Andrews, Edmund L. "Census Bureau to Dismiss Analyst Who Estimated Casualties," New


Andrews, Edmund L. "Census Bureau to Dismiss Analyst Who Estimated Casualties," New
York Times, 7 March 1992, p.9.

Beth Osborne Daponte, a data analyst with the Census Bureau, has been fired by the Bureau for
publishing materials which had not been "peer reviewed." The data published are considered
very sensitive by the Bush Administration especially now, during his reelection campaign. Here
he is trying to look good to the American people and she blows the lid on his lie that there were
but few civilian casualties during the Iraq War. By her estimate: 40,000 Iraqi solders and 13,000
civilians died in direct military conflict. "In addition, she estimated that 30,000 people had died
during the Shite and Kurdish rebellions after the war and that 70,000 people from health
problems caused by the destruction of water and power plants."

Daponte gave her estimates to a reporter for the Associated Press in January.

Her sources have been investigated by the Bureau. The conclusion is simple: "We just can't have
people saying that they know best and releasing things without defending their work in from of
the peers."

"In January, the Census Bureau discarded Ms. Daponte's draft report and replaced it with one
prepared by her superior, Frank Hobbs, chief of the population studies branch. That report
matches Ms. Daponte's estimates on the number of direct military deaths during the war, and the
number of deaths caused by post-war violence. The chief differences are that it sharply reduces
her estimate for the number of civilian deaths and omits some 5,000 deaths that Ms. Daponte
estimated occurred in military action after the war."

"The Census Bureau said today that it intends to dismiss an analyst who estimated that 13,000
Iraqi civilians were killed by American-led forces during the Persian Gulf War, an estimate that
was more than twice as high as the bureau's subsequent official calculation."

The Politics of Numbers are absolutely fascinating. And here is a bureaucratic bit of terrorism: if
the analysis provides figures which are embarrassing, fire the analyst and emend the numbers,
especially in an election year.