Discrimination is reduced when group identity can be falsified

Studies show that Democrats and Republicans treat copartisans better than they do non-copartisans. However, party affiliation is different from other identities associated with unequal treatment. Unlike race or gender, people can more easily falsify, i.e., lie about or conceal, their party affiliation. We use a behavioral experiment to study how people allocate resources to copartisan and non-copartisan partners when partners are allowed to falsify their affiliation and may have incentives to do so. When affiliation can be falsified, the gap between contributions to signaled copartisans and signaled non-copartisans is eliminated. This happens in part because some participants---especially strong partisans---suspect that partners who signal a copartisan affiliation are, in fact, non-copartisans. Suspected non-copartisans earn less than both partners who signal that they are non-copartisans and partners who withhold their affiliation. The findings reveal an unexpected upside to the availability of falsification: at the aggregate level, it reduces unequal treatment across groups. At the individual-level, however, falsification is risky.