Crackdown on Corruption: A Natural Experiment in Safe and Swing Districts

I investigate the theoretical and empirical effects of increased law enforcement on the equilibrium level of bribes for the case of Albania during the period of 2005-2010. My paper centers on "harassment" bribes, which consist of payments for public services that by law should be free. I model bribery behavior as a negotiation process between public oofficials and consumers. As enforcement increases, corruption prevalence can, in theory, increase or decrease. Recent policy changes in Albania o er a good natural experiment to test this empirically. Two events took place in 2007: local elections and an increase in fines against corruption. I examine how the 2007 fine increase for corrupt behavior impacts bribery. Data show that corruption is a bigger problem for poor people and since the left's political platform is more pro-poor, looking at the left-right governed district variation seems appropriate. Using a difference in difference methodology that compares safe left and right-governed districts, I find that a 10 percent increase in enforcement leads to a 4.38 percent drop in bribery frequency. As enforcement increases, quality of services does not improve and enforcement measures are less e ffective on the medical and education sectors.