QALYs and Policy Evaluation: A New Perspective

Document Type



This paper has been published as "QALYs and Policy Evaluation: A New Perspective," 6 Yale Journal of Health Policy, Law and Ethics 1 (2006).


“QALYs” (Quality-Adjusted Life Years) are a metric for health and longevity very widely employed by health researchers. Surveys are used to assign health states a quality ranking on zero-one scale, with zero representing a health state no better than death and one perfect health. The total QALY value of a health profile is calculated as the time spent in its component health states, each weighted by its quality. Until a few years ago, despite the huge academic literature on QALY measurement, this approach was seldom used by policymakers in the U.S. But there have been recent signs of governmental interest in QALYs, particularly by the FDA, which has now employed QALYs in almost twenty rulemakings.

This Article provides a new, welfarist defense of QALYs. Welfarists, who think that government should maximize social welfare, or that social welfare should at least be an important determinant of governmental choice, are often skeptical of QALYs. Why use a QALY scale rather than the willingness-to-pay (WTP) metric traditionally favored by economists? I answer that question by building on prior work with Eric Posner, which draws a distinction between Kaldor-Hicks efficiency and overall well-being. The sum of WTP amounts may correlate with Kaldor-Hicks efficiency, but this traditional cost-benefit procedure doesn’t perfectly track overall well-being. Under certain conditions, a QALY-based procedure can be a better proxy for overall welfare than traditional cost-benefit analysis (the sum of WTP amounts).

The Article therefore challenges the conventional wisdom among health economists that traditional cost-benefit analysis dominates QALY-based measurement. It also challenges the standard view of how QALYs might function in policy choice. Standardly, QALYs are seen as a measure of policy “effectiveness,” to be used in cost-effectiveness analysis. In cost-effectiveness analysis, health and longevity are measured on a nonmonetary scale such as QALYs, and cost-effectiveness ratios are used to identify the optimal choice. I reject this picture of QALYs. Rather, I argue, QALYs have a role to play in nontraditional cost-benefit analysis. The health and longevity impacts of a policy choice can be measured on a QALY scale, then converted to dollars using some conversion factor, and added to the monetized nonhealth effects of the choice, to determine its overall net monetized benefits. Indeed, the FDA has used QALYs in precisely this way: as an input to a cost-benefit analysis in which health and longevity are monetized by using some QALY-to-dollar conversion factor (for example, $300,000 per QALY), rather than as a component of cost-effectiveness analysis.

Part I of the Article reviews the current literature on QALYs. Part II shows that QALYs can be a better measure of the effect of health or longevity changes on overall welfare than the WTP scale. First, wealth effects, the “dead anyway” effect, and investment effects drive a wedge between the sum of WTPs and overall welfare, and QALYs can (under certain conditions) overcome these difficulties. Second, cognitive problems may interfere with the accurate elicitation of WTP amounts, and these problems can be circumvented (to some extent) by QALYs.

But QALYs are hardly a perfect welfarist measuring rod. Part III describes the limitations of QALYs. Part IV discusses the role that QALYs should play in policy choice, given their advantages as well as deficits relative to the WTP measure. I present a pragmatic view of the QALY-to-dollar conversion factor, and provide guidelines for choosing between WTPs and QALY-to-dollar conversions as an input to cost-benefit analysis. Among other things, my view of QALY-based cost-benefit analysis and WTP-based cost-benefit analysis as rough, proxy procedures for maximizing overall welfare helps illuminate a recent topic of great controversy: whether agencies should evaluate policy interventions that save lives by using a constant monetary “value of statistical life,” or rather by pricing life-years.

Date of Authorship for this Version

December 2005