Sabato's Crystal Ball

Tracking the 2008 Presidential Election

Alan I. Abramowitz, Guest Columnist July 24th, 2008

facebooktwittergoogle_plusredditpinterestmail

If you’re a political junkie like me, and if you’re reading this article there’s a pretty good chance you are, then you’re probably addicted to the Gallup and Rasmussen tracking polls. Where else can you follow the ups and downs of the presidential race every day? I can hardly wait to get my daily fix of Rasmussen every morning and Gallup every afternoon.

The folks at Gallup and Rasmussen are certainly providing a valuable service. Beyond just tracking the horserace, the vast numbers of respondents surveyed in these tracking polls have allowed for some fascinating analyses of the preferences of various subgroups within the electorate. Gallup, for example, is posting weekly breakdowns of candidate preference by age, race, gender, education, party, ideology, and church attendance on their website. It’s a real goldmine of information for election watchers.

Lately, though, something’s been bothering me about those tracking polls. I’ve had a nagging suspicion for a while that the results of the tracking polls are out of line with the results of other national polls on the presidential race. Not way out of line, but enough to be noticeable, at least to me. So I decided to check on my suspicion by doing a systematic comparison of the results of the Gallup and Rasmussen tracking polls with all other national polls that were conducted in May, June, and July. I got the Rasmussen poll results from their website and the Gallup and other poll results from pollingreport.com.

Here’s what I found. Since the beginning of May, over 74 days of polling, the Gallup tracking poll has shown Barack Obama with an average lead of 1.6 percentage points over John McCain. During the same time period, the Rasmussen tracking poll, over 76 days of polling, has shown Obama with an average lead of 1.8 percentage points. But during the exact same time period, 38 other national polls have shown Obama with an average lead of 5.2 percentage points.

It’s not a huge difference. But given the numbers of respondents interviewed in these polls-about 60 thousand in the Gallup tracking poll, 75 thousand in the Rasmussen tracking poll, and 40 thousand in the other national polls-it is certainly a statistically significant difference. More importantly, the Gallup and Rasmussen results give a different impression of the state of the presidential race from other national polls. A lead of less than two points suggests a much tighter race than a lead of between five and six points.

There’s also no correlation between the day-to-day, week-to-week, or month-to-month changes in the two tracking polls. Whether Obama’s lead goes up or down in one is unrelated to whether it goes up or down in the other. Today (Monday, June 21st), for example, Obama’s lead decreased from two points to one point in the Rasmussen tracking poll but increased from three points to six points in the Gallup tracking poll. Between June and the first three weeks of July, Obama’s average lead increased from 2.4 points to 3.5 points in the Gallup tracking poll but decreased from 4.8 points to 2.6 points in the Rasmussen tracking polls. The lack of any correlation between the two polls suggests that these changes are a result of random statistical noise rather than any real shift in the underlying preferences of the electorate. In fact, the results of other national polls indicate that there has been little or no change in voter preferences over the past three months: the average Obama lead was 5.1 points in May, 5.3 points in June, and 5.2 points in July.

I don’t have any explanation for the discrepancy between the tracking polls and other national polls. It is certainly possible that between now and Election Day, the results of the two types of polls will come together. In the meantime, however, these results indicate that political junkies should take their daily dose of Rasmussen and Gallup with a large grain of salt.


Alan I. Abramowitz is the Alben W. Barkley Professor of Political Science at Emory University, and the author of Voice of the People: Elections and Voting Behavior in the United States (2004, McGraw-Hill). He can be contacted via email at polsaa@emory.edu.