Venue: The Fuqua School of Business, Duke University, 1 Towerview Drive, Durham, NC 27708-0120

 

Presentation

Shifted deductibles for high risks: more effective in reducing moral hazard than traditional deductibles

Authors: Richard van Kleef (Erasmus University Rotterdam); Wynand van de Ven (Erasmus University Rotterdam); René van Vliet (Erasmus University Rotterdam)

Presenter: Richard van Kleef (Erasmus University Rotterdam)

Discussant: Roger Feldman (University of Minnesota)

Session: Cost-Sharing

Room: Seminar E

When: Monday 3:15 p.m. - 4:45 p.m.

Rationale: An important goal of deductibles in health insurance is to reduce moral hazard. Traditional deductibles, however, are not effective for high-risk individuals who know at the start of the contract period that their expenditures are going to exceed deductible amount d anyway. For these individuals we propose to shift the deductible range from [0,d] to [s,s+d] with starting point s depending on risk characteristics.

Objective: The goal of this paper is to theoretically and empirically illustrate the concept of shifted deductibles.

Theoretical concept: For high risks, e.g. those with a chronic disease, the effect of a moderate deductible on medical consumption will be low with s= €0, but also with s= €500,000. With s= €0 the probability of having maximum out-of-pocket expenditures is (very close to) 1; with s= €500,000 the probability of having no out-of-pocket expenditures is (very close to) 1. In both situations these individuals have no incentive to contain costs since lower costs will not result in lower out-of-pocket expenditures. Somewhere in between these extremes there will be more uncertainty about out-of-pocket expenditures, resulting in higher incentives for cost containment. A crucial question is 'How to find the optimal starting point?'. We assume the optimal starting point to be such that the variance in out-of-pocket expenditures is maximized.

Data and Method: We empirically illustrate the concept of shifted deductibles with administrative data from a Dutch insurance company. We use a four-step statistical procedure to find the optimal starting point at individual level: 1) Estimate an expenditure model where medical expenditures Y depend on relevant risk characteristics. Explanatory variables in this study are: age, gender and prior expenditures; 2) Calculate the expected value and variance of Y; 3) Calculate the expected value and variance of out-of-pocket expenditures for different values of s; 4) Find for each individual the starting point for which the variance in out-of-pocket expenditures is maximal.

Results: For the 10-percent highest risks in our data the optimal starting point (at individual level) of a €1,000-deductible is to be found (far) beyond €1,200. This corresponds to a deductible range of [€1200,€2200] or further. For this group, such an optimal shift reduces mean out-of-pocket expenditures by €400 and increases its variance to nearly five-fold. These effects are smaller with differentiation of the starting point at group level instead of the individual level. An age-related starting point, however, performs substantially better than a traditional deductible as well.

Conclusion: For high-risk individuals, shifted deductibles result in lower out-of-pocket expenditures and lower total expenditures (due to moral hazard reduction) than traditional deductibles. The level at which starting point s is differentiated implies a trade-off between transparency and incentives for cost containment.