Skip to main content
ABC News
Sept. 5: Looking Ahead After Charlotte

There was almost nothing worth talking about in the polls that were out on Wednesday. So let’s take a quick look forward instead — to about a week from now, by which point we will have plenty of data on what effect the Democratic National Convention had on the polls.

Measuring the bounce the incumbent party gets out of its convention is trickier than doing so for the challenger, which holds its convention first. That is because in a year like this one, when the conventions are so closely spaced, the polls will not have had any time to return to equilibrium after the challenger’s convention bounce.

The assumption that our forecast model makes, however, is that the incumbent’s polling is likely to be just slightly inflated — by a percentage point or two — after his convention.

Since the polls before the Republican convention had roughly a two percentage point lead for Barack Obama, what that means is that the forecast model will want to see Mr. Obama ahead by three or four points in the national polls next week to give him a score of par. If that is what we see in the polls, the forecast model’s conclusion will be that Mr. Obama emerged from the conventions in roughly as strong a position as he held before they began.

But par would be a decent score for Mr. Obama, for the simple reason that he was ahead. To belabor the golf analogy, we went into the conventions with Mr. Obama holding the equivalent of a one-stroke lead with four holes left to play (perhaps these holes represent the debates, the closing arguments, the major economic reports to come between now and November, and so forth). The closer you get to the end of the course, the more valuable it becomes simply to deny your opponent an opportunity to catch up.

There are other possibilities as well, of course.

If Mr. Obama holds a 5- or 6-percentage-point lead in the polls by the middle of next week, we would have to call that a birdie. That would be Mr. Obama’s strongest standing in the polls since February or March, when Mitt Romney’s numbers with the general electorate were suffering because of his struggle to win the Republican primaries. It would also be similar to where George W. Bush exited his convention in 2004, a race that parallels this one in many respects — and which is a favorable precedent for Mr. Obama.

If Mr. Obama’s lead is in the high single digits by next week, then he gets an eagle. I certainly do not expect this to happen, since the polls have been hard to move for any reason this year. But if Mr. Romney gets very little bounce and Mr. Obama gets a rather conspicuous one, it is simply going to look like Mr. Obama made the more persuasive argument to a majority of American voters. The election would not be over, but we might be at the point where Mr. Obama would have to make a substantial error in order to lose.

On the other hand, suppose Mr. Obama leads by only one to two percentage points in polls next week. That is not a total disaster for him. But remember that the model’s hypothesis is that Mr. Obama’s post-convention polls are likely to inflate his numbers by a point or two. In other words, it would treat these polls as showing the equivalent of a tied race, meaning Mr. Obama was slightly worse off than before the conventions. We would call that a bogey, with Mr. Obama having squandered his one-stroke lead.

Finally, if the polls after the Democratic convention actually show Mr. Romney ahead, we would score that as a double-bogey for Mr. Obama. Like the eagle scenario, this one seems unlikely — it might require both an extremely flat speech from Mr. Obama on Thursday night, and then a poor jobs report on Friday morning.

But it obviously would not be good news for Mr. Obama if it did occur. Although no challenger in modern times has won the election without leading in the polls after his convention, the same is also true of incumbents.

Nate Silver founded and was the editor in chief of FiveThirtyEight.