Do Brokers Know Their Voters? A Test of Guessability in India

Abstract

Theories of clientelism broadly depend on the untested assumption that brokers possess fine-grained information on voters’ political preferences prior to elections, and often can monitor their votes. As the first direct test of this assumption in a competitive democracy, I develop a measure, guessability, which measures the ability of local brokers to identify the partisan preferences of voters from their local area. I apply this method to elected village politicians in rural India who often perform brokerage functions. I find that these local leaders perform well at identifying the partisan preferences of co-partisans, but perform no better than a low-information benchmark that captures the level of guessability that outsiders can plausibly achieve with respect to non-core voters. This suggests that an electoral strategy rooted in quid-pro-quo exchange is extremely inefficient in rural India where the clientelism is believed to be feasible, with implications for prominent theories applied to other settings.

Keywords: clientelism, brokers, India, vote buyung