Background: In February, 2007, the Associated Press (AP) conducted a poll of 1,002 adults in the United States about their attitudes towards the war in Iraq. Respondents were remarkably accurate estimating the current death toll of US soldiers, yet were grossly inaccurate in estimating the current death toll of Iraqi civilians. We conducted a search of newspapers reports to determine the extent of the discrepancy between reporting Coalition and Iraqi civilian deaths, hypothesizing that there would be an over-representation of Coalition deaths compared to Iraqi civilian deaths.
Methods: We examined 11 U.S. newspapers and 5 non-U.S. newspapers using electronic databases or newspaper web-archives, to record any reports between March 2003 and March 2008 of Coalition and Iraqi deaths that included a numeric indicator. Reports were described as "events" where they described a specific occurrence involving fatalities and "tallies" when they mentioned the number of deaths over a period of time. We recorded the number of events and tallies related to Coalition deaths, Iraqi civilian deaths, and Iraqi combatant deaths
Results: U.S. newspapers report more events and tallies related to Coalition deaths than Iraqi civilian deaths, although there are substantially different proportions amongst the different U.S. newspapers. In four of the five non-US newspapers, the pattern was reversed.
Conclusion: This difference in reporting trends may partly explain the discrepancy in how well people are informed about U.S. and Iraqi civilian fatalities in Iraq. Furthermore, this calls into question the role of the media in reporting and sustaining armed conflict, and the extent to which newspaper and other media reports can be used as data to assess fatalities or trends in the time of war.