1




Paper prepared for the Annual Bank Conference on Development
                   Economics, Paris, June 2000




           COMMUNITY IMPERFECTIONS




                        Jean-Philippe Platteau




Department of economics and CRED (Centre de Recherche en Economie du
                 Développement), University of Namur
                         Rempart de la Vierge, 8
                        B-5000 Namur Belgium
               Email : jean-philippe.platteau@fundp.ac.be

                                                                                     2


1.     Introduction

Since the last decades have been characterized by a strong disillusionment with
the performances of the state, especially in countries of SubSaharan Africa, and
since market imperfections are a pervasive feature of the rural economy in
developing regions, it is perhaps not surprising that many professional and
academic economists have looked increasingly to rural communities as a
providential escape out of the state-market dilemma. In this way, they have got
nearer to their colleagues from other social sciences for whom personalized
relationships sealed by various forms of reciprocal exchanges in rural
communities allow villagers to effectively solve important problems and to
increase their welfare. Among the functions of communities that have received
prime attention among economists, the following are worth pointing out : (i)
communities supply informal insurance to their members in a world fraught with
pervasive uncertainties (the community as an insurance supplier); (ii) they allow
to establish trust (also a public good) in trade relations within a world where
written contracts do not exist owing to widespread illiteracy (the community as a
trust-enforcing agency); (iii) they help surmount the free rider problem inherent
in the production of local public goods (the community as a producer of
collective goods); (iv) and they ease the enforcement of property rights to local-
level natural resources (the community as a rights-enforcing agency).
       New strategies based on the idea of decentralized or participatory
development have gained increasing currency among all international
organizations (including the World Bank) and donor agencies. The attitude of
optimism which accompanies them is partly grounded in the belief that rural
communities can be an effective channel of development if genuine delegation
of powers and responsibilities takes place on a sufficient scale. This is certainly
a step in the right direction and it is not the purpose of the present paper to call it
into question. Yet, the very enthusiasm with which decentralized development
is embarked upon may carry serious risks if it arouses unrealistic hopes of rapid
success and makes this success too much dependent on the performances of
rural communities.
       As a matter of fact, communities are not the ideal social arrangements that
they are sometimes supposed to be.              They may be subject to several
dysfunctionings that are susceptible of diminishing their effectiveness in solving
the above-mentioned problems. In the following, the intent is to consider
several of their potential shortcomings with respect to these specific problems,
that is, the paper is concerned with community failures rather than community
successes. This is with a view to drawing attention to several areas where it may
be advisable for a decentralized approach to rural development to strengthen or
complement communities instead of relying on them blindfully.                    The
contemplated task is carried out in four successive sections corresponding to the
four community functions that have just been distinguished. In the final section,

                                                                                             3


a number of essential policy lessons are derived from the analysis of community
imperfections.


2.     Communities as insurance suppliers

There is a growing literature arguing that rural communities are a rather
effective device to provide decentralized insurance through reciprocal
contingent voluntary transfers. This is so because continuous and personalized
relationships among community members help create an interaction framework
resembling a repeated game of indeterminate or infinite duration. In such a
game, reputation effects are at work and punishments can be meted out to
willing free riders, thereby ensuring that promises to help fellow villagers hit by
adverse economic events are self-enforcing (Sugden, 1986 : chap. 7 ; Kimball,
1988 ; Coate and Ravaillon, 1993 ; Townsend, 1994 ; Kocherlakota, 1996 ;
Ligon, Thomas and Worral, 1997).                Because written and legally binding
contracts are not feasible (the arrangement is informal in this precise sense), an
incentive compatibility or participation constraint must indeed be respected to
ensure the self-enforceability of the risk-pooling mechanism.
       Such a mechanism has been remarkably well anticipated by David Hume
in the middle of the 18th century :

     "We can better satisfy our appetites in an oblique and artificial manner, than by their
     headlong and impetuous motion. Hence I learn to do a service to another, without
     bearing him any real kindness; because I forsee, that he will return my service, in
     expectation of another of the same kind, and in order to maintain the same
     correspondence of good offices with me or with others. And accordingly, after I have
     serv'd him, and he is in possession of the advantage arising from my action, he is
     induc'd to perform his part, as foreseeing the consequences of his refusal ...After these
     signs [i.e., promises] are instituted, whoever uses them is immediately bound by his
     interest to execute his engagements, and must never expect to be trusted any more, if he
     refuse to perform what he promis'd" [Hume, 1740: Book III, Part II, Sect. V, 521-2].

Pioneer empirical studies on reciprocal transfers, in particular the study by
Platteau and Abraham (1987) on a fishing village in South India (Kerala state)
and the study by Townsend on an agricultural village, again in South India
(1994), have reached rather optimistic conclusions regarding the ability of
agrarian societies to cope with risks within the framework of decentralized
mutual insurance mechanisms.              Yet, if Townsend's evidence (household
consumption varies with the aggregate income of the village but not with the
household income) is consistent with a substantial extent of risk-sharing, it also
lends itself to other interpretations. Indeed, his findings can be attributed to
other informal methods of consumption smoothing (such as borrowing and
lending, or saving and dissaving, including asset adepletion and replenishment)

                                                                                   4


than to group-based insurance per se (Alderman and Paxson, 1994 ; Ravallion
and Chaudhuri, 1997).         Platteau and Abraham' study, on the other hand,
provides direct evidence about pervasive reciprocal transfers and stabilized
intertemporal consumption flows, yet the sample is very small and transactions
are not genuine transfers but interest-free loans with flexible repayment
conditions.
       Recent evidence tends to indicate that informal mechanisms can not only
have a restricted scope, but also be fragile, unreliable in protecting the poorest
households, and susceptible of retarding capital accumulation and economic
growth (Platteau, 1991 ; Alderman and Paxson, 1994 ; Morduch, 1995). As
regards the first point, from the observation that reciprocal transfers are frequent
it is not legitimate to infer that they provide effective insurance to villagers. For
one thing, such transfers may involve only a limited fraction of the village
population and, for another thing, they may represent only a low proportion of
the income shocks in bad periods (Rosenzweig, 1988 ; Czukas, Fafchamps, and
Udry, 1998 ; Cox and Jimenez, 1990, 1997 ; Alvoet, 1997 ; Goldstein, 1999).
       Risk characteristics are an important determinant of the feasibility of
group- or community-based informal insurance. Thus, it is evident that low
frequency events as well as repeated shocks (shocks strongly correlated over
time) and covariate risks are difficult to insure against on a local level. This
observation carries the unfortunate implication that informal village mechanisms
may be particularly fragile when needed most (Morduch, 1999). In fact, it is the
combination of risk characteristics with the mindset and the associated
behavioural traits of the people in agrarian societies that is liable to reduce the
efficiency of group-based insurance arrangements.
       Just consider the case of low-frequency events.          Gertler and Gruber
(1997) have found that in Indonesia households can protect their consumption
levels against only about 30% of low-frequency shocks that seriously impair
performance over the long term (cited from Morduch, 1999). Platteau (1997)
has likewise found that, in Senegal, insurance within informal sea rescue mutual
groups of artisanal coastal fishermen is quite limited, at least as far as
compensations for damaged assets are concerned. As a matter of fact, the
indemnity paid to fishermen struck by an accident at sea represents only a small
portion of the cost of the damages suffered. Furthermore, insurance is highly
imperfect as attested by the frequency with which fishermen pull out of their
group on the pretext that their participation has entailed losses for them. It is
worth noting that the frustration of willing-to-leave participants does not appear
to originate in the classical incentive problems of moral hazard and adverse
selection. Emphasis is instead put on the unbalanced character of a situation in
which some people have contributed while others have benefitted from their
contributions. Revealingly, when a discontented fisherman decides to pull out
of his mutual aid group, not only does he ask for his past money contributions to
be returned to him (he does not ask to be compensated for his labour

                                                                               5


contributions as it is not the practice to compute the money equivalent of labour
time spent in assisting others), but the other members of the group tend also to
give in to such a demand because they consider it legitimate.
      What this example shows is that, contrary to what the economic analysis
of voluntary reciprocal transfers implicitly assumes, members of agrarian
societies may just be confused about what constitutes an exchange of risks. This
finding should not come as a surprise as soon as it is realized that the very
notion of insurance implies a rather abstract mode of thinking: people must be
able to conceive that there is a positive value in the reduction or outright
elimination not of a certain undesired future event but of what is merely a
possible negative outcome. This means that they must be ready to play in a
genuine game in which ex post there will be winners and losers and that, as a
result of mutual insurance, income will therefore be redistributed from lucky to
unlucky participants to the scheme.
      In point of fact, the norm which sets the standard of fair behaviour in
agrarian societies is that of balanced reciprocity and it typically applies to
transactions of risk exchange. According to this standard, any `gift' must be
returned at some future time, even if the exact date at which the `counter-gift'
will be made is not known in advance and if people do not necessarily insist on
receiving strict equivalents of what they have contributed. In other words,
people engage in relationships and condition their continued participation in
them on the expectation that net payments will more or less balance out over
time. When they make contributions, villagers thus strongly expect that, in one
form or another, there will be a tangible quid pro quo for their present
generosity. The idea that such a return is uncertain (it might come but it might
just as well not come) is unacceptable to them: balanced rather than conditional
reciprocity is the principle that motivates their actions (Platteau, 1997).
      Interestingly, discussions with informal (health) mutual insurance groups
in Kibera slum, Nairobi (November 1996) led me to the same conclusion of a
basic misunderstanding of the insurance concept.          Here too, it is common
occurrence that members decide to leave a mutual because they consider that
they have been too long on the giving side. And, upon departure, they are again
being paid back all the contributions they have previously made to the group.
Moreover, even when members are expelled for repeated misconduct (e.g.,
throwing unjustified suspicion on the executive committee's management,
causing vicious rivalries among members, attending general assemblies in a
drunken state), their entire contributions are returned to them as a matter of
principle. Evidence of similar behaviour has been reported to me in the case of
mutual insurance groups in rural Albania (personal communication of Anne-
Claude Creusot from CIRAD, Paris).
      In terms of the theory of reciprocal risk-sharing, the problem seems to
arise from the fact that the random shocks (e.g., accidents at sea) are not
frequent enough to ensure the sustainability of informal risk-pooling

                                                                                              6


arrangements. Since the expected length of time between realizations of the
shock is rather long, ­a characteristic formally reflected in relatively high
discount rates­, the participation constraint is likely to be violated or, if risk-
pooling is established, its extent will strongly diverge from the first-best
situation of full sharing. Upon a careful examination, however, it becomes
evident that the theory and the role played in it by the discount factor are not
able to account for the kind of observations considered here. As a matter of fact,
the theory can explain that villagers refuse to participate in a mutual aid network
from the very beginning (given a low discount rate, the participation constraint
is violated at each and every period since it is forward-looking), not that they
have chosen to engage in it and have later decided to pull out. The crux of the
matter is that their behaviour is time-inconsistent.
        To put it in another way, the failure of economic theory directly follows
from the fact that the rate of discount is a factor that affects the stream of the
agents' future utilities : it reflects the subjective way in which individuals
project their future situation. Yet, the main lesson from the above-reported
evidence is that, when deciding whether to continue to participate in a risk-
pooling network, individuals actually pay attention to the past history of
occurrences of natural hazards and to the corresponding past history of inter-
individual transfers.
        A new type of model of informal risk-sharing partly remedies this
difficulty, precisely because it allows for history-dependent decisions, that is,
strategies are non-stationary : conditional transfers between agents along any
equilibrium path depend not only upon the current state of nature and the past
history of defections and punishments, but also on the past history of inter-
individual transfers (Thomas, and Worral, 1994 ; Fafchamps, 1999 ; Ligon,
Thomas and Worral, 1997). When the transfers are history-dependent in this
sense, the constrained-efficient arrangement, assuming complete information but
limited commitment, is a simple updating rule that can be construed as a debt
contract with occasional debt forgiveness1.
        Even though the constrained-efficient insurance arrangement with non-
stationary strategies provides a much better account of the Senegalese story than
the constrained-efficient stationary arrangement, it still falls short of giving a
complete explanation. As a matter of fact, a debt contract with (occasional)
forgiveness is not fully compatible with the principle of balanced reciprocity. In

1 In the words of the authors, "the household that receives a bad shock receives income from
the other household, but thereafter `repays' this `loan' at a constant rate until another bad
shock is received by one of the households. At this point the resemblance to a standard debt
contract ceases. The household suffering the latest bad shock receives a 'loan' of the same
size as before, and starts repaying the following period. The previous history is forgotten, so
it doesn't matter who had previously 'borrowed' from whom; all that matters is who was the
last to receive a loan.     If both households simultaneously receive bad shocks then the
repayments continue, except they are reduced for that period, proportionately to the fall in
aggregate income" (Thomas and Worrall, 1994: 16-17).

                                                                                             7


terms of Thomas and Worrall's model, the Senegalese fishermen do not appear
to act rationally in so far as they consider that they must be roughly
compensated for their past contributions. It is therefore as though a social norm
embodied in the principle of balanced reciprocity would prevent people from
obtaining as much insurance as they could did they not obey this principle.
Divergence between Thomas and Worrall's updating rule and the logic of
balanced reciprocity is most apparent when the same asymmetric state of nature
repeats itself. Such a rule then implies that the agent who suffered a series of
bad shocks and therefore received successive transfers from the other agents
starts repaying only his last debt as soon as asymmetry of incomes has ended
and all agents receive the same income. Balanced reciprocity, by contrast,
implies that all the previous `debts' incurred in the form of transfers are
remembered and considered to be outstanding.
       From the fact that members of traditional agrarian societies tend to obey a
principle of balanced rather than conditional reciprocity, it does not follow that
all risk-pooling arrangements are unsustainable. When these arrangements meet
some conditions, indeed, they may be compatible with the requirements of
balanced reciprocity (see Platteau, 1997, for more details). First, voluntary
exchanges of non-covariate risks are likely to be observed in traditional
communities whenever realization of the random shock is so frequent that
everyone can reasonably expect to need others' assistance within a rather short
period of time (see the above-noted case of fishing incomes which vary widely
from one day to the next and across individuals). In these circumstances,
indeed, since the length of time elapsing between successive realizations of the
random shock is short, mutual insurance will quickly approach balanced
reciprocity. Second, if realizations of a given kind of random shocks are not
frequent enough, the obstacle can be overcome by pooling together a sufficiently
large number of risks. For example, in the agricultural villages of northern
Nigeria studied by Udry (1990, 1993, 1994), people are informally insured
against a wide variety of production and consumption random shocks (such as
flooding, wind damage, and infestation by insects or various household
emergencies comprising medical problems, rain damage to houses, etc)2.
       Third, if assistance takes on the form of contingent loans, as observed in
the aforementioned studies by Abraham and Platteau and by Udry, participants

2 The natural tendency for economists is to view the pervasive existence of multi-purpose or
multifunctional forms of informal insurance in traditional agrarian societies as a comforting
sign that their members have well understood the mechanism of collective insurance and the
principle of risk diversification in particular. As a matter of fact, by multiplying the set of
circumstances under which a member is eligible for assistance, multi-purpose arrangements
are expected to enlarge the scope of insurance through pooling of non-correlated risks. An
alternative or complementary interpretation is suggested here: villagers pool several types of
risk so that every participant can benefit from the network by the way of concrete assistance
within a rather short time horizon, thereby ensuring that debits and credits more or less
balance out over time.

                                                                                 8


can be assured that the relationships with other members of the risk-sharing
scheme are likely to be roughly balanced. This is because the insurance aspect
of such loans tends to vanish from sight due to the prominent character of quid
pro quo typical of a credit transaction.       This is especially so when, as
documented by Udry, the creditor does not significantly adjust the terms of the
loan (repayment date and interest rate) to ease the debtor's situation. We are no
doubt a long way from the constrained-efficient non-stationary rule according to
which debts should be completely forgiven in certain stages of the sequence of
hazard occurrences.
       Informal risk-sharing shemes are not only vulnerable to the balanced
reciprocity logic prevailing in agrarian societies, but they also tend to work
unsatisfactorily for the poorest households. There are several reasons for this.
First, relations within insurance networks are likely to be tense when all the
members are down on their luck as a result of a collective shock. In such
circumstances, indeed, holding onto whatever income is available is especially
tempting and the promise to share with others may be easily broken. This holds
true, in particular, for poor members who are close to their subsistence margin
and therefore run a relatively high risk of being excluded from the network
(Morduch, 1999). Second, as argued by Hoff (1997) within the framework of a
model assuming imperfect information but perfect commitment (the community
has the means to enforce commitments to share) and where a majority voting
rule among participants determines the extent of income pooling, a unique level
of endowment wealth exists such that households with endowments below that
threshold level will participate in the informal insurance arrangement, while
richer households will not. In other words, this arrangement may be a poverty
trap since the absence of richer households makes insurance inevitably less
effective for the poorer ones. Moreover, widening wealth differentials, say, as a
result of newly emerging outside income-earning opportunities, create
incentives for the better-off households to leave the risk-sharing scheme. As
they exit, the welfare gains from the scheme to those remaining in the system
fall, and it can unravel.
       Things may work in the other way around, though. As a matter of fact,
insurance may be provided as an auxiliary service in informal schemes that are
constructed around the principle of balanced reciprocity. Such is, for example,
the case of rotating credit-and-savings associations (ROSCAS) in which
participants make regular payments to a common pool (the `pot') periodically
allotted to each one of them in turn (Platteau, 1997). When the sequence of
assignment of the `pot' takes account of the relative needs of the participants,
and/or when mutual help is offered as an additional service, insurance is part and
parcel of the scheme and the poorest households are likely to be excluded. This
is because their low or unreliable incomes do not allow them to make regular
contributions to the association's fund. There is actually plenty of evidence

                                                                                  9


attesting to this sobering fact (see, e.g., André and Platteau, 1998 ; Abraham,
Baland and Platteau, 1998).
       In her aforementioned work, Hoff has assumed that people voluntarily
choose to participate or not in an informal risk-sharing scheme and freely decide
whether to stay or to pull out following a shock to the distribution of wealth. It
is far from certain, however, that this conventional assumption is a good
description of reality as far as tribal, lineage-based societies are concerned. In
point of fact, equalizing norms may exist that preclude an individual from rising
above other members of his group and, in particular, from leaving insurance
pools where everyone is expected to share income with others. Opposition to
withdrawals may not arise only from the fear that a reduction of the size of the
insurance network will diminish the group's ability to spread risk (Fafchamps,
1992). It also results from the highly personalized character of interindividual
relationships in small societies and from the invidious comparisons and envious
feelings that close proximity among human beings easily breeds (Bourdieu,
1992).
       In this new light, redistributive norms that enjoin economically rising
individuals to share their surplus with fellow community members appear as a
form of taxation designed to curb positional race for status, since the effort of
any individual to improve his lot generates positional externalities that
negatively affect the welfare of the community's other members (Frank, 1995,
1998). In traditional tribal societies, such norms are often backed by powerful
sanctions that include social pressures, constant harassment and the use of
effective mechanisms of ideological intimidation, most notably witchcraft
accusations and practices. Witchcraft accusations are especially effective when
they are based on the allegation that economic success of a particular individual
is the result of some hidden and treacherous exploitation of fellow villagers (see
Platteau and Hayami, 1998 ; Platteau, 2000a : chap. 5, for illustrations and
references).
       Village societies are thus far from being havens of peace and harmony
where people voluntarily come to help each other. In some sense, they are
repressive societies whose members must comply with norms that severely
sanction individualized entrepreneurship and capital accumulation conceived as
acts of social deviance. Mechanisms of informal insurance may therefore be at
the root of inefficiencies that undermine economic progress in the long term.
Interestingly, Landes (1998), Thompson (1993) and other historians have noted
that the same egalitarian logic actually prevailed among the European peasantry
and was even carried over into the guild system of medieval cities : "The aim
was an egalitarian social justice, but it entailed serious constraint on enterprise
and growth ­a safety net at the expense of income".            The principle was
"laudable but static" (Landes, 1998 : 42-43).
       In conclusion, two scenarios can be observed when new economic
opportunities arise that can potentially create social and economic differentiation

                                                                                              10


within rural societies. Under the first scenario, dynamic individuals succeed in
pulling out of their risk-sharing network ­through outmigration, religious
conversion, or by transforming themselves into local patrons (Platteau, 2000a :
212-25)­, and the remaining participants are less effectively insured against
random shocks than before. Under the second scenario, dynamic individuals are
constrained to stay within their risk-sharing group and valuable opportunities for
accumulation and growth are foregone. In neither case can the community-
based insurance arrangement be considered to adapt successfully to the
challenge of modernization.             Yet, it is noteworthy that, if the dynamic
individuals turn themselves into local patrons, reciprocal exchange gives rise to
an asymmetric form of insurance in the form of patron-client relations
(Fafchamps, 1992).          Such transformation is likely to reduce the poor's
bargaining strength since they will now be involved in bilateral vertical ties
rather than in multilateral and horizontal relationships. Also, a way out of the
above dilemma can be found if dynamic individuals are altruists of some sort
and succeed in creating a cooperative framework within which individual capital
accumulation can proceed (the rotating savings-and-credit associations are a
good example3), or if they devote a lot energy to starting projects that benefit
many people from their community (Platteau, 2000a : 227-230).


3.      Communities as trust-enforcing agencies

Economic transactions offer considerable scope for cheating and malevolence.
A once-for-all transaction can indeed be depicted as a one-period Prisoner's
Dilemma game and, as is well-known, the unique equilibrium outcome of that
game is universal cheating. Repeated game theory nevertheless suggests that the
Pareto-superior outcome characterized by honest behaviour on the part of all the
parties to the exchange is a distinct possibility. As a matter of fact, repeated
transactions, either with the same partner or with members of a given group


3"The subtlety about these associations lies in the fact that they enable all participants to
individually save and possibly invest part of their incomes (they actually provide a
coordinating mechanism that systematizes and disciplines the members' savings efforts) while
at the same time creating a sense of solidarity and togetherness well in tune with the old spirit
of village life. The latter effect obtains not only because all participants equally contribute
and benefit from the scheme, but also because many of such associations buid up an insurance
fund to help their members in difficulty (usually through special loans). Moreover, the
extremely tight discipline self-imposed by the members (regarding the times of payment of
the periodic contributions and the amounts to be contributed) within a clearly democratic
setup creates special bonds among them. Owing to this dual characteristic of both pursuing
customary modes of interpersonal relations and allowing private accumulation of income
surpluses (through a co-operative mechanism), rotating savings and credit associations can be
considered as new collective forms of organization that permit the emergence of
individualistic economic pursuits" (Platteau, 2000a : 228).

                                                                                            11


within which information circulates well, allow for reputation effects and
decentralized sanction mechanisms that have the desired effect of disciplining
transactors. Like in the case of informal insurance, the village community
appears as an ideal institution to achieve that effect. In the words of Hayami and
Kawagoe (1993) :

        "In the village community everyone is watching everyone. Gossip about one's
        misconduct is circulated by word of mouth faster than any other means of
        communication. In such an environment a significant cost would be incurred to a
        person who would violate a contract with a fellow villager, since not only would he
        lose benefits from the present contract but the resulting bad reputation would deprive
        him of future opportunities to enter into contracts with other villagers as well"
        (Hayami and Kawagoe, 1993 : 167).

The above-described mechanism based on ostracization strategies has been
named a multilateral reputation or sanction mechanism by Greif (1993, 1994,
1998). In game-theoretical terms, if everyone follows a strategy whereby he acts
honestly and moreover punishes (say, by cheating or boycotting) any agent who
has cheated some member of the group in the past, this strategy of multilateral
punishment is an equilibrium strategy and expectations are self-enforcing:
believing that all other participants have this strategy, any given individual has
an incentive to follow the same and at equilibrium nobody cheats. Greif calls
`collectivist' or `segregated' these societies where everyone expects everyone
else to respond to any act of dishonesty committed in any pairwise encounter
within the community space. The observed honesty equilibrium is thus the
spontaneous outcome of common cultural beliefs of the `collectivist' type. It
reflects a reputation mechanism among self-interested individuals whereby an
agent can establish ex ante that his most profitable course ex post is to be
honest: in other words, the transactor can credibly commit to honesty (Greif,
1989: 858, 866-67).
      A noteworthy feature of such a social system is that participants are
incited to follow the multilateral punishment despite the fact that the agent's
strategy does not call for punishing agents who violated the collective
punishment procedure (Greif, 1993: 535).               The self-enforcing character of
collective punishment obtains insofar as a cheater stands to lose nothing (he
expects anyway to be ostracized in the future) and is therefore inclined to cheat
again in new interactions. Others thus do better by avoiding new deals with that
individual and, in this way, the second-order public-good problem of
sanctioning is solved.
      Nothing nevertheless guarantees that the right predisposition to trust
prevails when the game starts and that it is correctly perceived, that is, nothing
ensures a priori that individuals will choose what Sugden has called strategies of
`brave reciprocity' whereby they decide to give honesty a try at the risk of being
cheated by opportunistic transactors.           As we have learned from the Folk

                                                                                             12


Theorem, there are indeed many possible Nash equilibria in the infinitely
repeated Prisoner's Dilemma game and the multilateral punishment strategy is
only one out of many possible equilibrium strategies. In the words of Gambetta
(1988) :

        " ... when the game has no history a cooperative first move is essential to set it on
        the right track, and unconditional distrust could never be conceived as conducive
        to this.... This problem may be circumvented by assuming the presence of
        uncertain beliefs and a random distribution which accommodates the probability
        of the right initial move being made and being `correctly' interpreted. Yet there is
        no reason why the appropriate conditional beliefs should typically be the case, and
        the optimal move may be hard to come upon by accident (while we may not want
        to have to wait for it to come upon us). If it is true that humans are characterized
        by a lack of fine-tuning and a tendency to go to extremes, the assumption that
        trust will emerge naturally is singularly unjustified..." (Gambetta, 1988: 227-228;
        see also Binmore, 1992: 434).

That communities can be pervaded by distrust and suspicious feelings is a fact
confirmed by the experience of the South Italian countryside (Banfield, 1958 ;
Putnam, 1993) and rendered more credible by the above-noted atmosphere of
envy and jealousy typical of close-knit societies. The point which I want to
stress is different : even assuming that individuals are ready to take the risk of
honest behaviour in trade relations, they may still fail to achieve the
`collectivist' outcome described by Greif. This is because information may not
circulate easily among community members even in the absence of physical
obstacles to communication. This situation happens when people are reluctant
to disclose facts of malfeasance of which they have been the victims for reasons
which arise from considerations of honour or face saving.
      Several reasons can explain such communication failures in small-scale
societies. (i) People may feel ashamed to confess that they have been fooled. In
a culture of honour, indeed, people's behaviour is driven by the need to avoid
giving the impression that they lack the strength to protect their own interests
effectively (Nisbett and Cohen, 1996 : 4-7). To admit that they have been
cheated is to recognize that they are somehow weak and vulnerable to attacks by
predatory individuals. This appears to be a dangerous step in so far as it is likely
to erode power of deterrence.            (ii) The victims of malfeasance may wish,
consciously or not, that fellow group members deal with the same deadbeats by
whom they have been burnt. Such a motivation is more likely to be observed in
a social context of rivalrous relations. (iii) The third reason, on the contrary, is
linked to more altruistic motives and therefore more characteristic of close
relations between people and families with long-established relationships. In
these circumstances, the victims of malfeasance may want to avoid bringing
dishonour to the malevolent actor and his family. As a matter of fact, to act in a
way that is susceptible of ultimately causing the quarantine and, possibly, the
ruin of a fellow villager whose family or lineage has resided over many

                                                                                      13


generations in the same location is a major responsibility which many would
like to escape from. When sanctions are gradual, people may still be reluctant to
take steps that may contribute to jeopardize the future well-being of some
member of their community.
       As an example of the first and third possibilities, one may refer to the
fishing and marketing town of Estancia in the Central Philippines. There, we
are told, local Filipinos do not convey to each other information about customers
who did not pay their debts, "for fear of exposing to others that they themselves
had been successfully taken advantage of", and also for fear of insulting their
debtors by suggesting that they might not be willing or able to meet their
obligations.      In contrast, Chinese merchants willingly exchange among
themselves credit information about Filipino customers, thereby reducing the
frequency of bad debts (Szanton, 1998: 256, 260). The same pride which
prevents Estancia's indigenous people from transmitting information about loan
delinquents has also been observed among the Moose communities of Burkina
Faso, a cultural group known for the considerable value attached to
considerations of honour and pride (Badini, 1994). Thus, in the Yalogo region
(located in the northeastern part of the country), a lot of secrecy surrounds all
loan transactions because local people are extremely wary of not bringing shame
(yandé in local parlance) on resident families with whom they are in continuous
interaction. When a credit-giver is not being repaid, he punishes the defaulter
by refusing to lend him any more in the future, yet does not act to destroy his
reputation4.
        Due to this sort of communication failures arising from cultural traits
rather than technical obstacles, bilateral rather than multilateral reputation and
punishment mechanisms are brought into play. Following Greif's terminology,
a bilateral reputation mechanism obtains when transactors punish only defectors
who have happened to hurt their own interests. As a consequence, misbehaviour
on the part of an individual in his dealings with another member of the group
only affects their mutual relationship without yielding any influence on other,
potential partners in the same community. The societies where such behaviour
and beliefs predominate are considered by Greif to belong to the realm of
`individualistic' cultures5. Greif himself considers that Third World societies
tend to be `collectivist' societies compared with the `individualistic' societies of
the advanced Western countries. The argument here aims at qualifying this
characterization by pointing to the possibility of `individualistic' cultures in the
rural areas of the southern hemisphere as well.
       In such areas, as illustrated above, the culture may be a `culture of
honour' with the result that the chain of information transmission is broken and
decentralized collective punishment becomes impossible. What must be added
now is that sanction mechanisms, whether bilateral or multilateral, may fail to be

4 I am indebted to Denis Dubuisson who made this field observation.
5 Greif himself tends to regard Third World societies as `collectivist' in nature.

                                                                                 14


established even in situations where information of the required kind is
available. This happens whenever people, out of a feeling of empathy or
`sameness', are reluctant to punish a fellow community member found guilty of
some rule violation. Such a conclusion actually follows from the third above
motive, since avoiding to bring dishonour to another family has been justified on
the ground that such dishonour may cause harm to that family.
       Sanctioning mechanisms are especially problematic when applied to kin
relations. Thus, in Singapore, some entrepreneurs prefer to abstain from trading
with close personal relationships to avoid the strains and tensions arising from
intra-kin dealings (Li, 1998: 161). In Ghana, selling on credit to relatives and
neighbours seems to be carefully avoided by manufacturing and trading firms
because it is considered as amounting "to the death warrant of the firm".
Payment problems would be frequent "as friendship and family ties get in the
way of pressurizing clients" (Fafchamps, 1996: 441). In a survey of fifty-eight
firms in Accra and Kumasi undertaken in 1993, no credit sales to relatives and
family members were actually recorded (ibidem; see also Ledeneva, 1998: 195-
200, for evidence on present-day Russia). In Vietnam, local entrepreneurs
apparently regard kin as more difficult to manage than non-kin because
criticizing or giving orders to kin is generally difficult and unpleasant
(Malarney, 1998: 282-83).
       What the foregoing discussion shows is that the existence of effective,
decentralized multilateral punishment mechanisms is far from certain in small-
scale societies where relationships are highly personalized. Different (subgame
perfect) equilibria may prevail and their characteristics may differ depending
upon the kind of internal relations that exist among the members of the
community or social group considered. A plausible hypothesis is the following
one : multilateral sanction mechanisms are more likely to be established and
reliable when the degree of proximity of intra-group relations has an
intermediate value. In other words, when relations are either too close or too
distant, it is doubtful that such mechanisms can operate effectively. When they
are too distant, information within the group may be imperfect either because of
objective barriers to communication or because the first two aforementioned
motives, (i) and (ii) may discourage people from conveying the information at
their disposal. When they are too close, on the other hand, the third motive, (iii),
may get in the way of effective information transmission or people may be
reluctant to punish group members (such as relatives) known to have indulged in
some fraudulent behaviour.         In-between these two situations, however,
informational conditions as well as motivations to punish may be conducive to
multilateral reputation and sanction mechanisms.
       Three remarks must be made at this stage. First, face-saving tends to be
an important consideration in small-scale societies where human relationships
are more or less continuous and highly personalized. Dealing with people who
are eager to save face in all circumstances is a difficult task that is typically

                                                                                         15


handled by an authority structure, often an elders' council or an elder figure. In
other words, decentralized mechanisms may prove unreliable to discipline such
people. Subtle negotiations under the auspices of an agency endowed with a
sufficient degree of legitimacy is likely to be the only effective manner to bring
rule-violators to heel in the kind of societies considered here. Erosion of the
influence and legitimacy of traditional authorities is therefore bound to
undermine a community's ability to regulate behaviour and control malfeasance
in agrarian societies. Contributing to such a process of erosion are several
forces that combine their effects, namely the rising levels of education among
the youth, their increased exposure to the ouside world, and their enlarged
opportunities to earn independent sources of income as a result of market
development.
   Second, assuming that multilateral sanction mechanisms work effectively, it
cannot be taken for granted that they are optimal whether in a static or a
dynamic sense. This point has not escaped Greif's attention. On the static level,
since the size of the group within which circulation of information makes such
mechanisms possible is exogenously determined by historical circumstances,
there is no a priori reason to believe that it is optimal given the available
trading opportunities (Greif, 1992 : 129-131). This conclusion continues to hold
true even while allowing for the possible extension of the sphere of honest
trading through the linking of multilateral reputation and sanction mechanisms
(see Aoki, forthcoming : chap. 4). On the dynamic level, on the other hand, the
existence of a multilateral sanction mechanism may discourage members of the
communities concerned from devising new institutions that may ultimately
prove more effective in overcoming trust failures in complex economies (where
the domain of exchange is wide, agents are mobile, economic transactions are
discontinuous, etc). In the other way around, the individualism prevailing in
societies which failed to establish multilateral reputation mechanisms may
provide an impetus for the development and perfection of third party
enforcement of claims.        The idea is therefore that societies are placed on
different institutional trajectories and that institutions which appear to be more
efficient at some point of the trajectory may well give rise, in a path-dependent
logic, to inferior institutions at a later stage.
       Third, honest behaviour has the best chances to prevail in close-knit
societies if social norms evolve that emphasize reciprocal fairness among
participants. According to the code of reciprocal fairness, `you behave the way
which you would like the others to behave, but only if they actually meet this
expectation' (Sugden, 1984: 774-75; Elster, 1989: 214). In other words, people
want to be kind to those who have been kind to them (the positive reciprocity
aspect) and to hurt those who hurt them (the negative reciprocity aspect) (Fehr
and Tyran, 1996 : 134 ; see also Rabin, 1993)6. These preferences underlie the

6 Note that reciprocal fairness is neither purely egoistic behaviour (since reciprocators
incur some personal cost in responding to other people's actions) nor purely altruistic

                                                                                               16


payoff structure of the so-called Assurance Game in which honesty becomes a
(non-unique) Nash equilibrium in the one-period play of the game7. Since the
nature of interactions in small groups is highly conducive to preplay
communication, individuals are better able to reveal and signal their intended
plans of actions and to learn about others' intentions. They have the opportunity
to reassure one another and form optimistic expectations about their mutual
behaviours so that honesty can be established (Baland and Platteau, 1996 : 92).
Moreover, reciprocity is more likely in long-term relationships since there are
more opportunities for reciprocation of both the positive and negative kinds
when parties interact longer (Fehr and Tyran, 1996). The important point is that
the socially efficient outcome is a possible equilibrium in a repeated Assurance
Game of finite duration8.
        This argument illustrates the proposition, supported by recent
experimental evidence, that the institutional context seems to affect human
behaviour (see Platteau, 2000a : chap. 7, for references). Thus, the impact of
reciprocity is negligible if contracts are completely specified. In other words,
competitive markets with completely specified contracts quickly converge to the
competitive equilibrium even if the resulting allocation is unfair by almost any
conceivable definition of fairness (except, of course, the notion of fairness based
on market-clearing as a means to prevent waste). By contrast, reciprocity plays
an important role when contracts are incompletely specified, even in the cases
where incomplete contract situations are embedded in a competitive market. In
other words, "the existence of opportunities for reciprocation may significantly
alter market outcomes" (Fehr, Kirchsteiger, and Riedl, 1998: 19 ; Fehr and
Gächter, 1998). Another way of formulating this conclusion is as follows :
market-like situations involving competition among many anonymous buyers
and sellers (and complete contracts) tend to induce self-regarding behaviour
while more personalized exchange settings tend to yield choices consistent with
other-regarding or relational preferences (Lane, 1991; Schotter, Weiss, and
Zapater, 1996; Bowles, 1998: 87-90).



behaviour (since "gifts are exclusively presented to people who gave something before, or
are supposed to give something back later") (Fehr and Tyran, 1996: 134).
7  In the extensive form of the game, honesty is the unique (subgame perfect) Nash
equilibrium.
8This result actually illustrates an important theorem of game theory which has been
proved by Benoit and Krishna (1985) following which, if a strategic-form game has
multiple equilibria (when it is not repeated) that give two or more different payoffs to each
player, then, under general conditions, the average payoffs in subgame-perfect equilibria of
long finitely repeated versions of this game are very close to any average payoffs attainable
in the infinitely repeated versions of the game (Myerson, 1991: 338). This theorem points
up the essential difference that exists between the Prisoner's Dilemma and other game
forms: while in the PD game, the results obtained under infinitely repeated versions
radically differ from those obtained in finite versions, this is not true of other game forms.

                                                                                            17


       It is probably too optimistic to assume that the required social norms are
properly internalized by everybody in the community.                     If there are some
inveterate opportunists (individuals with the payoff structure of a Prisoner's
Dilemma), social norms of reciprocal fairness can no more substitute for
sanction mechanisms : not only opportunists but also fair reciprocators
themselves may be tempted to cheat (out of a fear of meeting an opportunist). In
this situation, the ability of the community to quarantine opportunists appears as
a beneficial outcome that prevents the other individuals from being discouraged
to trade. Yet, this is only a partial equilibrium effect. As a matter of fact, the
existence of community mechanisms (or business networks) that expel
opportunists may produce a negative externality on the functioning of the
region-wide market. By absorbing honest agents, they actually increase the
density of dishonest individuals involved in anonymous market exchanges and
there is thus a possibility that informal trust-enforcing mechanisms are
inefficient in general equilibrium even though they enhance efficiency in partial
equilibrium (see Kali, 1997).


4.     Communities as producers of local public goods

The framework of human interactions is obviously different when a group or
community is confronted with the task of producing local public goods, whether
it is concerned with the building of collective infrastructure or the management
of common-property natural resources. Instead of pairwise interactions, what
we have here are strategic situations that involve numerous players at the same
time, that is, `collective-action' games (Dixit and Skeath, 1999 : chap. 11).
There are well-known arguments to the effect that small groups or communities
are comparatively apt to produce collective goods. One of them is the so-called
`incentive dilution' mechanism highlighted by Olson (1965). Free riding is a
strategy whereby an individual trades a reduction in his own effort, from which
he alone benefits, for reductions in the income of the whole group, which are
shared among all members. Therefore, as the size of the group increases, the
terms of this exchange become more and more favourable to the free rider (since
shares are diluted), and vice-versa when the size of the group decreases9.
Moreover, the advantage of small groups is not only that they prevent incentives
from being excessively diluted, but also that they allow for agreements to be
reached among the people concerned at low negotiation costs, which include the



9Note interestingly that the same incentive dilution process occurs with a Chicken game.
As the size of the group increases, the structure of the game modifies itself from a Chicken
game into that of a Prisoner's Dilemma, and, as a result, the collectively rational outcome
which was initially established degenerates into the Pareto-dominated non-cooperative
outcome (see Baland and Platteau, 1996 : chap. 89-90).

                                                                                   18


costs of communication and bargaining as well as, possibly, those of creating
and maintaining a formal organization.
       It therefore seems that small size is an unambiguous blessing. A closer
look at the issue none the less reveals that this is not necessarily so (see Baland
and Platteau, forthcoming, for more details). Thus, if we consider the classical
problem of the production of a pure public good (benefits are nonexcludable and
nonrival), an increase in the number of agents in the society which leaves the
wealth of the original agents unchanged does not reduce the aggregate provision
of the public good. In fact, when the additional agents are contributors, the
equilibrium level of the public good also increases, in a way similar to that
resulting from an increase in the aggregate wealth of the contributors. In other
words, it is possible that the original participants reduce their individual
contributions as a result of the entry of additional contributors, yet this reduction
will be smaller than the total contribution of the new agents. When the new
agents do not bring any wealth of their own but receive transfers from the
original agents, the total production of the public good may increase, decrease,
or remain constant.
       A second possible situation arises when users of the collective good,
typically a common property resource, share the benefits from joint use in direct
proportion of the relative amounts of their appropriation efforts which they
freely decide. In this case, as expected, an increase in the number of users is
bound to reduce the efficiency of the joint use of the resource by raising the
Nash equilibrium level of its exploitation. This is because any new agent
entering into the resource domain will always decide to exert a positive amount
of effort (since this is the only way of earning a share of the income flow yielded
by the resource). The original agents will react by diminishing their own level
of effort, but never to such an extent that the aggregate level of effort is being
reduced.
       Finally, it may be the case that users benefit from the collective good
produced (or the public bad avoided) in proportion of their pre-determined share
or `interest' in that good.       Shares are often directly related to the users'
ownership of the relevant factors of production (e.g., the number of boats which
a particular fisherman owns relative to the total fleet owned by local fishermen,
or the size of the irrigable landholding of a particular farmer relative to the total
service area operated under an irrigation system). In such conditions, the impact
of an increase in group size will depend a lot on whether the existing shares
have to be redistributed or not. Thus, if the entry of new users increases the
number of beneficiaries by adding new `shares', if it does not increase the costs
of producing the collective good (say, the maintenance cost of an irrigation
canal) and if, moreover, the new users voluntarily contribute a positive amount,
the new equilibrium level of aggregate contributions will be higher. If, on the
contrary, they decide not to contribute, this level will remain unchanged.

                                                                                  19


       Conclusions differ when the existing shares must be redistributed among a
larger number of users. The impact of such a redistribution on the aggregate
provision of the collective good will depend on the precise pattern of
redistribution that takes place. The important result is that, even in this case, an
increase in group size does not necessarily reduce the level of the public good in
the local economy. If, for example, the new users contribute in equilibrium and
have obtained their `shares' from small agents, who previously did not
contribute, the aggregate supply of the collective good will be raised.
       An analysis of the impact of wealth inequality on the provision of
collective goods can be conducted along the same line (for detailed arguments
and proofs, see Baland and Platteau, 1997, 1998, 1999, forthcoming). Such an
analysis calls into question the simple view according to which more equality is
conducive to collective action. At the same time, it disproves Olson's extreme
conclusion that `In smaller groups marked by considerable degrees of inequality
(...) there is the greatest likelihood that a collective good will be provided'
(Olson, 1965: 34). The fact of the matter is that the richer an agent (in income,
wealth, or `interest'), the larger his contribution to the collective good.
Voluntary contributions are increasing with wealth (for empirical confirmation,
see, e.g., Wade, 1988; White and Runge, 1995; Gaspart et al., 1998;
Shanmugaratnam, 1996). In this sense, Olson's intuition is partially confirmed:
"the greater the interest in the collective good of any single member, the greater
the likelihood that that member will get such a significant proportion of the total
benefit from the collective good that he will gain from seeing that the good is
provided, even if he has to pay all of the cost himself" (ibidem).
       The fact that contributions are positively related to wealth does not
however imply that an equalizing redistribution of wealth necessarily increases
the aggregate provision of the public good. Thus, when some poor agents do not
contribute, redistributing income from those agents to contributing agents
increases provision of the collective good. In contrast, redistribution between
contributing agents has ambiguous effects. Yet, it remains true that public good
provision is highest when a single agent concentrates all the wealth, as Olson
correctly hypothesized.
       The above conclusions have been derived from the analysis of several
situations constructed as non-cooperative games.         When the possibility of
regulation is contemplated, the impact of inequality ­like the influence of group
size on negotiation costs (see supra)­is found to be less ambiguous. In the
presence of inequality, regulation is more difficult to design and implement
because regulatory instruments are imperfect and are often limited to uniform
quotas or constant tax rates (Johnson and Libecap, 1982 ; Libecap and Wiggins,
1984 ; Libecap, 1990 ; Baland and Platteau, 1998, 1999 ; Gaspart and Platteau,
forthcoming). In particular, as resource users are more different, the regulated
solution tends to be less efficient. It is also more likely that some of the users
will be hurt by the regulation proposed, in the absence of compensatory transfer

                                                                                   20


schemes. As a result, if we require the regulated solution to Pareto-dominate the
ex ante unregulated situation, as inequality rises, the Pareto-dominating
regulation, if it exists, tends to be less efficient.
      On the other hand, it must be reckoned that wealth inequality may
encourage collective regulation if it favours the emergence of leaders committed
to the production of some collective good. This takes us back to the previous
point (the non-cooperative game-theoretical argument) since the problem of the
creation of a regulatory organization can be interpreted as a problem of
voluntary contribution to a collective good.          In this instance, though, the
collective good is subject to non-convexities in so far as a large part of the costs
involved are in the form of set-up costs. Moreover, agents are constrained in
their capacities to contribute, in so far as most contributions take on the form of
labour efforts spent with a view to setting up and developing the organization. In
such circumstances, users with the biggest interests in the resource contribute
more effort and take more initiative towards meeting that objective.
      The argument can be further extended to other important aspects of
collective regulation, namely, the revelation of correct information, the
enforcement of rules, the monitoring of collective activities, the implementation
of sanctions, etc. As a matter of fact, users with a larger interest are those who
can be expected to see to it that rules are effectively abided by and to set an
example of rule abidance, and to spend time and other resources in order to
ensure proper monitoring and sanctioning of users' behaviour. Large users may
actually find it in their interest to make excess contributions to the running of the
regulatory scheme for fear that it might collapse. Solid evidence actually exists
in support of this hypothesis that the costs of initiating and performing
regulatory tasks are often borne by the economic elite (see, e.g., Wade, 1987,
1988; Sengupta, 1991; Braverman et al., 1991; Menzies, 1994; Gaspart and
Platteau, forthcoming).
      On the above grounds, the idea that small and homogeneous (from the
standpoint of wealth distribution) groups or communities are the ideal vehicle of
collective action thus appears to be seriously challenged. It must nevertheless
be borne in mind that the underlying theoretical reasoning is based on the
description of Nash equilibria in one-period games or games of finite duration.
Like in the previous sections, it can be argued that the games played by
members of agrarian communities are closer to infinitely repeated games. In
such games, as we know, there is a plethora of equilibrium strategies and, in
particular, the efficient outcome can be possibly sustained in equilibrium. The
problem then becomes one of equilibrium selection.
      As argued in the previous sections, in so far as small-group settings
correspond to frameworks of (infinitely) repeated interactions with perfect
information among participants, `cooperation' can be sustained as an
equilibrium of the game (see, e.g., Ostrom, 1990 ; Sandler, 1992 : chap. 2 ;
Baland and Platteau, 1996 : chaps. 4-6 ; Dixit and Skeath, 1999 : chap. 11).

                                                                                           21


This is, of course, because repetition and perfect information make reputation
effects and (decentralized) punishment possible10. The efficient outcome is all
the more likely to obtain if community members have multiplex relationships
that get them involved in several games that are played simultaneously. In such
circumstances, indeed, a free-rider in one game may be punished in another
game (Ellickson, 1991 ; Aoki, forthcoming). Moreover, because people know
each other well, they can communicate easily and this ability, by making social
norms clear to all members (regarding the expected behaviour, the expected
sanctions, etc.), can help them coordinate on the `good' equilibrium. Easy
communication among people sharing a same culture (another, critical
dimension of homogeneity) can also help them avoid misunderstandings and, by
spreading information about the efficacy of the detection of free-riding
mechanism, it can help build trust within the group.
       Note that, if production of the collective good entails a coordination
problem or threshold effects, easy signalling of willingness to contribute by
members assuming leadership positions makes `cooperation' a more likely
outcome in small than in large groups, and this is obviously true even in one-
shot games. The same conclusion applies if, as suggested at the end of Section
3, members of small groups are fair reciprocators instead of opportunists so that
the problem is really one of mutual assurance. When the group is quite large (so
that the participation of one more person does not make much difference), it
may even be the case that an Assurance Game gets transformed in a PD so that
everyone wants to shirk11. Finally, because the fruits of their actions are more
closely and immediately visible, it may be easier in small groups to find
volunteers or benefactors ready to expend personal effort to benefit the whole
(Dixit and Skeath, 1999 : 374).

       Arguments to the effect that face-saving considerations or values centered
on honour and dignity may break reputation and punishment mechanisms (see
supra) are apparently less compelling in collective-action problems than in
situations involving pairwise encounters. An obvious reason is that secrecy for
the purpose of avoiding shame, either for oneself or for others, may not be
possible if actions are simultaneously observed by all participants. In other
words, a member of the community is not in a position to save a relative, a

10As pointed out by Dixit and Skeath (1999 : 373), in large groups one's person free-riding
may be hard to detect, and inflicting severe punishment on a person without being sure of his
guilt beyond a reasonable doubt is not only morally repulsive but is also counterproductive
("the incentive to cooperate gets blunted if cooperative actions are susceptible to punishment
by mistake").
11This can happen if the individual benefit from a pure public good as well as the individual
cost of participation are both a function of the number of participants (Dixit and Skeath,
1999 : 365-367). Note that the same sort of `incentive dilution' may occur when starting from
a Chicken Game since this game can also transform itself into a Prisoner's Dilemma as group
size increases (Baland and Platteau, 1996 : 89-90).

                                                                                           22


neighbour or a friend from the ordeal of disgrace because peer pressure is
exercised as soon as defection is detected. In such conditions, the fear of public
humiliation should act as an powerful deterrent to free-riding or rule violation.
This means that the more sensitive the people are to shame feelings the more
effective the punishment mechanism can be expected to be in multiperson
games. In the extreme, such feelings can be sufficiently strong to make material
sanctions unnecessary.         As pointed out with respect to traditional Moose
culture in Burkina Faso :

        "Activated by social rebuke and the accompanying public humiliation, the feeling of
        shame appears as the most formidable weapon in the service of the traditional Moose
        pedagogy (the moaga).       Above the individuals, indeed, this feeling asserts the
        supremacy of social judgement and constitutes a powerful regulating mechanism to
        which everybody submits... Since a person can exist only through collective opinion,
        it is collective opinion that rates people and rare are those who are willing to incur
        the risk to defy it. The point is that its verdict is merciless and without appeal"
        (Badini, 1994 : 146-47 ­my translation).

There is, however, another side of the coin. As a matter of fact, the efficacy of
small-scale collective actions is being dented as a result of the susceptibilities
and preference for pride that are characteristic of human beings living in the
context of close and highly personalized relationships. Thus, constant concern
to avoid offending people motivates group or community members to engage in
long discussions aimed at reaching consensus and convincing everyone that his
opinions have been attentively listened to. It is therefore not surprising that
unanimity rules tend to be preferred not only in traditional village communities
but also in small groups in general.
      In fact, behind people's concern about establishing consensus among
them there often lie strategic considerations. More precisely, they believe that
someone who does not agree with a decision or a rule feels perfectly entitled to
violate it once it is being implemented.                In order to ensure maximum
compliance, group members therefore deem important to enlist the support of all
of them even if this entails high transaction costs during the decision-reaching
process.     The underlying problem is evidently that the very concept of
democracy as implied by the majority rule is not accepted in traditional village
societies : the only principle that appears to make sense is that we can only
follow the rules which we are in agreement with. Under such circumstances,
serious tensions and conflicts are bound to arise at the implementation stage if
care has not been taken to convince everyone about the rightfulness of the rules
to be applied. Another, even more worrying implication, is that people do not
feel tied to rules decided at higher-than-village levels, such as laws and
procedures enacted by a central state (on this, see also the fascinating insights
into the mindset of the Hungarian peasantry in the first half of the 20th century,
in Janos, 1982 : 241-242).

                                                                                            23


       In support of the above, there is evidence that, when group size is
adjustable (such as in the case of self-help groups in peri-urban or urban
contexts) and when an increase in membership would provide economic benefits
(such as in insurance or credit groups where risk diversification requires a
sufficient number of members), members are reluctant to enlarge their group lest
they should lose the advantages of consensus-seeking procedures (Abraham,
Baland and Platteau, 1998 : 37-39).
       Also noteworthy is the fact that in traditional settings radical sanctions are
avoided whenever possible and protracted negotiations are typically engaged in
order to allow the deviant to amend himself in a honourable manner. A defector
must be an unrepentant rule violator or an inveterate troublemaker to be liable to
exclusion, and his exclusion occurs only after repeated warnings and
conciliatory attempts. Interestingly, from a recent study of community-based
effort regulation schemes in the fishing area of Kayar in Senegal, it is apparent
that fishermen resist too strong and rigorous inspection procedures and
sanctioning mechanisms on the ground that people should not act as policemen
vis-a-vis each other when belonging to the same community (Gaspart and
Platteau, forthcoming).
       So far, I have reasoned as though the community is egalitarian. This is
not true in many traditional societies where status or rank considerations play an
important role. Owing to substantial differences in local ranks, implementing
uniform punishment procedures is difficult, not to say impossible. As a matter
of fact, a given sanction is comparatively costly for high-status individuals since
they have to suffer a greater loss of dignity and self-respect than common
people. The former may not oppose the formal decision to apply an identical
treatment to all group members, yet, in actual practice, they take it for granted
that their traditional status should afford them a favourable treatment in all
circumstances. Such is the basis of justice in traditional setups. None the less,
such discriminatory attitude on the part of high-status individuals is not
problematic inasmuch as it is considered legitimate by the other members of the
village community12. Deemed equally legitimate is the fact that the traditional
elite repeatedly use their favourable rank position to abstain from exerting effort
in situations where the largest total social payoff is attained when some
community members are allowed to shirk, such as in the multiperson prisoner's
dilemma or the `chicken' forms of the collective-action problem (for a
presentation of such games, see Dixit and Skeath, 1999 : 363-374).
       Such advantages attached to the elite do not create problems as long as
they are regarded as the normal attributes of ascriptive power and status, or the
expected returns for leadership functions. If excessive privileges are being
extracted by traditional power figures, however, consensus around their

12This implies that commoners recognize the higher cost of punishment for high-status
individuals when formally identical sanctions are meted out to all participants irrespective of
rank and status.

                                                                                24


authority breaks down and communities or groups are plagued by all sorts of
incentive problems that ultimately cause their performances to flop. The latter
sort of predicament becomes more likely as the opening of village communities
to new market opportunities may tempt traditional power-holders to use their
privileged position to appropriate an unfair share of new collective incomes
(Baland and Platteau, 1996 : 270-78).
       Even while their behaviour is beyond reproach, the erstwhile local rank
hierarchy tends to be called increasingly into question by the young generations
who are attracted to democratic ideals and to individual-centered values through
education and other forms of exposure to modern lifestyles and outlooks. As
competition for leadership positions becomes more open, power struggles risk
undermining the cohesion of small groups. It is therefore essential that the
rivalrous tensions thereby generated are kept under check by means of some
mechanism or rules. A consensual mode of functioning is not very helpful to
appease such interpersonal tensions since they arise from the necessity to choose
some persons and to reject others. An interesting finding of the aforementioned
study of self-help groups in Kibera slum (Nairobi) by Abraham, Baland and
Platteau (1999) is the following : many groups apply strict rules that define a
code of politeness and courtesy for the meetings and are backed by effective
sanctions (fines and temporary exclusion).
       The avowed reason adduced by sample group members is that such rules
are badly needed to prevent resentments, frustrations and jealous feelings, such
as those born of unsuccessful contests, from being overtly expressed and from
disturbing the meetings and the decision-making processes. In particular, there
is often a formal requirement that the elected president of a group or association
be politely addressed to.      Moreover, some groups have adopted so-called
confidentiality rules that have been established with a view to avoiding outside
disclosure of internal disagreements. The concern is indeed that leakages about
internal problems could easily exacerbate intra-group tensions and threaten the
viability of the collective.
       Another important consequence of the diffusion of individual-centered
values amongst the younger generations is the claim that "it is morally wrong to
make exaggerated demands for obedience, or to manipulate or coerce the
individual beyond a certain point in order to achieve social or political ends"
(Stone, 1990 : 151). Moreover, considerations of honour and vulnerability to
public reprimands gradually lose importance. An immediate implication is the
progressive erosion of the influence of traditional social norms that serve to
instill into individuals feelings of identification with their community and to
enjoin them to sacrifice their self-interest for the collective good.
       Before turning to the role of rural communities as property-rights
enforcers, it is important to draw attention to a methodological pitfall. We have
pointed out above that group size may be constrained by sustainability
considerations arising, for example, from the need to maintain group cohesion

                                                                                      25


through consensus-seeking procedures. If there are scale economies of some
sort, a larger group would bring more welfare to the members provided, of
course, that the group's internal mode of operation can be modified so as to
allow effective functioning on a larger scale.           Since empirical studies that
attempt to relate group size with various criteria of performance, either on the
level of internal functioning or output delivery, typically fail to take these
dynamic considerations into account, wrong conclusions can be easily derived
from them. More precisely, if it is empirically observed that small groups
perform better than large ones ­which, incidentally, is not a systematic
finding13­ , it may not be concluded that small groups are the right vehicle of
collective action (see, e.g., Sengupta, 1991 ; Tang, 1992, 1994).                Such a
conclusion is unwarranted because the possibility cannot be ruled out a priori
that had they been able and encouraged to adjust their modus operandi
appropriately large groups could have performed much better than they actually
did whereas small groups have attained their maximum potential. The same
remark applies to conclusions to the effect that homogeneous groups perform
better than more heterogeneous ones.


5. Communities as property rights enforcers

       In land-abundant areas, land is traditionally the corporate property of the
whole village community whose representatives are in charge of allocating
rights of use and taking the steps deemed necessary for a proper management of
the resource. As land pressure rises, the system gradually evolves towards
individualization of tenure, implying that the prerogatives of village or lineage
authorities diminish while the rights of the land possessor increase accordingly.
Individualization makes its way through the gradual extension of use rights (for
example, the right to recultivate the same plot of land even before the normal
period of fallow has elapsed, or the right to plant trees and to bring other
improvements to the land, are increasingly recognized) and, above all, through
the addition of transfer rights. In an ascending order of hierarchical importance,
the latter comprise the right to lend the land along traditional lines (that is, as
part and parcel of a wider relationship of reciprocal exchange between two
families and lineages), the right to give it, to bequeath it, to rent it out (against
cash payments) and, eventually, to sell it.
       Individualization of land tenure is also reflected in the growing autonomy
enjoyed by farmers regarding their decisions of use and, particularly, regarding
their decisions to transfer land. Thus, in the initial stages of individualization,
the rights to rent out or to sell land parcels are seriously circumscribed by the
requirement that land ought to remain within the family or the lineage. It is the

13Some studies indeed reach the conclusion that a positive relationship prevails between
group size and various performance indicators (Esman and Uphoff, 1984 : chap. 5).

                                                                                  26


responsibility of the lineage heads to see to it that this condition is duly abided
by and this is why their permission is explicitly requires before any sale of land
can take place. Such condition is however gradually relaxed as land becomes
more scarce (see, e.g., Boserup, 1965 ; Noronha, 1985 ; Bruce, 1986 ; Barrows
and Roth, 1989 ; Migot-Adholla et al., 1991 ; Platteau, 1996, 2000a : chaps. 3-
4 ; 2000b). The fact remains, however, that formalization of land rights through
centralized registration and titling does not take place under this scenario since it
requires a deliberate intervention of the state in community affairs. In other
words, absent such a state intervention, the community remains the enforcer of
land property rights, whether they are of a rather communal nature when land is
abundant or of a rather individualized nature when it has become scarce.
       Reliance on local communities for allocating and enforcing land rights
offers significant advantages. First, contrary to formal procedures such as land
titling which are costly and impose definitive land rights, informal practices at
village level are cheap (they economize on information costs) and flexible.
Second, even though social differentiation is not to be underestimated, village
communities in land-abundant areas tend to provide social security to all their
members and to ensure that everybody can participate in new opportunities.
Such considerations of social security and equity usually dominate pure
efficiency concerns, which should be regarded as a positive contribution in a
generally insecure economic environment (Lawry, 1993: 73). Third, enduring
customary systems tend to receive remarkable consensus, in particular
consensus on the normative order justifying land claims (Saul, 1993). And,
fourth, contrary to a widespread view, informal tenure systems embedded in
community life do not necessarily hamper the development of land markets and
formal sanctioning of land rights does not necessarily stimulate such a
development. In other words, titling turns out to be neither a necessary nor a
sufficient condition for land market activation. In the same manner, tenure
rights backed by the informal village order do not seem to discourage
investment.
       Two recent studies, one by André and Platteau (1998) on western Rwanda
and another one by Baland et al. (1999) on Central Uganda, have thus found a
high incidence of informal land market transactions that appear to be well
secured by informal procedures (land sale transactions are typically attested by
written documents established in the presence of witnesses). In the village
studied by the former authors, almost 30% of land parcels owned by local
inhabitants (representing about the same proportion of total land owned) had
been acquired through market purchases. The closure of the land frontier in
Eastern Rwanda and the ending of migration possibilities in both Uganda and
Zaïre (causing reverse migration movements) combined with rapid population
growth suddenly caused an enormous amount of land hunger that was not
counterbalanced by an expansion of non-agricultural employment opportunities.
As a result, the local land market became increasingly active, thus

                                                                                  27


demonstrating that spontaneous individualization of land rights, unassisted by
any process of titling or registration at the state level, can be quite effective in
activating land market transactions (almost all land sales were actually illegal).
       In Uganda, informal land sales are even more frequent than in Rwanda :
close to 50 percent of the total land area owned by the sample households in 36
different villages have been purchased, which is probably one of the highest
recorded rates of land market activity in rural Africa. Clearly, what matter are
de facto rather than de jure rights and de facto rights are those sanctioned and
enforced by the village informal order.       This is again attested, albeit in a
different manner, in the case of Kenya. There, exercising the full prerogatives
formally attached to freehold tenure rights guaranteed by legal titles would
contravene important community norms that are embedded in the indigenous
tenure system. As a matter of fact, many owners of titled lands do not consider
that they are entitled to transfer their lands outside the lineage or to make
permanent transfers without having previously obtained the approval of their
family or community.
       It is equally worth noting that the District Land Control Boards in charge
of approving land sales are frequently reluctant to permit transactions which
would leave families (and their descendants) landless and destitute. That is why
they insist that all adult members of the household (including women) of the
title-holder are to be present at the hearing to indicate their agreement with the
sale. The government has actually sanctioned this de facto situation since a
presidential directive aimed at minimizing land disputes requires the agreement
of family members in addition to that of the title-holder prior to any sale or use
of land as collateral (Haugerud, 1983 : 84 ; Mackenzie, 1993 : 200 ; Pinckney
and Kimuyu, 1994 : 10). In Zimbabwe, likewise, a proposal by a land tenure
commission appointed by the government (October 1994) provides that
individual farmers should be given the right to own their land, but their right to
buy and sell it should be subject to the approval of the traditional village council
(the sabuku) which in pre-colonial days used to be vested with the prerogative of
allocating local lands (The Economist, 21-27 January 1995 : 49).
       On the other hand, there is growing evidence that basic informal tenure
rights are sufficient to provide investment incentives to the farmers (Bruce and
Migot-Adholla, 1994 ; Besley, 1995 ; Gavian and Fafchamps, 1996 ; Brasselle,
Gaspart, and Platteau, 1998 ; Baland et al., 1999) so that there is no need for the
state to establish a sophisticated and costly administrative machinery for the
purpose of formally securing land rights by means of titles. Enforcement of land
rights by village communities therefore seems adequate not only to allow the
development of land market transactions but also to induce investment
behaviour and promote productivity gains in agriculture when the scarcity of
land makes such behaviour and such gains necessary.

                                                                                    28


       What is remarkable about the above evidence is that it bears ample
testimony to the ability of village communities to evolve as the need arises.
When land scarcity requires that tenure becomes more individualized, they
continue to play their role as property rights enforcing agencies, yet on a smaller
scale and in a different way compared to times of land abundance. From there,
it would nevertheless be wrong to conclude that village communities do actually
solve the rights enforcement problem to complete satisfaction. Their main
limitation in this respect arises from their parochial character that tends to be
reflected in chauvinistic attitudes towards outsiders. Owing to fears aroused by
actual or anticipated scarcity of their critical productive asset, traditional
displays of tolerance and hospitality give way to crude and open manifestations
of discrimination against strangers (see Platteau, 2000b).
       Original occupants may not hesitate to use violence to oppose the transfer
of lineage lands to outsiders, committing acts of sabotage, looting, burning and
theft on the property and crops of the new landholders. An extreme example of
the excesses to which this opposition may lead is the so-called `Manifesto of the
Oppressed Negro-Mauritanian'. Written by an extremist group belonging to the
black (haalpulaar) community of Mauritania, it is an aggressive reaction to the
post-1983 introduction of private land rights conferred (by adjudication) upon
stranger owners over irrigated lands located on the right bank of the Senegal
river.   In this manifesto, the Negro-Mauritanians are invited to use any
conceivable means to prevent their customary lands from passing into the hands
of the Beydane elite (of Moorish origin), that is, "to boycott, ban, kill if needed,
all those who encourage the sale of land; destroy, burn the possessions of these
strangers who come to develop your lands while the land should belong to our
villages" (quoted from Bayart, 1989 : 82).
       Since cultivators on both sides of the river belong to the same
(haalpulaar) community14, and since some people residing on the left bank (that
is, on the Senegalese side) used to cultivate lands located on the right bank (that
is, on the Mauritanian side) over which they had customary access since
generations, the transfer of lands on the right bank to Moorish owners was
bound to create serious tensions between the two neighbouring countries. These
tensions could not be contained and led to a war between Senegal and
Mauritania in April 1989 (Leservoisier, 1994 : Chap. 10 ; Maiga, 1995 : 54-57).
       It is also worth noting that, on the left bank, there is growing resentment
and mounting opposition against transfers of land through `illegal' sales ­chiefs
or other powerful individuals sell customary land without the consent of the
community and unduly appropriate the proceeds­ to rich purchasers.
Opposition is especially directed at capitalists or civil servants from Dakar and
at the well-to-do (maraboutic) elite which acceded to economic prosperity
through its leadership role in the export oilseeds boom in the Sine-Saloum. It is


14The haalpulaar (`Toucouleur') community is a mix of Fulani and Serere ethnies.

                                                                                  29


easy to multiply examples of this kind. In the Peanut Basin itself, for example,
we learn that every year villagers "meet to trace the borders of the land and set
the rules so that everyone bands together to keep the land in the village" (Golan,
1990 : 15).
       In Ghana, as the frontier land became gradually exhausted, indigenous
(Akan ) ideology began to reassert with vigour "the inalienable rights of the
native custodians of the land, and the inalienable rights of individual usufruct"
(Robertson, 1987 : 77). In Western Burkina Faso, local residents fear a flood of
migrant settlers (mainly Moose) into their ancestral lands, but, so far, thanks to
the strength of their traditional social structure (based on agnatic lineages), they
have succeeded in effectively blocking further settlement on their territory (Saul,
1993: 81-2 and Laurent, 1995 ; see also Riddell et al., 1987 : 31 for Zaïre).
       Land alienation to strangers, whether through sale transactions or state
mandatory allocations, is not necessarily opposed by the use of open violence.
Milder and more subtle forms of hostility such as malpractices of deceit,
manipulations and double deals directed against strangers are frequent
occurrences. In Kenya, for example, "farmers who have pledged their land titles
as security for loans are sometimes tempted into selling their land to strangers
without informing them of the charges on the land", which leaves them free to
default on their loans if they wish (Shipton, 1988: 111). Also, some farmers sell
their land to several (stranger) buyers at once or agree to sell "in the expectation
that after they have collected their money, the sale will be ruled void by the local
land control board and district registrar and they will succeed in evading
repayment of the money paid by the buyer" (ibidem).
       Evidently, such discriminatory postures of indigenous communities
towards strangers entail both efficiency and equity costs. Efficiency costs arise
from all the transaction costs and the ensuing land market imperfections that are
created by distrust and opposition between indigenous and stranger farmers. As
illustrated above, land may be prevented from accruing to the latter, whether
through land sale, land rental, or land adjudication processes. To the extent that
strangers are more performing than indigenous members of the host
communities, efficiency losses will result.         Sometimes, resistance against
allocation of land to strangers does not take place immediately but only after
they have demonstrated their ability to manage land more successfully than the
original residents. Thus, in the Zorgho region in Burkina Faso, some areas of
suitable land were developed under the auspices of the PDLG (Projet de
Développement Local du Ganzourgou) and granted to immigrant farmers who
were interested in irrigated agriculture.        Indigenous farmers preferred to
continue their traditional rainfed agricultural practices and did not show any
interest in new agricultural methods. Yet, when the land improvement scheme
proved successful, indigenous farmers reacted by opportunistically claiming
back what they consider as their own `ancestral' lands and by demanding the

                                                                                            30


expulsion of all stranger farmers (personal communication of Hubert
Ouedraogo).
       Equity costs are the direct result of the aforementioned fact that certain
communities may retain surplus land for themselves (say, because they want to
have reserve land available for future generations) while farmers from other
communities may be land-hungry. A well-functioning land rental market might,
of course, overcome such imbalances in land endowments, at least for a certain
time period. None the less, the fear of losing ownership rights if land is thus
rented out to stranger farmers may act as an impediment to rental contracts.
       In the above circumstances where serious tensions develop between
indigenous communities and immigrant groups as a result of growing actual or
anticipated land scarcity, land markets may therefore be prevented from working
in a satisfactory manner. Or, more exactly, land transactions may carry low
transaction costs when made between people native of the resident community,
yet entail considerable transaction costs as soon as land deals involve
strangers15.


6. Policy discussion

A dominant view among international donor agencies is that communities or
groups are a powerful lever of economic and social development and that many
opportunities have been lost in the past by not tapping their potential. True, like
markets and the state, communities are fraught with imperfections, yet these
imperfections can be largely remedied through so-called capacity-building
programmes aimed at strengthening the collective action abilities of village
societies. In other words, to say that communities are `poor but efficient' may
be an exaggeration, yet the hope is entertained that they are at least potentially
efficient.
       The present paper has actually provided illustrations demonstrating the
need to upgrade the organizational skills of traditional communities or groups.
For example, there is substantial room for adapting and modernizing informal
insurance schemes by building awareness among members about the advantages
of going beyond the logic of balanced reciprocity, or, if that objective is too
difficult to achieve, by introducing credit as a substitute for insurance so as to
remain within the purview of such a logic. It is also possible to concentrate


15 Unfortunately, there is no guarantee that a centralized process of registration and
enforcement of land rights will prove more successful. Indeed, if the allocation of land is not
deemed fair and legitimate by the indigenous communities, important transaction costs will in
any event arise that impair the efficiency of land and other rural factor markets. Such costs
follow from the uncertainties created by a situation in which land rights granted to immigrant
farmers (through formal adjudication or through state-enforced market transactions) are likely
to be seriously disputed by a majority of local residents (Platteau, 2000b).

                                                                                31


efforts on imparting new management skills upon group members so that larger
insurance groups become viable or a wider variety of risks become insurable.
When the insurance pool is enlarged, greater risk diversification can be achieved
whereas insuring a wider variety of risks may have the effect of increasing the
frequency of damage occurrences and payments of indemnities enough to
maintain balanced accounts among individuals within the group.
       Institutional support measures can also help community-based groups and
associations to devise rules and sanction mechanisms that are susceptible of
achieving greater efficiency in the production of local public goods. Better
performances may thus require that group size be increased (if scale economies
exist), or that membership be made more heterogeneous (e.g., if people of
different ethnic affiliations have complementary skills). But to be viable, larger
and more heterogeneous groups require the adoption of more sophisticated rules
and procedures. Concomitantly, accounting skills and the discipline of regular
record-keeping are an important condition for the success of local organizations
in tackling the complex problems that result from greater market penetration and
the multiplication of economic opportunities (see, e.g., Datta and Kapoor, 1996 :
261-62).    For this strengthening process to take place, it is essential that
educated youth can get increasing access to leadership positions and
responsibilities.
       The above view is nevertheless tainted with excessive optimism. There
are indeed glaring instances of situations where imperfections of communities
call for active, complementary interventions by the state. For example, there is
considerable scope for a mediatory role of public agencies in the handling of
inter-community relations. This has been illustrated with regard to enforcement
of property rights. When land pressure becomes acute and there are stranger
landless farmers willing to cultivate customary fallow land, an awkward
situation may easily develop following the refusal of village authorities to
comply with the demand of these unfortunate farmers.             By decreeing a
redistribution of existing land rights, the state cannot be expected to solve the
problem thereby created. As a matter of fact, customary village authorities have
many means at their disposal to subvert state-imposed rules and formal property
rights when they run counter to their perceived legitimate interests. What is
therefore needed is an approach based on cooperation rather than confrontation
between the central state and the rural communities. Owing to the persisting
influence of traditional concepts of corporate land ownership and social identity,
it is clearly a more effective strategy for the state to negotiate acceptable
compromises with customary authorities so that the interests of stranger groups
or other communities are somehow taken into account.
       Finally, it must be reckoned that communities may have features that
make them unfit for the fulfillment of some tasks of modern economic
development. In such circumstances, they need to be replaced or substituted for.
Due to slow adaptability, the functioning of communities may thus become too

                                                                                  32


burdensome to effectively meet new emerging challenges, such as how to insure
fishermen against the increased risks of equipment damage and casualties
resulting from the motorization of canoes, or how to insure farmers against the
increased risks arising from the necessity to incur substantial cash expenditures
associated with the adoption of new agricultural technologies.
       For instance, consensual procedures may not give way quickly enough to
majority rules in spite of the multiplication of decisions and the continuously
changing character of the economic and technical environment. As a result of
this sort of institutional rigidity, young and dynamic elements of communities
may be induced to leave traditional organizations or to gradually lose interest in
them. The solution is obviously for them to form their own group through a
self-selection process. However, elders are then likely to feel that their erstwhile
authority is bypassed, that due respect is not any more paid to them, and that
their long-established wealth advantage is being threatened. If so, they will
oppose the actions of the new emerging groups and try to break their initiatives.
Dynamic elements may then be forced to choose between submitting and
complying with the elders' requirements, on the one hand, and escaping to an
urban setting where the repressive logic of traditional structures can no more
reach them, on the other hand.
       Only if village headmen and the like are enlightened enough to put the
interests of their community before their own, to overcome their frustrations and
to accept the need for change from a status- to a performance-based society can
one hope for constructive cooperation between the different strata of the
traditional society. Only then can one see the emergence of an atmosphere
conducive to the free blossoming of new initiatives and experiments by young
individuals who are relatively well educated and have been exposed to novel
ways of thinking and doing through their school and work experience. The
same conclusion applies, mutatis mutandis, when considering the potential role
of women in the economic and social development of rural areas.
       The existence of a strong power structure at local level may not only
undermine grassroot initiatives for the production of public goods but it may
also create a high risk that any central resource channelled through community
organizations will be captured by the village elite. The risk is especially high
when economic differentiation proceeds quickly and results in the emergence of
hierarchical and exclusive patron-client ties that succeed horizontal relationships
of mutual insurance and labour cooperation. Under these conditions, indeed, the
poor must accept increased personalized dependence to continue to ensure their
livelihood under bad shocks. Their bargaining strength falls as a result and it
then becomes easier for the new or the old elite to preempt state resources
entrusted to community organizations.
       There is actually some recent evidence that when the responsibility of
allocating central resources is delegated to local organizations (such as
panchayaths in Bangladesh), local elites tend to appropriate for themselves

                                                                                           33


whatever portion of the resources that they need and to let the poor have the
leftovers only (Galasso and Ravallion, 2000). It is obviously possible that the
informational    advantage      of    community-based         welfare     programmes        ­
communities know better than state agents which are the poor families within
themselves­ is outweighed by the aforementioned accountability or `capture'
problem ­communities may not be as accountable to their poor as higher-level
state agencies. In such instances, the poor are better targeted through a state
distribution system than through a decentralized mechanism (Bardhan and
Mookherjee, 2000).        Or, at least, as illustrated in the case of farmers'
associations in Taiwan, there is a need for government intervention to prevent or
offset dominance by local elites (Stavis, 1982-83).
       There is yet a second, more subtle sense in which it is true to say that "the
enthusiasm for community-based targeting in policy circles has clearly run well
ahead of the evidence" (Galasso and Ravallion, 2000). Indeed, in order to
minimize the costs of institutional support to local communities and
organizations, outside agencies frequently follow the practice of asking
communities to `elect' leaders. For a reason well explained by Esman and
Uphoff, such an approach is bound to produce perverse results and to be self-
defeating :

     "The most prominent members are invariably selected and then given training and
     control over resources for the community, without any detailed and extended
     communication with the other members about objectives, rights, or duties. Creating the
     groups through these leaders, in effect, establishes a power relationship that is open to
     abuse. The agency has little or no communication with the community except through
     these leaders. The more training and resources they are given, the more distance is
     created between leaders and members. The shortcut of trying to mobilize rural people
     from outside through leaders, rather than taking the time to gain direct understanding
     and support from members, is likely to be unproductive or even counterproductive,
     entrenching a privileged minority and discrediting the idea of group action for self-
     improvement" (Esman and Uphoff, 1984 : 249).

Here is therefore a tricky dilemma : acting through leaders enables outside
agencies to reduce costs of institutional support to tolerable levels, yet at the
same time it creates a serious risk to strenghten existing elites at the expense of
the grassroots. The risk is especially high if rural `strong men' are connected
with urban elites which are prompt to exploit the new policy of decentralized
development for their own benefit. This is why one should worry about the
spawning of local NGOs that are created at the (hidden) initiative of state
employees, whether they have been laid off as a result of structural adjustment
measures in the public sector or not. It is thus not hard to find instances where
opportunistic state agents have responded to reduced budget allowances by
initiating or gaining access to local organizations that are eligible for receiving
financial resources from outside agencies such as foreign NGOs, international
organizations and bilateral aid programmes.                 It would be ironical that

                                                                                34


decentralized development programmes intended for avoiding the heavy hand of
the state end up being (partly) captured by officials in the disguise of community
organizations.

                                                                                           35


References

Abraham, A., J.M. Baland, and J.P. Platteau, 1998, "Groupes informels de solidarité dans un
bidonville du Tiers-Monde : le cas de Kibera, Nairobi", Non Marchand (Editions De Boeck
Université, Bruxelles), N° 2, pp. 29-52.

Alderman, H., and C.H. Paxson, 1992, "Do the Poor Insure? A Synthesis of the Literature on
Risk and Consumption in Developing Countries", Policy Research Working Papers, WPS
1008, Washington, DC: The World Bank.

Alvoet, E., 1997, Les transferts privés en Indonésie, unpublished M.A. thesis, Department of
economics, University of Namur, Belgium.

André, C., and J.P. Platteau, 1998, "Land Relations Under Unbearable Stress: Rwanda Caught
in the Malthusian Trap", Journal of Economic Behavior and Organization, Vol. 34, N° 1, pp.
1-47.

Aoki, M., forthcoming, Towards a Comparative Institutional Analysis, MIT Press.

Badini, A., 1994, Naître et grandir chez les Moosé traditionnels, Paris & Ouagadougou : Sépia-
A.D.D.B.

Baland, J.M., and J.P. Platteau, 1996, Halting Degradation of Natural Resources ­ Is There a
Role for Rural Communities?, Oxford: Clarendon Press.

Baland, J.M., and J.P. Platteau, 1997, "Wealth Inequality and Efficiency in the Commons ­
Part I: The Unregulated Case", Vol. 49, N° 4, pp. 451-482.

Baland, J.M., and J.P. Platteau, 1998, "Wealth Inequality and Efficiency in the Commons ­
Part I: The Regulated Case", Vol. 50, N° 1, pp. 1-22.

Baland, J.M., and J.P. Platteau, 1999, "The Ambiguous Impact of Inequality on Local
Resource Management", World Development, Vol. 27, N° 5.

Baland, J.M., and J.P. Platteau, forthcoming, "Institutions and the Efficient Management of
Environmental Resources", in Mähler, K.-G., and J. Vincent (eds.), Handbook of
Environmental Economics, chap. 18, Amsterdam : North-Holland.

Baland, J.M., F. Gaspart, F. Place, and J.P. Platteau, 1999, "Poverty, Tenure Security and
Access to Land in Central Uganda: the Role of Market and Non-market Processes", Cahiers
de Recherche, Faculté des sciences économiques, sociales et de gestion, Development series,
University of Namur (Belgium).

Banfield, E.C., 1958, The Moral Basis of a Backward Society, Chicago: The Free Press.

Bardhan, P., and D. Mookherjee, 2000, "Decentralizing Anti-Poverty Program Delivery in
Developing Countries", mimeo, University of California, Berkeley.

                                                                                         36


Barrows, R., and M. Roth, 1989, "Land Tenure and Investment in African Agriculture:
Theory and Evidence", Land Tenure Center, LTC Paper N° 136, University of Wisconsin-
Madison.

Bayart, J.F., 1989, L'Etat en Afrique, Paris: Fayard.

Benoit, J.P., and V. Krishna, 1985, "Finitely Repeated Games", Econometrica, Vol. 53,No. 4,
905-22.

Besley, T., 1995, "Property Rights and Investment Incentives: Theory and Evidence from
Ghana", Journal of Political Economy, Vol. 103, N° 5, pp. 903-937.

Binmore, K., 1992, Funs and Games: A Text on Game Theory, Cambridge, Mass.: D.C. Heath
& Co.

Boserup, E., 1965, The Conditions of Agricultural Growth: The Economics of Agrarian
Change under Population Pressure, London: Allen and Unwin.

Bourdieu, P., 1992, The Logic of Practice, Stanford, Cal. : Stanford University Press.

Bowles, S., 1998, "Endogenous Preferences: The Cultural Consequences of Markets and other
Economic Institutions", Journal of Economic Literature, Vol. 36, N° 1, pp. 75-111.

Brasselle, A.S., F. Gaspart, and J.P. Platteau, 1998, "Land Tenure Security and Investment
Incentives: Further Puzzling Evidence from Burkina Faso", Cahiers de recherche, CRED
(Centre de Recherche en Economie du Développement), Department of Economics,
University of Namur, Belgium.

Braverman, A., J. L. Guasch, M. Huppi and L. Pohlmeier, 1991, "Promoting Rural
Cooperatives in Developing Countries: The Case of Sub-Saharan Africa", World Bank
Discussion Papers, 121, Washington, D. C.: World Bank.

Bruce, J.W., 1986, "Land Tenure Issues in Project Design and Strategies for Agricultural
Development in Sub-Saharan Africa", Land Tenure Center, LTC Paper No. 128, University of
Wisconsin-Madison.

Bruce, J.W., and S.E. Migot-Adholla (eds.), 1994, Searching for Land Tenure Security in
Africa, Dubuque, Iowa: Kendall/Hunt Publishing Cy.

Coate, S., and M. Ravallion, 1993, "Reciprocity Without Commitment: Characterization and
Performance of Informal Insurance Arrangements", Journal of Development Economics, Vol.
40, N° 1, pp. 1-24.

Cox, D., and E. Jimenez, 1990, "Achieving Social Objectives Through Private Transfers -- A
Review", The World Bank Research Observer, Vol. 5, N° 2, pp. 205-18.

Cox, D., and E. Jimenez, 1997, "Coping with Apartheid : Inter-household Transfers over the
Life-Cycle in South Africa", Boston College and World Bank, mimeo.

                                                                                           37


Czukas, K., M. Fafchamps, and C. Udry, 1998, "Drought and Saving in West Africa : Are
Livestock a Buffer Stock ?", Journal of Development Economics, Vol. 55, N°..., pp. 273-305.

Datta , S.K., and S. Kapoor, 1996, Collective Action, Leadership and Success in Agricultural
Cooperatives ­A Study of Gujarat and West Bengal, New Delhi & Calcutta : Oxford & IBH
Publishing Co.

Dixit, A., and S. Skeath, 1999, Games of Strategy, New York & London : W.W. Norton &
Cy.

Economist (The), 1995, "Out of Africa, a smoother ride", June 10th-16th.

Ellickson, R.C., 1991, Order Without Law- How Neighbors Settle Disputes, Cambridge,
Mass. and London: Harvard University Press.

Elster, J., 1989, The Cement of Society -- A Study of Social Order, Cambridge: Cambridge
University Press.

Fafchamps, M., 1992, "Solidarity Networks in Preindustrial Societies: Rational Peasants with
a Moral Economy", Economic Development and Cultural Change, Vol. 41, N° 1, pp. 147-
174.

Fafchamps, M., 1996, "The Enforcement of Commercial Contracts in Ghana", World
Development, Vol. 24, N° 3, pp. 427-448.

Fafchamps, M., 1999, "Risk Sharing, Quasi-credit and the Enforcement of Informal
Contracts", Stanford University, Journal of International Trade and Economic Development,
Vol. 8, N° 3, pp. 257-278.

Fehr, E., and J.R. Tyran, 1996, "Institutions and Reciprocal Fairness", Nordic Journal of
Political Economy, Vol. 23, N° 2, pp. 133-144.

Fehr, E., and S. Gächter, 1998, "How effective are trust- and reciprocity-based incentives ?",
in Ben-Ner, A., and L. Putterman (eds.), Economics, Values, and Organization, Cambridge :
Cambridge University Press, pp. 337-363.

Fehr, E., G. Kirchsteiger, and A. Riedl, 1998, "Gift Exchange and Reciprocity in Competitive
Experimental Markets", European Economic Review, Vol. 42, N° 1, pp. 1-34.

Frank, R., 1995, The Winner-Take-All Society, New York: Martin Kessler Books at the Free
Press.

Frank, R., 1998, "Social Norms as Positional Arms Control Agreements", in Ben-Ner, A., and
L. Putterman (eds.), Economics, Values, and Organization, Cambridge : Cambridge
University Press, pp. 275-295.

Galasso, E., and M. Ravallion, 2000, "Distributional Outcomes of a Decentralized Welfare
Program", mimeo, The World Bank, Washington.

                                                                                            38


Gambettta, D., 1988, "Can We Trust Trust ?", in Gambetta, D. (ed.), Trust-Making and
Breaking Cooperative Relations , Oxford: Basil Blackwell, pp. 213-237.

Gaspart, F., M. Jabbar, C. Melard, and J.P. Platteau, 1998, "Participation in the Construction
of a Local Public Good with Indivisibilities : An Application to Watershed Development in
Ethiopia", Journal of African Economies, Vol. 7, N° 2, pp. 157-84.

Gaspart, F., and J. P. Platteau, forthcoming, "Collective Action for Local-Level Effort
Regulation : An Assessment of Recent Experiences in Senegalese Small-Scale Fisheries", in
F. Stewart, R. Thorp, and J. Heyer (ed), Institutions and Development,    Oxford : Clarendon
Press.

Gavian, S., and M. Fafchamps, 1996, "Land Tenure and Allocative Efficiency in Niger",
American Journal of Agricultural Economics, Vol. 78, pp. 460-471.

Gertler, P., and J. Gruber, 1997, "Insuring Consumption Against Illness", Cambridge, MA :
NBER Working Paper 6035.

Golan, E.H., 1990, "Land Tenure Reform in Senegal : An Economic Study from the Peanut
Basin", Land Tenure Center, LTC Paper No. 101, Madison : University of Madison-
Wisconsin.

Goldstein, M., 1999, "Chop Time, No Friends, Intrahousehold and Individual Insurance
Mechanisms in Southern Ghana", University of California, Berkeley, mimeo.

Greif, A., 1989, "Reputation and Coalitions in Medieval Trade: Evidence on the Maghribi
Traders", Journal of Economic History, Vol. 49, No. 4, pp.857-882.

Greif, A., 1992, "Institutions and International Trade: Lessons from the Commercial
Revolution", American Economic Review, Papers and Proceedings, Vol. 82, No. 2, pp.128-
133.

Greif, A., 1993, "Contract Enforceability and Economic Institutions in Early Trade: the
Maghribi Traders' Coalition", American Economic Review, Vol. 83, No. 3, pp. 525-548.

Greif, A., 1994, "Cultural Beliefs and the Organization of Society: A Historical and
Theoretical Reflection on Collectivist and Individualist Societies", Journal of Political
Economy, Vol. 102, N° 5, pp. 912-950.

Greif, A., 1998, "Historical and Comparative Institutional Analysis", American Economic
Review, Vol. 88, N° 2, pp. 80-89.

Haugerud, A., 1983, "The Consequences of Land Tenure Reform among Smallholders in the
Kenya Highlands", Rural Africana, No. 15-16, Winter-Spring.

Hayami, Y., and T. Kawagoe, 1993, The Agrarian Origins of Commerce and Industry: A
Study of Peasant Marketing in Indonesia, London: Macmillan and New York: St. Martin's
Press.

Hoff, K., 1997, "Informal Insurance and the Poverty Trap", University of Maryland, mimeo.

                                                                                           39




Hume, D., 1740, A Treatise of Human Nature, (ed. L.A. Selby-Bigge), Oxford: Clarendon
Press (1888).

Janos, A., 1982, The Politics of Bacwardness in Hungary, 1825-1945, Princeton, N.J. :
Princeton University Press.

Johnson, R.N., and G.D. Libecap, 1982, "Contracting Problems and Regulation : The Case of
the Fishery", American Economic Review, Vol. 72, pp. 1005-1022.

Kali, R., 1997, "Endogenous Business Networks", Department of Business Administration,
Instituto Tecnologico Autonomo de Mexico, Mexico (mimeo).

Kimball, M.S., 1988, "Farmers' Cooperatives as Behavior Toward Risk", American
Economic Review, Vol. 78, N° 1, pp. 224-32.

Kocherlakota, N.R., 1996, "Implications of Efficient Risk Sharing Without Commitment",
Review of Economic Studies, Vol. 63, N° 4, pp. 595-609.

Lane, R.E., 1991, The Market Experience, Cambridge: Cambridge University Press.

Landes, D., 1998, The Wealth and Poverty of Nations ­Why Some Are So Rich and Some So
Poor?, London: Little, Brown and Cy.

Laurent, P-J, 1995, "Les pouvoirs politiques locaux et la décentralisation au Burkina Faso",
Cahiers du Cidep, N° 26, University of Louvain-la-Neuve.

Lawry, S.W., 1993, "Transactions in Cropland Held Under Customary Tenure in Lesotho", in
T.J. Bassett and D.E. Crummey (eds.), Land in African Agrarian Systems, Madison,
Wisconsin: The University of Wisconsin Press, pp. 57-74.

Ledeneva, A.V., 1998, Russia's Economy of Favours ­Blat, Networking and Informal
Exchange, Cambridge: Cambridge University Press.

Leservoisier, O., 1994, La question foncière en Mauritanie ­Terres et pouvoirs dans la région
du Gorgol, Paris: Editions L'Harmattan.

Libecap, G.D., 1990, Contracting for Property Rights, New York : Cambridge University
Press.

Libecap, G.D., and S.N. Wiggins, 1984, "Contracting Responses to the Common Pool :
Prorationing of Crude Oil Production", American Economic Review, Vol. 74, N° 1, pp. 87-98.

Li, T.M., 1998, "Constituting Capitalist Culture ­The Singapore Malay Problem and
Entrepreneurship Reconsidered", in Hefner, R.W. (ed), 1998, Market Cultures ­Society and
Morality in the New Asian Capitalisms, Boulder, Col.: Westview Press, pp. 147-172.

Ligon, E., J. Thomas, and T. Worall, 1997, "Informal Insurance Arrangements in Village
Economies", mimeo, Department of economics, Warwick University.

                                                                                          40


MacKenzie, F., 1993, " `A Piece of Land Never Shrinks': Reconceptualizing Land Tenure in
a Smallholding District, Kenya", in Bassett, T.J., and D.E. Crummey, (eds.), Land in African
Agrarian Systems, Madison, Wisconsin : University of Wisconsin Press, pp. 194-221.

Maiga, M., 1995, Le fleuve Sénégal et l'intégration de l'Afrique de l'Ouest en 2011, Dakar:
Codesria & Paris: Karthala.

Malarney, S. K., 1998, "State Stigma, Family Prestige, and the Development of Commerce in
the Red River Delta of Vietnam", in Hefner, R.W. (ed), Market Cultures ­Society and
Morality in the New Asian Capitalisms, Boulder, Col.: Westview Press, pp. 268-289.

Menzies, N.K., 1994, Forest and Land Management in Imperial China, London : St. Martin's
Press.

Migot-Adholla, S.E., P. Hazell, B. Blarel, and F. Place, 1991, "Indigenous Land Rights
Systems in Sub-Saharan Africa: a Constraint on Policy?", World Bank Economic Review, Vol.
5, No. 1, pp. 155-175.

Morduch, J., 1995, "Income Smoothing and Consumption Smoothing", Journal of Economic
Perspectives, Vol. 9, N° ..., pp. 103-114.

Morduch, J., 1999, "Between the Market and the State : Can Informal Insurance Patch the
Safety Net ?", World Bank Research Observer, Vol. 14, N° 2, pp. 187-207.


Myerson, R.B., 1991, Game Theory - Analysis of Conflict, Cambridge, Mssts & London:
Harvard University Press.


Nisbett, R.E., and D. Cohen, 1996, Culture of Honour ­The Psychology of Violence in the
South, Boulder, Col. : Westview Press.

Olson, M., 1965, The Logic of Collective Action , Cambridge, Mass. : Harvard University
Press.

Ostrom, E., 1990, Governing the Commons ­The Evolution of Institutions for Collective
Action , Cambridge : Cambridge University Press.

Pinckney, T.C., and P.K. Kimuyu, 1994, "Land Tenure Reform in East Africa: Good, Bad, or
Unimportant?", Journal of African Economies, Vol. 3, N° 1, pp. 1-28.

Platteau, J.P., and A. Abraham, 1987, "An Inquiry into Quasi-Credit Contracts: The Role of
Reciprocal Credit and Interlinked Deals in Small-Scale Fishing Communities", Journal of
Development Studies, Vol. 23, N° 4, pp. 461-90.

Platteau, J.P., 1991, "Traditional Systems of Social Security and Hunger Insurance: Past
Achievements and Modern Challenges", in Ahmad, E., J. Drèze, J. Hills, and A.K. Sen (eds),
Social Security in Developing Countries, Oxford: Clarendon Press, pp. 112-70.

Platteau, J.P., 1996, "The Evolutionary Theory of Land Rights As Applied to SubSaharan
Africa : A Critical Assessment", Development and Change, Vol. 27, N° 1, pp. 29-86.

                                                                                         41




Platteau, J.P., 1997, "Mutual Insurance as an Elusive Concept in Traditional Rural
Communities", Journal of Development Studies, Vol. 33, N° 6, pp. 764-796.

Platteau, J.P., and Y. Hayami, 1998, "Resource Endowments and Agricultural Development:
Africa versus Asia", in Aoki, M., and Y. Hayami (eds), The Institutional Foundations of East
Asian Economic Development, London: Macmillan, pp. 357-410.

Platteau, J.P., 2000a, Institutions, Social Norms and Economic Development, Harwood
Publishers.

Platteau, J.P., 2000b, "Allocating and Enforcing Property Rights in Land : Informal versus
Formal Mechanisms in Subsaharan Africa", Nordic Journal of Political Economy, Vol. 26, N°
1, pp. 55- 81.

Putnam, R.D., 1993, Making Democracy Work - Civic Traditions in Modern Italy, Princeton,
NJ: Princeton University Press.

Rabin, M., 1993, "Incorporating Fairness into Game Theory and Economics", American
Economic Review, Vol. 83, N°5, pp. 1281-1302.

Ravallion, M., and S. Chaudhuri, 1997, "Risk and Insurance in Village India : Comment",
Econometrica, Vol. 65, N° 1 , pp. 171-184.

Riddell, J.C., J.W. Salacuse, and D. Tabachnick, 1987, "The National Land Law of Zaire and
Indigenous Land Tenure in Central Bandundu, Zaire", Land Tenure Center, LTC Paper No.
92, Madison : University of Wisconsin-Madison.

Robertson, A.F., 1987, The Dynamics of Productive Relationships ­African Share Contracts
in Comparative Perspective, Cambridge: Cambridge University Press.

Rosenzweig, M., 1988, "Risk, Implicit Contracts, and the Family in Rural Areas of Low-
Income Countries", Economic Journal, Vol. 98, N°..., pp. 1148-1170.

Sandler, T., 1992, Collective Action ­Theory and Applications, Ann Arbor : University of
Michigan Press.

Saul, M., 1993, "Land Custom in Bare: Agnatic Corporation and Rural Capitalism in Western
Burkina", in Bassett, T.J., and D.E. Crummey (eds.), Land in African Agrarian Systems,
Madison, Wisconsin : The University of Wisconsin Press, pp. 75-100.

Schotter, A., A. Weiss, and I. Zapater, 1996, "Fairness and Survival in Ultimatum and
Dictatorship Games", Journal of Economic Behavior and Organization, Vol. 31, N° 1, pp. 37-
56.

Sengupta, 1991, Managing Common Property ­Irrigation in India and the Philippines, New
Delhi : Sage Publications.

                                                                                          42


Shanmugaratnam, N., 1996, "Nationalization, Privatization and the Dilemmas of Common
Property Management in Western Rajasthan", Journal of Development Studies, Vol. 33, N° 2,
pp. 163-87.

Shipton, P., 1988, "The Kenyan Land Tenure Reform : Misunderstandings in the Public
Creation of Private Property", in Downs, R.E., and S.P. Reyna (eds), Land and Society in
Contemporary Africa, Hanover & London : University Press of New England, pp. 91-135.

Stanfield, D., 1990, "Rural Land Titling and Registration in Latin America and The
Caribbean: Implications for Rural Development Programs", Land Tenure Center, Madison :
Unversity of Wisconsin-Madison.

Stavis, B., 1982-83, "Rural Local Governance and Agricultural Development in Taiwan", in
Uphoff, N.T. (ed.), Rural Development and Local Organization in Asia, New Delhi :
Macmillan, Vol. II, pp. 166-271.

Stone, L., 1990, The Family, Sex and Marriage in England 1500-1800, Harmondsworth :
Penguin Books.

Sugden, R., 1984, "Reciprocity: The Supply of Public Goods Through Voluntary
Contributions", Economic Journal, Vol. 94, pp. 772-787.

Sugden, R., 1986, The Economics of Rights, Co-operation and Welfare,          Oxford: Basil
Blackwell.

Szanton, D.L., 1998, "Contingent Moralities ­Social and Economic Investment in a
Philippine Fishing Town", in Hefner, R.W. (ed), Market Cultures ­Society and Morality in
the New Asian Capitalisms, Boulder, Col.: Westview Press, pp. 251-267.

Tang, S.Y., 1992, Institutions and Collective Action ­Self-Governance in Irrigation Systems ,
San Francisco : ICS Press.

Tang, S.Y., 1994, "Institutions and Performance in Irrigation Systems", in Ostrom, E., R.
Gardner, and J. Walker, Rules, Games, and Common-Pool Resources, Ann Arbor: The
University of Michigan Press, pp. 225-45.

Thomas, J.P., and T. Worall, 1994, "Informal Insurance Arrangements in Village
Economies", Department of economics, University of Warwick, mimeo.

Thompson, E.P., 1993, Customs in Common, London : Penguin Books.

Townsend, R.M., 1994, "Risk and Insurance in Village India", Econometrica, Vol. 62, N° 3,
pp. 539-91.

Udry, C., 1990, "Credit Markets in Northern Nigeria: Credit as Insurance in a Rural
Economy", World Bank Economic Review, Vol. 4, N° 3, pp. 251-69.

Udry, C., 1993, "Credit Markets in Northern Nigeria: Credit as Insurance in a Rural
Economy", in Hoff, K., A. Braverman, and J.E. Stiglitz (eds), The Economics of Rural
Organization, Oxford: Oxford University Press (Published for the World Bank), pp. 87-108.

                                                                                         43


Udry, C., 1994, "Risk and Insurance in a Rural Credit Market: An Empirical Investigation in
Northern Nigeria", Review of Economic Studies, Vol. 61, pp. 495-526.

Wade, R., 1987, The Management of Common Property Resources: Finding a Cooperative
Solution, The World Bank Research Observer, Vol. 2, N° 2, pp. 219-234.

Wade, R., 1988, Village Republics: Economic Conditions for Collective Action in South India,
Cambridge: Cambridge University Press.

White, T. A. and C. F. Runge, 1995, "The Emergence and Evolution of Collective Action:
Lessons from Watershed Management in Haiti", World Development, Vol. 23, N° 10, pp.
1683-1698.