Well-Being Science

Tyler DesRoches (Arizona State University) - There is no consensus on the normative foundations of behavioral welfare economics. Behavioral economists typically reject the traditional criterion of welfare – the satisfaction of every “raw” preference – because the empirical findings of their sub-discipline has demonstrated that, the framing effect, false beliefs, the endowment effect, etc., cause agents to make systematic mistakes when confronted with choices. Instead, most behavioral economists champion the satisfaction of “purified preferences” as their criterion of choice. Purified preferences are the preferences that the agent would have had if her reasoning had not been subject to various psychological distortions. However, purified preferences have been recently criticized on various (methodological, epistemological, and normative) grounds and while alternative have been proposed, there is still no widespread agreement on the criterion of welfare for behavioral welfare economics (Sugden 2018; Sugden et al. 2016; Hausman 2016; Sugden 2004). This paper develops an alternative foundation for behavioral welfare economics: values-based preferences. Similar to Dan Hausman’s (2012) view of preferences, the satisfaction of values-based preferences has an evidential relation (not a constitutive relation) to agent welfare or well-being. However, values-based preferences are distinctive because they are explicitly grounded in the agent’s own value commitments, which are normative, affective, and stable for the agent who possesses them (Tiberius 2008). Critically, the objects of values-based preferences are not individual choices, but patterns of consumption over a specified period of time. To have a value-based preference is to have a preference for a pattern of consumption that accords with one’s own pre-determined plan, which is ultimately grounded in one’s value commitments (Bratman 1987). In short, the agent’s value commitments pick out patterns of consumption that are compatible with those commitments. To have a plan and to act on it is to adopt a particular kind of strategy for consuming goods and services: agents who choose according to their plans are resolute choosers (McClennan 1990; Gauthier 1997). As for judgements over the correctness of any single instance of consumption, this determination cannot be made without first referring to the pattern of consumption. If, over time, the agent’s choices form a pattern that is incompatible with their own value commitments, then we can conclude that the agent has either abandoned their value commitment or made an error by their own subjective standards. It will be argued that there is a strong case to be made for values-based preferences to serve as the normative foundation of behavioral welfare economics.

Gil Hersch (Virginia Tech) - There are a variety of measures that are considered by their proponents to be good ways of operationalizing well-being. These include the traditional economic measures, such as Gross Domestic Product (GDP) per capita, Subjective Well-Being (SWB) measures, such as questionnaires about how one evaluates her life as a whole, as well as capabilities-based measures such as the OECD’s Better Life Index (BLI). But how does one decide which measure is a better operationalization of well-being? Advocates of different measures provide only a cursory account of why their chosen measure is an appropriate operationalization of well-being. I argue that this is a significant gap in the literature, which once addressed, will enable a much more reasoned way of choosing between different measures of well-being. The problem faced in the well-being literature can be described as similar to what Chang (2004) calls the “problem of nomic measurement,” which arises when we want to measure a quantity X that is not directly observable. We infer it from another quantity Y, which is directly observable, but we have no way of establishing empirically what the law that governs the relationship between Y and X is. In the context of well-being, proposed measures of well-being can be thought of as quantity Y whereas well-being itself can be thought of as quantity X. The question then is how to figure out if there is a function f, and what it is, without being able to measure quantity X directly. If we are ever to decide which of the different measures Y is a better operationalization of well-being X, we need to look at how the proponents of the different measures account for the link between well-being and the measure they champion. What we need from those advocating for various measures as measures of well-being as a whole is that they provide fully fleshed out accounts of why their chosen measure is indeed a measure of well-being. It would then be possible to seriously examine the merits of each measure as an operationalization of well-being. Without such positive accounts, all we are left with are negative arguments against different measures. While negative arguments are helpful in guiding us as to which measures seem to be poor operationalizations, in light of the variety of ways one can defend multiple measures as operationalizations of well-being, the process of elimination is a long one. Looking at the assumptions that support different accounts, we can judge the relative plausibility of these accounts. This is a comparative exercise. It is only meaningful to reject an account in favor of a given measure as an operationalization of well-being when a more plausible account is at hand.

Nicole Hassoun (Binghamton University) - How should we measure welfare? Doing so is important for many reasons. Policy makers have recently turned towards various measures and indicators of well-being in addition to purely economic measures, such as GDP per capita. The hope is that these indicators will help a variety of actors evaluate progress in ensuring that everyone lives good lives. Scientists have responded with a variety of new metrics from the Human Development Index, Genuine Progress Indicator, and Sustainable Economic Development Assessment to the measures employed in the Organization for Economic Cooperation and Development Framework for measuring well-being and progress and the World Happiness Report (International Panel on Social Progress, 2016; Helliwell, Layard, & Sachs, 2013). But which, if any, of these metrics should we use and for what purpose? This paper has three aims. First, to defend what some have called the “vending machine” approach to measuring well-being. It suggests that we should attempt to measure what we think matters. We can judge the success of the measure by seeing whether it approximates the correct theory well enough for a given purpose. Second, this paper aims to provide some useful guidance for actually creating measures by, for instance, distinguishing some of the main choices that researchers must make when constructing different kinds of measures. In doing so, I will establish that the process of trying to implement a theory can benefit not only the resulting measures but the theory as well. That is, we can improve the theory in the process. I will argue that determining the best measure is ultimately an empirical question, which depends on the purpose of the inquiry and the need for accuracy amongst other things. Third, to make the case that it is useful to have a well-worked out theory of well-being before attempting to measure it, this paper points out that if we lack such a theory, we may not only use the wrong measures for the wrong purposes, but we may use poor proxies as well. Moreover, we may fail to actually measure an interesting phenomenon. This paper makes these points by reflecting on how insufficient attention to theory has caused problems for scientists developing and employing measures of life quality and many other things in the past (Brey 2012; Alexandrova 2017).

Daniel Hausman (University of Wisconsin-Madison) - Well-being or welfare (which I take to be synonymous) play important roles in economics, psychology, and philosophy. Normative economists have linked welfare to command over consumption goods. Psychologists have been concerned with the sources of positive and negative affect. Philosophers have been concerned with what constitutes a good life. Economists have taken welfare to be indicated by – if not constituted by – preference satisfaction. Psychologists have regarded welfare as a matter of feelings. Philosophers have been ambivalent and conflicted, and they have often been skeptical of the views of economists and psychologists. If one identifies welfare with preference satisfaction, then its measurement is a matter of measuring preferences, which is in principle a task for a positive science. Similarly, if one identifies welfare with sets of feelings, those feelings can be studied by a positive science. Once one has accepted an evaluative thesis specifying what well-being is, then there seems to be no difficulty in principle in measuring well-being or studying its causes and effects. Of course, if one accepts some version of the open question argument, it is impossible to identify welfare with any non-evaluative state of affairs, be it preference satisfaction, feelings, or whatever; and even if one does not accept any version of the open question argument, the theories of well-being that are implicit in psychology and welfare economics are deeply flawed; and few firm conclusions can be drawn concerning welfare from measuring preferences or affect. On the other hand, it would be absurd to deny that economists and psychologists can learn anything about what promotes individual and social welfare from studying what satisfies preferences or causes certain affect. If neither preference satisfaction nor mental states constitute well-being, how can investigations of what satisfies preferences and what causes certain affects contribute to our understanding of what makes for a good life? The solution to this puzzle requires an account of how it is possible to understand what welfare is and to measure it, without possessing any acceptable philosophical theory of welfare. The solution lies in what I call “folk theories” of welfare. These consist of platitudes such as “people are generally better off if they are healthy” that are contentful and useful, although far from universal truths. Implicit acceptance of folk theories enables economists to make claims linking welfare to preference satisfaction, without taking preference satisfaction to constitute welfare, and it enables psychologists to make claims linking welfare to certain affective states, without taking those states to constitute welfare. The vagueness and ellipses in folk theories have their costs, however, and economic and psychological studies of welfare are inevitably tentative and contestable.