Sunday, September 26, 2021
HomePatriot DispatchesForecasting the 2012 Presidential Election with State-Level Economic Indicators

Forecasting the 2012 Presidential Election with State-Level Economic Indicators

There has been a lot of heartburn and rising blood pressures from the reporting of national polls by the malignant media. Some of it is fueled by establishment Republicans going “tsk-tsk, don’t call it sinister.” Recently, Rush Limbaugh mentioned two political science professors at the University of Colorado who forecast presidential elections using alternative criteria which does not include polling. Ken Bickers at the Boulder campus, and Michael Berry at the Denver campus, created a forecasting model that is based on the Electoral College; it is the only one of its type to include more than one state-level measure of economic conditions. To make their predictions, the professors comb economic data from all 50 states and the District of Columbia. The professors’ model includes state and national unemployment figures as well as changes in income. The new analysis includes unemployment rates from August and changes in per capita income from the end of June. Go here for more details on the methodology used.

The 2012 election is the 9th presidential election cycle using this state economic data and the electoral college. Of note, is that the previous 8 elections have all been successfully predicted based on this model. This different kind of analysis is similar to what Bill James “money ball” analysis is for baseball. In both cases the establishment despises the ideas that challenge their conventional wisdom. Below is a table with their estimates in the last 8 election cycles.

State-Level Economic Forecasting Model Diagnostics: 1980-2008

Year States correctly classified Democrat states(Estimated Actual) Democrat EV (Estimated Actual) Outcome correctly classified
2008 48 30   29 370   365 Yes
2004 46 15   20 217   251 Yes
2000 47 23   21 264   266 Yes
1996 45 34   32 402   379 Yes
1992 41 29   33 347   370 Yes
1988 45 11   11 171   111 Yes
1984 50 3   2 17   13 Yes
1980 42 8   7 68   49 Yes

The prediction, based on the model analyzing returns from the prior eight presidential elections is that the president will win 16 states, plus the District of Columbia for a total of 208 electoral votes. Mitt Romney will win 34 states for a total of 330 electoral votes. The table below displays Obama’s states. While this forecast is great for pumping up Republican spirits, it does not mean we can relax. Whether it is a forecast using traditional national polling, or a forecast with state-level economic indicators, or a shaman viewing the entrails of a goat, the only thing that really matters is an enthusiastic GOTV turnout. We must take nothing for granted.

State Percent
Michigan 51.38
New Jersey 51.91
Oregon 52
Maine 52.04
Nevada 52.32
Massachusetts 52.39
Washington 52.49
Illinois 55.17
Connecticut 55.56
Delaware 55.73
Maryland 55.82
California 56.52
New York 56.73
Rhode Island 57.94
Vermont 60.69
Hawaii 66.34
District of Columbia 86.36

I am retired after 36 years of being a state of Indiana employee. I enjoy writing and reading conservative blogs.


Leave a Reply


Must Read