When, Where, How: Planning to Vote

Organization : Behavioral Insights Group

Project Overview

Project Summary

Days before the 2008 presidential election, unlikely voters received phone calls in which they were encouraged to vote and prompted to plan their visit to the polling station.


Voter turnout among all individuals targeted by the implementation-intention script increased 0.9 percentage points, from 42.9% to 43.8%. In single-voter households, turnout increased 2 percentage points, from 40.7% to 42.7%.

Show Tags
Hide Tags


The project’s goal was to increase voter turnout among potential voters: registered voters who participated in one or fewer primary elections since 2000. Voting is a high-salience, socially important behavior that is essential to a functioning democratic society. While campaigns and organizations spend millions of dollars to increase voter turnout, recent research  suggests that simple interventions like facilitating plan-making can help people follow through on their intentions.



In the days before the 2008 presidential election, a professional firm contacted a sample of unlikely voters in Pennsylvania by telephone. Callers communicated with individual voters using  an implementation-intention script that included a reminder about the upcoming election and also asked three follow-up questions to facilitate plan making. The follow-up questions asked individuals to indicate the time at which they would vote, where they would be going to the polling station from, and what they would be doing before going to vote. Researchers believe that articulating the when, where, and how of following through on an intention creates cognitive links between an anticipated future situation and the intended behavior.

The implementation intention script differed from a standard script that encouraged voter participation by reminding individuals about the election and their civic duty to vote; and from an alternative self-prediction script that included a reminder about the upcoming election and also asked individuals whether they intended to vote.



Change in voter turnout for the 2008 election according to number of voters in a household

randomized evaluation found that asking voters about their intention to vote and prompting them to form voting plans increased overall voter turnout by 0.9 percentage points, from 42.9% to 43.8%. However, among the 23% of targeted households that were successfully contacted over the phone, turnout increased by 4.1 percentage points

The implementation-intentions script demonstrated greater effects across single-voter households than multiple-voter households. Overall turnout in single-voter households increased by 2 percentage points, from 40.7% to 42.7%. In the 22.1% of targeted single-voter households that were successfully contacted, turnout increased by 9.1 percentage points.

Voting can be time-consuming and require coordination among members of a household. The implementation-intentions script may have been rendered impotent in multiple-voter households because voters who live together may be more likely to make voting plans organically than those who live alone. Supporting this hypothesis, this study found that individuals in single-voter households were less likely to have made voting plans before receiving a phone call.


Implementation Guidelines

Inspired to implement this design in your own work? Here are some things to think about before you get started:

  • Are the behavioral drivers to the problem you are trying to solve similar to the ones described in the challenge section of this project?
  • Is it feasible to adapt the design to address your problem?
  • Could there be structural barriers at play that might keep the design from having the desired effect?
  • Finally, we encourage you to make sure you monitor, test and take steps to iterate on designs often when either adapting them to a new context or scaling up to make sure they’re effective.

Additionally, consider the following insights from the design’s researcher:

  • A professional firm conducted the phone calls between Saturday and Monday before Election Day using standardized scripts. The timing of the phone calls during the day was not important in this study.
  • All phone calls were under 2 minutes long, and were conducted using the software CATI.

For more guidance on implementing this design, select “I want to try this” from the left drop-down menu.


Project Credits

David W. Nickerson Temple University

Todd Rodgers Harvard University

Sign up for updates

Stay up to Date

Enter your email for occasional updates on new projects.