Skip to main content
Log in

The impossibility of rational consumer choice

A problem and its solution

  • Regular Article
  • Published:
Journal of Evolutionary Economics Aims and scope Submit manuscript

Abstract

In this paper we show that a rational consumer choice along the lines traditionally suggested might lead to paradoxical results if one considers multidimensional goods, which incorporate a series of incommensurable aspects. Thereby, we explore the similarity between the resulting paradox and Kenneth Arrow’s well known Impossibility Theorem. Based on these considerations we suggest a solution for the former problem along the lines of Herbert Simon and Amos Tversky, which might—if driven to its extreme—even provide a unique and arguably rational solution for consumer choice among multidimensional goods. Eventually, we argue that the resulting framework poses a potentially useful starting point for further developing an evolutionary theory of consumer choice.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Fig. 1
Fig. 2

Similar content being viewed by others

Notes

  1. Or, in a more modest sounding formulation, is able to rank any two given alternatives over the whole set.

  2. There is a vast amount of literature based on this basic idea, with various fields of application such as environmental economics or supply chain management (see exemplary Arrow and Raynaud 1986; Pohekar and Ramachandran 2004; Mendoza and Martins 2006; Wang et al. 2004). However, we have not come across a contribution which applies an ordinal variant of this framework on traditional consumer choice theory in full depth.

  3. Within heterodox economics Lancaster’s conception is used by a variety of authors including at least post-Keynesian (Lavoie 1994) and evolutionary approaches (e.g. Nelson and Consoli 2010).

  4. One might differentiate three types of multidimensional goods: (a) goods with different and competing ends (e.g. meat to feed me or my dog), (b) decomposable goods (a good with different parts, e.g. a suit decomposable into a shirt, a jacket, trousers...), where the different dimensions might be separated from each other and (c) non-decomposable goods (as the car in the example above). In this article we only refer to the third type where the different dimensions of a specific good are not decomposable (see Bianchi (1997) for further examples).

  5. See also Tversky and Shafir (1992).

  6. In an ordinal setup a weighting of the different dimensions does not represent a well-defined problem since it remains, due to the incommensurability between dimensional weights and concrete product performance within a certain dimension, unclear whether the decision should be based on either weight or concrete performance (with the exceptional case where all weights are exactly equal for all the relevant dimensions, which again reduces to our standard scenario).

  7. We would get a similarly paradoxical result if the individual cared equally about speed and design but not at all about safety. In this case A ≻ B because it is better in both dominant categories. Furthermore B ~C, since each of them is better than the other in one of the dominant categories. However for the same reason C ~A. In sum we get the rather impossible result A ≻ B~C ~A, which strongly resembles the results presented in Arrow (1950).

  8. Formally, this can be obtained within the above scheme by reiterating all but the last update of the criterion vector in the same order.

  9. “The characterisation of maximising behaviour as optimisation, common in much of economic analysis, can run into serious problems [...], since no best alternative may have been identified for choice. In fact, however, optimisation is quite unnecessary for ‘maximisation’, which only requires choosing an alternative that is not judged worse than any other. This [...] is also how ‘maximality’ is formally defined in the foundational set theoretic literature.” (Sen 1997, 746, Italics in original) Optimisation, on the contrary, implies the existence of some kind of mathematical extreme value problem.

  10. Learning would be another empirically relevant and more endogenous source of creating new wants and thereby ‘discovering’ new product dimensions (as heavily emphasized by Witt (2001)). Exploring this issue would, however, force us to retain the assumption of completeness, which is beyond the scope of this paper: our aim is to illustrate what happens if one drops the assumption of optimisation.

  11. Note that a heating facility or a television set may, by itself, primarily be a tool to satisfy a basic need (body warmth or cognitive arousal). However, when such tools become part of a car, they add novel dimensions to an established product. Since in most cases the underlying basic needs are already satiated by a variety of other tools and the introduction of such a multi-purpose good is not intended to satisfy a deprivated basic need, but mostly aims to make the satisfaction of an acquired want generally more pleasurable, these novel dimensions are also to be seen as related to acquired wants.

References

  • Ackerman F (1997) Consumed in theory: alternative perspectives on the economics of consumption. J Econ Issues 31(3):651–664

    Google Scholar 

  • Alexander L (2011) What are constitutions, and what should (and can) they do? Soc Philos Policy 28(1):1–24

    Article  Google Scholar 

  • Arrow KJ (1950) A difficulty in the concept of social welfare. J Polit Econ 58:328–346

    Article  Google Scholar 

  • Arrow KJ (1966[1951]) Social choice and individual values. Wiley, New York

    Google Scholar 

  • Arrow KJ, Raynaud H (1986) Social choice and multicriterion decision-making. MIT Press, Cambridge, MA

    Google Scholar 

  • Bianchi M (1997) Collecting as a paradigm of consumption. J Cult Econ 21:275–289

    Article  Google Scholar 

  • Condorcet MJ (1785) Essai sur l’application de l’analyse a la probabilité des décisions rendues a la pluralité des voix. L’Imprimerie Royal, Paris

  • Einav L, Levin J (2010) Empirical industrial organisation: a progress report. J Econ Perspect 24(2):154–162

    Article  Google Scholar 

  • Gigerenzer G, Todd P (1999) Simple heuristics that make us smart. Oxford University Press, New York

    Google Scholar 

  • Lancaster KJ (1966) A new approach to consumer theory. J Polit Econ 74(2):132–57

    Article  Google Scholar 

  • Lancaster KJ (1971) Consumer demand. Columbia University Press, New York

    Google Scholar 

  • Lavoie M (1994) Foundations of Post-Keynesian economic analysis. Edward Elgar, Cheltenham

    Google Scholar 

  • Leibenstein H (1950) Bandwagon, snob and Veblen effects in the theory of consumers’ demand. Q J Econ 64:183–207

    Article  Google Scholar 

  • Loasby BJ (2001) Cognition, imagination and institutions in demand creation. J Evol Econ 11:7–21

    Article  Google Scholar 

  • Mas-Colell A, Whinston MD, Green JR (1995) Microeconomic theory. Oxford University Press, New York

    Google Scholar 

  • May KO (1954) Intransitivity, utility and the aggregation of preference patterns. Econometrica 22(1):1–13

    Article  Google Scholar 

  • Mayhew A (2002) All consumption is conspicous. In: Fullbrook E (ed) Intersubjectivity in economics: agents and structures. Routledge, London, pp 43–55

  • Mendoza GA, Martins H (2006) Multi-criteria decision analysis in natural resource management: a critical review of methods and new modeling paradigms. Forest Ecol Manag 230:1–22

    Article  Google Scholar 

  • Metcalfe JS (2001) Consumption, preferences and the evolutionary agenda. J Evol Econ 11:37–58

    Article  Google Scholar 

  • Nelson RR, Consoli D (2010) An evolutionary theory of household consumption behavior. J Evol Econ 20:665–687

    Article  Google Scholar 

  • Peukert H (2001) On the origins of modern evolutionary economics: the Veblen legend after 100 years. J Econ Issues 35(3):543–555

    Google Scholar 

  • Pohekar SD, Ramachandran M (2004) Applications of multi-criteria decision making to sustainable energy planning—a review. Renew Sust Energ Rev 8:365–381

    Article  Google Scholar 

  • Rabin M (1998) Psychology and economics. J Econ Lit 36(1):11–46

    Google Scholar 

  • Reinstaller A, Sanditov B (2005) Social structure and consumption: on the diffusion of consumer good innovation. J Evol Econ 15:505–531

    Article  Google Scholar 

  • Savage LJ (1972[1954]) The foundations of statistics. Dover, New York

    Google Scholar 

  • Sen A (1997) Maximisation and the act of choice. Econometrica 65(4):745–79

    Article  Google Scholar 

  • Simon HA (1955) A behavioural model of rational choice. Q J Econ 69:99–118

    Article  Google Scholar 

  • Simon HA (1956) Rational choice and the structure of the evironment. Psychol Rev 63:129–38

    Article  Google Scholar 

  • Simon HA (1976) From substantive to procedural rationality. In: Latsis SJ (ed) Method and appraisal in economics. Cambridge University Press, Cambridge, pp 129–148

    Chapter  Google Scholar 

  • Simon HA (1978) Rationality as process and as product of thought. Am Econ Rev 68(2):1–16

    Google Scholar 

  • Stigler G, Becker GS (1977) De gustibus non est disputandum. Am Econ Rev 67(2):76–90

    Google Scholar 

  • Sugden R (1985) Why be consistent? A critical analysis of consistency requirements in choice theory. Economica 52:167–83

    Article  Google Scholar 

  • Sugden R (1991) Rational choice: a survey of contributions from economics and philosophy. Econ J 101:751–85

    Article  Google Scholar 

  • Tversky A (1972) Choice by elimination. J Math Psych 9:341–67

    Article  Google Scholar 

  • Tversky A, Shafir E (1992) Choice under conflict: the dynamics of deferred decision. Psychol Sci 3:358–61

    Article  Google Scholar 

  • Varian HR (1992) Microeconomic analysis. Norton, New York

    Google Scholar 

  • Veblen TB (1899) The theory of the leisure class. Macmillan, New York

    Google Scholar 

  • Wang G, Huang SH, Dismukes JP (2004) Product-driven supply chain selection using integrated multi-criteria decision-making methodology. Int J Prod Econ 91:1–15

    Article  Google Scholar 

  • Witt U (2001) Learning to consume—a theory of wants and the growth of demand. J Evol Econ 11:23–36

    Article  Google Scholar 

Download references

Acknowledgements

The authors are grateful to an anonymous referee for helpful suggestions and to Michael Landesmann for fruitful discussions.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Jakob Kapeller.

Appendix: Arrow’s theorem and the impossibility of consumer choice: an analogy

Appendix: Arrow’s theorem and the impossibility of consumer choice: an analogy

This section is devoted to a formal treatment of the aforementioned topics. We are going to define a rational ordering process and set down several axioms this process should satisfy. The treatment mirrors the discussion in Arrow (1950, 1951), but is presented completely independently. We aim to show that the problem of collective choice as posed by Arrow is formally equivalent to the problem of consumption choice among multidimensional goods. Table 2 provides a first overview by showing some basic terminological similarities between these two approaches.

Table 2 Some basic structural analogies between Arrow’s impossibility theorem and the problem of consumption choice among multidimensional goods

Linear, transitive orderings

We will demand that the ordering \(\succcurlyeq_{d}\) of the products within each product dimension d is linear and transitive. Linearity states that for any two products P and Q we either have

$$ P \succcurlyeq_{d} Q~~~\mbox{or}~~~ Q \succcurlyeq_{d} P. $$

This excludes the possibility that two products are incomparable within said dimension. Furthermore, we shall assume transitivity.

Rational ordering process and axioms

Assume now we are given a set of products and dimensions in which those products are ranked following a linear and transitive order relation. How would a rational consumer rank the products according to their desirability based on their dimensional rankings alone? A rational ordering function will be any process satisfying certain axioms which, based on the dimensional orderings \(\succcurlyeq_{d_{1}}, \succcurlyeq_{d_{2}}, \dots\) alone, extracts an ordering \(\succcurlyeq\) of the products reflecting their overall desirability.

Axiom 1

(Unrestricted domain) Although somehow already inherent in our discussion, we explicitly state that the rational ordering process should yield a product ordering for any possible orderings within the dimensions.

Axiom 2

(Account for ordinal superiority) So far, any universally applicable process is regarded as a rational ordering process, we therefore need to make certain that the dimensional orderings are reflected in the overall ranking. Suppose we are given m products in d dimensions and are given the dimensional rankings and a rational ordering process assigns the overall ranking

$$ P_1 \succcurlyeq P_2 \succcurlyeq P_3 \succcurlyeq \dots $$

Assume now furthermore that from this we create a second set of dimensional rankings by choosing one product P and setting up the new dimensional rankings by using the old rankings and moving the product P to the left or leave it at the same place while, at the same place, leaving all other interproduct relations the same. This corresponds to a manufacturer improving the product in some dimensions while leaving it untouched in other dimensions. Then, in the overall product ranking, any rational ordering process \(\succcurlyeq\) should rank the improved product P superior to all products to which it was superior before.

Axiom 3

(No prejudices) This axiom is in somewhat the same spirit as the last axiom and states that for any two distinct products P and Q, there are dimensional rankings such that

$$ P \succcurlyeq Q~~~\mbox{ and other rankings yielding }~~~Q \succcurlyeq P. $$

Axiom 4

(Independence of irrelevant alternatives) If the rational ordering process assigns an overall ranking \(\succcurlyeq\) to a given set of products based on their dimensional rankings \(\succcurlyeq_{d}\) and if then furthermore one product turns out to be unavailable and is then removed from the dimensional rankings, then any rational ordering process applied to the reduced dimensional relations should yield the same overall ranking as before with the single change of the unavailable product being removed.

The surprising result is now that although all these conditions are very natural within this framework, there is only a simple type of rational order function satisfying all these conditions.

Theorem 1

(Impossibility of rational order functions; Arrow 1950) Suppose we are given at least three products and a rational order process satisfying all the conditions above. Then this process is monomanical, i.e. there exists a ‘dictatorial’ dimension such that the process assigns to each possible set of dimensional relations an overall rating, which is identical to the product rating within this dimension.

Proof

We deduce the result from Arrow’s Impossibility theorem. The linearity and transitivity conditions imply Arrow’s axioms 1 and 2; the universality condition implies Arrow’s condition 1 [unrestricted domain], the second and the third axiom imply Arrow’s conditions 2 [positive association of social and individual values] and 4 [citizens’ sovereignty], while the axiom of independence of irrelevant alternatives is the same in both settings. Arrow’s theorem then implies that the social process must be dictatorial, which corresponds to our definition of monomanical. □

Remark (Two products)

(Two products) The assumption that at least three products be presented is a necessary one: there is a rational ordering process if we are only given two different products P and Q ordered in d dimensions according to \(\succcurlyeq_{1}, \dots, \succcurlyeq_{d}\) given by saying that product P is ranked superior to Q if it is ranked superior to Q in more or the same number of dimensions than Q is ranked superior to P. In the latter case both products are equal.

Rights and permissions

Reprints and permissions

About this article

Cite this article

Kapeller, J., Schütz, B. & Steinerberger, S. The impossibility of rational consumer choice. J Evol Econ 23, 39–60 (2013). https://doi.org/10.1007/s00191-012-0268-2

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00191-012-0268-2

Keywords

JEL Classification

Navigation