Also called “trend polls,” tracking polls are conducted primarily by news organizations and specialized polling firms such as Reuters/Ipsos, Public Policy Polling (PPP), Zogby, Rasmussen, and Gallup. They most commonly ask citizens whom they plan to vote for on Election Day (or, alternately, whom the respondent would vote for if they had to cast their ballot now). Such polls are designed to predict; for this reason, they occur at fixed (and generally evenly spaced) intervals, such as daily, once per week, or once every two weeks. At each interval, those sampled for the poll are asked a question that is worded identically to the one of the week before. News organizations, polling organizations, and candidates’ campaigns use tracking polls to determine which candidates are in the lead, which is integral to horse-race campaign coverage. Significant surges in the polls are viewed as a sign of a successful campaign strategy, whereas declines in support are viewed as harbingers of a losing campaign.
Political scientists are critical of tracking polls because they are often misinterpreted by journalists and political pundits. Every poll has a margin of error, and when the margins between candidates have overlap, journalists often portray one candidate as in the lead, treating apparent differences in support as statistically meaningful (when they are not). Moreover, while some movement in the polls may be due to campaign events, it is difficult to disentangle this from the short-term noise (measurement error) that is endemic to any poll. A more reliable estimate of candidate popularity can be found in a poll of polls, which is becoming a more common feature in political coverage of campaigns.
Another criticism of tracking polls is that they place excessive emphasis on the horse race itself, often at the expense of an in-depth analysis of the important issues of the day. Thomas Patterson argues that journalists treat elections as sporting events, in which every action by a candidate is geared toward winning the game. Polls are simply a means of keeping score in this type of coverage. Voters learn a great deal about which candidate is leading and which one is behind in the polls (that is, who is winning the game and who is losing), but they learn little else. Such coverage may mobilize voters who want to turn out to support their “team,” but it isn’t clear that horse-race coverage helps to mobilize voters who are not already strong partisans. When elections are viewed through the lens of a game, team loyalty becomes more important than other objectives such as good governance, and this may lead to partisan polarization, citizen distrust of government, and unwillingness by political leaders to discuss complex and controversial issues with voters.
Green, Donald P., and Alan Gerber. “How to Interpret Tracking Polls.” Campaigns and Elections 19 (August 1998): 23–27.
Patterson, Thomas E. Out of Order. New York: Knopf, 1993.
PBS Newshour. “Online Focus: Perplexing Polls.” Transcript, October 31, 2000 http://www.pbs.org/newshour/bb/media/july-dec00/polls_10-31.html. Accessed September 5, 2015.