Research

Local politicians implement many policies and form a talent pool for national office. With local newspapers in decline, how can they establish their national reputation and improve local outcomes? I study a two-period political accountability model in which incumbents can be policy-motivated (good) or corrupt (bad). They face re-election or promotion. They can use local information to improve policy or for personal gain, but only local watchdogs can distinguish this. Or they can ignore local information and play it safe, which is verifiable. If local news declines, good types signal their type by increasingly playing it safe: securing a good reputation with their party, but worsening policy outcomes. Voters might prefer outsider candidates who are more likely to be corrupt, but disavow national parties. The model rationalizes growing abuse and underuse of local office, the nationalization of local politics, and the severing of the pipeline between local and national office.

Conventional wisdom favors moderation, yet recent elections feature candidates who campaign, win, and govern as hardliners. We develop a model in which candidates motivated by ideology and office are better informed than voters about policy consequences. Before the election, candidates commit to a policy or remain ambiguous. Commitment sacrifices post-election flexibility, but credibly conveys information. Informative campaigns require moderate polarization. Then candidates commit only when evidence supports their preferred policy, persuading voters. With high polarization or increased office-motivation, candidates will not pay the electoral cost of letting their opponent persuade. They always commit to opposing policies: one conveying available evidence, one concealing it. Voters learn little, yet elections, not experts, determine policy. Examples from Eisenhower to today illustrate how rising polarization reshapes campaigns, voter learning, and governance.