Do Lenders Favor Politically Connected Firms? Rent Provision in an Emerging Financial Market

Corruption by the politically connected is often blamed for economic ills, particularly in less developed economies. Using a loan-level data set of more than 90,000 firms that represents the universe of corporate lending in Pakistan between 1996 and 2002, we investigate rents to politically connected firms in banking. Classifying a firm as “political” if its director participates in an election, we examine the extent, nature, and economic costs of political rent provision. We find that political firms borrow 45 percent more and have 50 percent higher default rates. Such preferential treatment occurs exclusively in government banks—private banks provide no political favors. Using firm fixed effects and exploiting variation for the same firm across lenders or over time allows for cleaner identification of the political preference result. We also find that political rents increase with the strength of the firm's politician and whether he or his party is in power, and fall with the degree of electoral participation in his constituency. We provide direct evidence against alternative explanations such as socially motivated lending by government banks to politicians. The economy-wide costs of the rents identified are estimated to be 0.3 to 1.9 percent of GDP every year.