In the horserace polling of the presidential race between President Barack Obama and challenger Mitt Romney, both candidates are essentially tied and neither is close to getting over the 50% finish line to victory. Thus, it is long past time we started paying attention to the undecided voters who will choose the winner in November.
Historically, when incumbent presidents run for reelection, they generally only get the ballots of voters who openly approve of them. 9 of 10 undecided usually break to the challenger because they really do not approve of the incumbent. If you ask your wife whether she approves of your marriage and she responds that she is undecided, start shopping for a divorce attorney.
The handful of polls – Rasmussen (lean GOP), Public Policy Polling (lean Dem) – which asked the undecided whether they approve of Mr. Obama found that these voters heavily disapprove of the President. However, PPP found that these voters also do not approve of Mr. Romney, which suggests that the undecided are not automatically breaking toward Romney as they have to past challengers.
Assuming that the undecided currently hold a pox on both of your houses opinion of the candidates, we need to identify who these voters might be in order to determine for whom they might begrudgingly cast a ballot.
In their horserace poll, Mitt Romney held a narrow 48% to 47% lead over Barack Obama with 5% undecided. Those undecided are:
> In the past, a plurality (44%) voted straight GOP, while 32% voted Democrat and 24% split their votes between the parties.
> A heavy majority (63%) of the undecided self-identity as conservative, 12% as moderate and an interestingly high minority (25%) say they are liberals.
> At the time of the poll, a plurality (44%) disapproved of Mr. Obama’s job performance (32% strongly), while only 29% approved (7% strongly) and 25% were unsure. These results were a bit better for the president than the Rasmussen and PPP polls linked above.
> The reelect cross-tab indicated that 28% of the undecided say they will vote to replace Mr. Obama, 27% will consider someone else, 44% are unsure and only 1% say they will vote to reelect the president.
> 77% of the undecided are white, which is just a bit larger than the white share of likely voters. Interestingly, 14% of the undecided are Hispanics, a much larger amount than their share of the electorate.
> The undecided are disproportionately male (55%), seven percent more than their usual share of likely voters.
> The undecided are also a church going group, 49% attending every week and 16% attending monthly. 41% are Born Again/Evangelical
In sum, the undecided appear to be predominantly white, male, conservative, Republican and Independent, married and church going folks who disapprove of and are at least considering firing the president.
However, there are two interesting subgroups of undecided who at least supported Mr. Obama in the past – the 25% who self identify liberal and the 14% who self identify hispanic. The Battleground cross-tabs do not go deep enough to allow us to determine to what extent these two groups overlap.
I would expect the liberals to go home to Obama in November. They will not vote GOP.
However, the Hispanics are a very interesting group. They are culturally conservative and have suffered double digit unemployment under Obama. Hispanic men disapprove of Obama to about the same extent as white women. Romney might have an opportunity to make some inroads with Hispanics, if they do not decide to stay home as they did in 2010.
Of course, the Battleground poll is just one snapshot in time, but it appears that Romney can expect the lion’s share of undecided to break for him in November.