Tortured beliefs: How and when prior support for torture skews the perceived value of coerced information

In the wake of recent revelations about US involvement in torture, and widespread and seemingly-growing support of torture in the US, we consider how people judge the value of information gained from informants under coercion. Drawing on past work on confirmation biases and moral judgments, we predicted, and found, that American torture supporters are more likely than opposers to see coerced information as relatively valuable and necessary in a scenario describing the foiling of an al-Qaeda terrorist attack. Judgments of coerced information value in the scenario also predicted endorsement of using the episode as a “success story” to justify torture in future cases. A second study shed light on an important boundary: Prior general support for torture predicted the perceived value of coerced information when the interrogated informant was an outgroup member (an al-Qaeda informant tortured by US operatives) but not when the informant was an ingroup member (an American soldier tortured by al-Qaeda). Overall, the results suggest that advocates for torture may readily interpret ambiguous evidence as implying the value and necessity of extreme interrogation techniques when used by the ingroup. Our findings also indicate that torture supporters often expect selective efficacy, whereby they see torture as more likely to yield valuable information when it is used by “us” compared to “them.” •US adult judgments of terrorism scenarios were skewed by prior support for torture.•Those previously supporting torture saw coerced information as more valuable.•Torture opposers did not show a bias for or against coerced information.•Results also revealed a “selective efficacy” boundary concerning informant identity.•Supporters privileged coerced information from outgroup but not ingroup informants.