The New York Times misreported its own meeting with the President-elect concerning his views on torture, and how much he has changed his mind. Here I want to focus on one aspect of Mr. Trump’s statement in his meeting with the Times. He said: “If it’s so important to the American people, I would go for it. I would be guided by that.”

That is a mistaken and improper path to follow. Recent social science research suggests that Americans’ support for torture turns on prejudice toward the suspect. A study by a team of political scientists asked individuals whether they would support torture if a suspect were involved in terrorist-related activity. Respondents’ support for torturing the suspect surged when his name was “Arab-sounding” compared to when he had a “Caucasian-sounding name.” The same effect held true when the situation described did not involve terrorism, but instead an ordinary crime. (See the graph above).

During the study, respondents were asked whether they would “agree or disagree with the decision to use interrogation tactics that would inflict severe pain or suffering.” In the graph above, the vertical axis represents their responses on a 7-point scale (a value of 1 indicates “strongly disagree;” and value of 7 indicates “strongly agree”). The center lines in each colored box represent the average response for the specific question posed. (One of the authors of the unpublished study, Professor Will H. Moore, has written about some of the findings elsewhere. The graph above, however, was not included in Moore’s write-up. With the permission of Moore and his team, I have included it here.)

It is important to spotlight another dimension of the findings. Although the following results lacked statistical significance, in the study’s sample: note that respondents were more willing, on average, to support torturing the “Arab” man suspected of a common crime than torturing the “Caucasian” man suspected of terrorism. Wrap your head around that for a moment.

Without statistical significance, of course, that particular finding is just suggestive, and it would be worth other researchers following up with similar tests. (Please email the author of the study if you would like a copy of the manuscript and supporting information.)

In sum, there is a good reason we do not commit questions like torture to referenda and majority vote, and that actions like waterboarding are strictly prohibited by international and domestic law (see posts by David Luban and Alex Whiting). Rather than conducting online and emailed polls for their favored policy positions, Mr. Trump and his team would be better advised to be guided by the laws that bind his office and our nation’s military forces.

Image: Carlos Latuff via Wikimedia Commons
Chart: Will H. Moore et al.