(Editor’s Note: We are extremely pleased to publish this significant strategic analysis by noted political analyst and TDS contributor Alan Abramowitz, the Alben W. Barkley Professor of Political Science at Emory University)
With five months to go until Election Day 2012, all indications are that the presidential race between Barack Obama and Mitt Romney is going to go down to the wire and that the outcome will ultimately be decided by voters in 10-15 battleground states where neither candidate has a significant advantage.
In deciding how to allocate money and other resources in these battleground states, the key question facing the Obama campaign is how much emphasis to give to voter registration and mobilization versus persuasion of undecided and weakly committed swing voters. The conventional wisdom about the 2012 presidential election, trumpeted by most pundits and media commentators, is that the outcome will be decided by the swing voters and that the candidate who is viewed as closest to the center will have the best chance of winning their support. However, the evidence presented in this article, based on recent polling data from the battleground states, shows that Democrats have little chance of winning over many swing voters but a much better chance of winning the votes of the unregistered if they can get them on the voter rolls and turn them out on Election Day.
Swing Voters: Unhappy with Obama but Unenthusiastic about Voting
In order to compare the potential payoffs of a strategy emphasizing mobilization compared with one emphasizing persuasion, I analyzed data from a March 20-26 Gallup Poll in twelve key battleground states: Colorado, Florida, Iowa, Michigan, Nevada, New Hampshire, New Mexico, North Carolina, Ohio, Pennsylvania, Virginia and Wisconsin. This was the most recent battleground state polling data available for analysis. A total of 1046 adults were interviewed on landline and cellular telephones including 871 registered voters.
One important finding from Gallup’s battleground state poll is that there were relatively few swing voters in these swing states. Among registered voters, 49 percent supported Barack Obama and another one percent indicated that they leaned toward Obama while 41 percent supported Mitt Romney and another two percent leaned toward Romney.
The March 20-26 survey was conducted at a time when Mitt Romney was still battling with Rick Santorum for the Republican nomination. Now that Romney has locked up the GOP nomination, Obama’s lead in these battleground states may very well be smaller. What is striking, however, is that as early as March, relatively few registered voters were unwilling to state a preference in a Romney-Obama contest. Even combining leaners with the undecided, swing voters made up less than 10 percent of the electorate in these twelve states.
Still, with the race between Obama and Romney expected to be very close, even a small group of swing voters could decide the outcome. So who were these swing voters? To answer this question, I compared the characteristics and political attitudes of swing voters (those who were undecided or only leaning toward a candidate) with the characteristics and attitudes of registered voters who were supporting either Obama or Romney. The results are displayed in Table 1.

The data in Table 1 show that compared with voters supporting a candidate, swing voters were disproportionately white and female. They were also much more likely to describe themselves as completely independent and much less likely to describe themselves as Democrats or independents leaning toward the Democratic Party than other voters. But the most dramatic differences between swing voters and voters supporting a candidate involved their opinions about President Obama and their enthusiasm about voting in 2012.
Swing voters had much more negative opinions of President Obama’s job performance than other voters. In fact their opinions were almost as negative as those of Romney supporters. Only 11 percent of swing voters approved of Obama’s job performance compared with 6 percent of Romney voters. In contrast, 92 percent of Obama voters approved of the President’s job performance.
But while swing voters were similar to Romney voters in their evaluation of President Obama’s job performance, they were much less enthusiastic about voting. Only 19 percent of swing voters described themselves as extremely or very enthusiastic about voting in 2012 compared with 47 percent of Romney supporters and 50 percent of Obama supporters. And 58 percent of swing voters described themselves as not too enthusiastic or not at all enthusiastic about voting compared with only 27 percent of Romney supporters and 21 percent of Obama supporters.
These findings suggest that efforts by the Obama campaign to persuade swing voters are likely to be unproductive and could even backfire. These voters have a decidedly negative view of the President and are very unlikely to vote for him. The best the Obama campaign can hope for is that most of these swing voters will stay at home on Election Day.
The Other Unknown in the Equation: Unregistered Voters
In addition to swing voters, there is another group in the electorate whose behavior has the potential to influence the outcome of a close presidential election–those who are not currently registered. In fact, in the Gallup battleground state poll there were almost twice as many unregistered voters as swing voters.

Not only did unregistered voters outnumber swing voters, but their characteristics and political attitudes were very different from those of swing voters or those of registered voters. Table 2 compares the characteristics and attitudes of unregistered voters with those of registered voters in the Gallup battleground state survey. Unregistered voters were disproportionately young and nonwhite and, in marked contrast with swing voters, had more favorable opinions of President Obama’s job performance than registered voters. Most importantly, when asked about their presidential candidate preference, unregistered voters chose Barack Obama over Mitt Romney by a better than two-to-one margin.
These findings suggest that the Obama campaign would be well advised to focus its efforts in the battleground states on voter registration and turnout rather than on trying to win over swing voters. However, unregistered voters, like swing voters, were rather unenthusiastic about voting. Getting them registered and to the polls could be challenging.

But while unregistered voters in general were unenthusiastic about voting, unregistered Obama supporters were considerably more enthusiastic than unregistered Romney supporters. This can be seen very clearly in Table 3. Fifty-nine percent of unregistered Obama supporters were at least somewhat enthusiastic about voting compared with only 34 percent of unregistered Romney supporters. These results suggest that a strategy that emphasizes turning unregistered Obama supporters into Obama voters could pay significant dividends for the President’s reelection campaign in the swing states.
A Note on the Results of the Wisconsin Recall Election:
Turnout Key to Walker Victory
The level of overreaction to the Wisconsin results, even by some usually sensible folks like Ezra Klein and Greg Sargent, is excessive. This one election does not mean that we are now in a new, “post-Citizens United” era in American politics. It is not necessary to diminish the importance of the Supreme Court’s decision to recognize that, empirically speaking it is simply not why Barrett lost. He lost because, as the exit polls revealed, a lot of Wisconsin voters were uncomfortable with the idea of recalling a sitting governor in the absence of evidence of misconduct in office and because the Republicans turned out in larger numbers than Democrats. Massive advertising certainly played a role in the election but it wasn’t the key factor.
An examination of the voting patterns and exit poll results in Tuesday’s Wisconsin recall election indicates that, in fact, turnout was a key factor in incumbent Republican Scott Walker’s victory over his Democratic challenger, Milwaukee Mayor Tom Barrett. While there was a heavy turnout for a special election, the final total of just over 2.5 million votes fell well short of the nearly 3 million votes cast in the 2008 presidential election. And Republicans appear to have done a better job of getting their voters to the polls. Turnout for the recall election was 91 percent of 2008 turnout in suburban heavily Republican Waukesha County, the largest GOP county in the state, but only 83 percent of 2008 turnout in Milwaukee County, the largest Democratic county in the state.
The same pattern was evident in the exit poll results. The 2012 recall electorate was noticeably older, whiter, more conservative and more Republican than the 2008 electorate. Voters age 65 and older outnumbered those under the age of 30 by 18 percent to 16 percent on Tuesday. In contrast, four years ago, 18-29 year-old voters outnumbered those 65 and older by 22 percent to 14 percent. Most significantly, on Tuesday Republicans outnumbered Democrats by 35 percent to 34 percent according to the exit poll. Four years ago, Democrats outnumbered Republicans by 39 percent to 33 percent.
Despite Scott Walker’s fairly easy win on Tuesday, Democrats apparently were able to retake control of the state senate by defeating one GOP senator. And Democrats can take heart from one result from the exit poll. Even with a Republican-leaning electorate, Barack Obama led Mitt Romney by 51 percent to 44 percent when exit poll respondents were asked how they would vote in the presidential election. These results suggest that, Obama should be considered a solid favorite to carry the state again, especially if Democrats turn out in larger numbers in November.