Previous theoretical analysis suggested that group heterogeneity facilitates collective action because agents with high stakes in a public good are also endowed with large amounts of wealth and thus will have greater incentives to contribute voluntarily to the public good. Results from this empirical study of Amazonian fishing communities do not support these former theoretical predictions. When increased wealth stratification was associated with greater availability of alternative economic opportunities, wealthier households chose not to contribute proportionally to the additional costs of a conservationist scheme. Organizational contributions were non-proportional or negatively correlated to household wealth in communities with less developed or failed management institutions. Larger and more homogenous communities were able to enforce rules about proportional contributions by their members and to bear the cost of creating and maintaining a fully developed local management institution.
According to the standard theory of economic individualism, voluntary collective-action institutions (e.g., local management institutions) should not exist (Barnes, 1992). However, one clear result from more than two decades of research is the finding that people do not always follow the inexorable logic of the "Tragedy of The Commons" metaphor (Berkes, 1985; Hardin, 1968; McKean, 1992; Simmons & Schwartz-Shea, 1993). Users of Common-pool resources, (1) including fishers, both in peasant and industrial societies, have repeatedly shown their capacity to organize themselves (Acheson et al., 1998). Members make credible commitments to monitor each other's behavior and impose sanctions on those who display inappropriate behaviors. Thus, the idea of decentralizing natural resource management to local communities is increasingly gaining acceptance in policy-making and in people-centered development projects (Carney & Farrington, 1998; Cernea, 1991; Ostrom, 1990).
There are, however, two kinds of problems that may make local self-regulation less likely to be successfully achieved than would appear to be the case at first sight. Regulation of appropriation level may seriously limit a household's income and for this reason is liable to generate strong opposition from some resource appropriators. Additionally, regulation may have organizational costs for participant households, thereby making the management scheme costly to achieve and maintain (Baland & Platteau, 1996). Despite an impressive growth of theory and empirical research during the last several decades, the origin and dynamics of collective action remain disputed (Heckathorn, 1996). As pointed out by Taylor (1992), it is necessary to provide a better explanation of why some groups of users are able to solve their own collective action dilemma endogenously--without external help--and other groups are not. While the relationship between group size and collective action is now well understood (the smaller the group the stronger its ability to perform collectively), the impact of wealth inequality remains a largely unsettled question (Baland & Platteau, 1997).
The artisanal fishing communities of the Middle Amazon River provide a "natural laboratory" in which to explore these issues. In this region, some communities have autonomous local resource management schemes to regulate their fishing practices while others do not. Of those that do, some control only access to fishing grounds, while others control both access and individuals' level of resource appropriation. In some communities, there is widespread adherence to the management scheme; in others, opposition threatens to destroy the management institution and to deplete the local fish stocks. This study aims to contribute to the understanding of the relationships between a group's internal economic inequality and the provisioning of community management institutions for governing common pool resources. …