Privatization of credence goods : theory and evidence from service contracting

This paper analyzes public sector contracting of credence goods, i.e., services for which the producer has private information whether a certain treatment is needed or not. I develop a model where a credence good is supplied by an agent who can invest in quality and exert effort to reduce costs. If quality is important but non-contractible, public agencies cannot commit to contracts with private firms that induce a truthful diagnosis. Privatization therefore increases costs due to overtreatment. In contrast, optimal contracts under in-house production entail weak financial incentives to induce treatment, implying that public sector managers use their private information to avoid unattractive tasks. I test the model on a data set of residential youth care facilities and find support for the theoretical predictions: Total cost is twice as high in private facilities due to a much longer length of stay. The fact that a teenager is troublesome has a much stronger negative impact on outcomes in public facilities. -- Privatization ; public sector contracting ; credence goods ; incomplete contracts ; contracting out ; residential youth care ; juvenile delinquency