Journal of Health Economics 28 (2009) 1039–1047 Contents lists available at ScienceDirect Journal of Health Economics journal homepage: www.elsevier.com/locate/econbase The predictive validity of prospect theory versus expected utility in health utility measurement Jose Maria Abellan-Perpi ˜ nan a , Han Bleichrodt b,∗ , Jose Luis Pinto-Prades c a Department of Applied Economics, University of Murcia, Murcia, Spain b Erasmus School of Economics & iMTA/iBMG, Erasmus University Rotterdam, Rotterdam, Netherlands c Department of Economics, University Pablo d’Olavide & Fundación Centro de Estudios Andaluces, Andaluces, Sevilla, Spain article info Article history: Received 20 November 2007 Received in revised form 23 July 2009 Accepted 10 September 2009 Available online 20 September 2009 JEL classification: D81 I10 Keywords: Health utility measurement Prospect theory Expected utility Standard gamble Time trade-off abstract Most health care evaluations today still assume expected utility even though the descriptive deficiencies of expected utility are well known. Prospect theory is the dominant descriptive alternative for expected utility. This paper tests whether prospect theory leads to better health evaluations than expected utility. The approach is purely descriptive: we explore how simple measurements together with prospect theory and expected utility predict choices and rankings between more complex stimuli. For decisions involving risk prospect theory is significantly more consistent with rankings and choices than expected utility. This conclusion no longer holds when we use prospect theory utilities and expected utilities to predict intertemporal decisions. The latter finding cautions against the common assumption in health economics that health state utilities are transferable across decision contexts. Our results suggest that the standard gamble and algorithms based on, should not be used to value health. © 2009 Elsevier B.V. All rights reserved. 1. Introduction A crucial ingredient in health care evaluation is measuring the utility of health states. There is much ambiguity in the existing measurement methods with different methods that have the same prima facie plausibility leading to systematically different results. A danger of this divergence in measured utilities is that health care evaluations come to depend on the measurement method used. This would not be a cause of concern if it were known which method should be preferred. Unfortunately, no consensus exists on this question. Most health care evaluations today still assume expected util- ity. For example, the standard gamble (SG), a widely used method for measuring health utility, is based on expected utility. Expected utility was the leading normative and descriptive theory of decision making in the 1980s when economic evaluations of health care took off. Since then, many descriptive deficiencies of expected utility have been documented (Starmer, 2000). It is now widely accepted ∗ Corresponding author. E-mail addresses: dionisos@um.es (J.M. Abellan-Perpi ˜ nan), bleichrodt@ese.eur.nl (H. Bleichrodt), jlpinpra@upo.es (J.L. Pinto-Prades). that expected utility is not valid as a descriptive theory of human decision making. Evidence of violations of expected utility for health outcomes includes Llewellyn-Thomas et al. (1982), Rutten- van Mölken et al. (1995), and Stalmeier and Bezembinder (1999). Bleichrodt (2002) argued that the violations of expected utility imply that the SG overestimates the utility of health. Bleichrodt et al. (2007) provided empirical evidence to sustain this argument. The descriptive deficiencies of expected utility have led to a multitude of alternative theories of decision under risk. Of these theories, prospect theory is currently the most important. Because prospect theory is descriptively superior to expected utility, it can lead to better health evaluations. Bleichrodt et al. (2001) proposed adjustments of existing health valuation methods under risk, in particular the SG, based on prospect theory. They showed that their adjustments lead to higher internal consistency in the sense that theoretically equivalent methods give similar results. Further sup- port for their adjustments is in van Osch et al. (2004), van Osch et al. (2006), and Bleichrodt et al. (2007). None of these studies examined the external validity of prospect theory, i.e. whether the adjustments are better able to predict people’s choices than the tra- ditional methods based on expected utility. The studies exploring internal consistency were all based on a common response mode, typically matching. It is possible that the adjustments correct for a 0167-6296/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.jhealeco.2009.09.002