I just read a John Sides post on Ten Miles Square about the impact of the first presidential debate on this fall’s national election, and posted this response:
The first debate did move the polls, but they were bound to tighten up anyway when the Romney campaign reached full swing.
The Pollster polling average is misleading about the impact of the first debate. The Pollster average curve is smoothed in a way that makes it look like a trend was developing several days before a poll-changing event. You can get around that by looking separately at periods before and after an important event, such as the first debate.
If you go here you see a steady four-point Obama lead in the week leading up to the first debate. Then, the week after it is a nearly flat, nearly tied race, with Romney up one by 10 Oct. But Romney never had a better week, so even that seeming trend is questionable. By late October Obama was edging ahead, even before Sandy.
My conclusion is that the event of the first debate did produce a Romney bump, and he kept most of that gain but not enough to win. So the debate mattered.
But I think the mere fact of the debate mattered more than whether Romney was seen as winning the debate. It was a couple hours of free TV for the candidate who was less familiar to the public and had been painted unfavorably in news media and advertising. Just getting on TV and making a good personal impression was all he had to do. I don’t believe Obama could have stopped that from happening.
In the later debates, though, there was nothing more for Romney to gain. His human face and rhetorical skills had already got out there. They were always going to get out there eventually, even if there weren’t any debates. But the first debate was a big opportunity for him to get them out there all at once, for free. He did that and tightened up the race, but that was all he could do, and it wasn’t enough.