By turning down nearly $85 million in public financing for the 2008 general election, Barack Obama’s campaign has put itself in a position to spend at least $200 million between the Democratic convention and Election Day. Much of that money, and millions more from state and national Democratic Party organizations, labor unions, and other allied groups will be devoted to voter mobilization: registration and get-out-the-vote (GOTV) efforts aimed at substantially increasing turnout among African Americans, Hispanics, and younger Americans, all of whom are expected to strongly support the Democratic candidate--if they vote. Hundreds of paid organizers have already been dispatched to key states such as Michigan, Missouri, and Ohio to prepare for a massive voter registration drive in the late summer and fall. Some three hundred paid organizers have reportedly been assigned to Ohio alone.
The Obama campaign is apparently convinced that a well-funded voter mobilization campaign can dramatically increase turnout among these pro-Obama demographics and potentially alter the outcome of the election in states including Colorado, Virginia, North Carolina, and even Georgia, where such groups make up a large proportion of the population and an even larger proportion of non-voters. But how realistic is this belief? And is this the best way for Obama to be spending his millions? Evidence from the 2004 presidential election and the findings of recent GOTV research indicate that mobilization campaigns work even if their effects are fairly modest. Dollar for dollar, money spent on voter registration and get-out-the-vote drives probably pays bigger dividends than money spent on further saturating already saturated media markets with 30-second TV spots
The best research on voter mobilization has been done by political scientists Donald Green and Alan Gerber of Yale University. Their results are generally considered the gold standard in this area because they are based on hundreds of carefully designed controlled experiments with real voters in real elections. And although their research thus far has focused exclusively on GOTV campaigns, their findings would appear to apply to voter registration campaigns as well. Anyone seriously interested in this subject should get a copy of the just-released second edition of their book, Get Out the Vote!
What Green and Gerber have found, in a nutshell, is that the most effective mobilization campaigns rely on door-to-door canvassing by well-trained, enthusiastic volunteers who are familiar with the community in which they are canvassing. Telephone canvassing, when done by skilled, enthusiastic volunteers, is a good second option because it is generally less costly than door-to-door canvassing. Other techniques, such as leafleting and mass mailings, have smaller effects. Finally, media-based campaigns (i.e. radio and TV ads urging people to vote) are generally ineffective, although Green and Gerber report in the second edition of their book that the 2004 “Rock the Vote” campaign appears to have had some success in raising turnout among its target audience of younger voters.
Still, the bottom line is that even the most effective GOTV campaigns only increase turnout by a few percentage points. Moreover, the impact of these campaigns on election outcomes would be even more modest, since not all of those who turn out will vote for the same party or candidate. It is almost impossible to mobilize supporters of your own party without mobilizing at least some supporters of the opposing party.
The results of the 2004 presidential election provide some additional, albeit indirect, evidence about the effects of voter mobilization campaigns. In 2004, the two major parties and allied organizations poured hundreds of millions of dollars into voter registration and GOTV campaigns. These efforts were heavily concentrated in a dozen battleground states whose electoral votes were considered up for grabs.
Turnout was substantially higher in the battleground states than in the rest of the country in 2004. According to data compiled by Michael McDonald of George Mason University, an average of 67 percent of eligible voters turned out in the twelve targeted battleground states compared with only 60 percent in the other 38 states. More significantly, voter turnout increased by an average of 6.7 percentage points in the battleground states compared with only 3.8 percentage points in the rest of the country. It seems likely that the millions of dollars poured into voter registration and GOTV campaigns in the battleground states were responsible for most, if not all, of this difference.
So who benefitted from increased voter turnout in 2004? Despite all of the claims about the effectiveness of the GOP’s “72 hour program,” a multiple regression analysis of the election results shows that the higher the voter turnout in a state, the better the Democratic presidential candidate did. For every one point increase in turnout, John Kerry’s share of the vote increased by between one-tenth and two-tenths of a percentage point. Unfortunately for Kerry, though, this effect was much too small to change the outcome of the election.
It has long been part of the conventional wisdom of American politics that voter mobilization campaigns are more important for Democrats than for Republicans. That’s partly because the Democratic Party’s electoral base is made up disproportionately of lower income and minority voters who may need an extra push to register and get to the polls on Election Day. In addition, however, Democrats have an advantage when it comes to voter mobilization because their core supporters are more geographically concentrated. This makes voter registration and GOTV campaigns more efficient because you’re less likely to inadvertently mobilize supporters of the opposing party.
Political science research and the results of the 2004 election demonstrate that pouring millions of dollars into voter registration and GOTV campaigns will not turn red states blue. However, a substantial investment in voter mobilization could increase the Democratic share of the vote by a point or two in a few swing states, and based on the last two presidential elections, that could well be the difference between victory and defeat.
Alan I. Abramowitz is a professor of political science at Emory University.