The conventional wisdom about Barack Obama’s path to reelection is that, though the president is unpopular, he will run a strongly negative campaign against the GOP nominee – tarring him as a radical or (in the case of Mitt Romney) an unprincipled flip-flopper. Thus, voters who might not be happy with him will nevertheless support his reelection because it’s better to back the devil you know than the devil you don’t know.

The template for this is supposedly the George W. Bush 2004 campaign, which is ironic because Democrats howled in protest over the allegedly extreme negativity of that contest. Regardless, Obama hopes to do what Bush did to John Kerry: attack him so relentlessly that people can’t support him.

What to make of this approach? Put simply: It is a weak political strategy with little historical evidence to validate it.

To start, we have to clear away the confusion generated by the conventional wisdom. And let’s begin with this simple chart, with data pulled from the 2004 exit poll.

This should really dismiss the notion that Bush won because he made Kerry unelectable. The election that year was a referendum on Bush: people who disapproved of him voted overwhelmingly for Kerry; people who approved of him voted overwhelmingly for Bush. In fact, the Bush approvers/Kerry voters were more numerous than the Bush disapprovers/Bush voters.

This pattern was repeated across the 50 states. Consider the following chart.

The takeaway point from this is that the strategy Obama plans to run in 2012 scored Bush exactly zero electoral votes in 2004. In most states, we see the same pattern as we did nationwide: Bush won between 5 and 10 percent of people who disapprove of him, far too small a portion to swing any states his way.

If anything, Kerry did a better job at peeling away voters from the “other” side than Bush did. Bush’s job approval among 2004 voters was 53 percent, yet he won just short of 51 percent of the vote – because about 10 percent of all Bush supporters chose to back Kerry. These Bush approvers/Kerry voters cost Bush four states.

This trend in 2004 is not particularly unique. The following data is from the American National Elections Study, which has tracked presidential job approval among voters since 1972.

Notice that the only two years where we see a substantial amount of support for the incumbent president among disapprovers are 1972 and 1980 – but neither election should be of much solace to Obama.

In 1972 the Democrats nominated George McGovern, and accordingly Richard Nixon picked up a substantial amount of support from Democrats who were alienated by his candidacy (the 1972 platform was arguably the most radical platform ever adopted by a major political party). That’s unlikely to happen this time around; Obama will not win much more than 7 or 8 percent of Republicans.

As for 1980, Jimmy Carter’s approval among Democrats that year was very tepid; the pre-election Gallup poll had him at just 61 percent among his own partisans. A good portion of those disapproving Democrats held their noses and voted for Carter, anyway. Again, this is not a good comparison to Obama, as it is unlikely that he’ll win anything less than 90 percent of Democratic voters.

So, can Obama win by running against the GOP nominee? I'd say no -- at least not if can't bring something positive to the table. If Obama is unpopular on Election Day, it will take a miracle (e.g. the GOP nominating an extraordinarily weak candidate) for him to win. History suggests that, if anything, the president is more likely to lose approvers than win disapprovers.

And note here that we are talking about the voters, who historically tend not to be as pro-Democratic, and thus not as partial to Obama, as the entire adult population. In fact, the exit poll in 2004, which had Bush's net approval at +7, showed him in better standing than almost all of the major media polls taken just prior to the election.

I recognize that the Obama campaign team needs to put out some sort of message about how it wins next year. I also recognize that the media is duty-bound to report on it, relatively uncritically. However, this is just not a viable strategy, at least not without a significant uptick in the president's own standing. If he is still as unpopular as he has basically been for the last sixteen months, he can’t win by running against his opposition. Barring the possibility that the GOP nominates a 21st century version of George McGovern, Obama wins if and only if a majority of voters think he’s doing a good job.

Next Page