Peter Bergen and Katherine Tiedemann produce some data:
Our analysis of the drone campaign is based only on accounts from reliable media organizations with substantial reporting capabilities in Pakistan.
We restricted our analysis to reports in the New York Times, Washington Post and Wall Street Journal; accounts by major news services and networks -- the Associated Press, Reuters, Agence France-Presse, CNN and the BBC -- and reports in the leading English-language newspapers in Pakistan -- The Daily Times, Dawn and The News -- as well as those from Geo TV, the largest independent Pakistani television network. (Links to all those individual reports can be found here.)
The news organizations we relied upon collectively for our data cover the drone strikes as accurately and aggressively as possible. And though we don't pretend that our study is accurate down to the last civilian death in every drone strike, we posit that our research has generated some quite reliable data on the number of militant leaders killed, a fairly good estimate of the number of lower-level militants killed and a reliable sense of the real civilian death rate.
It would be fascinating to see a comparison with the militant/civilian ration in Gaza. But 2008 seems to me a clear case of disproportionate civilian casualties - and the ratio seems to have improved under Obama. I suspect that if we want to look at why the US is losing the war, we start with those civilian casualties in 2008. You can't stop an insurgency by murdering civilians. Neither in Gaza nor in Afghanistan.