Why the GOP might just squeak out the 2016 election

Why the GOP might just squeak out the 2016 election — unless the nominee is terrible 

Goldman Sachs is out with a note about its political forecasting model, and how the economy might affect the 2016 White House race. According to the Goldman model, the fundamental factors that matter most are real GDP, real consumption, and real personal income. With the 2012 election added to the sample, the change in non-farm payrolls also seems to have acquired more predictive power. And the stock market? Not so much.

Timing matters, too. Goldman: “It is only around the current stage of the presidential cycle (i.e., late in the year before the election) that economic variables tend to become useful in predicting the outcome, and not until Q2 of the election year that many indicators reach their maximum predictive value.” The bank also includes whether the incumbent party has held office for at least two terms, kind of a “fatigue” factor I suppose.
So here is what the model churned out:

Regarding the more important fundamental indicators, our current economic forecast implies a close race. However, the outcome depends heavily on the model chosen. Considered in isolation (i.e., in a bi-variate regression), our forecasts for three indicators with the greatest predictive value—real GDP growth, real per capita personal income growth, and the change in non-farm payrolls—all suggest that the Democratic candidate should win just over 50% of the popular vote. However, adding the incumbent party dummy variable described above shifts the results by an average of 3 percentage points, transforming a narrow win for the incumbent party into a loss with around 47% of the two-party vote.

Several caveats are in order. First, we do not know who the candidates are, and early polling—which is itself not of much use—shows a wide dispersion of voter preferences depending on the individual candidates. Second, it is early enough in the campaign that the actual economic data—itself subject to revision—don’t give us much of an idea about the eventual winner. Instead, we must rely on our own forecasts of those variables, which are apt to change as events unfold. Third, it is the electoral college, not the popular vote, that determines the election outcome. There is normally a small “bias” toward one party in the electoral college, so that one party could lose some of its popular vote without actually losing any electoral votes. In a close election, this bias is typically worth less than one percentage point of the popular vote, but has occasionally been enough to swing the result (in 2000 and 1888, for example). Recent research suggests that the current concentrations of partisan support going into the 2016 election could favor the Democratic candidate by one to two percentage points. If the pattern holds, this would imply a closer contest than the figures above suggest. The upshot is that, if our economic forecast proves correct, the economic data should make for a fairly even playing field.

So if you buy the Goldman model and the economic forecast that feeds into it, Republicans should be considered the slight favorite — candidate quality  and economic revisions aside. And here is that Goldman econ forecast, by the way: “Based on developments in financial markets as well as some of our recent analytical work, we think another modest downward adjustment to our GDP forecasts is now warranted. Specifically, we are reducing our GDP growth forecasts for the next three quarters by 0.25pp each, taking the numbers to 2.5% for Q4 2015 and 2.25% for Q1 and Q2 of 2016.” So a same old, same old, new normal forecast is good news for the GOP’s political prospects. So choose wisely, Republicans!