CSB Working Paper, 2021, 5: Well-Being Measurement

The social welfare function (SWF) framework includes a well-being measure w(∙), for converting outcomes into vectors (lists) of well-being numbers. These well-being numbers are interpersonally comparable. This chapter discusses the construction of the well-being measure. It supposes that w(∙) operates on individual “histories,” a history being a combination of an attribute bundle a and a preference R. That is w(∙) = w(a, R). This setup is quite general. It encompasses preference-based well-being…

The social welfare function (SWF) framework includes a well-being measure w(∙), for converting outcomes into vectors (lists) of well-being numbers. These well-being numbers are interpersonally comparable. This chapter discusses the construction of the well-being measure. It supposes that w(∙) operates on individual “histories,” a history being a combination of an attribute bundle a and a preference R. That is w(∙) = w(a, R). This setup is quite general. It encompasses preference-based well-being measures (namely those that assign well-being numbers to histories containing different bundles but the same preference in deference to that preference), as well as non-preference based measures. The chapter covers both, although mainly focusing on the former. Here, two approaches are discussed: the “equivalence approach,” whereby an individual’s well-being hinges on her attributes and her ordinal preference; and the “vNM approach,” which uses lottery preferences rather than ordinal preferences.