The last post has briefly surveyed the methodological linkage between preferences and choice. The aim of this post – the 2nd in our posts on the Khavruta’s discussions in the past year – will elaborate on the normative consequences of current revealed preference theory.
Since the ordinal revolution in welfare economics, the disciplinary standard is to define welfare as the fulfillment of preferences. Under this perception, if an individual prefers x to y, this means that he is better off with x than he is with y. The assumptions of revealed preference theory bridges between choice and well-being. As economics Nobel laureates George Akerlof and Robert Shiller wrote: “A common precept of standard economics is that people only make the choices that maximize their welfare. This assumption even has a fancy name, “revealed preferences”: that people reveal what makes them better off by their choices” (Akerlof and Shiller, 2015, P. 170). The following statements, by Ian Little, reflect a similar view: “a person is, on the whole, likely to be happier the more he is able to have what he would choose” (Little, 1949, p. 98.) or “Policy x is deemed better than policy y for an individual if and only if, given the opportunity, the individual would choose x over y” (Gul and Pesendorfer 2008, p. 24).
Although these statements seem intuitive, they are not self-explanatory. Is it always right to say that individuals choose what is best for them? This question was presented by the economist Amartya Sen more than 40 years ago. Sen’s answer was a definitive no: People’s choices may reflect a compromise between a variety of considerations of which personal well-being is just one (Sen, 1973; Sen, 1977).
In other words, in order to define preferences as the determinant of choices it needs to include various considerations. However, if preferences include other factors than one’s well-being, it does not make sense to define well-being as the fulfillment of preferences.
Behavioral economics poses another challenge to the relationship between choice, preference, and well-being by questioning the assumption that individuals have stable and context-independent preferences. Psychologists and behavioral economists have shown that the context in which a choice is made influences people’s choices. For example, people tend to choose their default option, they have loss aversion (which means they would pay more to avoid the loss of something than to gain it from the beginning), they are influenced by framing effects (the setup of the problem they are facing), etc.
Relieving the assumption of stable context-independent preferences, the epistemic assumption of identity between observed choice and underlying preferences is no longer trivial. Although it is theoretically possible to define revealed preferences in terms of choices without assuming underlying stable and context-independent preferences, without this assumption it will be hard for the theory to carry the prescriptive and normative features it currently holds. For example, if people are systematically biased toward choosing the default option, it is possible to say that it reveals the fact they prefer to choose the default option, but it will be difficult to defend the claim that this fact itself guarantees us that the default option will always improve people’s well-being compared to the other option available.
Does this mean that economists need to abandon the conventional approach to welfare economics? This question is one of the issues that Philosopher Daniel Hausman asks in his seminal work Preference, Value, Choice, and Welfare. Hausman’s answer to this question is not a decisive yes or no, but one that takes a different route: Hausman rejects the definition of underlying preference solely in terms of choice, and claims that the connection between preferences and choices is constrained by beliefs. More relevant to our discussion, he rejects the notion of welfare as the fulfillment of preferences, even when preferences are informed, rational, and undistorted.
However, Hausman holds that there are cases in which methods of appraising policy based on preference satisfaction do make sense. This is because preferences are often a good proxy for determining what makes people better off. This is the case when it is reasonable to assume that: (1) people act to promote their own interests, (2) people are well informed and (3) people’s preferences are undistorted. In these cases economists can justifiably use choice as an indicator to people’s well-being and choices can be adequate for welfare analysis. According to Hausman, under those three assumptions, it is valid to say that people prefer what, according to their subjective evaluation, would make them better off, regardless to our assumptions about the determinants of their well-being.
However, in the real world, the majority of welfare decisions do not necessarily fulfill those three assumptions. Hausman argues that in some instances, it makes sense to take steps to purify people’s preferences from mistakes and distortions. Hausman does not explicitly explain what preference purification is, but this term became the target for a methodological critique posed by behavioral economists Gerardo Infante, Guilhem Lecouteux and Robert Sugden (Infante, Lecouteux & Sugden, 2016, henceforth: ILS)
ILS attached the term “preference purification” to a tendency they find in the work of some prominent behavioral economists (such as Richard Thaler, Douglas Bernheim and Andrei Shleifer) to treat deviations from choices as predicted by conventional rational choice models as mistakes. According to this view, individual well-being is given by the preferences that the agent would have revealed, had her decision-making not been affected by reasoning imperfections (like limitation of attention, information, cognitive ability or self-control, all are influenced by the set-up of the field in which the choosing takes place). For example, when considering agents’ tendency to choose their default option this approach would imply that people tend to do so because of limitation of attention or lack of information, and not because the default option is necessarily the option they prefer. Therefore, if a policy planner wishes to know what will improve people’s well-being, she needs to know what they actually prefer, so the task for the planner is to try to reconstruct individuals’ underlying preferences by simulating what they would have chosen, had they not been subject to those imperfections.
ILS accuse economists who embrace this approach as implicitly using a dualistic model of the human being, in which an inner rational agent is trapped in an outer psychological shell. In other words, although behavioral economics suggest that in many cases it is not legitimate to assume that people’s preferences are context-independent, this conventional approach to behavioral welfare economics actually assumes that people have rational context-independent preferences, but it is their psychological mechanisms that causes them to make mistakes when they come to choose.
What are the problematic aspects of this approach to welfare economics? Hausman, in his answer to the critique posed by ILS, identifies three problems ILS present in their article (Hausman, 2016). The first is methodological – this image of an inner rational agent is not based on any scientific psychological theory. The second is epistemological – it is not evident what those ‘real’ preferences are. The third is normative – it seems that for ILS, there is something normatively unacceptable in this approach to welfare economics, although they do not clearly explain what is it that bothers them.
One claim that is implied by ILS is that when examining choice data, an outside observer must decide what is considered as a mistake, which is itself a value judgment. They repeatedly claim that if we take seriously the understanding that there are no context-independent preferences it means that there is no reason to think that there are real preferences at all. Although not explicitly expressed, it seems that for ILS the fact that people’s choices are a complicated process has dramatic consequences. If preferences cannot be simply described as maximizing personal utility, this alone should encourage economists to abandon the traditional view of choice as indicating welfare and adopt different normative approaches. As Hausman sais, “It seems that ILS object to normative economics that imposes the economist’s or policymaker’s judgment of what is good for people rather than simply furthering people’s choices.” (Hausman, 2016, p.29)
Hausman’s view is different from the one ILS present. He too rejects the traditional view of welfare as the satisfaction of preferences and rejects it as a sound philosophical ground for normative economics. However, he does not say that this rejection by itself requires economists to embrace a new view. In fact, throughout his book, Hausman makes a distinction between the way economists describe their practice and the practice itself. As he declares: “My quarrels are less with how economists employ the concept of preferences than with how economists have described what they are doing” (Hausman, 2011, p. 7). Hausman book clarifies several misconceptions about the concept of preferences, but argues that: “Even though preferences are central to mainstream economics, the precise interpretation of preferences is sometimes of little importance” (p. 136).
To conclude, Hausman’s view on the preference-choice relations raises two major questions that lie in the core of philosophy of economics. First, is the demarcation between rhetoric and practice warranted? In other words, is methodological soundness the only thing that matters for describing economists’ practice? Economists have roles in public convincing, arguing, judging and other public aspects of their work in which they influence what can be considered as a legitimate proposition. Some of the prestige that allows them this public role stems from the rigorous positivistic rhetoric they use. If so, how does it changes the legitimacy of Hausman’s demarcation?
Following Hausman’s demarcation, a second question concerns the implications on normative economics. ILS, as well as Sen, believe that ethical considerations should play a role in normative economics and that normative analysis is impossible without a justifiable moral grounding. Hausman, on the other hand, does not believe this sort of justification is imperative for normative economics, but rather emphasizes the methodological aspects of normative economics, is it enough? Should discussions about the effects of different ethical perceptions on policy questions be considered under the scope of normative economics?
Although these questions are an essential part of the reciprocal discussion between philosophers and economists, they are beyond the scope of this post. However, we plan to explore some of those discussions in our future meetings, which we will share in this blog.
*** Special thanks to Victor Levi for assisting in the editing and proofreading this post.
Akerlof, G. A., & Shiller, R. J. (2015). Phishing for phools: The economics of manipulation and deception. Princeton University Press.
Gul, F., & Pesendorfer, W. (2008). The case for mindless economics. The foundations of positive and normative economics: A Handbook, 1, 3-42.
Hausman, D. M. (2011). Preference, value, choice, and welfare. Cambridge University Press.
Hausman, D. M. (2016). On the econ within. Journal of Economic Methodology, 23(1), 26-32.
Infante, G., Lecouteux, G., & Sugden, R. (2016). Preference purification and the inner rational agent: a critique of the conventional wisdom of behavioural welfare economics. Journal of Economic Methodology, 23(1), 1-25.
Little, I. M. (1949). A Reformulation of the Theory of Consumer’s Behaviour. Oxford Economic Papers, 1(1), 90-99.
Sen, A. (1973). Behaviour and the Concept of Preference. Economica, 40(159), 241-259.
Sen, A. K. (1977). Rational fools: A critique of the behavioral foundations of economic theory. Philosophy & Public Affairs, 317-344.
Thaler, R. H., & Sunstein, C. R. (2009). Nudge: Improving decisions about health, wealth, and happiness. Penguin..
 For Hausman, preferences are subjective states that jointly with beliefs determine choice. Hausman severely criticizes the attempt to define preferences in terms of choices. However, he adds that in fact most of what is called “revealed preference theory” is not actually committed to this assumption on preferences, and economists are aware of the importance of beliefs.
 He actually refers to it in just one paragraph in his book (Hausman, 2011, p.102)
 For example, the term libertarian paternalism, coined by Cass Sunstein and Richard Thaler, refers to the attempts to influence choices in a way that will make choosers better off as judged by themselves (Thaler & Sunstein, 2009, p. 5).
 Amartya Sen also argued that normative economics should change if we accept that the traditional view of choice as maximizing welfare is wrong. Accordingly, Sen formulated a new theory of welfare, known as The Capability Approach.
 As was mentioned, Hausman rejects the possibility to define preferences in terms of choice and the definition of welfare as preferences satisfaction. In addition he rejects the claims that preferences are matters of taste, that they are based only on self-interested benefit, and that economists have nothing to say about the formation and modification of preferences (p. 8).