Volunteer Phone Calls Can Increase Turnout

• Date: 01 May 2006
• Published date: 01 May 2006
• Author: David W. Nickerson
• DOI: 10.1177/1532673X05275923
• Subject Matter: Articles

10.1177/1532673X05275923American Politics ResearchNickerson / Volunteer Phone Calls Can Increase Turnout

Volunteer Phone Calls

Can Increase Turnout

Evidence From Eight Field Experiments

David W. Nickerson

University of Notre Dame

Gerber and Green argue that get-out-the-vote phone calls do not increase turn-

out based upon field experiments testing nonpartisan professional phone

banks. This article argues that the quality of the phone calls matter and that

brief, nonpartisan phone calls can raise voter turnout if they are sufficiently

personal. Totest this hypothesis, a series of eight volunteer nonpartisan phone

campaigns to mobilize voters were studied using randomized, controlled

experiments. The campaigns targeted voters across six different cities in 2000

and 2001. Contra Gerber and Green, the phone calls are found to boost turnout

3.8 percentage points. Based on these estimates, volunteer phone calls produce

one vote for every $26 per vote, which is cost competitive with door-to-door

canvassing. Differences between the professional phone banks previously

studied by Gerber and Green and the volunteer phone banks herein are also

discussed.

Keywords: election campaigns; political participation; voter turnout; field

experiments; phone banks; natural experiments; voting

In the 2004 presidential election, telephones played a crucial role in the get-

out-the-vote (GOTV) strategy for both politicalparties. In Ohio alone, the

Democrats claimed that 3 million phone calls had been made by volunteers

prior to Halloween, whereas Republicans boasted 2.5 million volunteer

phone calls (Polman, 2004). In the 96 hours leading up to election day, both

parties claimed to make nearly 400,000 phone contacts per day (Balz &

Edsall, 2004). In an election that came down to the wire, both parties were

relying heavily upon telephone conversations to move supporters to the

polls.

271

American Politics Research

Volume 34 Number 3

May 2006 271-292

© 2006 Sage Publications

10.1177/1532673X05275923

http://apr.sagepub.com

hosted at

http://online.sagepub.com

Author’s Note: The author gratefully acknowledges helpful comments from Don Green and

anonymous reviewers and the cooperationof the Youth VoteCoalition. The author would also

like to thank the PewCharitable Trusts and the Institute for Social and Policy Studies for funding

the research.

GOTV phone calls now constitute a large part of American electioneer-

ing; however,the efficacy of phone calls as a means of increasing voter turn-

out has been recently questioned. In 1998, Gerber and Green conducted two

large-scale field experiments to test the utility of nonpartisan phone calls as a

means of increasing voter turnout in New Haven and West Haven, Connecti-

cut (Gerber & Green, 2000b, 2001). In contrast to the numerous smaller

experiments (Adams & Smith, 1980; Eldersveld, 1956; Miller, Bositis, &

Baer, 1981) and observational studies (e.g., Rosenstone & Hansen, 1993),

Gerber and Green conclude that phone calls do not move voters to the polls.

Contrary to expectations, both Gerber and Green experiments detected a

noisy decrease in voter turnout from receiving a brief, nonpartisan GOTV

call from a professional phone bank (–2.0 percentage points for New Haven

and –1.3 percentage points for West Haven, each with a standard error of

roughly 2).1Pooling the two Gerber and Green experiments, the m obiliza-

tion effect of the phone calls is estimatedto be –1.6 percentage points with

a standard error of 1.4. The odds of the Gerber and Green calling campaigns

conducted in 1998 being effective at mobilizing the vote (i.e., having a posi-

tivetreatment effect) are roughly 13 in 100. If Gerber and Green (2001, p. 82)

are correct when they describe phone calls as ineffective and increasing turn-

out “not at all” (Gerber & Green, 2000b, p. 653), then both political parties

made a costly mistake, financially and politically.

The expectations of both political parties and the Gerber and Green null

finding can be reconciled if the assumption that all phone calls are equally

effective is relaxed. The random assignment to treatment and control groups

assures the internal validity of the two 1998 experiments, however,the exter-

nal validity of the finding remains an open question.2There are two primary

reasons to believe that phone calls can be effective in mobilizing voters and

that Gerber and Green detected the effectof a particular type of phone call.

First, phone calls are effective across a wide range of settings.

Telemarketing would not havedeveloped into an industry with $100 billion

in annual sales if phone calls were not effective at motivating consumers (“A

New Internet Democracy?”, 2003). Phones are also used to encourage chari-

table donations and volunteer work in a wide range of settings. For instance,

a randomized field experiment found phone calls using a “foot in the door”

message to be effective in eliciting blood donations (Hayes, Dwyer,

Greenwalt, & Coe, 1984). If a phone call can motivate monetary and blood

donations, why should voter turnout differ?

Second, door-to-door canvassing and direct mailings are effective at

mobilizing voters (Gerber & Green, 2000b; Gerber, Green, & Green, 2003;

Green, Gerber, & Nickerson, 2003). The reason typically given for the

increased efficacy of face-to-face efforts over mailings is the more personal

272 American Politics Research