- Take Action
- Reformer's Blog
- Campus Training
- Publicity Resources
- Activism Resources
- Advice for Student Leaders
- Manuals and Writings
- Programs and Services
- Legal Organizations
- In-State Policy Organizations
- Biased Textbooks
- Politically Neutral Occupy Technology
- Live Webcasts
- Activism Grants
- Journalism Grants
- Resources for Student Leaders
- Conservative Heroes
Campus Reform's Levi Fox exclusive interview with Colorado Professor, Dr. Michael Berry:
CR: When can we expect to see the September numbers?
Berry: September state unemployment data is scheduled to be released at the end of this week, but the updated forecast using August unemployment numbers will be the final forecast we make before the election. You can read the CU press release about the updated forecast here: http://www.colorado.edu/news/releases/2012/10/04/updated-election-forecasting-model-still-points-romney-win-university
CR: Do you foresee any major change compared to the August results?
Berry: We generate predictions for each candidate’s share of the two-party popular vote for all 50 states and DC. Of course, when you reestimate the model with more contemporary data, there will be variations in these projections from the prior estimates. For the purposes of generating an Electoral College forecast, I doubt that September unemployment data would dramatically change our prediction for how states will ultimately allocate their Electoral College votes. For example, in our published article, we projected that President Obama would receive 48.2 percent of the vote in Colorado. Using more current economic data, we project that Obama will receive 46.7 percent of the state vote. In terms of the Electoral College, the forecast is only altered when a state crosses the 50 percent threshold in either direction. In our updated forecast, only one state–New Mexico–switched from one candidate’s column to the other.
CR: What did you find as the most surprising or intriguing results within this election’s study?
Berry: Aside from generating the forecast itself, what I found most interesting were the list of variables we found that did not generally affect presidential election results. For example, we modeled the site of the party conventions, the partisanship of the governor, and the home-state of the vice-presidential candidate, among others, and none exerted a significant effect on how a state votes for president. So for 2012, our analysis tells us that Florida and North Carolina are no more likely to vote for Romney or Obama respectively than if the conventions hadn’t been held in those two states.
CR: Have you received any negative criticism by academic peers?
Berry: None that I’ve heard.
CR: How certain are you that Mitt Romney will win on November 6th?
Berry: We didn’t calculate a specific confidence level for the Electoral College result, but our model indicates that Governor Romney has a 77 percent likelihood of winning the popular vote.
Follow Levi Fox on Twitter @levithefox
CampusReform.org is a project of the non-profit Leadership Institute. Please consider supporting our efforts.
Share your comment