Too Close to Call? Uncertainty and Bias in Election-Night Reporting

To read the full-text of this research, you can request a copy directly from the author.


Previous research has found that early election-night projections can have a depressing effect on turnout in presidential elections. Although this effect has been found to be small in the past, it may be enough to sway state outcomes and, potentially, the final outcomes of close presidential and other elections. Therefore, this article analyzes the election-night presidential projections of the three major cable news networks in 2000 and 2004 to examine the forces that lead to the timing of election-night calls. Copyright (c) 2007 Southwestern Social Science Association.

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the author.

The principle of simultaneity—that citizens should vote as far as possible at the same time—is more significant than is usually appreciated. It is based on a fundamental value of democratic theory, and it has substantial implications for electoral practice. It also invites further normative and empirical research. The most important value that simultaneity expresses is a form of equality—equal respect. First, if citizens have only information they would have had if they were voting at the same time, the value of each citizen's choice is no greater than that of any other citizen. Second, when citizens go to the polls on the same day, publicly participating in a common experience of civic engagement, they demonstrate their willingness to contribute to the democratic process on equal terms. Taking simultaneity seriously has implications for electoral practices. It would limit the practice of media projections of election results, including the reporting of exit polls, and the calling of elections before all the polls close. It also casts doubt on the increasingly widespread use of early voting—absentee ballots and voting by mail. Early voting weakens the value of the experience of participating in a civic activity. Voting alone may be worse than bowling alone.
Full-text available
This article analyzes candidate competition and attrition in presidential primaries from 1912 to 2000. We use a modified market concentration index to measure the number of effective candidates entering presidential primaries and to calculate winnowing across presidential primaries. We test competing hypotheses about the increasing number of candidates entering primaries. We find that the number of effective candidates increased following the reforms of the 1970s. We reject the hypothesis that the number of candidates entering primaries increased gradually as a function of long-term factors. We also find similarities in the winnowing of candidates in each party during the prereform and postreform eras, which suggests the timing of candidate withdrawal owes to more than differences in each party’s delegate allocation rules. Republican presidential primaries more efficiently winnow candidates than do Democratic presidential primaries in both the prereform and postreform eras.
Full-text available
We measure media bias by estimating ideological scores for several major media outlets. To compute this, we count the times that a particular media outlet cites various think tanks and policy groups, and then compare this with the times that members of Congress cite the same groups. Our results show a strong liberal bias: all of the news outlets we examine, except Fox News' Special Report and the Washington Times, received scores to the left of the average member of Congress. Consistent with claims made by conservative critics, CBS Evening News and the New York Times received scores far to the left of center. The most centrist media outlets were PBS NewsHour, CNN's Newsnight, and ABC's Good Morning America; among print outlets, USA Today was closest to the center. All of our findings refer strictly to news content; that is, we exclude editorials, letters, and the like. “The editors in Los Angeles killed the story. They told Witcover that it didn't ‘come off’ and that it was an ‘opinion’ story.… The solution was simple, they told him. All he had to do was get other people to make the same points and draw the same conclusions and then write the article in their words” (emphasis in original). Timothy Crouse, Boys on the Bus [1973, p. 116].
Full-text available
Using voting data for presidential elections from 1976 to 2000, this paper documents an unusual and large drop off in Republican voting rates for Florida’s western Panhandle during the 2000 General Elections. Little change appears to have occurred in the rate that non-Republicans voted. The results appear more consistent with the early call reducing Republican voting rates than the networks discouraging all voters from voting by incorrectly calling the polls closed in the western Panhandle. Copyright Springer Science + Business Media, Inc. 2005
Here is an accessible, up-to-date guide to event history analysis for researchers and advanced students in the social sciences. The foundational principles of event history analysis are discussed and ample examples are estimated and interpreted using standard statistical packages, such as STATA and S-Plus. Recent and critical innovations in diagnostics are discussed, including testing the proportional hazards assumption, identifying outliers, and assessing model fit. The treatment of complicated events includes coverage of unobserved heterogeneity, repeated events, and competing risks models. The authors point out common problems in the analysis of time-to-event data in the social sciences and make recommendations regarding the implementation of duration modeling methods.
This report gives the results of the analysis of data collected from a subset of the respondents in the 1980 Presidential Election Study, which ascertained what time of day the person voted, what election night news they heard, and when they heard that news. This analysis indicates that many people heard projections of the presidential election outcome before the local polls closed, that hearing news of the projected outcome decreased the likelihood of voting among those who had not already voted, and that exposure affected the likelihood of turnout among Republicans more than among Democrats (although Democrats were less likely than Republicans to vote late on election day).
The Nightly News Nightmare, Third Edition, examines news coverage of presidential nomination and election campaigns from 1988 to 2008. The book focuses on changes in the amount, tone, and focus of news coverage in these different electoral contexts. In addition to network news, the authors examine online news, cable television, talk radio, candidate campaign discourse in these election years. Farnsworth and Lichter find that the news media, despite the wide variety of outlets, have consistent problems in terms of fairness and focus on substantive matters rather than the horse-race reporting of the latest polls. The authors conclude that online news had many of the same problems found in mainstream news coverage.
D.R. Cox has suggested a simple method for the regression analysis of censored data. We carry out an information calculation which shows that Cox's method has full asymptotic efficiency under conditions which are likely to be satisfied in many realistic situations. The connection of Cox's method with the Kaplan-Meier estimator of a survival curve is made explicit.
This study is an attempt to disentangle the influence of exposure to several distinct kinds of television content during the last few days prior to the 1972 election on information levels, candidate preference, and reports of several kinds of voting behavior. Using post-election interviews from a national sample of the American electorate (and some pre-election data), the study demonstrates that last-minute exposure to television news, special programming, paid programming, and election projections are associated with somewhat different patterns of behavior. Simple "reinforcement" notions of media effects, commonly employed in voting studies, are inadequate, and clear distinctions must be maintained in future studies concerning the kinds of exposure (within a medium), its quality, and the conditions under which exposure occurs. A molecular, rather than molar, approach to the study of television and politics is clearly warranted.
Methodological details are presented of a survey research technique known as exit or election day polling. Using this technique, major American news organizations collect and analyze voting and attitude data from samples of persons who have just cast ballots. On the basis of the 1980 elections, differences in polling strategies and performance of the exit poll method are examined. How election day survey data are used by journalists is discussed.
This study analyzes time of voting effects on the results of exit poll data collected in a Cleveland municipal tax referendum. The data reveal significant differences both in the times different groups vote and in the reported levels of support for the tax levy by time of day. The temporal variation in the reported levels of support for the tax levy are then shown to be directly related to the times when different support groups cast their ballots. Collectively, these findings suggest the need for day-long interviewing in the design of exit polls.
This article summarizes the findings of the effect of exit polls on voting behavior. Both macro and micro methods have been used, and no methods are perfect. Exit polls appear to cause small declines in total voting in areas where the polls close late for those elections where the exit polls predict a clear winner when previously the race had been considered close.
By the weekend following the election of 2000, two possible frames were available to the press covering developments in Florida. In the first, Gore had won the popular vote and the outcome in the electoral college was uncertain. In the second, Bush was ahead in the vote in the state that would determine the results in the electoral college and, as such, the presumed victor until Gore proved otherwise. Elite discourse as revealed in Sunday morning talk shows eventually settled into the second frame, but not until the certification of the Florida vote by Florida Secretary of State Katherine Harris. Elite discourse was not, however, beneficial to Bush in the early weeks of the protest phase of the election.
The purpose of this study is to assess the influence of corporate media owners over news content. In particular, we address the claim that the financial interests of corporate owners lead America's news bureaus to downplay the significant issues surrounding the growing concentration of ownership of the country's mass media. To do so we examine newspaper coverage of one aspect of the 1996 Telecommunications Act: the loosening of restrictions on television ownership. We compare coverage of this aspect of the Telecommunications Act in newspapers owned by companies that stood to gain from the loosening of these restrictions, with coverage in newspapers owned by companies which did not stand to gain. We find substantial differences in how newspapers reported on these proposed regulatory changes depending on the financial interests of their corporate owners.
In recent years presidential charges of maltreatment by the press have become commonplace. Various scholarly research into political communication appears to confirm the validity of these charges. However, a number of issues prevent one from inferring bias from the high levels of unfavorable presidential news these studies report. The research reported here is designed to overcome these problems and allow us to test the bias hypothesis more conclusively. Applying this design to the three networks’ evening news programs during the years 1990 through 1995, we find qualified support for the bias hypothesis but even more compelling evidence that changes in presidential approval, whether favorable or unfavorable, drive news coverage of the president's public support. We also find surprising differences in the networks’ routines and patterns of coverage that call into question the common assumption of homogenous network behavior.
The election projection of the 1980 presidential contest by NBC raised much speculation concerning its possible impact on voting in states where the polls were still open. Research on the subject has started from different assumptions, used different data and methods, and come to different conclusions concerning the real-world effects of such early calls. Using district-level voting and demographic data and focusing on deviations from normal voting patterns, this study finds the early call to have had a small but measurable impact on presidential and congressional turnout, and a somewhat larger impact on depressing the vote for Democratic candidates at both levels. In addition, higher income, white collar, and better educated populations appear to have been affected to a greater extent. While the overall impact was too small to have affected the outcome of the presidential race, at the congressional level as many as fourteen races were won by margins smaller than the estimated impact of the early call in those districts.
The influence of election night television broadcasts on the voting behavior of persons who have not yet gone to the polls is only one of many potential effects of TV that increasingly arouse both curiosity and concern. Like all previous studies of this particular issue, the present study, carried out during the 1968 presidential election, concludes that the broadcasts had no measurable effects on either turnout or candidate choice.
We investigate the market for news under two assumptions: that readers hold beliefs that they like to see confirmed, and that newspapers can slant stories toward these beliefs. We show that, on the topics where readers share common beliefs, one should not expect accuracy even from competitive media: competition results in lower prices, but common slanting toward reader biases. However, on topics where reader beliefs diverge (such as politically divisive issues), newspapers segment the market and slant toward the biases of their own audiences, yet in the aggregate a conscientious reader could get an unbiased perspective. Generally speaking, reader heterogeneity is more important for accuracy in media than competition per se.
Bush Camp Tries to Assume a Winner's Pose
  • Yardley
  • Michael Jim
  • Cooper
Yardley, Jim, and Michael Cooper. 2000. ''Bush Camp Tries to Assume a Winner's Pose.'' New York Times November 9:B. Uncertainty and Bias in Election-Night Reporting
The Unexplained Poll Discrepancy University of Pennsylvania Available at h Ownership and News Bias: Newspaper Coverage of the 1996 Telecommunications Act
  • Freeman
  • Steven
Freeman, Steven F. 2004. The Unexplained Poll Discrepancy. University of Pennsylvania. Available at h Gilens, Martin, and Craig Hertzman. 2000. ''Corporate Ownership and News Bias: Newspaper Coverage of the 1996 Telecommunications Act.'' Journal of Politics 62(2): 369–86.
Coup Attempt Reveals Real Gore
  • Will
  • F George
Will, George F. 2000. ''Coup Attempt Reveals Real Gore.'' Times Union (Albany, NY) November 12:Editorial.
A Night at the Races: Response Latency in Declaring Election WinnersCandidate Competition and Attrition in Presidential Primaries
  • Norrander
  • Lee Barbara
  • Bradford S Sigelman
  • Jones
Norrander, Barbara, Lee Sigelman, and Bradford S. Jones. n.d. A Night at the Races: Response Latency in Declaring Election Winners. Unpublished manuscript. Steger, Wayne P., John Hickman, and Ken Yohn. 2002. ''Candidate Competition and Attrition in Presidential Primaries, 1912–2000.
How Does CNN Make Election Projections? Available at hhttp:// www.cnnScooping the Voters? The Consequences of the Networks' Early Calls of the 1980 Presidential Race
  • Cbs News
CBS News. 2000. Did Networks Discourage Voters? Available at h stories/2000/11/10/politics/main248524.shtmli. 2004. How Does CNN Make Election Projections? Available at hhttp:// Delli Carpini, Michael X. 1984. ''Scooping the Voters? The Consequences of the Networks' Early Calls of the 1980 Presidential Race.'' Journal of Politics 46(3):866–85.
Averaged Nonpartisan Polls Available at h 2004. Electoral Vote Projections
  • Electoral-Vote
  • Com 2004. Averaged Nonpartisan Polls. Available at h 2004. Electoral Vote Projections. Available at h htmi.
Election Night Bias Charged GOP: Al Aided by Fast Calls
  • Richard Huff
Huff, Richard. 2000. ''Election Night Bias Charged GOP: Al Aided by Fast Calls.'' New York Daily News November 17:TELEVISION.
The Election Polls: An American Waterloo
  • Ladd
  • Everett
Ladd, Everett C. 1996. ''The Election Polls: An American Waterloo.'' Chronicle of Higher Education November 22:A52.
University of Pennsylvania
  • F Steven
  • Freeman
  • Hoffstetter C. Richard