How the media chooses to report the numbers of dead and dying in Iraq may play a role in shaping the public's knowledge about the consequences of the war. The AP poll showed that respondents were substantially more accurate in identifying the number of Americans who had died as a result of the Iraq war than Iraqis. There are several striking findings in our study. First, by even the most conservative estimates, at least 20 times as many Iraqis have died as Coalition members, but in virtually all American papers there are far more articles describing coalition death events and tallies. Second, this relationship was reversed for three of the four Middle-Eastern papers and the one British paper included. Third, there are substantial differences between U.S. newspapers.
To the best of our knowledge, there is no other study that has quantified how the media has represented deaths in the time of war, or compared US newspapers to newspapers in the Middle East. The discrepancies we have observed should encourage further study into how media representation affects public knowledge, attitudes towards wars, and public debates about the consequences of a war. Our study cannot and does not explain how or why there is a variance in reporting and nor does it measure accuracy per se. Certainly all newspapers can only report some fraction of killings in a country as vast and violent as Iraq. But specific factors seem to be influencing the level and the relative reporting of victims from the different nationalities involved. These factors may include: local interests; local sensitivities, such as residents who have lost relatives; wire service affiliations; access to embedded journalist reports; the economic and political agendas of the various editorial boards, and other political pressures.
This study poses a challenge to journalism. Mark Twain once said to a class of Army cadets, "By the etiquette of war, it is permitted to no one below the rank of newspaper correspondent to dictate to the general in the field." [6] While his joke suggests that journalists tend to assume an expertise they may not deserve, there is an important point contained therein: in a democracy, the military operates on behalf of the people, who in turn are in a large part informed by their media outlets. Journalists risk their lives to report on wars and face numerous ethical challenges - from the ethics of what to report to the practicalities of being embedded. However, despite what individual journalists may achieve, there can be an overall effect from the media as a whole, and our study appears to show that Americans will likely be exposed to one story (of American fatalities) with much less information about another story (of Iraqi fatalities). Given the discrepancy between the U.S. and the non-U.S. newspapers, it would appear that political and cultural factors may be playing a role in how these stories are being told.
This study has a number of significant limitations. We cannot assume that media reports of death events/tallies correlates with overall knowledge, especially if people obtain their news from numerous sources. Respondents do not only get their news from newspapers. At about the same time as the AP poll was conducted, Harris Interactive conducted a poll in which they asked about how Americans get their news [7]. The vast majority (77%) said they watch local broadcast news and 71% said they watched network broadcasts or cable news several times a week or daily. 63% reported reading a local daily newspaper and 18% read a national newspaper several times a week or daily. 64% go online to get news (which may be from newspapers online). Thus, a limitation of our study is a focus on newspapers, rather than broadcast news, and our exclusion of national newspapers such as Herald Tribune or USA Today. A review of TV and radio reports would be far less easily quantified because length of segments varies widely and visual images of killed or injured may be far more memorable or exert more influence.
In terms of study design, this study involved a number of participants each tallying scores from a newspaper; this required individual judgment as to what constituted an "event" and a "tally"; discussions amongst the participants, usually over possibly ambiguous reports, suggested that although this remains a limitation, the reliability and validity of these terms was high with little variation between reviewers. Furthermore, while it is not typical of study designs to have a flexible set of "definitions", we feel this review method captures the at-home readers' experience more robustly and best interprets the narrative component of war reporting.
While we believe our search methods capture a majority of the articles written about deaths in the Iraqi war, our search language did not include every mention of the war. It excludes non-specific, general remarks such as "the civilian death toll is rising". We explored many different search terms initially and these two pairs gave us a comparatively high number of hits with a low fraction of articles with no events or tallies. Of note, we were not expecting to capture every article, just the vast majority with a consistent process that could detect trends over time and between papers. Finally, it is worth remembering that other means of communicating in the newspaper, including editorials and photographs, may be discussing or evoking the war compellingly but without the numeric indicators that would place them within our study.