USC logo

All Swings Considered

Published by USC Bedrosian Center on

by David Gastwirth

Swing voters took center stage at this week’s [Wed, October 31st – one week before the election] Road to the White House. Given the undivided attention being paid to swing voters in the final stretch of the race, the topic is quite fitting for our last discussion leading up to the election. The panel included Michael Davidson, CEO of Gen Next; Giuseppe Robalino, Member of USC College Republicans; Teddy Davis, Press Secretary for Los Angeles Mayor Villaraigosa; and Aaron Taxy, President of USC College Democrats. Much of the conversation focused on identifying swing voters and understanding the rationale for the designation. Identifying and influencing the “swing vote” is the modus operandi for professional politicos today. Michael O’Brien of NBC News aptly describes this phenomenon with a recent story headline: “Independents’ day: Romney looks to swing voters for salvation.” For better or worse, swing is king and the swing voter strategy seems almost beyond question.

But the “swing voter” is, in many respects, a nebulous and elusive concept. Daron R. Shaw – author of a scholarly chapter on “Swing Voting and U.S. Presidential Elections” – suggests that we lack common definitions and metrics for measuring swing voters. As a result, the news media will “parrot the sometimes banal and often ill-defined demographic explanations offered by political consulting companies” (p. 75). The multitude of demographic groups (including college educated Whites, youth, Latinos, and married females) and particular geographic locales (7 states) that were discussed by panelists highlights this reality.

What exactly is a “swing voter”? Various terms and labels are thrown around – persuadable voters, preference changers, late deciders, undecided voters. The scholarly social science community offers some guidance on definitional concerns, albeit with a strong emphasis on the individual, and not group, as the unit of analysis. William G. Mayer, Professor of Political Science at Northeastern University, edited a particularly comprehensive volume on the nature and function of swing voters in American politics – especially presidential elections. Mayer defines a swing voter as “a voter who is not so solidly committed to one candidate or the other as to make all efforts as persuasion futile” (pg. 2). To operationalize this concept, Mayer refers to the feeling thermometer – which asks respondents to indicate their feelings towards candidate on a 0 to 100 scale – included on the American National Election Studies (ANES). A study of elections from 1972-2004 suggests that only those voters with pre-election feeling differences of less than 15 degrees are influenced by the campaign, thus making them swing voters. Swing voters are said to refuse to designate one particular party as the clear “embodiment of all virtue” (pg. 2). Using a method highly correlated with Mayer’s approach and included in the same volume, Jason E. Campbell concludes that precampaign swing voters have ranged from 18% to 30% of the electorate between 1956-2000. However, the importance of definition is clear when one considers that undecided voters rarely account for more than low single digits in most presidential election polls.

Sometimes swing voters are equated with political independents. However, Mayer suggests that independents – which may total 1/3 of the total electorate – cannot be assumed to be devoid of political party preferences. In fact, the majority of so-called independents do feel closer to a particular party and 25% of “pure independents” vote a straight party ticket. Additionally, a sizable number of self-declared partisans (19% of Democrats and 10% of Republicans, on average) break partisan allegiance when casting a presidential ballot. Another way in which academics, in particular, have tried to define swing voters is based on party fluidity. However, only 41% voters who switched party affiliation between elections were found to be persuadable during the general election campaign.

Another categorization for swing voters is those voters who do not name a preferred candidate when asked by pollsters. According to Mayer, undecided voters are “not only those who are literally undecided but also those who have some current vote intention but are weakly committed to that choice” (pg. 14). Mayer argues that this is the most appropriate marker of a “swing voter,” but it is extremely difficult to accurately and reliably identify this population as a result of survey design (especially question wording) and interview protocols. Additionally, undecided voters have been found to be “unusually fluid” (pg. 15).

Given the significant attention and resources (both by the media and campaigns themselves) dedicated to winning over swing voters, one would assume that there is strong evidence supporting swing voter impact on election outcomes. But James E. Campbell suggests otherwise in his chapter in Mayer’s book; “The notion that precampaign swing voters swing elections is a myth” (pg. 127). Analyzing ANES data, Campbell finds that only one successful candidate – Jimmy Carter in his 1976 race – relied upon precampaign swing voters to win the popular vote, and winning presidential candidates between 1976 – 2000 needed, on average, only 17% of the precampaign swing vote to achieve a popular vote majority. However, successful presidential contenders commonly do win the swing vote. In fact, Al Gore (2000) and George W. Bush (2004) are the only presidential candidates to top the popular vote count without winning over the majority of swing voters.

But much of discussion of swing voters – and campaign-related activity aimed at influencing the targeted population – emphasizes the demographic characteristics and group affiliations that are associated with undecided voters. Daron R. Shaw describes “cross-pressured” groups based on the “contrary ideological pulls of different components of their identities” (pg. 89). Various groups face this situation, including voters with college education, suburban and rural dwellers, and the infamous “soccer moms,” waitress moms” and “office park dads.” (pg. 89). Our panel discussed the cross-pressures facing Latinos, particularly how they are aligned with the GOP on social issues like abortion and gay marriage, but differ from party positions on the issue of the immigration reform. One panelist suggested that the Republican Party would have to dedicate significant energy in the next few years (especially if Romney loses) to the immigration issue, potentially via outreach by Marco Rubio or Jeb Bush. While some political pundits suggest that President Obama’s policies towards Israel may create a cross-pressure that swings Jews over to Romney’s side, Amitai Etzioni, a noted professor and social commentator, published an essay downplaying the importance of Israel in the political decision-making calculus of most American Jews. A similar discussion has taken place about the potential impact of Paul Ryan’s association with Medicare reform on the elderly vote.

So does the swing vote neatly align with identifiable groups? After analyzing characteristics of swing voters from 1968-1976 and 1996-2004, Shaw concludes that group identity had a stronger influence on swing voting in the earlier era, and many of the groups commonly targeted as swing voters – including Soccer moms, Latinos, and rural dwellers – do not actually behave as such. However, independents, men, waitress moms, and office park dads do have a slightly higher probability of being swing voters. Nevertheless, other individual-level psychological characteristics are far more important. Political independence positively influences swing voting, while individuals with greater political information have a lower probability of being a swing voter. Based on these findings, Shaw offers the following somber assessment: “Campaigns have limited resources, and the commitment necessary to reach independent, inattentive swing voters may be prohibitive” (pg. 98).

At least a few campaign observers have been spared the Kool-Aid. Rebecca Berg of The New York Times published a sober assessment of swing voters in this year’s presidential election. She concludes the following: “Myths about the behavior of these voters are pervasive and persistent: For example, that undecided voters break for the challenger as Election Day nears. (Data have shown this is often not the case.)” Even Paul Begala, a noted Democratic strategist, offers a reality check when it comes to campaign swing voter hysteria:

“Don’t expect this swing voter to move any time soon. She knows the election isn’t until November, and she’s not riveted to every gaffe and poll like political junkies are. She’ll watch clips of the conventions, snippets of the debates, and on Nov. 6 she’ll ask herself which candidate can fix our economy for middle-class families like hers. All in all, my guess is she would prefer if the candidates canceled their commercials and just gave her the $2,181.87.”

Findings from the Pew Research Center paint an interesting portrait of swing voters in this election, suggesting that they may not be as important to electoral fortunes as commonly thought. Using a definition of a swing voter as an individual who has either not decided on a candidate or may change their mind prior to the election, the Pew study reveals that the swing vote this cycle – 23% of the electorate – is 10 percentage points smaller than it was in 2008. However, this decline echoes a pattern associated with elections featuring the incumbent as a candidate.

But what about geography? Our presidential election is not a popularity contest with the winner merely amassing the most votes in aggregate. The Electoral College makes place considerations critically important. Whole states – including Iowa, Nevada, Ohio, Florida, Wisconsin, Colorado, Wisconsin, North Carolina, and Virginia – and even regions such as the Mountain West – assume “swing” status. The campaigns don’t care about all swing voters, only swing voters in swing places. Both Paul Begala and Rebecca Berg note that the majority of swing voters will receive little to no attention because they live in uncompetitive states like Texas or California. The Associated Press even manages to drill things down one more level, identifying 106 counties in the 9 swing states that are likely to determine the outcome of the election. Importantly, however, swing counties are defined based on the aggregate election results from prior cycles. Those counties that voted for Bush in 2004 but switched to Obama in 2008 made the cut. But, according to the AP story, there are few undecided voters in these locales and the counties share very little in common; “There is no single reason to explain why these counties seem to shift with the political wind. Their voters are far from monolithic, having little in common other than their voting patterns.” But, as a result of the perceived “swing” status, the AP reports that voters in the Tampa, Cincinnati, and Northern Virginia television viewing areas have been the target of 1/5 of all campaign spending.

Linda Killian offers detailed descriptions of state-specific swing voter groups in The Swing Vote: The Untapped Power of Independents (2012: St. Martin’s Press). New Hampshire is home to NPR Republicans, affluent and educated voters who are fiscally conservative, moderate to libertarian on social issues, and not strongly driven by religious concerns. The key swing voter in Colorado is defined as the Facebook Generation, twenty and thirty-somethings that never developed partisan ties. They are liberal on environmental and social issues, but distrustful of the government’s handling of the budget and economic issues. Virginia is home to the Starbucks Moms and Dads, middle and upper-middle-class suburban voters in the counties bordering Washington, DC. They are fiscally conservative, socially moderate, and ethnically diverse. They care about national security, education, and environmental issues, reflective of their general concern with the future world for their children. Rather than voting by party, these voters focus on individual candidates and particular issues of concern to them. Ohio has American First Democrats, blue-collar and middle-class voters with strong union ties that have been hit hard by recent economic challenges. They are described as being both populist and protectionist. They are more conservative on social issues than most Democrats, but they benefit from government programs and do not take a hard stance on government spending. Many voted for Ronald Reagan. They are said to “base their political decisions on whom they feel comfortable with and might like to have a beer with more than they do on which party could do them the most good” (pg. 162).

The two campaigns clearly believe that winning over swing voters in swing states is the key to victory, and they are making every effort to identify effective means of targeting these voters. A recent survey of “persuadable voters” in 5 key swing states conducted by Global Strategy Group and Public Opinion Strategies offers the most current, nuanced view of how the campaigns should – and likely are – targeting swing voters. The memorandum states that 49% of persuadable voters use the Internet for news, 58% use the Internet to obtain additional candidate-related information, and 62% trust information from the Internet (equal to print news and only 5% below broadcast news). The survey report authors offer the following advice to campaign insiders:

“As campaigns fight for persuadable voters’ attention in the weeks leading up to Election Day, this data suggests that the Internet is a key channel on which to reach them at a time when they are looking for information and are willing to listen. Persuadable voters are online. They are engaged and ready to listen. And they are looking for answers to the questions that will help them make up their minds come November 6.”

It may be that cost considerations are overstated, particularly if it is not necessary to pay for expensive television advertisements to reach the targeted groups. Additionally, this finding sheds some doubt on the idea that swing voters are unreachable or disengaged. But unlike television and print advertising, geography plays less of a role in the online world. Money may be saved, but will it be worth it? It may be harder to actually reach the target voters without a captive audience. But this survey does suggest that a swing voter outreach strategy rightly understood might be just the trick. In an era of highly competitive presidential elections, campaigns do not need to move mountains, just win over a few moms in a few places. Undecided moms, persuadable moms, late deciding moms, independent moms, Starbucks moms, soccer moms, waitress moms, or any other kind of moms will do.


Independents’ day: Romney looks to swing voters for salvation by Michael O’Brien, NBC News, October 31, 2012
“Swing Voting and U.S. Presidential Elections” by Daron R. Shaw in  The Swing Voter in American Politics by William G. Mayer 2007 Brookings Institutions Press, via Google Books
The Swing Voter in American Politics by William G. Mayer 2007 Brookings Institutions Press, via Google Books
Israel doesn’t swing Jewish voters by Amitai Etzioni, Special to CNN,  November 4, 2012
In Swing States, Obama Leads on Handling of Medicare by Michael Cooper and Allison Kopicki, The New York Times, November 1, 2012
Few Voters Are Truly Up for Grabs, Research Suggests by Rebecca Berg, The New York Times, August 16, 2012
Paul Begala on the Swing Voters Who Will Pick the President, The Daily Beast, July 16, 2012
23% – Percentage of Swing Voters Declines Compared to Four Years Ago, Rew Research Center for the People & The Press
America’s New Swing Region: Changing Politics and Demographics in the Mountain West, Brookings Institution Press, 2012
In Race to 270, It May Come Down to 106 Counties by Thomas Beaumont and Mike Schneider, Real Clear Politics, October 21, 2012
The Swing Vote: The Untapped Power of Independents by Linda Killian, St. Martin’s Press, 2012
Memo: RE: Persuadable Voters, Global Strategy Group, October 10, 2012

Bedrosian Center