Yale economist Ray Fair is one of the pioneers of modern-day election forecasting (Fair, 1978). His presidential vote equation, which was first used to predict the 1980 election, is still applied today, although, not surprisingly, it underwent several revisions in the nearly 40 years since its first publication. Basically, the model is built on four major assumptions or theories of voting:
- Incumbent presidents running have an advantage.
- Voters like change. Therefore, parties in office for two or more consecutive terms have a disadvantage.
- There is a slight but persistent bias favoring the Republican Party.
- The state of the economy affects the incumbent party vote.
The presidential vote equation predicts the Democratic two-party popular vote. In its latest specification, it is formulated as (Fair, 2009; 2018):
V = 48.06 + 0.673 (
I) – 0.721 (
I) + 0.792 (
I) + 2.25 (
DPER) – 3.76 (
DUR) + 0.21 (
I) + 3.25 (
The variables used in the model are defined in the following table:
The model, and earlier version thereof, has made forecasts for U.S. presidential elections since 1980. The table below shows all forecasts that could be found.
- Fair, R. C. (2018). Presidential and congressional vote-share equations: November 2018 Update.
- Fair, R. C. (2009). Presidential and Congressional Vote‐Share Equations. American Journal of Political Science, 53(1), 55-72.
- Fair, R. C. (1978). The effect of economic events on votes for president. The Review of Economics and Statistics, 60(2), 159-173.