Fair's equation to predict the 2016 presidential election is

VP = 42.39 + .667*G - .690*P + 0.968*Z

On the left-hand side of the equation, VP is the Democratic share of the presidential vote. Given that a Democrat is in office, a legacy of economic growth should tend to favor the Democratic candidate, while inflation would tend to work against the Democrat. On the right-hand side, G is the growth rate of real per capita GDP in the first 3 quarters of the election year (at an annual rate); P is the growth rate of the GDP deflator (a measure of inflation based on everything in the GDP, rather than just on consumer spending as in the better-known Consumer Price Index); and Z is the number of quarters in the first 15 quarters of the second Obama administration in which the growth rate of real per capita GDP is greater than 3.2 percent at an annual rate.

Obviously, some of these variables aren't yet known, because the first three-quarters of 2016 haven't happened yet. But here are Fair's estimates of the variables as of late April: G=0.87; P=1.28; Z=3. Plug those numbers into the formula, and the prediction is that the Democratic share of the two-party presidential vote in 2016 will be 44.99%.

Fair offers a similar equation to predict the 2016 House elections. The formula is

VC = 44.09 + .372*G - .385*P + 0.540*Z

where VC is the Democratic share of the two-party vote in Congressional elections. Plugging in the values for G, P and Z, the prediction is 45.54% of the House vote for Democrats.

Obviously, some of these variables aren't yet known, because the first three-quarters of 2016 haven't happened yet. But here are Fair's estimates of the variables as of late April: G=0.87; P=1.28; Z=3. Plug those numbers into the formula, and the prediction is that the Democratic share of the two-party presidential vote in 2016 will be 44.99%.

Fair offers a similar equation to predict the 2016 House elections. The formula is

VC = 44.09 + .372*G - .385*P + 0.540*Z

where VC is the Democratic share of the two-party vote in Congressional elections. Plugging in the values for G, P and Z, the prediction is 45.54% of the House vote for Democrats.

Of course, these formulas raise a number of questions. Where do these numbers and this formula come from? Why use these variables about economic growth rather than, say, the unemployment rate? Why measure inflation with the GDP deflator rather than with the Consumer Price Index? Where did the coefficient numbers come from?

The short answer to all these questions is that Fair's equations are chosen so that, if one looks back at historical election data from 1916 up through 2014, this equation is both fairly simple and does a pretty good job in predicting all the elections over time with the smallest possible error. The long answer to why these specific variables were chosen and how the equation is estimated is that you need to read the research papers at Fair's website.

Is there reason to believe that a correlation between the macroeconomy and election outcomes has existed during the last century or so of national elections, it will also hold true in 2016? Of course, Fair isn't making any claim that the macroeconomy fully determines election outcomes. Every election has lots of idiosyncratic factors related to the particular candidates and the events of the time. Correlations are just a way of describing or summarizing earlier patterns in the data. Fair's equation tell how macroeconomic factors have been correlated with election outcomes, based on the past historical record, but it doesn't have anything to say about all the other factors in a national election. For example, the predictions of the equation for the Democratic vote were way low in 1992, when Bill Clinton was elected, and also in 2004, when George W. Bush was re-elected. On the other side, predictions from the equation of the Democratic share of the vote were too high in 1984 and 1988, when Ronald Reagan was re-elected and then George Bush was elected.

At the most basic level, Fair's equation is just saying that a slow rate of economic growth during 2016, along with the fact that there haven't been many rapid quarters of economic growth during the Obama presidency, will tend to make it harder for Democrats to win in 2016. But correlation doesn't prove causation, as Fair knows as well as anyone and better than most, and he would be the last one to overstate how much weight to give to these kinds of formulas. Back in 1996, Fair provided a nontechnical overview of this work in "Econometrics and Presidential Elections," appearing in the

*Journal of Economic Perspectives*(where I work as Managing Editor). He wrote there:

"The main interest in this work from a social science perspective is how economic events affect the behavior of voters. But this work is also of interest from the perspective of learning (and teaching) econometrics. The subject matter is interesting; the voting equation is easy to understand; all the data can be put into a small table; and the econometrics offers many potential practical problems. ... Thus, this paper is aimed in part at students taking econometrics, with the hope that it may serve as an interesting example of how econometrics can be used (or misused?). Finally, this work is of interest to the news media, which every fourth year becomes fixated on the presidential election. Although I spend about one week every four years updating the voting equation, some in the media erroneously think that I am a political pundit—or at least they have a misleading view of how I spend most of my days."