http://fairmodel.econ.yale.edu/vote2004/index2.htm

In essence, using economic events, they have produced a linear model describing voting in a two party system. Their results in the past have been very interesting using this model. It has fit the elections since 1916 fairly well: http://fairmodel.econ.yale.edu/vote/vot0101.htm

As with any model, of course, there are a lot of potential problems and this one has the very real potential of generating aphysical results due to fitting (and I haven't done an analysis on their residuals), but this is still very interesting stuff.

Anyways, barring the chaotic factor of the system (or depending on them to cancel out), Bush is predicted to win by a small landslide thanks to the economy.

This all seems very interesting, but I find it hard to believe that it is solely economic factors that influence the outcome of an election. Where is the party platform and the candidate's stand on issues factored into these equations? I think that they make more of a difference than Yale seems to think they do. I'm no mathematical wizard, but another thing that bothers me is, by knowing the past outcomes of elections and the related statistics, he could have made an equation to simply fit the known results. I realize that this is how data prediction is done, by analyzing past results, but if you are doing it to fit your agenda, that is the wrong way to do it. Looking at some of the other pages on the site, the original idea was drawn up in 1978, meaning that the equation has actually only predicted the outcome of seven elections. I don't know the probability of that being coincidence or not. But overall, it's a very interesting concept and I'll be checking back to see how the model does in the 2004 election.

I am pretty sure that his results are beyond statistical chance, because if they weren't he probably wouldn't have gotten published.

I find it hard to believe that it is solely economic factors that influence the outcome of an election.

VOTE = Incumbent share of the two-party presidential vote.
PARTY = 1 if there is a Democratic incumbent at the time of the election and -1 if there is a Republican incumbent.
PERSON = 1 if the incumbent is running for election and 0 otherwise.
DURATION = 0 if the incumbent party has been in power for one term, 1 if the incumbent party has been in power for two consecutive terms, 1.25 if the incumbent party has been in power for three consecutive terms, 1.50 for four consecutive terms, and so on.
WAR = 1 for the elections of 1920, 1944, and 1948 and 0 otherwise.
GROWTH = growth rate of real per capita GDP in the first three quarters of the election year (annual rate).
INFLATION = absolute value of the growth rate of the GDP deflator in the first 15 quarters of the administration (annual rate) except for 1920, 1944, and 1948, where the values are zero.
GOODNEWS = number of quarters in the first 15 quarters of the administration in which the growth rate of real per capita GDP is greater than 3.2 percent at an annual rate except for 1920, 1944, and 1948, where the values are zero.
The equation has VOTE on the left hand side and the other variables plus a constant term on the right hand side. It is linear in coefficients. The sample period begins with the 1916 election. The equation is estimated by ordinary least squares.

In other words, they use many other variables besides economic ones. When you compute it yourself, though, you are only prompted for economic variables because they are the only variables that change within an election cycle. The others are pre-programmed in.

if you are doing it to fit your agenda, that is the wrong way to do it

I agree, and so does Mr. Fair.

Finally, a point about me as a social scientist trying to explain behavior versus me as a citizen. As a social scientist I am trying to do the best I can explaining the percentage share of the two-party vote. In this capacity I don't care who wins or loses, but how close the equation comes to explaining the actual share [of the vote]...Also, I shouldn't let my political views affect my scientific work. This is an exercise in trying to explain behavior, not shape it. As a citizen, however, I obviously care who wins or loses.

I don't think Fair is even a Republican (although I couldn't be sure). I think he is just calling the facts like they are, trying to predict the results of the election. Remember, last time he predicted that Gore would win the popular vote, and he did. This time he's predicting that Bush will win the popular vote. So how would this promote a political agenda?

I think this series of papers and equations is very interesting. Whether Bush wins or not, it will be neat to see if the predictions pan out. Until November (or Dec. or Jan. if we have to go through the recount process again), I guess we'll just have to hold our breaths and wait and see, and congratulate Fair if he's right.

<3 Mary

PS - Shouldn't the WAR variable be a 1 value during this election too, or does it not count because we're technically not at war anymore?

This all seems very interesting, but I find it hard to believe that it is solely economic factors that influence the outcome of an election.


SangReal posted the factors that make up the model, and even those aren't all of the factors involved--just the ones that are being used to calculate the vote share.

These account for some percentage of the variance in the system. It is not 100% (all models are wrong), but according to these results (assuming that they are not aphysical, a very real risk--I also haven't checked to see if the error is normally distributed) a very large percentage of the variance is made up by these factors (some models are useful).


Where is the party platform and the candidate's stand on issues factored into these equations?


"It's the economy, stupid."

The fact that they aren't factored in yet such a nice fit is reached should indicate that such variables (generally) make up a very low percentage of the variance in comparison to the economic variables, which make up a very high percentage of the variance.

This year I think a higher percentage of the variance from other factors will matter more than they have in the past, and so the Bush estimate right (0.58-0.60) now is likely high. That said, if economic factors account for large percentage of the variance (which they seem to) then such other factors either tend to cancel out or likely won't be able to overcome the huge difference between the two.



I'm no mathematical wizard, but another thing that bothers me is, by knowing the past outcomes of elections and the related statistics, he could have made an equation to simply fit the known results. I realize that this is how data prediction is done, by analyzing past results, but if you are doing it to fit your agenda, that is the wrong way to do it.


What agenda?


Looking at some of the other pages on the site, the original idea was drawn up in 1978, meaning that the equation has actually only predicted the outcome of seven elections.


It has also been revised multiple times since then. It predicted the last election to within a half of a percent.

SangReal posted the factors that make up the model, and even those aren't all of the factors involved--just the ones that are being used to calculate the vote share.
Yeah, I see that now. It's just so heavily based off of the economy that it didn't seem right at first.
The fact that they aren't factored in yet such a nice fit is reached should indicate that such variables (generally) make up a very low percentage of the variance in comparison to the economic variables, which make up a very high percentage of the variance.
See, I'm not good with the math action. Looking at the formula that should have been obvious.
This year I think a higher percentage of the variance from other factors will matter more than they have in the past, and so the Bush estimate right (0.58-0.60) now is likely high. That said, if economic factors account for large percentage of the variance (which they seem to) then such other factors either tend to cancel out or likely won't be able to overcome the huge difference between the two.
I agree with SangReal here, in that the war should be a one point factor in the equation, simply because it's such a hot button issue. It will be interesting to see how the formula adapts and how close the prediction is without the war being factored in.
What agenda?
As far as I know, there isn't one. It's the possibility of bias that bothers me. For all I know though, he's completely objective. I've never met him in person so I can't claim to know one way or the other.

In conclusion, I've been pwned. I'm out of my league here.:D