Policy Herding as a Challenge to Political Persuasion∗

Can rationally ignorant voters and informed politicians lead to persistently inefficient policies in a democracy? A long tradition of thought in political science and economics has answered negatively, due to the averaging out of randomly uninformed voters or persuasion attempts in a competitive setting. Rather, explanations for the persistence of inefficient policies has focused on interest group politics. In this paper, I present an argument for why voters may come to support inefficient policies. In particular, support may come from policy herding. The existence of a policy implies that choices were made to support it. If the policymaking process sometimes filters inefficient policies but rationally ignorant voters cannot distinguish when, they will incorrectly weigh the efficiency of an observed policy. Politicians without vested interests have incentives to persuade voters to support a more efficient policy, but policy herding biases against these persuasion efforts. On the other hand, politicians with vested interests have incentives to maintain the policy, further counterbalancing persuasion efforts. The longer a policy is in place, the further the herding effect is compounded. So even if vested politicians eventually leave the polity, the policy may persist due to voter support. I formalize this logic through an electoral model with incomplete information, and discuss the types of policies and institutions that make policy herding more likely.