Department of Statistical Science
Duke University

presents:

Joseph Lipscomb
Sanford Institute of Public Policy, Duke University

"Time Preference For Health In Cost-Effectiveness Analysis"

Abstract:

The federal government's Panel on Cost-Effectiveness in Health and Medicine recently recommended that:

(1) In cost-effectiveness analyses from the societal perspective, both the costs and health consequences of competing programs should be expressed in terms of their present value to society.

(2) In particular, monetary outcomes should be discounted to present value at a rate consistent with the shadow-price-of-capital (SPC) approach to evaluating public investments. Current empirical evidence is consistent with adoption of a real, riskless rate of 3% in the base case.

(3) Health consequences should be discounted at the same rate (and using the same standard exponential formula) as costs. Doing so guarantees that the marginal rate of substitution between health and wealth is time invariant -- leading to health resource allocation recommendations that are "time neutral" across cohorts and generations.

(4) However, because both the theory and empirical evidence regarding the relation between individual rates of time preference for health and market rates of interest remain unsettled, sensitivity analysis is strongly recommended. Specifically, one can consider certain two-stage procedures that acknowledge individual rates of time preference within cohorts while continuing to reflect societal time preference across cohorts.

In this discussion, the theoretical and empirical arguments leading to these recommendations will be critically reviewed.

December 6, 1996

4:00 pm - 5:00 pm

116 Old Chem Building