Cooperative strategies outside the laboratory — evidence from a long-term large-N-study in five countries

Abstract

Research on human cooperation as an evolutionary adaptation is largely based on results from the laboratory, e.g. public goods games. However, it is debated whether these results of human cooperative behavior extend beyond such settings and whether they are valid in other contexts. Critical issues include the absence of context, the very short period of play and possible observer effects. This article presents data from an alternative controlled, but context-rich setting – a public goods game in an online browser game – with around 18,000 players from five countries over a period of ten months without observer effects. This article focuses on the robustness of previous findings about cooperative strategies and whether different types of cooperative behavior extend beyond the laboratory setting. Thus, the data presented provides external validity to existing laboratory experiments. The results suggest some important qualifications to previous work, since less high cooperators and a differing proportion of conditional cooperation are found. However, this data confirms the reported proportion of free riders at about 25% of the population. Cooperative strategies appear to be stable over time, individually fixed and rather independent of environmental parameters.