National Mail Voter Registration Form
And so the whole shootin’ match comes down to around 4 percent of the voters in six states. [...] Four percent of the presidential vote in Virginia, Florida, Ohio, Iowa, New Mexico, and Colorado is 916,643 people. That’s it. The American president will be selected by fewer than half the number of people who paid to get into a Houston Astros home game last year. [...]
That's Democratic strategist
Paul Begala writing for the Daily Beast, bemoaning the campaigns' quest for swing voters and the outrageous sums of money that will be spent on convincing those few persuadables in the few states that can tip the election. In fact, according to his math, the $2 billion that will be spent on the election—in his mind, all of which is spent purely to sway those fewer-than-one-million minds—works out to $2,181 per swing voter.
Recent polling suggests that Begala isn't exaggerating too much, though. Swing voters have always been a small segment of the population, but at first glance it seems like there are fewer and fewer of them than ever. Polling back in spring of 2012, a point in the campaign where you might reasonably expect a lot of people to still be undecided, showed the vast majority of votes already locked down. Pew found that only 7 percent were truly undecided and not leaning in one party's direction or the other, while the first day of Gallup's tracking poll this year still found Barack Obama and Mitt Romney already taking over 90 percent of each of their party bases.
While it may be interesting to speculate on why there are so few swing voters any more—certainly the sorting-out of the parties into much clearer ideological and regional camps in the last few decades (with the gradual disappearance of conservative southern Democrats and moderate northeastern Republicans) has helped clear up a lot of people's uncertainty about where they belong, while the growth in news outlets with transparent partisan agendas helps reinforce existing political leans—the more important question becomes whether it's worth spending all that money on them. While no campaign should simply pretend swing voters don't exist and ignore them, eventually you reach a point of diminishing returns when trying to reach them (a point that's got to be somewhat lower than the $2,181 per swing voter cited by Begala), especially when there are other potential sources of votes which are not only potentially more cost-effective to tap but also potentially larger.
Alan Abramowitz, a political scientist from Emory University, is one of the most prominent swing-voter skeptics; he's come out with several articles in the last few months arguing that not only are swing voters are overrated as a voting bloc but that it's a better use of Democrats' time and money to focus on unregistered voters instead. We'll look at both of those arguments, starting with the idea that there are a lot fewer swing voters than we think there are (or at least than the news media encourage us to think there are).
To make this case, Abramowitz looks at a study from the 2008 election between Barack Obama and John McCain. The study, by American National Election Studies, followed a panel of voters for more than a year to see when and how they made up their minds. In the first survey the panel took, they were forced to choose between Obama and McCain; there was no "undecided" or "other" option, but they were asked to indicate whether they were extremely sure, very sure, moderately sure, slightly sure, or not sure at all about their choice. Seventy-five percent were extremely or very sure ... but 25 percent were either moderate or slightly sure or not sure at all. That's a lot of swing voters, right?
However, very few of the voters in that second pool—who you'd think were likely to switch, since their hand had been forced in having to choose someone in the first round—wound up changing their minds. All the mind-changing was basically reversion to the norm; Obama benefited slightly, gaining only an additional percentage point in support along the way. As Abramowitz puts it:
Only 8% of respondents switched candidates between June and November. These switches basically canceled each other out: 9% of McCain supporters switched to Obama while 7% of Obama supporters switched to McCain. Ninety-two percent of respondents ended up voting for the same candidate in November that they supported in June. [...] Nine percent of voters in the swing states switched candidates between June and November compared with 7% of voters in all other states.
The people who were likeliest to switch were, simply, those persons whose initial choice was out of whack with their own party identification. Only 1 percent of Democrats who supported Obama in June wound up voting for McCain, while 4 percent of Republicans who supported McCain in June wound up voting for Obama. On the other hand, 32 percent of Democrats who supported McCain in June wound up voting for Obama, while 39 percent of Republicans who supported Obama in June wound up voting for McCain. In other words, the people who in June seemed likeliest to swing, by the end, wound up not swinging at all, but just coming home to their usual party.
Abramowitz uses the data to break the electorate down into four groups: the stayers (who didn't change between June and November), who make up 92 percent of the electorate, returning partisans (voters who initially planned to swing but reverted to their usual party) at 5 percent, and departing partisans (voters who initially planned to stay loyal but then swung to the other party) at 2 percent. That leaves truly "swinging independents" (the group that the news media would have you believe hold the nation in the balance), who accounted for a total of 1 percent of the electorate. Most self-described "independents," Abramowitz points out, started out with a preference and stuck with it, consistent with many other studies' findings that "independents" actually are partisans, just ones who don't want to get saddled with a partisan label.
Over the fold, we'll talk more about the contention that Democrats should focus less on those few swing voters and more on voter registration and mobilization.
For this analysis, Abramowitz looked at Gallup's March 2012 poll of 12 battleground states. Even at that early point in the game (the GOP primary wasn't settled at that point), only 8 percent of the registered voters sampled expressed no preference between Obama and Romney.
A look under the hood at this slice of swing voters (the 8 percent undecideds, plus the 3 percent who were only "leaning" toward a candidate) revealed that they were disproportionately middle-aged (38 percent were 30-49, compared with 32 percent among non-swing voters), female (63 percent, compared with 51 percent of non-swing voters), and white (89 percent, compared with 78 percent of non-swingers). The two most important numbers, though, are that they disproportionately also disapproved of Barack Obama's job performance (75 percent, compared with 44 percent of non-swingers), and had low enthusiasm about voting (58 percent, compared with 25 percent of non-swingers).
The good news is that these swing voters held Mitt Romney is similarly low esteem. Nevertheless, Abramowitz concludes that between the Obama approvals and the lack of enthusiasm, "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."
On the other hand, he also takes a look at the demographic crosstabs on the unregistered voters that they survey reached. (It was a poll of 1046 adults, of whom 871 were registered.) This slice of the sample was disproportionately young (38 percent of unregistered voters were 18-29, compared with 18 percent of registered voters) and non-white (37 percent were non-white, compared with 21 percent of registered voters). In addition, they disproportionately approved of Obama (55 percent, compared with 49 percent of registered voters) and expressed a stronger preference for Obama (61 percent, compared with 50 percent of registered voters).
The one downside for the unregistered voters: They were just as unenthusiastic about voting as the swing voters were. Fifty-two percent reported "low" enthusiasm about voting, compared with 28 percent of registered voters. That points to the main challenge in finding and activitating them as a group: Even if they were successfully registered, that may still not get translated into a vote in November without a lot of follow-up. Here's that data in table form:
| Demographic |
Swing voters |
Non-swing voters |
Unregistered voters |
Registered voters |
| Age 18-29 |
14 |
18 |
38 |
18 |
| Age 30-49 |
38 |
32 |
38 |
33 |
| Age 50-64 |
27 |
28 |
14 |
28 |
| Age 65+ |
22 |
22 |
14 |
28 |
| White |
89 |
78 |
63 |
79 |
| Non-white |
11 |
22 |
37 |
21 |
| Obama approve |
11 |
53 |
55 |
49 |
| Obama disapprove |
75 |
44 |
38 |
46 |
| High enthusiasm |
19 |
48 |
15 |
45 |
| Moderate enthusiasm |
23 |
27 |
31 |
27 |
| Low enthusiasm |
58 |
25 |
52 |
28 |
One other statistic gives some hope that it's still worth the effort of trying to activate these voters: Unregistered voters who support Obama express considerably more enthusiasm about voting than do unregistered voters who prefer Romney. Obama-backing unregistered voters report 17 percent high enthusiasm, 42 percent moderate enthusiasm, and 41 percent low enthusiasm, while Romney-backing unregistered voters report 14 percent high enthusiasm, 20 percent moderate enthusiasm and 66 percent low enthusiasm.
The difficulty in activating unregistered voters in likely Democratic demographics, or voters who are registered but are disengaged and rarely vote but vote Democratic when they do, is one of the factors that leads one of the Democrats' smartest strategists, Ruy Teixeira, to propose a whole new definition of "swing voter." It's a definition that's more inclusive than the traditional media definition that focuses merely on independent, centrist voters who might vote for one or the other of the two parties; instead, it should also include voters who are already partisans, for whom the choice is whether to vote for one party or to not vote at all. As he puts it, "Swing voters are scattered throughout the social structure."
Activating the inactive across the spectrum takes a lot of work, which is why campaigns are loath to do it. It requires more manpower to find unregistered people, register them, and follow up with them to make sure they get to the polls. And it requires more and better targeting, with different pitches to different groups, instead of just coming up with one lowest-common-denominator, mass-mediated message and constantly repeating it with millions of dollars in ad buys behind it. Given the ever-escalating cost of reaching an ever-shrinking pool of persuadable centrists, though, it may start becoming a necessity.
