Reports of Palestinian civilian casualties during Israel’s Operation Protective Edge are known to lack objectivity because they are provided by Palestinian sources, which have incentives to exaggerate total numbers and misclassify combatants as civilians. Reports from the BBC and New York Times utilize simple statistical methods to show that misclassification is almost certain to be endemic, rendering these reports both meaningless and deceptive.
The New York Times’ Jodi Rudoren reports results of an analysis of fatality data obtained from the Palestinian Ministry of Health covering the period July 6-31:
The Times analysis, looking at 1,431 names, shows that the population most likely to be militants, men ages 20 to 29, is also the most overrepresented in the death toll: They are 9 percent of Gaza’s 1.7 million residents, but 34 percent of those killed whose ages were provided. At the same time, women and children under 15, the least likely to be legitimate targets, were the most underrepresented, making up 71 percent of the population and 33 percent of the known-age casualties.
The graph on the right displays these data. As a statistical matter, it is extremely unlikely that among civilian deaths 20-29 years old such a large proportion would be male. Thus, it is virtually certain that the Palestinian Health Ministry misclassifies many (and perhaps most) combatants as civilians.
Anthony Reuben, head of statistics for the BBC, also recognizes the disproportionate number of young adult makes among purportedly civilian casualties and finds explanations provided by the United Nations and various NGOs unpersuasive:
[I]f the Israeli attacks have been “indiscriminate”, as the UN Human Rights Council says, it is hard to work out why they have killed so many more civilian men than women.
Indeed, if these data were true, Israel attacks must be highly discriminate — i.e., targeted at male civilians 20-29 years old. To be indiscriminate is to be indistinguishable from random, and a strategy of randomness very nearly minimizes military effectiveness.
As to the casualty data, Both Rudoren and Reuben implicitly excuse the United Nations from responsibility for disseminating erroneous data as if they have no responsibility for information quality. In contrast, since 2001 agencies of the US Government have been required to ensure that the information they disseminate is objective, transparent and reproducible, as set forth in guidelines issued by the Office of Management and Budget. Thus, the Department of State is required to report casualty data objectively, even if the UN is not. Moreover, if the State Department were to disseminate UN figures in a way that conveyed the impression of endorsement, it would be obligated to validate the data before disseminating them.