This working paper (PDF) on discrimination against Muslims by Claire Adida, David Laitin and Marie-Ann Valfort is interesting in the wake of today’s decision by France to ban the full veil. The authors set up a set of experiments (using various twists on the dictator game) to try to figure out what drove French discrimination against Muslims in the workplace and elsewhere. One might attribute this discrimination either to rational beliefs about Muslims, or to innate bias. It might be that French people were disinclined to cooperate with Muslims because they rationally feared that their cooperation would not be reciprocated – a set of beliefs that might perhaps have come about as a result of discrimination but not be reducible to them (those who are discriminated against may reasonably have less interest to reciprocate trusting behavior on the rare occasions that it is extended to them, as Russell Hardin’s account of trust makes clear). It might alternatively be that French people discriminate against Muslims because of a ‘taste’ for discrimination – in less anodyne terms, because they don’t like being around them.
Adida, Laitin and Valfort conduct a variety of field experiments, and find that ‘rooted’ French people (those with four native grandparents) are less likely to be generous when the ‘salience’ of Muslims in the group increases. They furthermore suggest that this is best explained by a taste for discrimination rather than some rational system of beliefs about how Muslims will reciprocate or not reciprocate towards them. Finally, they suggest that this discrimination by French people seems to go together with a norm – that is, it is in part driven by the belief that other French people will also discriminate more against Muslims. The latter may account in part for the (to my mind rather less convincing) discussion of how this impacts job market discrimination.
Adida, Laitin and Valfort suggest that employers discriminate against Muslims because of their beliefs that more Muslims in the workplace will lead to lower workplace efficiency, because of French workers’ animus towards them. Occam’s razor would suggest that without further direct evidence, it is simpler to assume that much of the time French employers will share the motivations of other French people – that is, that they will frequently be directly biased against Muslims themselves. Notably, this bias seems distinguishable from simple anti-black racism – the experiments find differences in how French people respond to Senegalese immigrants when they are presented as being Muslim and Christian. Obviously, this does not make this apparently irrational bias any more normatively defensible than racism simple.