Greater socioeconomic inequality is associated with higher crime rates. If this association is causal, it is unclear how the population-level variable, inequality, affects decisions to offend in individuals’ heads. I will present a recent theoretical model in which individuals strive to remain their resources above a threshold of desperation that is set by their social context. Greater inequality means more individuals who are at or below this threshold. It becomes rational for them to offend as a risky strategy to leap clear of it. This produces a link between population-level inequality and individual decision-making. Moreover, we show that increasing punishment severity under these assumptions should not generally expected to reduce offending. I present a framework for studying the assumptions and predictions of the model in a multi-player incentivized economic game. Preliminary data are consistent with the predictions of the model. However, they are also consistent with simpler but still relevant hypotheses that do not use the assumption of a desperation threshold, such as that loss compared to some mental reference point leads to frustration and anger. We are currently attempting to test between these alternatives. We hope that the experimental framework, regardless of which way the results fall out, is useful for understanding antisocial motivations.