Navigation – Plan du site
Revue des livres
Essais critiques

Taking Preferences Seriously

Koen Decancq
p. 139-146
Référence(s) :

Daniel M. Hausman, Preference, Value, Choice and Welfare, Cambridge, MA.: Cambridge University Press, 2012, 168 pages, ISBN: 978-1107015432

Texte intégral

1Over the years, mainstream economics has slowly drifted away from its philosophic roots. Nowadays, economists–I consider myself one of them–seem to juggle routinely with concepts such as preferences, beliefs, welfare and self-interest without spending much attention on how these concepts should be interpreted in a philosophically sound way. Obviously, this trend is scientifically regrettable and even potentially dangerous. Preference, Value, Choice and Welfare offers a great help to any social scientist who wants to explore the philosophical roots of the economic discipline. It reviews in an accessible way some of the ideas gathered by Daniel Hausman during more than three decades of studying the economic method. Hausman dissects with subtlety and surgical precision the notion of preference, which is central to any neo-classical model of economic behaviour. Without the concept of preferences, there would be no game theory, no welfare economics, no behavioural economics, … in fact there would be no economics at all. Yet, Hausman shows by many examples how this central building block of economic modelling is plagued with misconceptions and how its different interpretations even conflict with each other.

1. Preferences as total comparative evaluation

2Amartya Sen (1977, 329) identifies two central meanings of the concept of preference. Preferences may be a ranking of alternatives in terms of expected advantage or a representation of individual choice. Sen warns against confounding expected advantage rankings and choice rankings, but defends the coexistence of a multiplicity of preference concepts to discourage an overly simple view of evaluation and choice. Hausman disagrees with Sen and advocates one single concept of preference, notably the one of a total comparative evaluation (Hausman, 2012, 21). The book does not contain explicit formal definitions, but I take total comparative evaluations to be rankings of alternatives that take into account every consideration the agent judges to be relevant. Hausman starts his discussion of this interpretation by scrutinizing both preference concepts listed by Sen.

3Hausman argues first that preferences are not rankings of expected advantage. Even if it is often sensible to take people’s preferences as guides to what benefits them personally, there are very few people who do not distinguish between the questions “What do I prefer?” and “What is best for me?”. Preferences are not exclusively self-interested and therefore cannot be defined by advantage, argues Hausman. This claim is supported by recent empirical research. Benjamin et al. (2012), for instance, illustrate the important role of other-regarding considerations in preferences by identifying “the happiness of your family” as a major component of individual welfare. As many aspects such as moral or aesthetic considerations, and even passions and whims influence preferences, restricting them to a ranking of expected self-interested advantage leads to a too narrow perspective.

4Second, preferences cannot be defined in terms of actual or hypothetical choice. At first, this claim may come as a surprise for many readers, especially for economists involved in revealed preference analysis. Gul and Pesendorfer (2008, 7), for instance, equate ‘utility maximization’ with ‘choice’. Hausman argues that only in conjunction with beliefs, choices are determined by preferences. As beliefs may change, preferences cannot be equated with choices. He gives the following evocative example of how the same choice reflects different preferences when beliefs differ:

At the end of Romeo and Juliet, Romeo enters the tomb of the Capulets and finds Juliet apparently dead. She is in fact alive, but he does not know that she took a potion that merely simulates death. Unwilling to go on living with Juliet dead, Romeo takes poison and dies. The alternatives he actually faces are death or eloping with Juliet—although, tragically, he does not know that. From the set of alternatives actually available to him {death, eloping with Juliet}—Romeo chooses death. If choice defines preference, then Romeo prefers death to eloping with Juliet. But, of course, he does not prefer death to life with Juliet. His choice does not reveal his preference, because he is mistaken about what the alternatives are among which he is choosing. (Hausman 2012, 27-28)

5This example illustrates that—more prosaic—empirical derivations of utility functions from choice behaviour can only be obtained under some assumptions on the beliefs. Yet, as most of these empirical applications are restricted to situations where beliefs are fixed, it remains unclear to me how important the role of beliefs is for the daily practices of economists who equate choices and preferences.

6After having refuted both of Sen’s interpretations of a preference, Hausman develops his interpretation of a preference as a total comparative evaluation in more detail. Contrary to rankings of expected self-interested advantage, preferences are total and they do not leave out anything that is regarded by the individual as relevant to the evaluation of the alternatives at hand. Contrary to choice rankings, they are to be combined with beliefs to determine choices. In the so-called standard model of choice, preferences as total comparative evaluations can then be employed to predict and explain choices in interplay with current beliefs and constraints on feasibility. Hausman advocates furthermore that the standard model should study the sources and formation of preferences, without which its predictions and explanations would be rather trivial. This again may sound surprising for many economists who consider preferences as exogenously given and who have some division of labour in mind between economics and other social sciences. For instance, Robbins ([1932] 1984, 86) writes “Why the human animal attaches particular values in this sense to particular things, is a question we do not discuss. That is quite properly a question for psychologists or perhaps even physiologists.”

7Hausman not only advocates that economists should be involved in studying preference formation, he claims furthermore that economic models already make substantive claims about the content of preferences and their formation. To illustrate that, Hausman introduces a distinction between two kinds of preferences. There are preferences among alternatives (final preferences) and preferences over properties and consequences of these alternatives (distal preferences). Distal preferences are taken as exogenously given, whereas final preferences can be explained by an extended standard model. Jointly with beliefs over properties and consequences of the alternatives, the distal preferences determine the final preferences. Final preferences, in turn, determine with beliefs and constraints the alternative chosen. Hausman gives three examples of what he calls consequentialist models of final preference formation: expected utility theory, multi-attribute utility theory and game theory. In his interpretation of expected utility theory, expected utilities of alternatives are derived from preferences among the consequences of these alternatives and the probabilities of the states of the world. Similarly, multi- attribute utility theory conceives alternatives as constituted of a bundle of attributes. Multi-attribute utility theory assigns values to these attributes and derives final preferences from these values. Along the same lines, Hausman interprets game theory as the theory of the formation of final preferences over strategies based on beliefs and distal preferences over outcomes of a game. These are fresh re- interpretations of existing economic theories as models of preference formation. Yet, it may seem as if the problem of preference formation is just shifted one level further, as one may wonder where the distal preferences come from and who is going to provide theories on their formation.

2. Making sense of welfare economics

8So far, the discussion has only involved the positive or descriptive side of economics. Yet, preferences also play an important role in the normative or prescriptive branch of economics, i.e. welfare economics. Welfare economists often take welfare to be the satisfaction of preferences. But, as individuals are prone to make bad choices—the reader can easily come up with examples, I believe—guiding policy to promote preference satisfaction may in fact not enhance welfare at all! This apparently daunting observation is the starting point for Hausman’s discussion of welfare economics and the role of preferences therein. Again Hausman defends a subtle position. On the one hand, he refutes preference satisfaction theories of welfare which are popular among welfare economists, but then he attempts to rescue welfare economics by arguing that welfare economics presupposes no theory of welfare at all. Rather than constituting it, preference satisfaction can serve as evidence of what is best for the individual, regardless of the underlying theory of welfare.

9Broadly speaking, there are three competing theories of welfare (Adler, 2012, 159-170 offers a recent and excellent overview). A first view is that only mental states are intrinsically good and that welfare consists of pleasure and happiness. Despite the potential discrepancy between subjective mental states and the objective reality, the recent booming literature on happiness builds on a revival of this Benthamite view (see Kahneman and Krueger, 2006 and Layard, 2005 for some examples of this revival and Fleurbaey et al., 2009 for a critical appraisal). Contrarily, some authors have argued for an objective approach of what contributes to welfare. Martha Nussbaum (2011), for instance, has proposed an objective list of ten aspects that contribute to human flourishing (eudaemonia). Yet, it is unclear how these items should be weighted and how the heterogeneity in the different conceptions of the good life should be taken into account (Decancq and Lugo, 2013). As a consequence, paternalism is lurking behind the corners of any objective list. Therefore, a third view relies on the preferences of the individuals themselves to determine what contributes to their welfare. Rawls (1971, 366) is one of the main proponents of this view, arguing that welfare should reflect the “plan that would be decided upon as the outcome of careful reflection in which the agent reviewed, in the light of all relevant facts, what it would be like to carry out these plans and thereby ascertained the course of action that would best realize his more fundamental desires.” Hausman refutes this perspective of welfare as preference-satisfaction based on a series of arguments. First, there is the argument previously developed that people may care about other things in addition to their own welfare. Consequently, a state of affairs is not good for an individual because it satisfies her preferences. In fact, satisfying preferences may have no bearing on welfare at all. Hausman refers to Derek Parfit’s example that the recovery of an ill stranger may be something which is preferred, but which cannot plausibly be argued to make one’s own life better. In addition, preferences may be unstable, inconsistent, antisocial or manipulated. The latter problems can be accommodated by focusing on laundered or idealized preferences, but—according to Hausman—the first counter-argument that a state of affairs is not good for an individual because it satisfies her preferences, can never be laundered away and forces us to abandon preference-satisfaction based theories of welfare. Personally, I had a hard time grasping this argument, as Hausman does not precisely define what it means for a state of affairs to be good for an individual.

10In fact, Hausman argues in the next chapter that for welfare economics to make sense it is not even necessary to define what it means for a state of affairs to be good for an individual. Although what people prefer does not determine what is good for them, it can be evidence concerning what is good for them, states Hausman. An alternative is preferred by self-interested and well-informed people because it is good for them, not good for them because they prefer it. Hausman compares his argument to Plato’s Euthyphro, where Socrates says of piety “that it is loved by the gods because it is holy rather than holy because it is loved by the gods.” (Hausman, 2012, 89). With this epistemic, evidential view on preferences, preferences are a good guide to what benefits well-informed and self-interested people and hence there is no need for welfare economists to commit themselves to one single theory of welfare. Moreover, this evidential view justifies the use of tools such as cost-benefit analysis that rely on information concerning preference satisfaction.

11Hausman takes a rather reductionist view on normative economics by focusing exclusively on the efficiency aspect of policy evaluation and eschewing deliberately any equity or distributional consideration. Hausman advocates a division of labour where the “job of welfare economists is to advise policy makers on how to expand the ‘pie’. The ethical question of how to distribute the ‘pie’ can be left to others to address.” (Hausman, 2012, 95). Yet, as in the redistribution of the ‘pie’ typically some crumbs get lost—due to people changing their behaviour, for instance—its size (efficiency) and distribution (equity) seem intimately linked. Hausman is not explicit on why welfare economists should abandon distributional analysis and to whom they should leave the field. Based on his earlier work (Hausman, 1995), I presume that Hausman’s worries on interpersonal comparisons of welfare underlie this suggestion of a retreat of welfare economics from equity questions. I don’t share these worries. Recent breakthroughs in fair social choice provide ethical justification for some (implementable) methods such as money-metric utilities to make interpersonal comparisons based on purely ordinal data on preferences (see Fleurbaey, 2009 and Fleurbaey and Maniquet, 2011). Additionally, one may wonder whether the notion of extended preferences proposed by Harsanyi and Arrow to enable interpersonal comparisons, should not be discussed in the book as well.

3. Challenges from behavioural economics

12In the last part of the book, Hausman deals with recent challenges to the standard model of choice raised by cognitive psychologists and behavioural economists. Not only do real-world people diverge in unsystematic ways from the axioms of choice, there are also systematic differences between people’s behaviour and the textbook model, caused by context-dependence such as loss-aversion and framing effects, for instance (see Kahneman, 2011 for a recent survey). These findings bring decision theorists to a crossroad. Either they continue requiring that preferences—in interplay with beliefs and constraints—determine choice and then the context should be seen as a complex factor affecting preferences, or they take the other road and regard these factors as separate determinants of choices beyond preferences. In line with his rejection of Sen’s call for a multiplicity of preference concepts, Hausman pleas convincingly in favour of the first alternative. This choice arguably requires richer and thicker theories of preference formation.

13How to derive these theories is the topic of the last chapter, in which Hausman defends the view that preferences are subject to scrutiny against the commonplace that preferences are mere matters of taste or comparative enjoyment, beyond rational criticism or discussion. Understanding and modelling preferences as outcomes of a rational process in the presence of ‘behavioural anomalies’, is certainly challenging, admits Hausman. It will require multidisciplinary efforts in which economic theory may play a key role. As Hausman illustrates, the nihilistic alternative of seeing preferences as purely a matter of tastes risks even to lead to mistakes when it comes to eliciting them with contingent valuation techniques.

14This thoughtful and thought-provoking book offers an important and welcome bridge between the social sciences. Non-philosophers—like me—may need a bit of time to get acquainted with the style of argumentation based on a discussion of examples rather than on formal derivations, but after some pages they will discover this book to be a clear mirror, reflecting sharply (upon) their own methods, results and interpretations.

I thank the editor, Matthew Adler, Guido Erreygers, Marc Fleurbaey, Jeroen Horemans and Priscilla Toffano for their comments and insightful discussions. All shortcomings are my own.

Haut de page

Bibliographie

Adler, Matthew D. 2012. Well-Being and Fair Distribution: Beyond Cost-Benefit Analysis. New York, NY: Oxford University Press.

Benjamin, Daniel J., Ori Heffetz, Miles S. Kimball, and Nichole Szembrot. 2012. Beyond Happiness and Satisfaction: Toward Well-Being Indices Based on Stated Preference. NBER Working Paper No. 18374.

Decancq, Koen and Maria Ana Lugo. 2013. Weights in Multidimensional Indices of Well Being: An Overview. Econometric Reviews, 32(1): 7-34.

Fleurbaey, Marc. 2009. Beyond GDP: the Quest for a Measure of Social Welfare. Journal of Economic Literature, 47(4): 1029-1075.

Fleurbaey, Marc and François Maniquet. 2011. A Theory of Fairness and Social Welfare. Cambridge: Cambridge University Press.

Fleurbaey, Marc, Erik Schokkaert and Koen Decancq. 2009. What Good is Happiness? CORE discussion paper 2009/17.

Gul, Faruk and Pesendorfer, Wolfgang. 2008. The Case for Mindless Economics. In Andrew Caplin and Andrew Shotter (eds), The Foundations of Positive and Normative Economics. Oxford: Oxford University Press: 3-42.

Hausman, Daniel M. 1995. The Impossibility of Interpersonal Utility Comparisons. Mind 104(415): 473-490.

Kahneman, Daniel. 2011. Thinking Fast and Slow. New York, NY: Farrar Straus and Giroux.

Kahneman, Daniel and Alan B. Krueger. 2006. Developments in the Measurements of Subjective Well-being. Journal of Economic Perspectives, 20(1): 3-24.

Layard, Richard. 2005. Happiness: Lessons for a New Science. London: Penguin.

Nussbaum, Martha. 2011. Creating Capabilities: The Human Development Approach. Cambridge, MA: Harvard University Press.

Rawls, John. 1971. A Theory of Justice. Cambridge, MA: Harvard University Press.

Robbins, Lionel. [1932] 1984 An Essay on the Nature and Significance of Economic Science. New York, NY: New York University Press.

Sen, Amartya K. 1977. Rational Fools: A Critique of the Behavioral Foundations of Economic Theory. Philosophy and Public Affairs, 6(4): 317-344.

Haut de page

Pour citer cet article

Référence papier

Koen Decancq, « Taking Preferences Seriously », Œconomia, 3-1 | 2013, 139-146.

Référence électronique

Koen Decancq, « Taking Preferences Seriously », Œconomia [En ligne], 3-1 | 2013, mis en ligne le 01 mai 2014, consulté le 29 juin 2017. URL : http://oeconomia.revues.org/639

Haut de page

Auteur

Koen Decancq

Centre for Social Policy Herman Deleek, University of Antwerp; Centre for Economic Studies, University of Leuven; Core, University of Louvain-la-Neuve. Koen.Decancq@econ.kuleuven.be

Haut de page

Droits d’auteur

Licence Creative Commons
Les contenus d’Œconomia sont mis à disposition selon les termes de la Licence Creative Commons Attribution - Pas d'Utilisation Commerciale - Pas de Modification 4.0 International.

Haut de page
  • Logo Association Œconomia
  • Logo CNRS
  • Les cahiers de Revues.org