Make It Personal

Get Out the Vote: How to Increase Voter Turnout (Second Edition) by Donald P. Green and Alan S. Gerber, Brookings Institution Press, 225 pages, $18.95

The late Alan Baron, sometime political consultant and full-time political wag, used to tell a story about a campaign kibitzer in Des Moines in the autumn of 1964. Every day as the election drew closer, the kibitzer would turn up at Democratic Party headquarters and implore the directors of the get-out-the-vote (GOTV) operation to use sound trucks. Such trucks, he insisted, if properly used--touring the streets of Des Moines while amplified voices urged the folks on the sidewalks to vote Democratic--were guaranteed to tilt the election in the Democrats' favor. And on Election Day, in no small part just to shut the guy up, the field directors did dispatch a couple of sound trucks to ride around town.

That day, the Democrats won a victory of historic dimensions. Powered by Lyndon Johnson's epochal landslide over Barry Goldwater, Iowa Democrats won all their usual offices and a slew of others they'd never won before.

The next morning, the kibitzer resurfaced in Democratic headquarters. "What did I tell you?" he gloated. "Sound trucks."

One of the problems for political candidates and their campaign managers is that seemingly more respectable versions of Baron's kibitzer are to be found in countless (though by no means all) political-consulting firms across the country. They tell prospective clients how their mailings, phone banks, or TV spots have boosted their previous clients' turnout, and if their prospective clients would just sign this contract, they too will know the thrill of victory.

To which Donald Green and Alan Gerber, both professors of political science at Yale, respond in no uncertain terms: Oh, yeah?

What Green and Gerber have done would seem conceptually obvious--except, no one has done it before. Working with academic colleagues and a range of political and civic groups and campaigns across the nation, they ran more than 100 experiments in elections over the past decade, testing which get-out-the-vote tactics--direct mail, phone banks, door-to-door canvassing, radio and television ads--actually turned out more voters. They designed all manner of GOTV efforts and employed them on groups of randomly selected voters while not employing them on a control group of other randomly selected voters, then checked after the election to see who'd voted and who hadn't, and whether those results had any correlation to the respective GOTV drives. They worked with nonpartisan good-government groups, with groups trying to mobilize African American, Latino, low-income, or environmentally inclined voters, and occasionally, and, remarkably, with candidates' campaigns--remarkably, because not many candidates will respond favorably to establishing a control group of voters who don't get canvassed or phoned or mailed on his or her behalf.

Get Out The Vote is not a theoretical consideration of political engagement and alienation. It does not advocate shifting our system toward, say, proportional representation or parliamentary forms of government as a way to increase popular participation in elections. Green and Gerber are concerned simply and totally with the actual electoral system we have saddled ourselves with, and their goal, as they put it, is to produce a "shopper's guide" for candidates pondering whether to use robocalls or canvassers. The great question they hurl at the reader in their very first sentence bears no trace of the controversies over postmodernist theory: "What are the most cost-effective ways to increase voter turnout?" So relentlessly practical are their concerns that at times, these two eminent Yale political scientists sound almost disconcertingly like your mother. "Along with clipboards containing maps and address lists," they write, "canvassers should carry plastic covers in case of rain."

But the sheer number and scale of the experiments they've run make Get Out The Vote a signally important tool to campaigners trying to figure out how best to campaign. It is also a signally important challenge to portions of the political- consulting industry, most particularly those consultants whose GOTV campaigns rely on recorded phone calls, paid phone bankers, or typical direct mail.

What Green and Gerber have found, in brief, is that the personal touch matters. "Door-to-door canvassing by friends and neighbors is the gold-standard mobilization tactic," they write. It's the contact itself that's the key: the kind of message that the canvassers delivered--whether they handed voters a position paper or a potholder--in itself had no effect on turnout rates. Phone banks staffed by genuinely enthusiastic and chatty volunteers worked as well. In test after test, however, a series of mailings to voters, or recorded phone calls from notables, had no measurable effect on voter turnout. The kind of canvassing and phone banks that work well, of course, are not easy to assemble, so Gerber and Green endeavor to price out the cost per vote of any number of these options, since not every campaign can mount an effective canvass (even though at $16 per vote, it is the most efficient way to turn out voters).

Green and Gerber may have produced the first rigorous academic study of how to move voters, but their work closely corresponds to some studies conducted outside academia over the past quarter-century. In the 1984 Democratic presidential primary in New Hampshire, for instance, the AFL-CIO campaigned for its endorsed candidate, Walter Mondale, through a program of recorded phone calls from the federation's president, Lane Kirkland, urging a vote for Mondale. The assumption that New Hampshire union members knew who Kirkland was, or cared whom he backed, was delusional, as the exit polls that showed Gary Hart beating Mondale among union members rather dramatically demonstrated. But in Berlin, the isolated mill town at the northern end of the state, a young union organizer had tried out a pilot program in which union stewards personally campaigned for Mondale among their fellow union members--and when Mondale swept Berlin (which consisted of little else besides the mill), the organizer, whose name was Steve Rosenthal, felt he was on to something.

In the late 1990s, when Rosenthal was the political director of John Sweeney's AFL-CIO, he was in a position to do more ambitious experiments on what moved union voters. He concluded what Green and Gerber have now concluded: Personal peer contact is indisputably the best way to turn out voters. Rosenthal's experiments are one reason why unions' GOTV efforts are generally successful.

In all their surveys of direct-mail programs, Green and Gerber found only one kind of mailing--a highly unusual kind--that pushed people toward the polls. The mailing included a report on how frequently the recipient had voted in recent elections and how frequently his or her neighbors had voted. Combine their findings on the efficacy of such peer-group pressure with their findings on door-to-door contacts, and the optimal Green and Gerber voter-turnout program would seemingly be based in a neighborhood organization whose members knocked on doors and kept tabs on the voting habits of their neighbors. There's a name for such an organization, or, at least, there was once: Tammany. One hundred years after Charley Murphy and George Washington Plunkitt ran New York, political science has determined that they were doing something right.