Polling Incumbents

This article by Guy Molyneux appearing in the American Prospect online explains why polling reports can often be misleading. It argues that the key statistic to focus on in presidential races is the incumbent's percentage, not the challenger's. The incumbent usually drops about a half a percentage point between the final polls and the actual vote, while the challenger picks up most of the undecided voters. In the current campaign, assuming a third party vote of approximately 2 percent, Molyneux argues, Bush must score 49 percent of the vote to win. Furthermore, Molyneux argues, even if Bush is doing very well in traditional Republican states, in most of the key battleground states, Bush is holding on to around 47 percent of the vote, which means that he will lose these states.

Indeed, if the scenario Molyneux describes holds-- big victories in the most Republican states, narrow losses in the battleground states-- Bush might even win the popular vote and lose the electoral college.

If that were to happen, two things might occur. First, both parties might finally see an advantage in getting rid of the electoral college, that antiquated relic of slavery. Second, Kerry would likely be a one-term president, for a Bush loss in 2004 would continue an unbroken trend in which candidates who win the Presidency without winning the popular vote (J.Q. Adams, Hayes, Harrison) do not get a second term.