FOR MANY YEARS, ECONOMISTS HAVE BUILT MODELS designed to predict the outcomes of presidential elections in the United States, and some of these models have been remarkably accurate. The electorate definitely tends to “vote their pocketbooks,” and the current and expected condition of the economy generally is a dominant factor in election outcomes.

Incumbency Matters The predictive models built by economists generally begin by establishing the impacts on election outcomes of incumbency, length of time in office, and party affiliation of the sitting president. Once these factors are established, the impacts of economic performance are assessed.

Incumbency clearly is an advantage for presidential candidates. Republican candidates have a slight edge in presidential elections (everything else being equal), and the party in power is well positioned unless it has held the White House for two or more terms (voters eventually tire of a political party if it has been in office too long).

This set of factors should give President Bush a leg up. A one-term Republican incumbent ordinarily starts out with a decided advantage.

Economics Rules The current and expected performance of the economy carries a lot more weight in election outcomes than incumbency, party affiliation, and time in office. Anumber of powerful models have incorporated various measures of price inflation, real economic growth, employment and unemployment, and personal income growth.

A recent model variation, developed by the highly regarded Macroeconomic Advisors (MA), was able to accurately predict the outcome of the popular vote in every election since 1952—including the 2000 cliffhanger. The model leans heavily on the growth of real disposable personal income over the three quarters prior to the election, as well as on the pace of inflation during the first 15 quarters of the current presidential term.

These two economic variables are fairly traditional. But the MA model departed from tradition by introducing housing starts into the model—an experiment that turned out to be highly successful.

Why Housing Starts? MA analysts discovered that housing starts turn out to be a good summary measure of how voters view the economy. Starts have much more explanatory power than do traditional measures of consumer confidence.

According to MA analysts: “In contrast to confidence measures which are based on surveys, housing starts are hard economic data that reflect significant economic decisions of home builders and home buyers based on their sense of well-being and confidence in the future.”

Putting Ittogether The MA model, incorporating economic variables consistent with the NAHB's forecasts (including housing starts in the election quarter), predicts that President Bush will capture a solid majority of the popular vote in November.

It's important to note, of course, that models grounded on economic factors know nothing about things such as the United States' involvement in Iraq, the military records of the two candidates, and international and domestic terrorism. But history shows that the overall economy and the housing market are powerful determinants in U.S. presidential elections.

David Souders
Chief Economist, NAHB Washington, D.C.