A March 15 piece in the Washington Post tells us that the UN's special human rights envoy found that the CIA's drone strikes in Pakistan violate that country's sovereignty. It also told readers that the drones had "resulted in far more civilian casualties than the U.S. government has recognized."
Unfortunately, that message was muddled by reporter Richard Leiby's he said/she said approach to the question of civilian deaths:
Estimates of total militant deaths and civilian casualties vary widely. Independent confirmation is difficult in part because the strikes often occur in remote, dangerous tribal areas where Taliban insurgents and Al-Qaeda and its allied militants are active.
The Bureau of Investigative Journalism in London has estimated that at least 411 civilians–or as many as 884–were among some 2,536 to 3,577 people killed in the CIA strikes in Pakistan. But Sen. Dianne Feinstein (D), who chaired the Senate Select Committee on Intelligence hearings last month that confirmed new CIA Director John O. Brennan, put the number of civilian deaths considerably lower.
"The figures we have obtained from the executive branch, which we have done our utmost to verify, confirm that the number of civilian casualties that have resulted from such strikes each year has typically been in the single digits," she said.
So, on the one hand, the Bureau has done extensive work documenting drone strikes. But then again you have a senator who heard from the government that it's much lower.
There is, of course, a way to report the difference between Feinstein's claim and other estimates. Conor Friesdorf did so in the Atlantic (2/11/13), contrasting the Bureau's totals with those of the New America Foundation and other researchers. None of these projects supports Feinstein's claim. His conclusion:
There is no reason to treat Feinstein's claim about civilians killed as if it is credible. All the publicly available evidence is arrayed against her position.
Yet she's treated by the Post as one of two sides of the drone deaths debate.