Information Accumulation and the Timing of Voting Decisions

59 Pages Posted: 27 Jan 2020 Last revised: 4 Nov 2020

Date Written: January 16, 2018


Surveys, polling data and media reports indicate that voters often choose who to vote for at different stages in the political campaign: some voters know from start of a campaign who they will vote for, while others decide on the last day. This dispersion of the timing of voting decisions can swing election results, thus are of importance to campaigns and policymakers. In this paper, I develop a model of costly information acquisition that rationalizes these observations. The model implies a key tradeoff between the cost of acquiring information, and the gain such information brings. The solution of this problem is an optimal stopping time decision, which is structurally estimated. I find that the information voters receive is noisy, with the standard deviation of a signal being around 25\% of the support of the ideology variable. I show evidence that later deciders are different on observable characteristics such as age, religiosity, education and political knowledge, and discuss how these differences arise in voter beliefs and costs. There is significant unobserved heterogeneity in voters' costs of acquiring information, which explains different voting decisions by observationally similar citizens. Under this framework, I consider the implications of the widely used policy of information blackouts (i.e. forbidding campaigns or polls just before the election). Although it is often thought to promote fairness, I find that such a policy harms voters with a 1-2% welfare loss, as the restrictions to information affect only those to whom it would benefit.

Keywords: Voting Choices, Information, Learning, blackouts

Suggested Citation

Canen, Nathan, Information Accumulation and the Timing of Voting Decisions (January 16, 2018). Available at SSRN: or

Nathan Canen (Contact Author)

University of Houston ( email )

4800 Calhoun Road
Houston, TX 77204
United States

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Abstract Views
PlumX Metrics