We are in a holding pattern as we await polls that fully reflect the effects of the Republican National Convention.
But from the standpoint of our forecast model, the most important points on Thursday were economic rather than rhetorical. Two of the seven data series that make up the model’s economic index were updated on Thursday, and they both had somewhat better news about the condition of the economy.
One of these is real personal income, versions of which are common in election forecasting models. The income numbers were extremely poor throughout 2011 and have been one of the better reasons to bet on President Obama being defeated.
However, the income figures have picked up so far in 2012, and that trend continued when the government released its July estimates of income growth on Thursday. Income grew at an annualized rate of 3.7 percent in July, according to the government’s initial estimates. This closely matches the pattern throughout the first seven months of 2012; personal income has grown at an annualized rate of 4.2 percent since January, adjusted for inflation.
The other measure is personal consumption expenditures, which make up about 70 percent of gross domestic product. The growth in consumption has been extremely sluggish over the past year — perhaps reflecting poor consumer confidence. But the July numbers were considerably better, with consumption growing at an annualized rate of 5 percent.
These measures do not receive as much attention as job growth or gross domestic product, but they are broad-based and reflect some of the many ways that voters experience and participate in the economy. They point toward what could be a stronger second half of 2012 for the economy, after a discouraging second quarter during which G.D.P. grew by just 1.7 percent.
As a result of the new economic data, Mr. Obama’s chances of winning the Electoral College improved slightly in Thursday’s forecast, to 71.6 percent from 70.4 percent on Wednesday.
Already a Convention Bounce?
By contrast, Mr. Obama’s standing declined in our “now-cast” on Thursday — our estimate of what would happen if the election were held today. The “now-cast” looks only at polls, ignoring any economic data.
In addition, the “now-cast” does not adjust for the fact that polls usually overrate the standing of a candidate in the midst of a “convention bounce.” (The Nov. 6 forecast builds in an adjustment for these effects.)
No polls have yet been released that reflect the public’s reaction to Mitt Romney’s acceptance speech on Thursday, but some sought to gauge reaction to the first two days of the Republican convention.
Mr. Romney has received encouraging news from the polling firm Ipsos, which has run an online tracking poll during the convention. On Thursday, Mr. Romney led by two percentage points among likely voters in that poll, a six-point improvement from Ipsos’s preconvention benchmark, when Mr. Romney trailed Mr. Obama by four points.
The caveat is that online surveys may suffer from more selection bias than telephone polls do, perhaps reflecting the views of voters who are more politically engaged (and therefore more likely to watch the conventions) than average. In addition, Mr. Romney has shown less improvement among registered voters in the poll, with whom he still trailed Mr. Obama by three percentage points. Polls of likely voters can sometimes be more sensitive to the effects of a party convention, since one of the purposes that conventions serve is to make political partisans more enthusiastic.
Mr. Romney has also shown some improvement in the Rasmussen Reports tracking poll, which reverted to a tie on Thursday after having shown Mr. Obama ahead by two percentage points in its last poll before the conventions. Mr. Romney’s position has receded slightly, however, in the Gallup tracking poll — although Gallup reports its results over a seven-day window, so very few of its interviews yet reflect voters who watched any of the convention.
By Monday, we should have some richer data on the magnitude of Mr. Romney’s convention bounce. The forecast model will treat a net gain of more than four percentage points as a bullish sign for Mr. Romney and a gain of fewer than four points as a bearish one.