Pondering the orientation of economic theory in 1939, John Hicks stated that “[i]t has to be recognized that a general abandonment of the assumption of perfect competition, a universal adoption of the assumption of monopoly, must have very destructive consequences for economic theory.”Footnote 1 Despite the unrealistic assumption of perfect competition Hicks maintained that, moving forward, the discipline should assume that real markets do not differ greatly from this idealization.
Today, the concept of rationality plays as central a role in economic theory as perfect competition did back then. In contrast to Hicks, Maurice Lagueux does not shy away from attacking the inadequacy of the mainstream conceptualisation of rationality. Not only does he powerfully demonstrate the weaknesses associated with viewing rational economic agents as consistent maximisers, but he also provides a constructive assessment of the alternative—a minimal rationality that in fact has deeper roots in the history of economic thought than its contemporary rival.
Among the book’s many impressive features, two aspects that deserve to be singled out are its focus and its breadth. Concerning the former, every section of the text is geared toward supporting the central thesis of the book, i.e., the defence of a minimal notion of rationality. There are no argumentative detours. As to the latter, while many attacks on rationality in economics are well argued, they are frequently rather narrow in scope. In contrast, Lagueux’s analysis is comprehensive. Be it the worry that a minimal rationality is too vague a theoretical foundation for the discipline of economics, be it an attempt à la Milton Friedman to legitimize the unrealistic character of rationality as maximisation and consistency, or be it various forms of the claim that economics can do without a notion of rationality altogether—Lagueux defuses all of these potential objections to his position. It is high time that economists heed this call for a methodological reorientation.
Having expressed my sympathy for Lagueux’s argument and my admiration for its presentation, I shall devote this review to raising a series of questions both regarding Lagueux’s portrayal of explanatory kinds of rationality in economics as necessarily instrumental (10) and concerning the exact contours of the minimalism he advocates.
These remarks are in part motivated by the conclusion Lagueux reaches at the end of Chapter 4, where he states that “we should forget the idea of clearly isolating irrational actions in order to more easily characterize rational ones.” (107) The reason for this, according to Lagueux, lies in the fact that “irrational actions cannot be separated from rational ones on the basis of their material description, unless this description excludes any interpretation that would make them rational.” (102) In other words, provided there is some potential motivation for an action, even though it might strike us as irrational prima facie, it can be redescribed and “rationalised.” One of Lagueux’s examples is a woman who engages in wishful thinking that seems irrational. However, when we learn that it enables her to overcome a suicidal depression, it becomes rational. Thus far, I entirely agree.
Now, Lagueux suggests that our inability to unambiguously delineate rationality from irrationality “hardly raises a problem for economists who want to explain general economic phenomena.” (108) This conclusion strikes me as optimistic. To see why, recall the content of the minimal rationality principle that Lagueux presents as animating economic models: “Human beings … are not stupid, and they normally make decisions and perform actions that make sense given their goals.” (5) But what if the notion of an action that makes sense is underdetermined, even given the agent’s goals? What if, in other words, not only the line between rationality and irrationality is blurred, but—partly as a consequence—the domain of rational actions is too large to allow any conclusion as to which one will be selected? It remains unclear how economists would then be able to explain economic phenomena, let alone predict them. While Lagueux acknowledges both the vagueness of his minimal rationality and its poor record when it comes to making predictions (e.g., see p. 237), this admission is clearly less bothersome for the philosopher of economics than for the economist.
I will suggest that economists have reacted to this problem by adapting their concept of rationality in two ways, only one of which is on the radar of Lagueux’s analysis. First, as Chapters 1 and 2 eloquently document, economists have shifted over the last two hundred years from a minimal rationality to a notion of rationality defined by maximisation and consistency, and this for a variety of reasons ranging from their urge to formalise and emulate the natural sciences to their reluctance to engage in interpersonal comparisons. Among other things, this move allows them to tackle the underdetermination of instrumental rationality albeit, as Lagueux spells out, at considerable costs in terms of the empirical adequacy and plausibility of their theories.
Yet, there is a second way to reduce the underdetermination of minimal rationality, and that is by working on the specification of the goals of economic agents, as opposed to the identification of the best available means to achieve them. This is what economists have done in the development of the various branches of rational choice theory—individual choice, interdependent choice (game theory), and social choice. As Lagueux acknowledges in his discussion of game theory in the Epilogue, choice theory proceeds by “attributing” rationality to individuals. The key question here is what kind of rationality we are talking about.
Lagueux argues that this rationality, “which typically implies a powerful deductive capacity and an iterative common knowledge of a kind that is rarely realised even by the most brilliant minds” (240), represents a parallel case to the axioms of maximisation and consistency in neoclassical theory. He maintains that game theory (and, one might conceivably add, choice theory in general) should equally do away with this unjustifiably strong notion of rationality and replace it with a version of the weak rationality principle he defends. While there is some truth to this analysis, I think it proceeds too quickly. Let me tackle the two claims of the previous sentence in reverse order.
Though he does not explicitly say so, Lagueux presents the rationality attributed to economic agents in game theory as an instrumental rationality. I think this analysis is mistaken or at best incomplete. To see why, consider a distinction between two complementary aspects of rational choice theory that are frequently conflated. First, rational choice theory can be understood as “a particular attempt to impose an intellectual order on social analysis.”Footnote 2 This is achieved by imposing certain logical constraints on the choices of economic agents. Second, rational choice theory can be seen as requiring “a fairly detailed account of the content of agents’ desires and beliefs before it can be used to predict actions, or the political outcomes that those actions generate.”Footnote 3 The first aspect represents a logic of choice, whilst the second puts forward a science of choice.Footnote 4
How is this relevant? In two ways. First, Lagueux’s discussion of game theory suggests that he perceives it as exclusively putting forward a logic of choice. And when addressing this aspect of game theory (or, I add once more, of rational choice theory in general), it may well be possible to argue that the strong notion of rationality relying on impressive deductive capacities as well as coherence should be replaced by a weak principle of rationality à la Lagueux. Given that game theory is not the focus of his book, Lagueux only sketches the outlines of such an argument.
Second, when understood as a science of choice, however, the rationality in rational choice theory does not only refer to the instrumental kind. As indicated above, the task of a science of choice is to make certain assumptions regarding individual desires and beliefs to then analyse what follows from them in the context of a given model. As an example, think of utility maximisation as the premise of one science of choice often used in economics. While a science of choice built on the assumption of utility maximisation has rightly been criticised, it is important to note that utility maximisation is not a constitutive feature of rational choice theory but merely one particular way of attributing preferences to individuals.
This brings us back to the second way of responding to underdetermination of a minimal rationality. When positing a science of choice, economists are effectively narrowing down the goals economic agents pursue in their model by attributing a specific “rationality” to them. They are stipulating a kind of what Lagueux calls value-rationality. Arguably, rationality in this sense should not be classified as minimal.
You might object that it is unfair to attack the book on an issue—game theory or, more generally, choice theory—that is clearly peripheral to its argument. In reply, I submit that the above considerations have implications that are central to the book. First, take Lagueux’s assertion that value-rationality is “not really explanatory” (10). If one accepts that the explanatory and predictive powers of a weak (instrumental) rationality principle are limited due to underdetermination, and if one accepts that this predicament can be overcome by stipulating a particular value-rationality, then it seems to me that the latter does at least contribute to the explanation of something: it explains behaviour under certain assumptions about the motivation of economic agents.
Second, if I am right in anticipating economists’ rejection of Lagueux’s weak rationality principle for its lack of precision and of predictive power, then pointing to the potential combination with a science of choice or value-rationality to overcome this difficulty will only make Lagueux’s argument stronger. After all, the above considerations take nothing away from the book’s critique of rationality as maximisation or rationality as consistency. They merely point out that not all forms of rationality in contemporary economics that deserve the label “explanatory” are of the instrumental kind, and assumptions about value-rationality can help to overcome the underdetermination of instrumental rationality considered in isolation. Note that the formulation of an adequate science of choice also establishes the link to the research programme of behavioural economics.
To conclude, these remarks have only scratched the surface of Lagueux’s rich and stimulating book. For any economist who is not negligently ignorant of the methodological foundations of his discipline, it would be irrational in Lagueux’s sense not to read this bookFootnote 5.
