Cross-Cultural Variation in Cooperation: A Meta-Analysis

Abstract

Impersonal cooperation among strangers enables societies to create valuable public goods, such as infrastructure, public services, and democracy. Several factors have been proposed to explain variation in impersonal cooperation across societies, referring to institutions (e.g., rule of law), religion (e.g., belief in God as a third-party punisher), cultural beliefs (e.g., trust) and values (e.g., collectivism), and ecology (e.g., relational mobility). We tested 17 pre-registered hypotheses in a meta-analysis of 1,506 studies of impersonal cooperation in social dilemmas (e.g., the Public Goods Game) conducted across 70 societies (k = 2,271), where people make costly decisions to cooperate among strangers. After controlling for 10 study characteristics that can affect the outcome of studies, we found very little cross-societal variation in impersonal cooperation. Categorizing societies into cultural groups explained no variance in cooperation. Similarly, cultural, ancestral, and linguistic distance between societies explained little variance in cooperation. None of the cross-societal factors hypothesized to relate to impersonal cooperation explained variance in cooperation across societies. We replicated these conclusions when meta-analyzing 514 studies across 41 states and nine regions in the United States (k = 783). Thus, we observed that impersonal cooperation occurred in all societies – and to a similar degree across societies – suggesting that prior research may have overemphasized the magnitude of differences between modern societies in impersonal cooperation. We discuss the discrepancy between theory, past empirical research and the meta-analysis, address a limitation of experimental research on cooperation to study culture, and raise possible directions for future research.

Did cooperation among strangers decline in the United States? A cross-temporal meta-analysis of social dilemmas (1956–2017).

“…Furthermore, the study samples were mostly college students, and college demographics have changed in the United States over the past several decades, to include more women and minorities (Anderson, 2003; Becker et al, 2010). That said, prior research has not found that cooperation varies by gender (Balliet et al, 2011; Spadaro, Jin, et al, 2022) or according to ethnicity in the United States (meta-analytic summary provided in the CoDa; Spadaro et al, in press). In addition, we found no significant difference in cooperation between student and nonstudent samples, and that the overall pattern of an increase of cooperation over time did not change when controlling for student sample (see the ).…”