The Effect of Romney’s Nomination on President Obama’s Chances for Re-election: Some Evidence From Intrade
Awhile ago, I used data to from Intrade to show that the there is a strong negative relationship between the probability of a recession in 2012 and the probability that President Obama is re-elected. In this post, I ask the question what is the relationship between the probability that President Obama is re-elected and the probability that Romney is nominated controlling for economy. If this relationship between the Obama and Romney probabilities is negative and statistically significant than this suggests that Romney is perceived by the market to be a strong candidate relative to other challengers.
First a graph of the overall data:
The three data series are (1) President Obama’s re-election chances in red with a dashed trendline, (2) Romney’s chances of getting the nomination in blue with a solid treadline, and (3) the chances that the economy goes into recession in 2012 in green with a dotted trendline. The data are intrade closing prices from Jan. 1, 2011 – Jan. 5, 2012.
Just from casual observation, two things are noticeable. First, the President’s chances of re-election have decreased over the year as the probability of Romney’s nomination has increased and the probability of a recession in 2012 had increased.
To parse this relationship more rigorously, I ran a regression of the natural log of the President’s re-election chances on the natural log of Romney’s nomination chances and the natural log of the economy going into recession. Using the natural log allows the regression coefficients to be interpreted as percentages. These are the results:
Linear regression Number of obs = 370
F( 2, 367) = 1074.45
Prob > F = 0.0000
R-squared = 0.8355
Root MSE = .03819
lno | Coef. Std. Err. t P>|t| [95% Conf. Interval]
lne | -.222992 .0082341 -27.08 0.000 -.2391839 -.2068001
lnr | -.0385798 .0063549 -6.07 0.000 -.0510763 -.0260833
_cons | 4.890662 .0194852 250.99 0.000 4.852345 4.928978
Both coefficients are negative and highly statistically significant, indicating that an increase in the probability of a recession and an increase in the probability of Romney’s nomination decrease the probability that President Obama is re-elected. Specifically a 1% increase in the probability that the economy goes into recession results in a -0.22% decrease in the probability President Obama is re-elected and a 1% increase in the probability that Romney is nominated results in a -0.04% President Obama is re-elected.
Now what’s clear here is that although both effects are statistically significant, the Romney effect is quite small. This suggests that the economy plays a much larger role in the chances of President Obama’s re-election than the choice of the Republican candidate. On the other hand, since increases for Romney come at the expense of the other Republican candidates it also suggests that Romney is a stronger choice for the nomination than other Republican candidates in aggregate.