This meta-analysis reveals that Americans’ level of cooperation among strangers has increased over the 61-year period 1956–612017. This finding challenges the idea that social capital and norms of cooperation have declined in the American society over time. Changes in American society support greater cooperation among strangers, which could underlie an ability for Americans to cooperate to solve present and future challenges (eg. solutions to climate change, pandemics, and sustainable resource consumption).
[supplement; OSF] Cooperation among strangers has been hypothesized to have declined in the United States over the past several decades, an alarming trend that has potential far-reaching societal consequences.
To date, most research that supports a decline in cooperation has relied on self-report measures or archival data. Here, we use the history of experimental research on cooperation in situations involving conflicting interests (ie. social dilemmas).
We meta-analyzed 511 studies conducted 1956–612017 with 660 unique samples and effect-sizes involving 63,342 participants to test whether the average level of cooperation observed in these studies had declined over time.
We found no evidence for a decline in cooperation over the 61-year period. Instead, we found a slight increase in cooperation over time. In addition, some societal indicators (eg. income inequality, societal wealth, urbanization level, and percentage of people living alone) measured 10–5 years prior to measures of cooperation were found to be positively associated with cooperation, suggesting that they may be potential societal underpinnings of increases in cooperation.
These findings challenge the idea that social capital and civic cooperation among strangers have declined in the United States over time, and we offer directions for future research to understand causes of an increase in cooperation.
[Keywords: cooperation, social capital, social dilemmas, meta-analysis, time]
Figure 1: Historical Changes Over Time in the Mean Cooperation Rate in Social Dilemmas. Note: The unstandardized regression coefficient (β) and p-value are from the simple meta-regression model (Model 1). The blue line represents average model predictions converted back from logits to cooperation rates. Data points represent study means and the size of the data point is proportional to study (inverse variance) weighting. Larger dots are equated with means that have a smaller variance. The shaded blue (gray) region indicates the 90% prediction intervals based on average model predictions and residual heterogeneity; on average, 90% of true cooperation rates will fall within this region.