Predicting the Election

Voting Machine (Aaron Webb)

Late last week, I contacted the “Political Science Young Guns,” a self-described nickname for three young professors in the political science department at Baylor.

With poll numbers and predictions in every news article, it can be hard to sort through the information overload. In order to provide the Baylor community with a short way to understand the different possible outcomes, I assigned each professor a specific scenario. Each professor had 250 words to “make the best case possible” that their scenario would in fact occur. Dr. Flavin will be arguing the factors indicating an Obama victory, Dr. Nichols a Romney victory, and Dr. Bridge a tie in the electoral college.

If you have not voted yet, I encourage you to do so today. Enjoy the summary of tonight’s possibilities!

Dr. Pat Flavin was tasked with making the best case possible for an Obama victory:

“Since World War II, there is a remarkable correlation between economic growth during a president’s first term and his chances of winning reelection.  It makes sense, then, why Mr. Romney’s campaign has sought to make the 2012 election a referendum on Mr. Obama’s performance during his first term by pointing to high unemployment and weak economic growth during the last four years.

Unfortunately for Mr. Romney, the current state of the American economy is not quite weak enough to doom Mr. Obama’s reelection chances – unemployment is just low enough and economic growth is just strong enough.  The election will be close, perhaps within one percentage point for the national popular vote, but the advantages of incumbency will again prevail and Mr. Obama will be reelected.

How will this happen?  Due to our unique procedure for electing the president, the race will essentially be decided by the outcome in eight “battleground” states: Colorado (9 electoral votes), Florida (29), Iowa (6), Nevada (6), New Hampshire (4), Ohio (18), Virginia (13), and Wisconsin (10).  By my count, Mr. Obama heads into Election Day with 237 “safe” electoral votes, while Mr. Romney has 206.  Based on my assessment of polling data, demographics, and campaign advertising and organization in these eight states, I predict Mr. Obama will win Iowa, Nevada, New Hampshire, Ohio, and Wisconsin, while Mr. Romney will win Colorado, Florida, and Virginia.  Adding these up, the final score in the Electoral College (270 needed to win) will be: Obama 281, Romney 257.”

Dr. Curt Nichols was tasked with making the best case possible for a Romney victory:

“The best argument that Mitt Romney will win the election is based on several possibilities. The first is that incumbents just aren’t reelected when the economy is doing as poorly as it is right now.

The second possibility is that most pollsters have gotten it wrong this year.  Pollsters must estimate what the electorate will look like on Election Day in order to get their survey results.   For example, pollsters have to make an educated guess – will 19% of all voters be young (18-29 years old) or will it be 17%?  If pollsters guess wrong, the poll results will be wrong.

The third possibility is that last minute ‘undecideds’ will vote against the incumbent by a 3 to 1 ratio (as they normally do).   If all of these possibilities come true, the economy has Obama down, Romney is actually doing two to three points better in the polls than is suggested, and will win most of the late deciders… he might be expected to win 51.7% of the vote (to Obama’s 47.8%) and carry almost all of the swing states… let’s say North Carolina, Florida, Colorado, Virginia, New Hampshire, Iowa, Ohio, and Wisconsin for a total of 295 electors.  Obama holds Nevada for a total of 243 electors.  Mitt Romney is the 45th President of the United States.”

Dr. David Bridge was tasked with making the best case possible for a tie in the electoral college:

“I predict that Barack Obama and Mitt Romney will tie in the electoral college vote.

Because of the way the electoral college is set up, presidential elections are contested at the state level. It is not that national polls do not matter. But the best presidential election predictions go state-by-state, and then add up the number of electoral votes for each candidate.

When I fill out my electoral map, I give Pennsylvania to Obama. And I give Florida and North Carolina to Romney. That leaves Colorado, Iowa, Ohio, Nevada, New Hampshire, Virginia, and Wisconsin as true “toss-up” states. They could go either way.

I believe Obama will win Ohio, New Hampshire, and Wisconsin. I believe Romney will win Colorado, Iowa, Nevada, and Virginia.

In the end, the total number of electoral votes for each candidate equals 269. Granted, this overall prediction requires predicting all 51 state (including DC) contests correctly. It leaves no margin for error, and is therefore highly unlikely. However, when I go state-by-state—which is how presidential elections are contested—I come up with a tie.

Additionally, there are a handful of other combinations among those seven states that would also result in a tie. Given that they are “toss-up” states, I would not be surprised to see them swing in some other way that results in 269-269.”

A variation of this article was first published in the Baylor Lariat on 11/6/12.

One response to “Predicting the Election

  1. …I really think one of the three above will turn out to be the true. (Obama wins, or Romney wins, or a Tie.) I just feel it in my bones. I’d even be willing to bet money on it. Lots and lots of money (if I had it…) Now…if someone will just give me those correct lottery numbers…

Comments are closed.