Jay DeSart: forecasting the outcome

Reading Time: 3 minutes The months before a presidential election are anxiety packed for most Americans. It is a time of great change. People stand on edge, and slander fly as the November deadline approaches. But what if you could know with some certainty which candidate was going to win before the actual vote? That is a prediction, and probable fact, that one Utah Valley University professor has granted the public.

Reading Time: 3 minutes

The months before a presidential election are anxiety packed for most Americans. It is a time of great change. People stand on edge, and slander fly as the November deadline approaches. But what if you could know with some certainty which candidate was going to win before the actual vote? That is a prediction, and probable fact, that one Utah Valley University professor has granted the public. Through the DeSart and Holbrook Model, Jay A. DeSart, a political science professor, claims that Barack Obama will come out victorious by 3.6 percent. DeSart is our hero for the week. Here is what he had to say.

Q: When was this model first created?

A: Our forecast model was first published in the International Journal of Forecasting in 1999. Since that time we’ve refined it and published an update to the model in 2003 in Social Science Quarterly. The model that we used to generate this year’s forecast was the product of a further refinement that we presented in a paper at the Western Political Science Association conference this past March.

Q: What was your inspiration and/or reasoning for making this forecast model?

A: It came from research that we were doing several years ago on the accuracy of statewide election polls. We’d found that those polls taken during the month of September were a fairly good indicator of how the election would turn out in each of the states. Using them to generate a prediction of how the election would turn out was just a natural extension. Another reason we were interested in doing it was because while there are many forecast models that are used to predict the national popular vote, ours is one of the only models that also generates a prediction of the all-important Electoral College vote. We’re able to do that because we are looking at state data.

Q: What did you hope to accomplish through your election model?

A: Well, certainly one of the things we were attempting to accomplish is to come up with a really accurate forecast model. But our interest goes well beyond that. Much of the research in election studies over the past several decades has basically suggested that what happens during the course of the fall campaign has a minimal effect on the outcome of the election. That notion is reinforced by the fact that we can generally predict, usually with amazing accuracy, how the election will turn out a month or two in advance of the election.

Q: As a political science professor, how do you view the political system found in Utah (specifically UVU)?

A: Well, moving from Florida to Utah as I did in 2004, it was quite a shock to the system. I’d moved from one of the most electorally competitive and important states in U.S. Presidential elections, I lived in Palm Beach County during the 2000 election mess, to quite arguably the least competitive state in the country. As a result, I don’t get to see as much of the campaigns here as I did in Florida.

That being said, I’ve been impressed by the level of interest and involvement that I’ve seen among the student body at UVU, even in this very electorally uncompetitive environment.

Q: Do you think that students, faculty and staff are sufficiently educated?

A: I think that there are plenty of opportunities for students, faculty and staff to become educated about the candidates and issues. In my four years here I’ve been very pleased to see the large number and wide range of events that UVU has put on. The challenge here, as it is everywhere in the country, is to get more people interested and involved. One of the big challenges for me is to get my students to see that politics and government are not the remote things that have very little to do with their everyday lives and to show them that the people we elect are making decisions that ultimately impact virtually every aspect of their lives.

Q: What should be done to change or improve the political system here on campus or in the community?

A: That’s the $64,000 question. I think what we need to do is demonstrate to the community that a vibrant democracy is dependent upon having a diversity of viewpoints being allowed to be heard. In that respect, as heated and difficult as the Michael Moore controversy was a few years ago, I think it went a long way to opening up a dialog, among students and faculty but also between UVU and the community at large, about such issues as the importance of political tolerance and the freedom of speech. I can honestly say that in my 17 years of teaching American Government, I’d never had a more lively, interesting and educational discussion of the First Amendment in my classes than those that I had in the fall of 2004 when Michael Moore came to campus.