These pieces used what the polling showed to be the most powerful and relevant issues with our targeted voters: Mitt Romney’s threat to veto the DREAM Act, the need to stand with President Obama and quotes from Harris County Sheriff Garcia saying that voting was essential to protect the community. The final piece in the program returned to emphasizing that the recipients were indeed registered to vote.

Typically, communications plans to swing voters are built upon the understanding that repetition is key to message penetration. Too often, plans for mailings to infrequent base voters ignore this principle. The Harris County Hispanic Turnout Project avoided this pitfall and sought to persuade the targeted voters with repeated, emotional and relevant messages.

The mail was backed up by a program that included door-to-door, phone and broadcast radio. Building on already existing civic engagement activities, the effort met with startling success: Of more than 140,000 low-propensity Hispanic voters in Harris County who were mailed, 58 percent voted. High-propensity Hispanic voters in the county (none of whom were mailed) voted at a nearly identical rate of 58.2 percent.

The turnout rate of the county’s low-propensity voters who were not mailed was only 25.6 percent. In short, low-propensity voters who received mail mirrored the performance of high-propensity voters and more than doubled the turnout of low-propensity voters who did not receive the mail.

What’s more, the 58 percent turnout of low-propensity Hispanic voters in Harris County who received the mail compares to a 48 percent turnout among Hispanic voters nationwide and a 39 percent turnout among Hispanic voters statewide in Texas.

The Harris County election results included a 971-vote victory for President Obama, a comfortable reelection for Sheriff Adrian Garcia and the protection and election of 14 Democratic District Judge candidates, many by razor-thin margins. The narrowest victory margin was 1,694 votes, five races were decided by fewer than 10,000 votes and only one of the 14 victories was by a margin greater than 27,000 votes.

The differential turnout between low-propensity Hispanic voters who were mailed and those who did not receive the mail would translate to an increased turnout of over 46,000 votes. At a minimum, it is likely that the program drove an additional 25,000 to 30,000 votes. The program clearly made the difference in most, if not all of the 14 District Judge victories.

A highly integrated campaign

While some have recently raised questions about the efficacy of mail in driving base Democratic voter turnout, the results of the 2012 Harris County Hispanic Turnout Project provide strong evidence that powerful, emotional and effective mail—combined with the newly discovered need to inform some voters of their eligibility to vote—can motivate voters and increase base turnout.

Indeed, the impact of these mailings (as part of a program that integrated mail, field, phones and radio) appears to far outpace previously reported turnout effects of social pressure messaging, which has only furthered our shared skepticism towards the social science research behind that messaging.

Our skepticism has been partially based on the limitations inherent in the systematic experimentation of a single variable in the inherently multivariate real world of campaigns. In the specific case of social pressure messaging research, we have also been struck by the limitations of the testing to date.

We do not find it surprising that social pressure messaging has outperformed mailings that appeal to civic duty. Given how common appeals to civic duty are at election time, if such messaging were effective, infrequent base voters would not be infrequent voters. Nor should it be surprising that social pressure messaging outperforms typical partisan get-out-the-vote messaging that views the process as a mechanical one of voter identification and mobilization.

Rather, what these results show is the need to test social pressure messaging against mail created to persuade infrequent base voters that voting is vital to them and their families. In addition, the phenomenon of significant numbers of base voters being unaware that they are registered and eligible to vote is an important discovery that we believe warrants further testing.

Dave Gold is President of Gold Communications, a Democratic direct mail firm. Andrew Myers is President and CEO of Myers Research, a Democratic polling firm.