At a time when behavioural economics has assumed centerstage as one of the most discussed areas in economics, Chris Dillow points to two papers that caution against indiscriminate generalizability of findings about human psychology and behvaiour based on narrow samples from Western societies. Both papers draw attention to the considerable variations in human nature on various basic domains across cultures that have the potential to generate different equilibriums when economics and psychology intersect.
Joseph Henrich, Steven J. Heine, and Ara Norenzayan, reviewed human responses to visual perception, fairness, spatial reasoning, moral reasoning, thinking‐styles, and self‐concepts, and finds "substantial variability in experimental results across populations" in these domains. For example, results of ultimatum game experiments differ hugely across tribes, with Americans making much higher offers than the Hadze in Tanzania or Tsimane in Bolivia.
Roberta Dessi and Xiaojian Zhao examined numerous studies in behavioural psychology that contrasts North Americans to Japanese and found that while North Americans value self-esteem and are prone to overconfidence, the Japanese are much less so to both. In contrast, shame appears to play a much more important role among Japanese than North Americans. It means that Americans and Japanese will respond differentially to incentives that plays on self-esteem and shame. Americans are more likely to respond positively to incentives that are tailored around promotion of self-esteem whereas Japanese are likely to react similarly to incentives structured around avoiding shame.
Does this mean that, fundamentally, all cognitive biases will have to be filtered through the lens of human cultural traits before proclaiming their universal applicability? Many commonplace cognitive biases do have a universal resonance. Human beings are not known for their ability to correctly predict what we would want in the future; are subject to self-control problems; are more averse to losses than attracted to similar sized gains; have problems evaluating complex choices; struggle to take decision when faced with an over-load of choices, and so on.
However, there are other findings which could end up going awry when filtered through certain cultural environments. One possible area relates to the outcome of the interaction between incentives and human behaviour, especially in the magnitude of incentives required to generate the desired outcomes. The outcome of monetary incentives for kidney donations could vary widely depending on respective cultures. Similarly, incentives to reward pro-social behaviour (like getting children to help at home) could generate a broad spectrum of outcomes based on the host cultural environments.
On same lines, studies have found that human response to financial incentives for behavioural changes vary depending on their respective socio-economic backgrounds. Jo Phelan and Bruce Link have found that those in lower socio-economic classes have poorer smoking cessation rates when responding to financial incentives.