Comparing Forest Decentralization and Local Institutional Change in Bolivia, Kenya, Mexico, and Uganda

Summary

In this paper we assess the institutional and environmental impacts of forest decentralization in Bolivia, Kenya, Mexico, and Uganda. We develop theories of institutional impacts based upon the specific content of decentralization reforms. We classify each country’s reforms in terms of the creation/change in local user group empowerment and accountability mechanisms. Using data from the International Forestry Resources and Institutions Program, we estimate the effects of forest decentralization on local forest investments, rulemaking, wealth inequality, and forest conditions in the four countries. Some results support our theory, but the theory is insufficient to explain the full range of outcomes.

Introduction

Forest decentralization programs have rapidly spread in developing countries in the last 20 years (Agrawal, Chhatre, & Hardin, 2008; Andersson, Gibson, & Lehoucq, 2004). There is now a large literature that examines case studies of decentralization and develops theoretical frameworks to explain the causes and consequences of decentralization (Boone, 2003, Falleti, 2005, O’Neill, 2003, Ribot et al., 2006). In this paper we use these frameworks to develop a theory of how changes in accountability and user empowerment that result from decentralization policies impact user group behavior and forest conditions. We then test this theory using a unique over-time dataset on forest resources and institutions in Bolivia, Kenya, Mexico, and Uganda, which allows us to test theories across a broader range of cases than has been possible in previous work. We find, consistent with previous theories, that in countries where reforms increase both upward and downward accountability as well as empower forest users, there are more likely to be positive results both in terms of intermediate outputs such as user group collective action and in terms of outcomes such as improved forest conditions and decreased income inequality. These effects, however, are weak and inconsistent. The theory explains the data most strongly in Mexico, the country with the most well-established and democratic decentralization in the study area, but the results from Bolivia and Kenya are mixed. Surprisingly, Uganda, the country with the least stable forest governance system, saw the greatest increase in one important measure of local collective action—the making of rules about forest governance. The mixed nature of these findings indicates that the theories we test are useful guides, but are insufficient to explain the full range of decentralization outcomes.

Research on forest decentralization, like much of the broader literature on decentralization, is plagued by analytical problems. First, decentralization is not a single, well-defined policy but rather a general term that is applied to a diversity of policies that may include some combination of (a) moving bureaucrats from central locations to sites closer to resources; (b) increasing the decision-making discretion of local level bureaucrats; and (c) increasing the decision-making authority of local users (Cohen & Peterson, 1996). Second, while there are a large number of theoretical arguments relating to the benefits and costs of decentralized political orders, under closer scrutiny these fail to generate consistent explanations of observed outcomes (Andersson et al., 2008, Treisman, 2007). Third, decentralization policies interact with numerous other pressures to change governance institutions, forest user behavior, and resulting forest conditions and livelihood outcomes (Andersson et al., 2008). The connection between national policy changes and changes in local level behavior are mediated by complex processes that inhibit policy implementation in even the best of circumstances (Pressman and Wildavsky, 1978, Sabatier, 1986). Forest management, an activity frequently undertaken in remote and politically marginal areas of poor countries, is not a promising candidate for implementation success.

Not surprisingly, evaluations of forest decentralization have reported disappointing results. The most influential theoretical work in this area is Agrawal and Ribot’s (1999) framework, which emphasizes the importance of actors, powers, and accountability on the influence of decentralization reforms on governance institutions.¹ Work in this tradition has led to the pessimistic conclusion that decentralization reforms have reinforced the power of the central state (see also Boone, 2003, Larson and Ribot, 2007, Ribot and Larson, 2005, Ribot et al., 2006). This literature has emphasized evaluations of the political consequences of decentralization, examining whether local actors have in fact gained political power, but has largely neglected the impact of decentralization policies on forest conditions and local level collective action around forest management (Andersson & Gibson, 2007). Most of this literature draws on relatively small numbers of cases, limiting the generalizability of the findings.

This paper seeks to address this gap by explicitly analyzing changes among forest users on the ground, placed within the larger political context of national decentralization policies. Rather than assuming uniform effects, we follow the approach suggested by Agrawal and Ribot (1999) and Larson (2003) by examining how particular decentralization policies affect local actor incentives, drawing on the extensive literature on decentralization in the four countries we study. We move a step further, however, by examining how these altered incentives affect both local forest user collective action (in terms of investment decisions and rulemaking) as well as the broader objectives of forest decentralization reforms (in terms of wealth inequality and forest conditions as seen by local user groups). Using the relatively large number of cases available in our dataset allows us to test the generality of theories developed by previous authors using smaller samples.

The paper is organized as follows. In the next section we review the history of forest decentralization reforms in each of the four countries. These case histories provide the basis for our theoretical predictions of the impact of decentralization on our outcome variables. In Section 3 we draw on existing theory to derive predictions of the effects of decentralization in each of the countries. In Section 4 we describe the data we use for our analysis as well as summary statistics for each of the four countries. In Section 5 we report our empirical results. In Section 6 we discuss these findings and in Section 7 we conclude.