Corruption in the Lab and in the Field: Evidence from a Market for Credence Goods

In many markets with asymmetric information, ex-post verification of product quality or service, even in the long run, is very difficult. Because sellers in such markets have a strong incentive to cheat their customers, economic theory posits poor functioning of markets. Yet, this is not always the case. How can we explain this? We investigate the role individual motivation for honesty in the functioning of such markets. Our study takes place unorganized milk markets in Delhi (India). This is a unique setting because extrinsic factors like liability, verifiability, competition, and reputation, which are known to affect market functioning, are naturally controlled for. We combine a novel behavioral measure of honesty, which allows us to detect partial cheating not considered in previous studies and a quantitative measure of cheating in the milk market, which is not available to customers.

We find that individuals with a higher motivation for honesty add substantially less water to milk. A comparison of our refined behavioral measure and the conventional statistical measures reveals that ignoring partial cheaters would lead to underestimation of the effect of honesty on market expansion. These results hold at both extensive and intensive margins and are robust to a variety of alternative specifications including the use of neighborhood fixed effects, which capture potential differences in the institutional environment. This finding is not just statistically but also economically very significant. Overall, milkmen who cheat in our experiment earn about 160 Euros per month from adding more water to milk.