How does the introduction of local elections affect the pattern of public spending and revenue generation at the local level? The relationship between institutions of political accountability and government performance remains a key concern for analysts and practitioners of public policy design. This article analyses how institutional design for electoral accountability affects public sector spending choices and service delivery in districts of Indonesia. It finds that electoral reforms had positive effects on expenditures, mainly due to expenditures brought about by the election of non-incumbents.
Indonesia’s political transition to democratization and greater devolution provides a rare opportunity to empirically assess the impact of reforms in political institutions on government behaviour. Starting in 2005, mayors were directly elected, but this change was introduced in a staggered way, once the terms of the incumbents had come to an end. The phasing in of direct elections was effectively exogenous.
In order to analyse the impact of changes in political selection and electoral accountability on local government performance, these reforms are compared with evidence on local spending patterns. Findings include the following:
- Per capita total expenditures in districts with direct elections in 2005 increased significantly from the pre-election to post-election period, relative to districts without elections.
- There was a positive and significant impact of democratic reform on the amount of revenue generated from own sources by the district governments. This contradicts the assumption that local electoral reforms lead to lower taxation. It also contradicts the assumption that deficits at district level are likely to increase based on the expectation that they will be absorbed by central government.
- The higher expenditure after elections seems to be driven by districts outside Java and Bali. Results in the pooled sample also seem to be driven by districts where non-incumbents were elected.
- There also seems to be an ‘anticipation effect’ at work in districts that are having or are about to have elections. In these districts, fiscal expenditure trajectories are very similar to those of districts that have already held elections.
- The data used has too short a time horizon to measure the impacts of expenditure increases. However, overall there is no strong evidence that changes in expenditures of district governments are responsive to the observable needs of the districts.
In the context of literature on decentralisation, these findings suggest that political decentralization can strengthen incentives of local governments to use opportunities provided by fiscal decentralisation more effectively. However, a number of issues require further research:
- The finding that expenditures rise in anticipation of elections has the positive implication that even expectation of an election changes local government incentives. However, it may also have the perverse effect of inducing spending cycles aligned with the electoral cycle.
- While the impact of electoral reforms on expenditure is distributed across a number of sectors, the health sector is an exception. It is unclear whether this is a reflection of citizen priorities, or whether other factors make it hard for governments to raise investments in health.
- The lack of evidence that changes in expenditure are responsive to observable needs may have other explanations than an absence of impact. It is important to track a more expanded set of outcome indicators over a longer period of time.
