The Logic of Subnational Public Goods Allocation in the Developing World

• In Event: Decentralization and Sub-National Development

Abstract

What explains the subnational variations in access to public goods in the developing world? Understanding the distributive logic of public goods and services of who gets what, why, and how, is increasingly important, and has implications on the lives and livelihoods of the developing world. The supply-side literature has provided important answers to this canonical question from a range of angles, including electoral logic and ethnic favoritism, and highlighted the incentives of elites' logic in the allocation of public goods. However, this literature has been less precise on why political elites choose the exact type of public goods they provide and how they are provided. Public goods are assumed to be a uniform entity that, regardless of which public good it is, undergoes the same allocative decision-making process. This is evident in the empirical literature on distributive politics where works focus on a single public good of choice, when in fact, public goods provision is a result of politicians' choice from a menu of public goods with varying characteristics. Therefore, I argue, adding to works like Kramon and Posner (2013) and Batley and Mcloughlin (2015), that there needs to be a more nuanced understanding of the types of the dimensions of public goods. Further, I argue that there needs to be a deeper understanding of how these dimensions interact with political elites' incentives. A growing number of works show that politicians pursue a mixed strategy in the allocation of public goods, rather than the long-assumed binary strategy of core versus swing or coethnic versus non-coethnic. However, only a few examine how politicians' incentives interact with the characteristics of public goods and how the interaction determines "what, where, and how" public goods are allocated to the voters.

To develop a more in-depth understanding of the distributive logic of public goods, I develop in this paper a theoretical argument that reconciles the political incentives of the elites that derive from electoral geography they face. I make the argument that the political geography elites face determines their time horizons, which in turn influence the type of public goods they provide, to whom, and how. I argue that the main dimensions of public goods political elites take into account are visibility, targetability, and the speed of delivery. I theorize that higher levels of electoral competition that lead to shorter time horizons are associated with the provision of universal public goods with larger catchment areas, higher visibility public goods, and quicker delivery, to maximize electoral gains within a shorter period of time. Lower levels of electoral competition and subsequently longer time horizons are associated with public goods that are targetable in implementation and provided to rich core voters—visibility and the speed of delivery matter less to the political elites facing lower political competition.

To test the empirical implications of the theoretical argument, I use the case of Indonesia. Indonesia is a country with a high level of electoral competition despite being a nascent democracy and has large subnational variations in access to public goods, making it an ideal case to test the theoretical argument. Using Indonesia's village potential survey (PODES) for years 2004-2019, I show that the political elites' decision on where to allocate universal public goods (public schools and public health centers) and targetable public goods (piped water) is a function of the level of electoral competition they face. More specifically, I show that when there is a higher level of electoral competition, the distance to closest public schools and health centers decreases, indicating that electoral competition is associated with the building of universal public goods. For targetable public goods, for which I use piped water, the results show a null effect. Unlike what the theoretical argument predicts, lower electoral competition is not associated with a greater allocation of piped water. This, I speculate, could be due to the measurement issue (which is currently a binary indicator of the presence of piped water for bathing) or due to an issue with the characterization of piped water as a targetable public good. That is, piped water may have greater spillover effects beyond the households that are targeted due to theft or due to the nature of the infrastructure and hence is not conceived by the political elites as a 'targetable' public good to provide to rich core voters.

In sum, the paper contributes to a growing area in the distributive politics literature that considers the specific characteristic of public goods and how they factor into the political elites' decision-making over allocation. The empirical test on Indonesia furthermore adds to an understudied area of subnational patterns of public goods provision in the country.

Author

• Soo Min Oh, University of Birmingham