Poli-Sci Perspective is a weekly Wonkblog feature in which Georgetown University's Dan Hopkins and George Washington University's Danny Hayes and John Sides offer an empirical perspective on the issues dominating Washington. In this edition, Sides looks at the state of public opinion on drones. For past posts in the series, head here.

Rand Paul's filibuster on Wednesday drew renewed attention to the U.S. government's program of drone warfare. Paul's focus -- whether Obama believed that he could legally authorize a drone strike on a U.S. citizen on American soil -- ultimately earned a direct response from Attorney General Eric Holder.

But, as Dylan noted in his FAQ, the main targets of drones have been mostly foreigners living in countries like Afghanistan, Pakistan and Yemen. The irony, given all the attention and some plaudits given to Paul's filibuster, is that most Americans support the use of drones to fight terrorists abroad. While Paul inveighed against a hypothetical killing, the actual killings that do happen are not that controversial in the minds of most Americans. An open question, however, is whether their minds could be changed.

Only last month, the Pew Center asked a random sample of Americans whether they supported "the United States conducting missile strikes from pilotless aircraft called drones to target extremists in countries such as Pakistan, Yemen, and Somalia?" A majority, 56 percent, approved while 26 percent disapproved and 18 percent were not sure -- numbers similar to two 2012 polls.

In fact, drone strikes attracted roughly similar amounts of support from across the partisan spectrum: 68 percent of Republicans approved, as did 58 percent of Democrats and 51 percent of independents. A pattern of relative bipartisanship is not all that common in public opinion today, but it is predictable in this case. When leaders in the two parties don't really disagree on something, there is no reason for partisans in the public to disagree either.  In John Zaller's magisterial account of how public opinion is formed and evolves, he refers to a pattern of bipartisanship like this one as a "mainstream effect." Like it or not, drone warfare has become so common that "mainstream" does not sound inapt.

Thus, there is little reason to expect public opinion about the drone program to change without concerted and prolonged dissent from political leaders. That does not seem to be forthcoming. Paul's dissent -- which didn't even emphasize foreign targets of American drones -- was met with harsh rebuttals from Lindsay Graham, John McCain and the Wall Street Journal editorial page. Democrats were not exactly rushing to stand with Paul either.


Would dissent from Capitol Hill make any difference? Actually, it might. Some evidence suggests public support for drone warfare is soft. The Pew survey provides hints of that. The main concern about drones -- one that 53 percent of the public was "very concerned" about -- was civilian casualties, which occur with some regularity although drones may present a lower risk of civilian casualties than some other kinds of military action.

Even more telling is a survey experiment conducted by political scientist James Igoe Walsh in 2012. He described an U.S. attack on a terrorist training camp in one of four ways:

  • as a drone strike with a high chance of success, no U.S. military casualties, and no civilian casualties
  • as a strike with military personnel that had a high chance of success and no civilian casualties, but 25 U.S. military casualties
  • as a drone strike with a low chance of success, but no casualties
  • as a drone strike with a high chance of success and no U.S. military casualties, but civilian casualties on the ground

Relative to first scenario -- the successful mission with no casualties -- respondents were less supportive of all three of the other scenarios. This is to be expected, especially for missions with a low chance of success or that would result in American casualties.


But more surprising was that the public was the least supportive of the mission that would result in Yemeni civilian casualties. As Walsh notes, "This is a real surprise, since it means that respondents attach as much or more value on the lives of foreign civilians as they do on U.S. military personnel."

There is reason to treat these findings with some caution. It is just one experiment and the subjects were not a random sample of the American public. One also could imagine other scenarios: would the public be willing to tolerate civilian casualties if known terrorist leaders were killed? Nevertheless, this experiment's findings dovetail with the Pew survey, and suggest that majority support for drone strikes may be fragile precisely because they put civilians at risk.

But support is unlikely to soften if there is no one to lead the opposition. And when Republicans like Graham are saying that drone strikes against terrorists are "one of the highlights of President Obama’s first term," I wouldn't expect opposition leaders to emerge any time soon.

Further reading: