Forecasting with Mixed Economic Signals: A Cautionary Tale

Abstract

In every respect, the 2000 election defied what had become the conventional wisdom about American presidential elections—that the electorate rewards the incumbent administration for good economic times and punishes it for bad economic times. Occurring at a time when the incumbent president enjoyed fairly high levels of approval and the nation enjoyed relative economic prosperity, it seemed the 2000 election had all the markings of an election in which voters would support the status quo and vote to return a Democrat to the White House. Indeed, a group of election forecasters gathered at the 2000 Annual Meeting of the American Political Science Association and offered predictions that Gore would win a plurality of the two-party popular vote that ranged from 52.8% to 60.3%.The forecasts were as follows: Alan Abramowitz, 53.2%; James Campbell, 52.8%; Thomas Holbrook, 60.3%; Michael Lewis-Beck and Charles Tien, 55.4%; Brad Lockerbie, 60.3% (revised after the APSA roundtable); Helmut Norpoth, 55.0%; Christopher Wlezien and Robert Erikson, 55.2%. All forecasts are for the incumbent party candidate's percent of the two-party vote. Of course, we now know that although Gore did manage to win the popular vote by the slimmest of margins, he lost the electoral vote by an equally slim margin, and George W. Bush was elected president.