Physician Practice and MCO Negotiation: The Impact of Time Sensitive Supply and Demand

When health care providers and managed care organizations (MCOs) bargain, the main tool providers have is the threat to refuse to be in the MCO’s network. In fact, anecdotal evidence indicates that a major mechanism that practices employ to maximize profits in the face of differing insurer reimbursements, limited capacity and stochastic demand is to choose insurers discriminately. Providers do not accept patients from every MCO, however, providers do not exclusively accept the most profitable MCO. In this paper, I apply these institutional facts to a Nash cooperative bargaining framework to develop a bargaining model that explicitly models the provider’s disagreement point with the MCOs. In doing this, I am able to solve analytically for the interdependence of prices between MCOs and add to previous bargaining models by making the value of a MCO to a provider more explicit. This model shows the impact of MCO market structure on prices. By introducing provider capacity constraints, I am able to model two important provider-side considerations: the risk capacity will be unused, and the risk that a low paying patient will displace a higher paying patient. Neither of these two effects have been previously captured in the bargaining literature, which typically has featured marginal costs as the limiting factor for providers contracting with MCOs. I also show how predictions in my model match empirical observations and estimates from other work. I demonstrate a strong negative association between MCOs’ market power and negotiated prices, and show that the degree of market level price differences predicted by this model is similar to what has been observed. Finally, recent empirical work has found that that price increases for Medicare are positively associated with private MCOs’ prices and that this impact is stronger in areas with more concentrated insurers, and areas in which Medicare patients represent a larger share of the market. My model analytically makes these predictions and can explain the underlying mechanisms.