Do Longer Constitutions Corrupt?

Abstract

Tsebelis and Nardi (2016) and Tsebelis (2017) report that constitutional length correlates with lower levels of GDP per capita. They argue that this may be the case because longer constitutions lead to greater corruption. However, uncovering a causal relationship between constitutional length and corruption is difficult. On the one hand, political elites may pressure drafters to include specific provisions that facilitate their rent-seeking efforts. On the other hand, constitutional drafters may be responding to corruption by including a large number of specific safeguards. Our aim in this paper is to explore whether there is a causal effect of constitutional length on corruption. We utilize data from the Comparative Constitutions Project (CCP) to identify 5 cases when a country experienced a ≥ 50% increase in constitutional length. For each of those cases, we compare the subsequent change in corruption to that of a synthetic control. We report evidence of a significant post-treatment increase in corruption for 3 out of 5 cases (Ecuador in both 1997 and 2008; Venezuela in 1999). However, the 2008 Ecuador result is not robust to a placebo test; and in the case of Venezuela it is difficult to distinguish a constitutional length effect from a “Chavez effect” (Grier and Maynard 2016). The evidence that longer constitutions corrupt is weak.

Keywords: Constitutions, constitutional amendments, constitutional length, corruption

JEL Classification: K00, K20, P50, P16, Z13, D73