Immune from news stories like Special Counsel Hur’s report or President Joe Biden’s press conference, Professor Ray Fair’s 2024 economy-driven vote share model shows Biden gaining ground since the last quarterly economic data was released.
A summary of Professor Fair’s two-party vote-share model can be found in a previous article and here is Fair’s description of the model updates:
Democratic Vote Percentage ~ I + DPER + DUR + WAR + (G * I) + (P * I) + (Z * I)
The variables represent:
- I = 1 if there is a Democratic presidential incumbent at the time of the
- election
- DPER = 1 if a Democratic presidential incumbent is running again
- DUR = 0 if either party has been in the White House for one term
- WAR = 1 if a World War is occurring, 0 otherwise
- G = growth rate of real per capita GDP in the first 3 quarters of 2024 (annual rate)
- P = growth rate of the GDP deflator in the first 15 quarters of the Biden administration, 2021:1-2024:3 (annual rate)
- Z = number of quarters in the first 15 quarters of the Biden administration in which the growth rate of real per capita GDP is greater than 3.2 percent at an annual rate
Applying the model above to election results and economic data from 1916 to 2020 gives us the following coefficients to apply to the variables for the 2024 Presidential election:
Democratic Vote Percentage = (I * -0.85) + (DPER * 2.11) + (DUR * -3.45) + (WAR * 3.90) + Intercept + 0.708*G – 0.606*P + 0.865*Z
Democratic Vote Percentage = 49.48 + 0.708*G – 0.606*P + 0.865*Z
The model coefficients (-0.85, 2.11, -3.45, 3.90, 0.708, 0.606, and 0.865) are values telling us how much their respective variables change the vote percentage. If the coefficient for G is 0.708, it means a one percentage point increase in G leads to a 0.708 percentage point increase in the Democratic vote percentage. A positive number means the Democratic vote percentage goes up as the coefficient increases, and a negative number means the Democratic vote percentage goes down.
The constant 49.48 is derived from the first part of the formula in yellow (the 1 or 0 “dummy” variables and the model intercept) about the political situation. It will not change as long as President Biden is running for re-election and there is no world war:
- Democratic presidential incumbent at the time of the election (I = 1)
- Democratic presidential incumbent is running again (DPER = 1)
- Neither party has been in the White House for more than one term (DUR = 0)
- No present world war (WAR = 0)
You can think of the model Intercept as the Democratic Vote Share when all the model variables are set to zero.
Constant = (I * -0.85) + (DPER * 2.11) + (DUR * -3.45) + (WAR * 3.90) + Intercept
Constant = (1* -0.85) + (1 * 2.11) + (0 * -3.45) + (0 * 3.90) + 48.22
Constant = -0.85 + 2.11 + 0 + 0 +48.22
Constant = 49.48
Now we will look at the economic forecasts from the second half of the model in green. The economic numbers improved from the last quarter and continue to (slightly) favor President Biden:
- G increased to 1.80 (previously 1.40)
- P decreased to 4.55 (previously 4.78)
- Z = 4 (no change from the previous quarter)
The updated Democratic Vote Share model is:
Democratic Vote Percentage = 49.48 + 0.708*G – 0.606*P + 0.865*Z
Democratic Vote Percentage = 49.48 + (0.708 * 1.80) – (0.606*4.55) + (0.865*4)
Democratic Vote Percentage = 49.48 + 1.2744 – 2.7573 + 3.46
Democratic Vote Percentage = 51.46
The model currently predicts (with data as of 1/25/2024) that the Democratic share of the two-party vote will be 51.46% in the 2024 Presidential election, up from 51.03% three months ago.
Again, the model relies on top-line economic forecasts and doesn’t factor in any polling or news events.
Previously, we’d compared the error from Professor Fair’s model with the Iowa Electronic Markets (IEM) two-party vote-share prediction market prices. We found that IEM was slightly more accurate than the Fair model, while the average of both predictions was more accurate than each alone. Here is the Fair model output with the IEM price from January 25, 2024
Biden (or the Democratic nominee) receiving 58% of the vote seems like a stretch; no candidate has reached 58% since Reagan in 1984. One possible explanation is the lack of liquidity on IEM: only 46 shares were traded in the month of January. That price also seems to be in keeping with past prices (and liquidity) on IEM this far out from Election Day:
- In January 2020, 366 shares traded, and the Democratic vote share was 56.2% on January 25th. The result was 52.3% (IEM +3.9).
- In January 2016, 27 shares traded, and the Democratic vote share was 59.8% on January 25th. The result was 51.1% (IEM +8.7).
Professor Fair’s updated model can be found here.