Information and sorting in the market for obstetrical services

Using a statistical model and a partial equilibrium economic search model, we develop a methodology for appraising the value of consumer information about the quality of health care providers and apply it to information about physicians' predispositions to perform cesarean section deliveries. There are three primary results. First, information's value is roughly proportional to a simple statistical metric of its accuracy; the constant of proportionality can be imputed from knowledge of consumer search methods and consumer preferences. Second, the function governing the production of information from data 'inputs' can have surprising economic properties, such as economies of scale, that usefully inform the efficient production of consumer information. These properties are robust even when the value of information cannot be precisely determined. Third, information's value is enhanced by the way physicians are sorted into hospitals but greatly attenuated by the presence of other dimensions of physician heterogeneity. Under plausible assumptions, the value of information in our empirical application is small enough that it is not surprising that consumers do not utilize it when selecting a physician.