Our economic model, which looks at a wider array of economic indicators than the other ones do, is a bit more equivocal. Essentially, it suggests that Mr. Obama’s narrow lead in the polls is broadly consistent with the economy’s lukewarm performance.
There are two things to keep in mind when interpreting this evidence. First, Mr. Obama’s lead in the polls — about 2 points — would reflect a below-average performance for an incumbent. Since World War II, elected incumbent presidents have won by an average of 7.5 points when seeking a second term.
Second, the historical evidence seems quite definitive that the public reacts more to the change in economic performance — that is, the rate of growth in crucial economic indicators in the year or so leading up to the election — than to how strong the economy is in an absolute sense. Over the last 100 years, there has been literally no relationship between the unemployment rate on Election Day — a measure of how completely the economy is fulfilling its productive potential — and the election result.
But there have been positive and statistically significant relationships between the growth rate in the economy — as measured by G.D.P., jobs, income and other indicators — and the election result.
The recent rate of growth — about 2 percent across an average of indicators, which is below average but not recessionary — points to a below-average but not necessarily losing position for Mr. Obama. Mr. Obama may also be helped somewhat by the fairly low rate of inflation, which receives some weight in our model — and, we think, also, in voters’ perceptions about how the economy is performing. It’s a little easier to adjust to a new but mediocre “normal” if grocery prices aren’t rising every time you go to the supermarket.