An Economist's Perspective on Well-Being Analysis and Cost-Benefit Analysis
Graham, Carol, Duke Law Journal
As a starting point, it is important to note that I come to this from the blunt-tooled perspective of an economist, rather than from the more refined conceptual approach of lawyers. Lawyers take much more time to unbundle underlying conceptual definitions into statements that can pass legal argument. Economists worry much more about translating questions of inquiry into equations which, when run on empirical data, can produce statistically and methodologically rigorous results. Although these two approaches are markedly different, I think there is mutual benefit to be gained from the professions talking more to each other. I surely benefited from reading these excellent articles.
That said, before we can think about sound legal arguments for utilizing these metrics (and the related concepts that they invoke), we should recognize that there is still a lot that we do not know in terms of what we are actually measuring when we discuss well-being. Although we have a clear sense of the various dimensions of well-being and which metrics measure each dimension, there is still an important dimension of well-being that falls into the economist's category of unobservables. Well-being is driven by variables between aspects that are observable, such as socioeconomic and demographic traits, and those that are unobservable, such as innate character traits, genes, and other such things that are difficult to measure. Yet those interactions are fundamental to well-being and to the related behavioral outcomes that we are trying to understand. There is some exciting new work that tries to disentangle the genetic components of well-being from what is determined by the environment, (1) for example, as well as experimental work that compares predicted well-being to actual choice behavior. (2) Yet even in that work, there are still questions about what the "it" that we are actually trying to measure is. Income, in contrast, is much simpler to define, if also less interesting.
I raise that point as a note of caution in terms of our ability to be as precise as these articles suggest that we should be about the underlying concepts of well-being. The economist's blunt-tooled approach seeks regularities in large scale data sets and then teases out the precise relationship underlying them, while perhaps falling short on some of the conceptual clarity that lawyers are asking for. This approach has some merit in helping us understand what is still in the realm of the unknown. Well-being is an exciting area precisely because it seeks to define human welfare more broadly than in simple income terms, yet that also raises a gray area where it is less clear exactly what it is we are defining.
That caveat aside, there is a consensus among economists and psychologists studying well-being on its two distinct dimensions--hedonic and evaluative. Indeed, answering the question of how each of these dimensions is relevant to policy and to our national statistics is the task for the new National Academy of Sciences panel (of which I am a member).
Coming from that perspective, I have some genuine criticism of both articles. I disagree with Professor Matthew Adler's conclusion that only experienced well-being is appropriate for policy, (3) and I am concerned that the article by Professors John Bronsteen, Christopher Buccafusco, and Jonathan Masur does not unbundle well-being into its distinct dimensions. (4) There is good information and analysis in both articles, and I will highlight that in my comments, but I would like to provide a bit more detail on the two dimensions of well-being--and how they relate to policy--first.
The first dimension, hedonic well-being (HWB), which I have categorized in past work as "Benthamite," is related to the environment or context in which people live--the quality of their jobs, their immediate state of health, the nature of their commute to work, and the nature of their social networks--and is reflected in positive and negative affective states, among other things] Daily experience is linked to health status and other outcomes via channels such as worry and stress on the one hand, and pleasure and enjoyment on the other. …