Is the bad apple or the bad basket to be blamed? Distinguishing the role of rational self-interest and conformity during the contagion of dishonesty

Learning from others' dishonesty involves motivations like pursuing self-interest and conforming to a dishonest group norm. However, these processes are rarely distinguished, hindering our understanding of how individual dishonest acts spread in a community. Through ethnographic analysis, we found that self-interest and dishonesty co-evolved in pre-industrial societies. We built a comprehensive model based on reinforcement learning (RL) to explore self-interest and dishonesty conformity, examining their role in dishonesty contagion among diverse samples (N=2349). Our findings show that self-interest is the primary driver of dishonest behavior propagation, while the effect of dishonesty conformity is context-dependent and more evident in those with lower self-interest propensity. In an fMRI experiment, we found that the cognitive control network mediated dishonest decisions when participants utilized dishonesty conformity to serve their self-interest. Frequent cheaters exhibited stronger activation in the vmPFC and anterior superior temporal sulcus when learning about the dishonest group norm, reflecting reward valuation and group norm interpretation.. DCM analysis identified the dACC as a hub that integrates these different cognitive processes involved when people learn from others’ dishonesty, and revealed the distinct neural circuits that underline the two processes of dishonesty.