تعیین پارامترها در تابع رفاه اجتماعی (SWF) توسط داده های ترجیحی بیان شده: کاربردی برای سلامت

ABSTRACT Determining the parameters in a social welfare function using stated preference data: an application to health

An important consideration when establishing priorities in the public sector is the amount of benefit generated by alternative allocations. As a result, there has been considerable research effort devoted to developing technologies that allow the benefits from a range of public services to be measured and subsequently valued. If benefits were the only consideration, then the objectives of public policy could be defined in terms of the maximization of these benefits. However, policy-makers, as well as the general public, are also likely to be concerned with how benefits are distributed.

Although standard economic models assume that people do not care about inequalities, there is increasing interest in the economics literature in people’s preferences regarding fairness (Rabin, 1993; Fehr and Schmidt, 1999). These models focus on self-centred inequality aversion in the sense that people care only about their own outcomes relative to those of other people, but they do not care about inequality amongst other people per se. This has important implications for how Social Welfare Functions (SWFs) are constructed and how preferences are aggregated (see e.g. Quesada, 2003). In this article, we are concerned with an individual’s distributional preferences as they relate to the treatment of other people. Specifically, our inquiry is into whether an SWF can be constructed from people’s other-regarding, or social, preferences (Menzel, 1999). The SWF is, in principle, a powerful device for determining how best to balance these competing objectives of efficiency and equality. However, in practice, there has hitherto been only limited success in developing an SWF that is operationally useful.

There have been some attempts to estimate the parameters of SWFs from the stated preferences of individuals. For example, Amiel and Cowell (1999) have asked respondents to choose between different distributions of income across population groups. Similar attempts have been made to estimate the SWF for health, using relatively small samples of students (Dolan and Robinson, 2001). In this context, a policy that maximizes population health might be of relatively less benefit to less healthy groups, or a policy that reduces inequalities might forego the opportunity to improve the health of the relatively healthy. This article demonstrates how the stated preferences of a sample of the general public can be used to estimate the parameters of an SWF in the domain of health. The issues addressed in this article are of real policy concern in many countries, such as the UK, Australia and New Zealand, which have put into place policies that seek to improve overall population health and reduce health inequalities (Department of Health, 1999; Rice and Smith, 2001). In order for the SWF approach to be operationally useful, two main questions need to be answered: (1) what type of SWF is to be employed and (2) how is the shape of the SWF to be determined? Sections II and III deal with each of these questions in turn and Section IV presents the design of an empirical study that elicited the public’s preferences over two health programmes, one that maximizes health and one that reduces inequalities in health between particular population subgroups. Section V shows how these data can be used to derive a set of relative weights to be given to a unit health gain to people from different population subgroups, and hence to estimate the shape of the SWF. Section VI discusses the implications of the results.