Republicans have money; Democrats have bodies
Whether you lead a group, staff a campaign, or are an activist, odds are you’d like to know how you can most efficiently use volunteer time for results at the ballot box.
What are the most effective methods of getting out the vote? Setting aside cost, where volunteers provide almost free labor, what volunteer activities are most likely to send voters to the polls?
This article summarizes large-scale tests, studies, and studies of studies, measured by political science professors, often working with voter mobilization organizations such as the League of Women Voters.
Executive Summary
Turnout increase
Commit Cards
- 5% to 11%. Phone or in-person contact plus pledge card can increase turnout by 7.9%, although signed pledges are notoriously challenging to secure.
- Regarding optimal language, see Green, McGrath & Aronow, at 36. This research found "I will vote because _______" increased the rate of return.
Canvassing
Personal phone calls
- If interactive and with a focused message: 1.9% to 3.3%
- If nonpartisan, as high as 3.8%
- Scripts asking where and when recipients intended to vote can add as much as 0.9%
Mass Mailings
Email
- Personal emails appear to have a strong GOTV effect on acquaintances who open them.
- Mass emails from strangers may have a weakly negative effect.
Text Messages
- To recipients who previously agreed to receive such reminders can increase turnout as much as 3.0 to 4.1%.
Postcards
- Mass produced: approximately 0.7% to 2%
- Handwritten: while studies indicate impact on voters may be as much as 0.4%, a recent study indicated writing postcards energizes volunteers, helping activists stay engaged and feel more likely to take future political action.
- As with email, the effect on friends and neighbors may be even greater, particularly with appeals to civic duty and the social norm of voting.
Robocalls
Sources and Exclusions
Sources. This summary drew in large measure from analyses by political science professors working with the League of Women Voters and other voter contact groups on large-scale tests, studies, and studies of studies reviewing voter contact and related public voting records. I am indebted to the Yale Institution for Social & Policy Studies and authors at numerous major universities for making many of these studies available freely online. Hyperlinks in this message take you to the sites where source analyses resided at the time of drafting this article. Useful overviews by primary authors are available at http://isps.yale.edu/node/16698 and http://www.mcmcgrath.com/documents/2013GreenMcGrathAronow.pdf. All errors are mine alone, and corrections welcomed.
Exclusions:
- This discussion does not address fundraising, voter registration, or volunteer-mobilization activities, which can have very beneficial effects.
- It does not address the efficacy of viewpoint persuasion methods.
- It does not address in detail the effects of timing during a campaign.
- It does not analyze the cost effectiveness of methods.
Detailed Discussion
Canvassing increases turnout by approximately 4.3 - 4.8% of voters who answer the door.
- For example, a "Michelson (2006) study found that canvassing increased turnout among young Latino voters from 7.0% to 9.4%. Given a contact rate of 50%, the estimated CACE was 4.8 percentage points."
- The best appeals are personal, unhurried, and exert social pressure such as "very strong appeals to the social norm of voting, presentation of information from the turnout rolls, information about turnout levels in the voter’s neighborhood, or mention of the practice of recording turnout information in administrative records".
Commit / pledge cards. Convincing people to sign a pledge is challenging. The rewards can be significant, however.
- One study showed that commit cards increased turnout by 5%, from 5% (voters receiving only a reminder) to 10%.
- Another study showed that average turnout for nonprofit pledge card voters was 11 points higher than turnout of all registered voters (59% vs. 48%).
- This report on a study of phone or in-person contact plus getting voters to sign a pledge card (note, this is a double-contact method) increased turnout by 7.9%.
Mass mailings in general have little to no effect on mobilizing voters , studies show, unless the message applies social pressure:
- Where mass mailings contained strong social pressure, such as listing the voting records of the household and neighbors and promising a future mailer with updates, if sent to voters with low to middling propensity to vote, the likelihood of voting increased by 3.3 to 3.6%.
- The downside, however, was the negative reaction from recipients, which in turn generated negative press coverage and angry complaints.
- Accordingly, positive peer pressure can be preferable, such as thanking people for voting in the last election, asking people to pledge that they will vote, and cordially reminding them voting is a matter of public record.
- Mailings can also increase name or face recognition of the candidate or opponent, which can meet campaign needs separate from mobilizing the vote.
- Mass mailings can be cost effective contrasted with, for example, paid canvassers. When canvassing labor is provided by volunteers, however, cost becomes less of an issue.
Postcards
- Mass produced postcards can be approximately 0.7% to 2% effective.
- The impact of hand written postcards are challenging to study, due to the lack of uniform messaging and timing. Neverthless, it appears the impact on volunteers is beneficial.
- Again, messaging is key. Before a primary, for example, "messages that emphasized the social norm of voting (what sort of person does or does not vote) were twice as effective as a basic GOTV mailing, and some messages that focused on why primary voting was especially important seemed more effective than standard GOTV messaging". Appeals to civic duty can be helpful.
Email
- Mass emails appear to have little productive effect and potentially a weakly negative effect, and emails from the local registrar of voters only increased voting by 0.5%.
- Personalized emails to acquaintances, in contrast, appear to have a strong CACE among those who open the emails. See Davenport, T. C. (2012). "Unsubscribe: The effects of peer-to-peer e-mail on voter turnout: Results from a field experiment in the June, 6, 2006 California Primary Election." Unpublished manuscript, U.S. Naval Academy, cited in Donald P. Green, Mary C. McGrath & Peter M. Aronow, "Field Experiments And The Study Of Voter Turnout," 23 J. Elec.s, Pub. Opinions & Parties 1, pp. 27-48 (2013), available at http://www.mcmcgrath.com/documents/2013GreenMcGrathAronow.pdf
Text messages to cell phones whose recipients previously agreed to receive this type of reminder raised voting likelihood by approximately 4.1%.
Leading to this fact: personal communication with voters tends to be most effective.
- Robocalls have limited positive effect and can even decrease turnout. According to a 2013 overview, the average CACE (complier average causal effect) is only 0.980% with a 95% interval ranging from 0.504 to 1.456%.
- Phone calls by individuals, in contrast, if interactive and with a focused message, can increase turnout by 1.9% to 3.3%. If nonpartisan, effects can climb as high as 3.8%. Calls asking where and when recipients intended to vote increased turnout, for individuals living alone, by as much as 0.9%.
- Acquaintance Contact. “Maybe the most effective is a close friend or coworker who says ‘let’s walk to the polls together’” — Prof. Donald Green. Similarly, sending text messages to friends on Election Day can have significant effect.
* Early Voter Contact: although early canvassing and calls can have some persuasive effect, it decays over time. See, e.g., Green, McGrath & Aronow, supra, at 34-35. Visiting voters before the last few days of a campaign is nevertheless useful for identifying local volunteers, getting voters excited about the campaign, identifying likely voters for each candidate, and identifying voters who may need assistance.