High cost pool or high cost groups—How to handle high(est) cost cases in a risk adjustment mechanism?

Abstract Competitive social health insurance systems (at least) in Western Europe have implemented systems of morbidity based risk adjustment to set a level playing field for insurers. However, many high cost insured still are heavily underfunded despite risk adjustment, leaving incentives for risk selection. In most of these health care systems, there is an ongoing debate about how to deal with such underpaid high cost cases. This study develops four distinct concepts by adding variables to risk adjustment or by setting up a high cost pool for underpaid insured besides the risk adjustment system. Their features, incentives and distributional effects are discussed. With a data set of 6 million insured, performance is demonstrated for Germany. All models achieve a substantial improvement in model fit, measured in terms of R2 as well as CPM. As the results of the various models are different in different dimensions, the trade-offs that have to be dealt with and should be addressed, when implementing a model to reduce underfunding of high cost cases.