Sunday, November 6, 2011

Some forecasts

The 2012 presidential election forecaster up at fivethirtyeight.org is great fun, and you could get lost in the numbers for a few hours there. Basically, Nate Silver has made assumptions about the Republican candidates' ideological positions and plugged those into a forecast model, along with growth in gross domestic product and Obama's approval ratings. Just for giggles, I put together the following graph based on the forecasting data used in that model. I have assumed Obama's approval rating remains at 43 percent.
The official forecast for economic growth is 2.7%, by the way, which would put Romney and Obama at a near dead-heat.

Now, here are some reasons not to take these numbers too seriously:
  • First, Obama's approval rating is not independent of economic growth. If the economy begins to grow at a quicker clip, his approval rating will probably rise into the 50s. It will likely drop into the 30s or worse if we experience an actual recession.
  • Second, even if economic growth stays right where it is today, Obama's approval rating is likely to change somewhat as a function of the campaign. Nearly all Democrats will come to approve of his performance, even if they may have reservations today. And nearly all Republicans will come to disapprove, although they're probably doing that already.
  • Third, perceptions of the Republican nominee's ideological stances may well change by next year. It's very hard to make realistic projections of Cain's governing ideology since he's never governed before. Perry would be facing a more liberal electorate than he's ever faced, and Romney would be facing a more conservative one. Plus, given Romney's history, there should be substantially large error bars on either side of his line.
Update: Graph label fixed.

6 comments:

Matt Glassman said...

Seth ---

Nice post. One comment:

I think Nate's model is unconcerned with Obama's approval rating going forward. The 43% is an estimate of the approval rating one year out; adjusting that bar is for if you have a different estimation of the 1-year-out approval.

I don't think there's a claim about the approval remaining static, so I *think* the model is supposedly robust to future changes in approval rating drawn from economic changes.

I think.

matt

prison rodeo said...

Sorry to be pedantic, but that's not a scatterplot. #dontbaitmewithtwitter

Seth said...

It's a scatterplot with lines connecting data points that have been rendered invisible. #youwouldntlikemewhenimangry

alex said...

How certain is the link between a candidate's ideology and performance?

Could it be due to selection effects - say, a relatively moderate candidate has to be stellar in other respects to convince his party (most of which is more extreme than he is) to give him the nomination, and this leads him to perform better in the general election?

metrichead said...

Seth, why does Huntsman have the most favorable probability of all GOP candidates in the graph if he's barely even a third-tier candidate in the race now?

Seth said...

Metrichead, Silver's data are based on three criteria: economic performance, Obama's job approval, and some measure of Republican candidate ideology. It doesn't include any measure of viability or likelihood of getting nominated or campaign skills or anything else like that. What appears to make Huntsman so electable in this analysis is simply that, by all observations, Huntsman appears to be the most moderate of the Republican candidates. That probably gives him an advantage in the general election over the other Republican contenders.

By the same token, of course, being so moderate makes him almost un-nominatable.