The question after last week’s presidential debate was not who had won — there was a clear consensus that Mitt Romney had got the better of Barack Obama — but how much difference it would make to the race. Going into the debate on Wednesday night, Obama was the clear favourite to win re-election, with a five-point lead in the polls. Nate Silver’s Fivethirtyeight forecast gave him an 86 per cent chance of victory. But, thanks to his strong debate performance and the President’s uncharacteristically weak one, Romney has narrowed that gap.

Different polls paint different pictures of exactly what effect the debate had on the race, though most show it moving towards Romney initially before going back slightly in Obama’s direction. Gallup’s tracker found Romney pulling even with Obama at 47-47 in the three days after the debate, from a five-point deficit in the three days leading up to it. Their daily updates are a seven-day average, and so the latest (showing a five-point Obama lead) still includes pre-debate responses, but it does tell us that Obama did better in the polling conducted on the Sunday after the debate than he had on the Sunday before it. Rasmussen’s three-day tracker had shown a 2-point lead for Obama just before the debate, which became a 2-point lead for Romney afterwards — but it now shows the two candidates level-pegging at 48-48. Ipsos had shown Obama up by around five-points before the debate, but its post-debate poll shows Obama’s lead cut to two points. RAND (which like Gallup publishes a seven-day average) has Obama’s lead narrowing only slightly, from five or six points pre-debate to four points now.

But the best number for Romney came yesterday, in a Pew Research poll that showed him four points ahead of Obama. It’s a particularly dramatic result because Pew has tended to produce some of the most Obama-friendly numbers and because it represents a 12-point swing towards Romney from their previous poll last month (which showed him eight points behind). And that was followed today by a poll from another pollster that usually produces relatively Obama-friendly numbers (this time Public Policy Polling) showing Romney leading by two points, whereas they had him trailing by four last week. On the other hand, some of the swing state polls, such as two from Rasmussen in Colorado and Iowa, actually show Obama gaining since last month.

Overall, Silver’s model — which aggregates all the national and state polls — suggests that Romney has roughly halved Obama’s lead from five points to 2.4 points (although it has yet to process today’s polls, such as the one from PPP). It’ll take a while, and a few more polls, to tell exactly what the race looks like in the aftermath of the debate. The picture has already been further complicated by Friday’s jobs figures, which showed the unemployment rate dropping to 7.8 per cent, providing Obama with some good news and possibly helping him recover slightly (as Gallup and Rasmussen seem to suggest he has). And it may soon be mixed up further by Thursday’s vice-presidential debate between Joe Biden and Paul Ryan, and next week’s ‘town hall’ debate between Obama and Romney.

But with just four weeks to go before election day, Obama remains the clear favourite — albeit not as strong a favourite as he was last week. The economic fundamentals favour him, and his strong approval and favourability ratings do not seem to have been hit by his poor debate performance. Silver’s model now rates his chances at 74.8 per cent — not as good as last week, but still better than at any time before the conventions.

Tags: Barack Obama, Mitt Romney, Polls, presidential debates, US election