Working Paper: CEPR ID: DP3352
Authors: Min Shi; Jakob Svensson
Abstract: This Paper uses a large new panel data set to examine the relationship between elections and fiscal policy. We find clear evidence of political business cycles in macroeconomic policy: spending increases before elections while revenues fall, leading to a larger deficit in election years. We also show that there are large systematic differences between developed and developing countries in the size and composition of the electoral policy cycles. We propose a moral hazard model of electoral competition to explain these differences. In the model, the sizes of the electoral budget cycles depend on the rents of remaining in power and the share of informed voters in the electorate. Using suitable proxies, we find that these institutional features explain a large part of the difference in policy cycles between developed and developing countries.
Keywords: Developing countries; Dynamic panel estimation; Political budget cycles
JEL Codes: D72; E62; P16; P26
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Elections (K16) | Increased government spending (H59) |
| Elections (K16) | Decreased revenues (F69) |
| Elections (K16) | Larger fiscal deficits during election years (H69) |
| Electoral environment and institutional features (D72) | Differences in size and composition of electoral budget cycles (D72) |
| Higher rents associated with remaining in power (R21) | Larger electoral budget cycles (D72) |
| Greater proportion of informed voters (D72) | Larger electoral budget cycles (D72) |