Paluck, E. L., Green, S. A. and Green, D. P. (2019). The contact hypothesis re-evaluated. Behavioural Public Policy, 3(2), 129–158. doi:10.1017/bpp.2018.25. Preprint: SocArXiv, 24 May 2017, doi:10.31235/osf.io/w2jkf.
Read in full on 2026-10-29 (NOTE-1393).
Why this is here
The exchange cited it as the check on Pettigrew and Tropp: randomized contact studies generally reduce prejudice, but the evidence for adults’ racial or ethnic prejudice is limited. Disagreement, of degree: the abstract says not that the evidence is limited but that studies addressing adults’ racial or ethnic prejudice are absent, and separately that interventions on ethnic or racial prejudice show substantially weaker effects. It also stresses the 27 studies’ small number and the untested scope conditions. The exchange dated it 2019 (the issue); it was online in 2018 and posted in 2017. Bears on CLAIM-275, as the caution on how far the contact evidence carries.
Note of 2026-10-29: matched in Pettigrew and Tropp’s Table 8
Pettigrew and Tropp 2006, LIT-1413, is now read (NOTE-1407). This paper disagrees with its near-equality of effects across targets, and with its claim that the most rigorous studies give the largest effects. On its headline figures the 2006 paper does give racial and ethnic targets the same mean as the rest (−.218 and −.220). But its own Table 8 shows what this paper found: among experiments the effect is −.221 (k = 9) for racial and ethnic samples against −.377 (k = 27) for other targets, so the rigor gradient is carried by the other targets. Structured programs raise the effect significantly for the other targets and not for racial and ethnic ones. The weaker ethnic and racial effects among the more rigorous studies are therefore already in the 2006 data, though its abstract softens the point. The status is unchanged.