Decentralization and Popular Democracy: Governance from Below in Bolivia

4. Decentralization and Responsiveness across Bolivia: A 21-Year View

chapter 4 Decentralization and Responsiveness across Bolivia: A 21-Year View Chapter 1 presented simple but powerful descriptive statistics implying that decentralization changed sectoral and spatial patterns of public investment. Decentralized governments invested in different places and different sectors than centralized government had before, shaping national investment patterns that became more responsive to real local needs. But such descriptive statistics, though often compelling, are analytically rudimentary. Deep qualitative evidence in chapters 2 and 3 showed that decentralized governments were as capable of corruption and ineptitude as of ef‹ciency and responsiveness . What is the balance of decentralization’s effects on the nation as a whole? Can more rigorous evidence tell us whether decentralization made government systematically more responsive to local needs, less responsive , or enough of both that there are no systematic effects? This chapter takes up the challenge with econometric tools and an original municipal-level database that contains yearly information on local investments for all Bolivian municipalities for the period 1987–2007. This doubles the size of the sample on which my previous work on Bolivia1 was based (1987–96). Hence the estimations provided below incorporate far more information than any of my previous work on Bolivian reform. More interesting, the longer time span examined here allows us to explore how decentralization’s impacts changed over time. In addition to ‹nancial data, the database includes demographic, social, economic, and geographic data, local electoral results, and institutional and procedural data from the innovative Censo Municipal. Unfortunately, changes in the wording of many census questions between 1992 and 2001 make it impossible to compare a number of speci‹c answers across time, leaving us with only one observation for a number of key sectoral variables 133 during the 20-year period. How then can we investigate decentralization’s effects given good time-series data on national and local investment ›ows, but cross-sectional-only data for some key variables? This chapter provides an econometric methodology for doing so. It also develops a theoretical model in which public goods are jointly provided by central and local governments. This serves as an aid for interpreting the econometric results, and as a consistency framework for thinking about some of the key trade-offs between local versus central provision of public goods more generally. The question of decentralization and responsiveness to need brings us back to the curious discrepancy discussed in the introduction between enthusiasm for decentralization among governments and policy analysts across the world, and the weak and contradictory evidence on its effects. In particular, the literature records little econometric analysis of the effects of decentralization on government responsiveness to local needs, and relevant anecdotal evidence is ambiguous. This is especially surprising given that increased responsiveness constitutes one of the central claims in favor of reform . Furthermore, I have not found any studies in the literature that follow the process of decentralization with rigorous evidence over any signi‹cant span of time. This chapter seeks to ‹ll these gaps by examining decentralization’s effects on government responsiveness to local need over 21 years in a careful, methodical way using abundant data from Bolivia. Focusing on one country allows me to control for political regime, external shocks, and other exogenous factors more systematically than a cross-country approach can. And Bolivia is particularly appropriate since reform was comprehensive and sustained, and so constitutes a social experiment. The remainder of the chapter is organized as follows. The second section models the joint provision of local public goods as a Stackelberg follower game in which political competition provides local governments with better information on heterogeneous local preferences, but central government is more productive. The third section presents the data and develops a methodology for estimating the effects of decentralization on government responsiveness to need when ‹nancial data is extensive, but social, demographic, institutional , and other data are not. The fourth section tests whether decentralization changed public investment patterns across Bolivia’s 311 municipalities and then examines the determinants of these changes focusing on variables of need. The ‹nal section concludes. 134 decentralization and popular democracy The Model As in many countries, Bolivia’s local services are jointly provided by central and local governments. The evidence in the fourth section below thus focuses on the effect that local institutions and control of local budgets have on the responsiveness of municipal investments to real local need. But before delving into the empirics of the question, it...