The New “F” Word: Indexed Out of the Election Debate

by: Bill Herman / University of Pennsylvania

If the Watergate story broke tomorrow, would it get any traction? I somehow doubt it. Rather, I fear Woodward and Bernstein would be shouted down as partisan hacks and conspiracy theorists, and the mainstream media would offer airtime to any experts who would discredit them.

The least-respected, and potentially biggest, story of 2004 is not about one political party breaking into another’s headquarters. Serious, documented allegations of electoral anomalies, malfeasance, and even electoral fraud in the presidential election are widely available online, but most are absent in the mainstream media. I do not use this paper to advance these theories; readers may follow the links themselves and draw their own conclusions. Rather, I argue that newsworthy allegations of fraud are systematically marginalized by major news outlets. First, I will describe W. Lance Bennett’s theory of indexing. Then, I will briefly explain four newsworthy fraud-related stories, two statistical and two anecdotal, describing how both have been indexed out of most news coverage. Finally, I will briefly indicate where the research should go from here.

The media do not give equal weight to all voices in political debates. As Bennett argues, the media “tend to ‘index’ the range of voices and viewpoints in both news and editorials according to the range of views expressed in mainstream government debate about a given topic.”[1] That indexing generally reflects material conditions of power. “An essential journalistic norm is to survey the recognized power players in a political conflict and to try to give representation to the various players in proportion to their relative power.”[2] As a corollary, if there is no debate among powerful players, there tends to be no story.

Two Statistical Analyses: Exit Polling and Florida Counties
Almost everybody who watched televised election night coverage thinks they know the exit polling story: early exit polls showed a clear Kerry victory, but final vote tallies showed that those exit poll numbers were wrong. The part that is rarely mentioned is that Kerry still led in the final exit polls. Dr. Steven F. Freeman, visiting faculty at the University of Pennsylvania, discusses this using uncorrected exit poll data taken from in the early morning hours of Wednesday, November 3.[3] Exit polls are a reliable predictor of electoral outcomes, used to demonstrate fraud in elections in Georgia (the former Soviet republic), and currently being offered as proof of election fraud in Ukraine (even by the same mainstream US press that ignores the story at home). Of eleven swing states, only one tabulated a final result in line with exit polls; the other ten states broke remarkably for Bush. Freeman argues that the odds that this discrepancy occurred merely due to random error are 1 in 250 million. He argues not that this is proof of fraud, but he places it on the list of possibilities and insists that this monumental anomaly must be reconciled. First and foremost, he calls for the polling company to release the full data so that scholars can analyze them fully.

Self-appointed experts, some of whom cannot competently evaluate (and may not have read) Freeman’s argument, have marginalized this story out of the headlines as just another “internet conspiracy theory.” Among such self-appointed experts is Douglas Chapin, whose website ( monitors election irregularities and reform initiatives. On November 18, 2004, Chapin was a guest on a Philadelphia-area call-in radio show.[4] One caller mentioned Freeman by name, citing his statistical estimate of 1 in 250 million, and asked “Dr. Chapin” what he thought about the paper. Chapin begins his answer by explaining that (unlike Freeman, who has a PhD from MIT’s Sloan School of Management) he is not a doctor. Qualifications aside, he provides a mini-lecture on the flawed methodology of the exit polls and says that he has “yet to see any data that leads me to believe that we’ve got a really rigorous study of that problem.” When the next caller talks about anecdotal stories of disenfranchisement and fraud, but not statistics, Chapin takes another swipe at what he calls “weak statistical analysis of election returns.” One more caller insists that there were hints of fraud, and the show’s host interjects that the Kerry campaign’s failure to contest the results should satisfy her that there was no malfeasance. Where there is no official debate, reporters see no debate at all.

At least one other statistical analysis concludes that an investigation is necessary. A study done at UC Berkeley compared the 2004 results in Florida with those in 2000.[5] The study concluded that Bush got 130,000 to 260,000 extra votes in heavily democratic urban counties with electronic voting machines. Here as well, pundits and self-appointed experts ignore or dismiss this study. With no debate between powerful players, the media limits this discussion to how best to deal with “isolated” problems. When these parameters are challenged, the “internet conspiracies” are discussed shallowly and compared unfavorably with the non-debate between the two dominant political parties.

Human Stories: Bev Harris and a Huge Reward
Bev Harris directs Black Box Voting, Inc., a nonprofit organization dedicated to “consumer protection for elections.”[6] The group, one of the loudest pre-election critics of unauditable direct recording electronic (DRE) voting machines, has conducted the largest Freedom of Information action in history. Unlike the implicit charge that fraud is conceivable based on statistical analyses, Black Box Voting has alleged that election fraud absolutely did take place. “It’s okay to use the ‘F’ word,” they argue, and the website has several preliminary reports to indicate that such a charge needs to be investigated.[7] Harris and others are in Florida right now, investigating on their own time with money largely raised by PayPal donations. “While some Florida counties have been attentive to the public interest and have promptly complied with our public records requests, … other counties have stalled, stonewalled, failed to comply in a timely manner, or outright refused to provide the records.”[8] In Volusia County, for instance, officials refused to turn over proper copies of voting machine tabulations; in a strange turn of events, Harris’s team found some of the original, signed vote tallies in the garbage and destined for the shredder. This is just one tidbit of a long-running saga that, if Harris is right, could send shockwaves through US politics for decades. Even if she is wrong, this story is more important to more people than, say, the Scott Peterson verdict.

If Harris is right, she and her team will have more than PayPal donations for their efforts. Consider another newsworthy, utterly invisible effort to uncover fraud: “On November 6, 2004, Justice Through Music,, posted a $100,000 reward for specific evidence of vote fraud in the presidential election. On November 16, that reward was doubled to $200,000 in light of the rampant voter irregularities reported online and in the media over the past two.”[9] Even though both of these stories (especially the Black Box investigation) have been staples in the blogosphere, no major television or print sources have covered them at all. Even within the election reform movement, Black Box Voting has little political power; Justice Through Music was not even part of the discussion until this month. This lack of political power therefore supports the non-coverage of an investigation and a cash reward of historic proportions.

Future Research and Beyond
The most important research question here is what actually happened on Election Day; most communication researchers are ill equipped to do this. The US needs investigative journalists with mainstream political capital to sort through all the smoke and tell us about the fire. Yet communication researchers can contribute in at least two significant ways. First, those who have experience with quantitative methodology (especially surveys) and data analysis can add to the statistical and methodological debate. Second, communication scholars with experience analyzing news content could help unpack the process that has kept the “F” word off the air. This paper is a preliminary analysis at best, and I bring far more political interest than scholarly gravitas.

We can do the most as a community of media critics and citizens by raising our voices loudly to demand fair media coverage, electoral reform, and free access to exit polling data. We must act as citizens today because the process of scholarly production moves at a glacial pace. If one’s goal is to write a paper about this for beefing up one’s curriculum vitae, it will almost certainly not be published before the election system for 2006 is finalized. That is too long to wait when the integrity of our voting system is in doubt. As media critics and scholars, we have the professional credibility to insist that even preliminary evidence of malfeasance is newsworthy; let us use what mainstream credibility we have together, even if it means saying the “F” word on television.

W. Lance Bennett, “Toward a Theory of Press-State Relations in the United States,” Journal of Communication 40 (1990): 106.
David Levin, “Structure of News Coverage of a Peace Process: A Test of the Indexing and Zero-Sum Hypotheses,” Press/Politics 8:4 (2003): 9.
Steven F. Freeman, The Unexplained Exit Poll Discrepancy, 2004 (23 November 2004).
Radio Times, Hour One, 18 November 2004 (23 November 2004).
Michael Hout, Laura Mangels, Jennifer Carlson, Rachel Best, Working Paper: The Effect of Electronic Voting Machines on Change in Support for Bush in the 2004 Florida Elections (23 November 2004).
Black Box Voting, Black Box Voting, 2004 (23 November 2004).
Justice Through Music, $200,000 REWARD for evidence of vote fraud in the presidential election (23 November 2004).

CNN Election Results
Exit Polls From Edison Media Research and Mitofsky International
Black Box Voting

Please feel free to comment.