Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
We study a model in which poor individuals of different types request a service and may suffer from the discrepancy between the service provider's beliefs and their true type. The utility loss resulting from such a discrepancy is referred to as an identity cost. Types are private information but service providers ideally prefer to favour individuals considered high type. They may choose to condition social service on an observable characteristic that is correlated with types but that can be manipulated by low type individuals at a cost. We show that whether conditionality enhances social welfare depends on this cost. If the cost is too low, high type individuals refrain from requesting the service because of the resulting identity cost, bringing about a so-called non-take-up equilibrium. In this case, unconditionality of the service provision is a Pareto improvement.