Partisan Mail and Voter Turnout in the United States

Researchers:
Matthew Green
Fieldwork by:
Location:
Connecticut and New Jersey, United States
Sample:
29,454 households in two states
Timeline:
1999 - 1999
Target group:
  • Voters
Outcome of interest:
  • Electoral participation
  • Voter behavior
Intervention type:
  • Nudges and reminders

In the United States, political campaigns increasingly rely on more impersonal and professionalized campaign methods, including direct mail, commercial phone banks, and television advertising. Direct mail may alter voter behavior, and its effects may vary depending of the content of the messages and the target audience. Yet the impact of direct mail campaigns on voter behavior is not well understood. Through a randomized evaluation, researchers examined the effects of partisan mail campaigns on voter turnout in state and municipal elections in Connecticut and New Jersey. Results indicate that partisan direct mail campaigns do little to stimulate voter turnout.

Policy issue

In recent decades, there has been a steep increase in both the amount of money spent on political campaigns and the professionalism with which those campaigns are carried out in the United States. However, increased levels of professionalism and campaign spending have not led to increased rates of voter participation. Voter turnout could be influenced by the way voters are mobilized. One of the hallmarks of recent campaign tactics is a shift away from personal forms of mobilization to more impersonal methods, like direct mail, phone calls, and television advertisements. These tactics are relatively less expensive than personal forms of mobilization like door-to-door canvassing and can be more quickly deployed at a large scale.

Survey data from the American national election study (ANES) have shown a strong positive correlation between direct mail and voter participation, but there is little evidence using experimental designs on whether direct partisan mail increases voter turnout. On the one hand, direct mail may increase turnout if it raises awareness about the existence of a candidate, provides information that reinforces partisan preference, persuades the undecided to vote for a candidate, or converts voters. On the other hand partisan direct mail could potentially shrink turnout if it discourages supporters whose candidate is being attacked or results in a general disenchantment with both candidates. What is the impact of partisan direct mail campaigns on voter behavior?

Context of the evaluation

In the United States, political campaigns increasingly rely on more impersonal and professionalized campaign methods. Party and campaign activists have been replaced by specialists in campaign management and mass communication, and direct mail, commercial phone banks, and television advertising are being more widely used than face-to-face canvassing. Direct mail is considered an important tool for reaching voters in large electoral districts and is widely used in state and local elections, where campaigns often lack the necessary financial resources to afford television advertising.

This shift in campaigning methods has taken place over a time of long-term decline in voter turnout. From a relatively high turnout rate in the early 1960s of about 63 percent, voter turnout to national elections had declined by six percentage points (to about 57 percent) by 2008. Local elections have followed a similar trend: turnout to Connecticut’s State and Municipal elections decreased from 81 percent in 1988 to 42 percent in 2003. Similarly, turnout to New Jersey’s General Assembly elections declined from 73 percent in 1961 to 47 percent in 2009.

The evaluation focused on two local campaigns: Connecticut’s mayoral race, and New Jersey’s legislative elections, both held in 1999. In Connecticut’s mayoral election, the race between the Republican incumbent mayor and the competing Democratic candidate was expected to be relatively close. In New Jersey, the two Democratic incumbents conducted a joint campaign and were predicted to have a fairly easy win.

Details of the intervention

Researchers examined the effects of direct mail campaigns on voter turnout during two elections in 1999: Connecticut’s mayoral election and New Jersey’s state legislature race. The campaigns both relied on direct mail as their principal means of communicating with voters, as none of the candidates could afford a television advertisement campaign.

The Connecticut evaluation included households with registered partisan voters, registered unaffiliated voters with a voting history, and households with newly registered voters. In total there were 9,900 households in the sample. In New Jersey, the sample comprised over 19,500 households, including partisan voters with a record of high turnout rates, partisan and independent voters with lower turnout records. Using mailing lists, researchers randomly assigned households to treatment and comparison groups. The participants received partisan mailings varying in number and content, some receiving up to nine pieces of mail during the final weeks of the campaign. The tone of the two campaigns was different: the non-incumbent candidate in the Connecticut mayoral election waged a negative campaign focusing on the opposing mayoral candidate’s weaknesses and the New Jersey incumbents’ campaign had a positive tone, highlighting the candidates’ legislative activities and supporting broadly popular causes. The comparison groups did not receive any mail from these campaigns, although voters did received some mail from other campaigns.

Researchers used voting records to compare the turnout rates for voters included in the mailing campaigns to those in the comparison groups and to calculate the marginal impact of each type of mailing.

Results and policy lessons

In both the Connecticut mayoral election and the New Jersey state legislature election, partisan campaign mail did little to stimulate voter turnout. Partisan campaign mail had a negligible effect on turnout regardless of the previous turnout history and affiliation of voters. In New Jersey, partisan mail with a positive tone had no significant impact on voter turnout, regardless of voters’ affiliation or previous turnout rates. In Connecticut, where the candidate waged a negative campaign, the turnout rate for voters that received the mailers slightly (though insignificantly) decreased, perhaps suggesting that effect of mailings may be contingent on the tone of their content and that a negative campaign may depress voter turnout. Precisely why direct mail has so little impact on voter behavior is still an open question: impersonal contact may make the information less salient and memorable, may not credibly convey the urgency of the request, or may not create the necessary personal connection between voters and the electoral process.

Partisan direct mail is intended to be an inexpensive way of reaching voters, but given the negligible impact partisan direct mailings have on voter turnout, it may be a less efficient use of campaign money for increasing turnout of voters who might otherwise abstain, relative to other mobilization tactics. In another study, researchers found that non-partisan personal canvassing was highly effective in increasing turnout. The effectiveness of face-to-face canvassing coupled with the limited impact of direct mail provide support for the hypothesis that the declining turnout rates in the United States may be connected to the shift towards more impersonal mobilization methods.

Gerber, Alan S., Donald P. Green, and Matthew Green. 2003. "Partisan Mail and Voter Turnout: Results from Randomized Field Experiments." Electoral Studies 22(4): 563-579.