: Apres l'orage" (special section). pp. 77-128.

J onckers, Danielle. "Le Mythe d'une Tradition Communautaire Villageoise dans la
Region Mali-Sud." In Jean-Pierre Jacob and Philippe Lavigne Delville, eds., Les
Associationss Paysannes en Afrique: Organisation et Dynamiques. Paris:

APAD/Karthala/IUED, 1994, pp. 121-134.

Jones. William I. 1976. Planning and Economic Policy: Socialist Mali and Her
Neighbors.

Kane, Cheikh Hamidou. 1961. L'aventure ambigué. Paris: Julliard.

Keane, John. "Despotism and Democracy: The Origins and Development of the
Distinction Between Civil Society and the State, 1750-1850." In John Keane. ed..
Civil Society and the State (London: Verso, 1988).

Kennedy, Peter. 1992. A Guide to Econometrics (3rd Ed.). Boston: MIT Press.

Kim, Jae-On, and Charles W. Mueller. 1978a. Introduction to Factor Analysis: What It
Is and How to Do It. Newbury Park, CA: Sage.

258

Kim. Jae-On, and Charles W. Mueller. 1978b. Factor Analysis: Statistical Methods and
Practical Issues. Newbury Park. CA: Sage.

Keel. Kathie. Lynn Carter. et al. Peijformance r’lIonitoring Plan for USAID/Mali's
Strategy (1996-2002). June 1996.

Komhauser. Arthur. 1959. The Politics ofMass Society. New York: Free Press.

Korten. David C. 1990. Getting to the 21 st Century: Voluntary Action and the Global
Agenda. West Hartford, CT; Kumarian Press.

Lachenmann, Gudrun. "Rural Development in Mali: Destabilisation and Social
Organization." Journal of International Agriculture. 25, 3, 1986, pp. 217-233.

Leonard, David K. 1977. Reaching the Peasant Farmer: Organization Theory and
Practice in Kenya. Chicago: University of Chicago Press.

Leonard, David K. 1982. "Analyzing the Organizational Requirements for Serving the
Rural Poor." In David K. Leonard and Dale Rogers Marshall, eds., 1982.
Institutions of Rural Development for the Poor: Decentralization and
Organizational Linkages. Berkeley: Institute of International Studies, University
of California, pp. 1-39.

Leonard, David K. 1982. "Choosing among Forms of Decentralization and Linkage." In
David K. Leonard and Dale Rogers Marshall, eds., 1982. Institutions of Rural
Development for the Poor: Decentralization and Organizational Linkages.
Berkeley: Institute of International Studies. University of California, pp. 193-226.

Le Roy, Etienne, Jacques Moineau, and Jean-Claude Rocaboy. "La CMDT et les
institutions rurales dans le cadre de Mali-Sud III." Paris: APREFA-LAJP, March
1991.

Lewis. Peter M. "Political Transition and the Dilemma of Civil Society in Africa."
Journal of International Aflairs. Summer 1992, 46, 1. pp. 31-54.

Leys. Colin. 1975. Underdevelopment in Kenya: The Political Economy of Neo-
Colonialism. London: Heinneman.

Lipset, Seymour Martin. 1960. Political Man: The Social Bases of Politics. Garden
City, NY: Doubleday.

Lucas, John. 1994. "The State, Civil Society, and Regional Elites: A Study of Three
Associations in Kano, Nigeria." Afiican Aflairs. 93: 21-38.

259

Mamdani. Mahmood. 1996. Citizen and Subject: Contemporary Africa and the Legacy
of Late Colonialism. Princeton. NJ: Princeton University Press.

Manheim. Jarol B. and Richard C. Rich. Empirical Political Analysis: Research Methods
in Political Science. Fourth Edition. White Plains. NY: Longman. 1995.

Milbrath. Lester W. and ML. Goel. 1977. Political Participation. 2nd Ed. Chicago:
Rand McNaIly.

Mill. John Stuart. 1991. (,‘onsiderations on Representative Government. Buffalo. NY:
Prometheus Books.

Miller. AH. 1974. "Political Issues and Trust in Government: 1964-1970." American
Political Science Review. 68, pp. 951-972.

Ministere de I'Agriculture, République du Mali. Revue du Secteur Agricole du Mali.
Bamako: Ministere de I'Agriculture, 1987. ‘

Monga, Célestin. 1996. The Anthropology of Anger: Civil Society and Democracy in
Africa. Boulder: Lynne Rienner.

Monga, Célestin. 1995. "Civil Society and Democratisation in Francophone Africa."
The Journal of Modern African Studies, 33.

Morduchowicz, Roxana. Edgardo Catterberg, Richard G. Niemi. and Frank Bell. 1996.
"Teaching Political Information and Democratic Values in a New Democracy: An
Argentine Experiment." Comparative Politics. 28, 4 (July). pp. 465-476.

Moss. Todd J. 1995. "US. Policy and Democratisation in Africa: The Limits of Liberal
Universalism." The Journal of Modern African Studies. 33, 2, pp. 189-209.

Muller, Edward N. and Mitchell A. Seligson. 1994. "Civic Culture and Democracy: The
Question of Causal Relationships." American Political Science Review. 88: 63 5-
652.

Muller, Edward N., Mitchell A. Seligson, and Ilter Turan. "Education, Participation, and
Support for Democratic Norms." Comparative Politics. (October 1987): pp. 19-
33.

New York Times. 29 April 1996. "Democracy in Mali" (editorial). p. A10.

New York Times. 16 October 1996. "In One Poor African Nation, Democracy Thrives."
p. A3.

260

New York Times. 11 January 1997. "Despite Setbacks. Democracy Gains in Africa." p.
A3.

New York Times. 7 September 1997. "Mali's Slips Reflect Stumbling African
Democracy." p. A3.

Niemi. Richard G. and Mary A. Hepburn. 1995. "The Rebirth of Political Socialization."
Perspectives on Political Science. 24: 7-22.

Nyang’oro. Julius. 1989. "The State of Politics in Africa: The Corporatist Factor."
Studies in Comparative International Development. 24, I: pp. 5-19.

Nyang’oro. Julius. 1986/1987. "On the Concept of 'Corporatism’ and the African State.
Studies in Comparative International Development. 21: pp. 31-55.

Ostrom, Elinor. 1990. Governing the Commons: The Evolution of Institutions for
Collective Action. Cambridge, UK: Cambridge University Press.

Oue'draogo, Bernard Le'dea. Entraide Villageoise et développement: groupements
paysans au Burkina F aso. Paris: L'Harmattan, 1990.

Ouédraogo, Bernard Lédéa. "Culture et Democratic." Paper presented at Fondation Paix
et Développement, Dakar, Senegal, December 1992.

Painter. Thomas M. 1990. "Cooperatives, Empowerment, and Rural Development in
Africa." Chapter 5 in Miriam S. Chaiken and Anne K. F Ieuret, eds., Social
Change and Applied Anthropology: Essays in Honor of David W. Brokensha.
Boulder: Westview Press, pp. 57-68.

Patterson, Amy S. 1997. "Making the Private Realm Effective: Rural Senegalese
Women and Local Decision-Making Venues." Paper presented at annual
conference of Midwest Political Science Association, Chicago, IL, 10-12 April.

Poulton, Robin and Michael Harris. 1988. Putting People First: Voluntary
Organizations and Third World Organizations. London: Macmillan.

Putnam, Robert D. 1993. Making Democracy Work: Civic Traditions in Modern Italy.
Princeton: Princeton University Press.

Putnam, Robert D. 1995. "Bowling Alone: America’s Declining Social Capital."
Journal of Democracy. 6: 65-78.

 

 

 

261

Remmer. Karen. 1989. "Neo-Patrimonialism: The Politics of Miltary Rule in Chile.
1973-1987." Comparative Politics. Vol 21, No. 2, January.

Renshaw, Laura. 1994. "Strengthening Civil Society: The Role of NGOs."
Development. No. 4, pp. 46-49.

République du Mali. Mission de Décentralisation. Undated. La Commune en
Questions... Bamako. Mali: Mali Communication et Conseil.

Ribot. Jesse C. 1998a. Political-Economic Analysis of Cooperatives Reform in Mali:
‘The State is the Best Hen. ' Report to USAID/Mali DGSO Team, September 30.

Ribot. Jesse C. 1998b. "Decentralization and Participation in Sahelian Forestry: Legal
Instruments of Central Political-Administrative Control." Harvard Center for
Population and Development Studies Working Paper No. 98.06, April.

Roberts. Richard L. 1996. Two Worlds of Cotton: Colonialism and the Regional
Economy in the French Soudan, 1817-1946. Palo Alto: Stanford University
Press.

Roberts. Richard L. 1995. "The Coercion of Free Markets: Cotton, Peasants, and the
Colonial State in the French Soudan, 1924-1932." In Allen Isaacman. and
Richard Roberts, eds. 1995. Cotton, Colonialism, and Social History in Sub-
Saharan A fiica. Portsmouth, NH: Heinemann, pp. 221-243.

Rokkan, Stein and Angus Campbell. 1960. "Norway and the United States of America."
International Social Science Journal. 12, l.

Rustow, Dankwart. 1970. "Transitions to Democracy." Comparative Politics. 2 (April).

Sabatini, Christopher A., Gwendolyn G. Bevis, and Steven E. Finkel. 1998. "The Impact
of Civic Education Programs on Political Participation and Democratic Attitudes."
Report to US. Agency for International Development. Washington, DC:
Management Systems International, February.

Sadowski, Christine M. 1994. "Autonomous Groups as Agents of Democratic Change in
Communist and Post-Communist Eastern Europe." In Larry Diamond, ed.
Political Culture and Democracy in Developing Countries. Boulder, CO: Lynne
Rienner, pp. 155-187.

Sandbrook, Richard. 1996. "Transitions Without Consolidation: Democratization in Six
African Cases', Third World Quarterly. 17.

262

Sandbrook. Richard. The Politics ofAfrica's Economic Recovery. Cambridge. UK:
Cambridge University Press, 1993.

Sandbrook. Richard. 1986. The Politics of African Economic Stagnation. Cambridge,
UK: Cambridge University Press.

Schmitter. Philippe C. 1974. "Still the Century of Corporatism?" In Frederick Pike and
Thomas Stritch. eds., The New Corporatism. South Bend, IN: University of Notre
Dame Press. pp. 85-131.

 

Schmitter, Phillippe C. and Terry Lynn Karl. 1991. "What Democracy Is... and Is Not."
Journal of Democracy. 2: 75-88.

 

Schuman, Howard and Stanley Presser. 1996. Questions and Answers in Attitude
Surveys: Experiments on Question F orm, Wording, and Context. Thousand Oaks,
CA: Sage.

Seligson, Amber. May 1999. "Civic Association and Democratic Participation in
Central America: A Test of the Putnam Thesis." Comparative Political Studies.
32, 3, PP. 342-362.

Sen. Amartya K. 1973. "Behavior and the Concept of Preference." Economics. 40. pp.
214-259.

Shaw. Timothy M. 1982. "Beyond Neo-Colonialism: Varieties of Corporatism in
Africa." Journal of Modern A frican Studies. 20: pp. 239-261.

 

Sigel, Roberta S. 1995. "New Directions for Political Socialization Research: Thoughts
and Suggestions." Perspectives on Political Science. 24, 1 (winter), pp. 17-22.

Sigel, Roberta 8., ed. 1989. Political Learning in Adulthood. Chicago: University of
Chicago Press.

Sigel, Roberta S. And Marilyn Brookes Hoskin. 1977. "Perspectives on Adult Political
Socialization: Areas of Research." In Stanley Allen Renshon, ed. Handbook of
Political Socialization: Theory and Research. New York: Free Press.

Sklar, Richard. 1979. "The Nature of Class Domination in Africa." Journal of Modern
Afiican Studies. 17, 4: 531-552.

Smith, Zeric Kay. 1997a. “Can African Democracy Survive a Vibrant Civil Society?
The Case of Malian Students.” Paper presented at the African Studies
Association Annual Conference, Columbus, Ohio, 13 November 1997.

263

Smith. Zeric Kay. 1997b. “‘From Demons to Democrats’: Mali's Student Movement
1991-1996." Review of African Political Economy. 72: 249-263.

Sorensen. Georg. 1993. Democracy and Democratization: Processes and Prospects in a
Changing World. Boulder, CO: Westview.

Stepan. Alfred. 1988. Rethinking Il/Iilitary Politics: Brazil and the Southern Cone.
Princeton, NJ: Princeton University Press.

Stouffer, Samuel C. 1955. Communism, Conformity. and Civil Liberties. New York:
Doubleday.

Sudman, Seymour. Applied Sampling. New York: Academic Press, 1976.

Sullivan. John L., George E. Marcus, Stanley F eldman, and James E. Piereson. 1981.
"The Sources of Political Tolerance: A Multivariate Analysis." American
Political Science Review. 75: 92-106.

Sullivan. John L., James Piereson, and George E. Marcus. 1982. Political Tolerance and
American Democracy. Chicago, IL: University of Chicago Press.

Talbott, Strobe. 1996. "Democracy and the National Interest." Foreign Affairs. 75: 47-
63.

Thomson, James T. and Chéibane Coulibaly. "Common Property Forestry Management
Systems in Mali: Resistance and Vitality Under Pressure." Unasylva. 4, 1(1995):
pp. 16-22.

Thomson, James T. and Djeidi Sylla. Local Environmental Practices and Orientations
for Rural Forestry in Mali 's Fifth Region, A report prepared for USAID/Mali.
Burlington, VT: Associates in Rural Development.

Tocqueville, Alexis de. 1945(1846). Democracy in America. New York: Vintage.

Tripp, Aili Mari. 1994a. "Gender, Political Participation, and the Transformation of
Associational Life in Uganda and Tanzania." African Studies Review. 37, 1, pp.
107-131.

Tripp. Aili Mari. 1994b. "Rethinking Civil Society: Gender Implications in
Contemporary T anzania." In John W. Harbeson, Donald Rothchild, and Naomi
Chazan, eds. Civil Society and the State in Africa. Boulder, CO: Lynne Rienner,
pp. 149-168.

 

264

Tripp. Aili Mari. 1992. "Local Organizations. Participation, and the State in Urban
Tanzania." In Goran Hyden and Michael Bratton. eds. Governance and Politics
in Africa. Boulder. CO: Lynne Reinner, pp. 221-242.

United Nations Development Program (UNDP). Human Development Report 1995.
New York: Oxford University Press, 1995.

United Nations Development Program (UNDP). 1996. Human Development Report
I 996. Oxford, UK: Oxford University Press.

Uphoff. Norman T. 1991. "Fitting Projects to People." In Michael Cemea, ed., Putting
People First: Sociological Varaibles in Rural Development. New york, NY:
World Bank and Oxford University Press, pp. 467-511.

USAID. Civil Society, Democracy and Development in Africa: Proceedings of a
Workshop for Development Practitioners. Washington, DC: USAID Bureau for
Africa, June 1994.

USAID/Mali. [no date]. Results Plan. Bamako, Mali: USAID.

Vengroff, Richard. "Governance and the Transition to Democracy: Political Parties and
the Party System in Mali." Journal of Modern African Studies. 31 :4(1993), pp.
541-562.

Vengroff, Richard. "The Impact of the Electoral System on the Transition to Democracy
in Africa: The Case of Mali." Electoral Studies. 13:1(1994), pp. 29-41.

Verba, Sidney. 1965. "Comparative Political Culture," in Lucian W. Pye and Sidney
Verba (editors), Political Culture and Political Development. Princeton:
Princeton University Press. 512-560.

Verba, Sidney, Kay L. Schlozman, and Henry E. Brady. 1995. Voice and Equality: Civic
Voluntarism in America. Cambridge, MA: Harvard University Press.

Verba, Sidney, N. Nie, and Jae-On Kim. 1971. The Modes of Democratic Participation:
A Cross-National Comparison. Beverly Hills, CA: Sage.

Walker, S. Tjip. "The Political Problem of Privatization in Africa." Paper presented to
the Annual Meeting of the Mid-West Political Science Association, April 18-20.
1996, Chicago, Illinois.

265

\lI/"allerstein. Immanuel. I964. "Voluntary Associations." In Political Parties and

National Integration in Tropical Africa. eds. James Coleman and Carl G.
Rosberg. Berkeley: University of California Press. pp. 318-339.

Widner. Jemiifer. 1994. "The Rise ofCivic Associations Among Farmers in Cote

d’Ivoire." In Civil Society and the State in Africa. eds. John W. Harbeson. Donald
Rothchild. and Naomi Chazan. Boulder. C 0: Lynne Rienner.

Witte. Robert S. and John S. Witte. 1997. Statistics (fifth edition). Fort Worth and
Orlando: Harcourt Brace.

Wood. Ellen M. 1991. "The Uses and Abuses of Civil Society," in Ralph Miliband. ed.,
The Socialist Register I 990. New York: Monthly Review Press.

Woods. Dwayne. 1994. "Elites, Ethnicity, and ‘Home Town’ Associations in the Cote

d’lvoire: An Historical Analysis of State-Society Links." Africa. 64, 4, pp. 465-
483.

World Bank. 1996. World Development Report I 996. Oxford, UK: Oxford University
Press.

Young, Crawford. 1994. “In Search of Civil Society.” In John W. Harbeson, Donald
Rothchild. and Naomi Chazan. eds., Civil Society and the State in Afiica.
Boulder, CO: Lynne Rienner. pp. 33-50.

Young. Tom. 1993. "Elections and Electoral Politics in Africa." Africa. 63, 3, pp. 299-
312.