Predicting Winner in the Next US Elections

Who Will Win?

Can we make predictions of the next elections? The answer is probably. Ray Fair (Yale University Professor) studied all previous US elections and arrived at the following model for predicting elections:

V = 0.468 –0.34 I + 0.047 I.d + 0.0065 g3 . I – 0.0083 p15 . I. (1 – d) + 0.0099 n . I. (1 – d) + 0.052 depr – 0.024 dur

Variables are defined as follow:

V = democratic share of the two party vote
I = 1 if democrats are in white house, -1 if republicans are
depr = 1 if president himself is running and is democrat, -1 if republican and 0 otherwise
p15 = average inflation over the past 15 quarters
g3 = growth rates of per capita real GDP in the past 3 quarters before the elections (i.e., corrected for US population growth)
n = number of quarters of the first 15 quartes of each period of a presidential administration in which growth rate is greater than 2.9%
DUR = 0 if the incumbent party has been in power for one or two consecutive terms, 1 if the democrats have been in power for 3 terms and –1 if republicans.
d = 1 if it is a war year and 0 otherwise

The equation has a correlation coefficienct of 96%and has a standard error or 1.9%, which means it is quite accurate.

Let us now try to apply it for the 2008 elections.

I = -1 since republicans are in white house
G3 = ¾ * (average growth rate over past 3 quarters % – 1.0) = 0.37, where 1.0% is the approximate annual growth rate of population in the United States, and correction by ¾ because we only need growth over 3 quarters and not annualized.
(obtained from http://www.economagic.com/em-cgi/data.exe/var/rgdp-qtrchg and assuming 3rd quarter to be average of previous 3 quarters)

n = 2-3, there are only 2-3 quarters in the past 15 quarters with good news.

Finally, inflation data are obtained from: http://data.bls.gov/PDQ/servlet/SurveyOutputServlet?data_tool=latest_numbers&series_id=WPSSOP3000&output_view=pct_1mth


p15 = approximately = 12.5%


The results are telling us that in case d = 1, democrats get 45%. In case d = 0, democrats get 58%. Therefore, if US public perceives they are in time of war, democrats lose and in case they focus on economy, then democrats will win in a landslide.

The table above tells us that all the economic factors combined can be neutralized by the war effect. It by itself has about 14% of leverage for the Republicans.

Sensitivity analysis shows that if economy turns real bad, the democrats can win up to 63.8% assuming the US population discounts the war effect. If the economy becomes great, the democrats will still win 52% still assuming the war effect is discounted. If the war effect is not discounted, then democrats get only 44% assuming economy turns better or get 46% assuming the economy turns worse.

The only contour for the Democrats to win is to neutralize the war effect and amplify the economic issues. Otherwise, they barely win. The reason the 52% contour was selected to demarcate a democratic win is because it represents 50% + margin of error. So, at 52% democrats win barely.

Conclusions:

1- According to current predictions, Democrats should win about 52% of popular vote on average, with a standard error of about 1.9%. Therefore, democrats can win by a narrow margin or it will be in a reality a toss-up given the margin of error of about 2%. Therefore on average both parties have have to adopt a very local strategy to capture the electoral college vote.

2- If the economy continues to turn even more south, democrats can win by a landslide up to 62%.

3- Portraying the economy in better shape doesn’t help the Republicans much. It gives them only a minor advantage and makes it still more local game and still becomes a contested election. Also, it will be very hard for republicans to turn the economy or even portray it in a positive light. This explains why the republican strategy in the convention ignored the economy completely. It just is not a good lever for them at all. They skirmished around it with things like America has largest and most prosperous economy in the world. Talk of better economy just manages the disaster for them so instead of a landslide it becomes a narrow loss or bring it back to being contested.

4- The biggest lever republicans have is the war. If they successfully focus on it they can bring significant advantage and neutralize the democrats. The strategy becomes simple for republicans: bring someone attractive (Sara Palin) to energize the message and hammer home that it is country in war with slogans like: “country first”. The strategy is working so far as evident by the jump in the polls. McCain can talk economy for ever and it won't really make a difference.

5- Democrats strategy has to focus on the poor economy and how they will fix it. This moves them to have a landslide. There is no easy strategy for them on war simply because Americans like victory and anytime republicans will show they are better positioned for Victory they can win (although narrowly). It will not matter who brought us to this position. The party that will show it can bring victory with no surrender will win. This is why “the economy stupid” is better strategy for democrats and then can say that a better economy will enable the US to win abroad. Democrats have to level with Americans on this point and clearly indicate they want to win and will find the best strategy to win and that past Pickering was just that pickering, and that republicans made mistakes but we are where we are and have to devise a strategy to win and not cut and lose.