II. THE BORDERLANDS OF ECONOMICS AND PSYCHOLOGY

In a letter to J. M. Clark of 1 March 1951 Lionel Robbins wrote that “the Nature and Significance was always intended to be a sort of preliminary manifesto designed to forestall the criticism that I did not know where the borderline between the different disciplines really lay.” The disciplines Robbins referred to were “history, psychology, and political philosophy,” which he studied at the London School of Economics under the influence of the left-wing political theorist Harold Laski, and “economic theory,” which he followed “as an outsider” in Edwin Cannan's class (quoted from Howson Reference Howson2004, p. 417).

In chapter IV of the Essay, devoted to the “nature of economic generalisations,” Robbins set out to manage the boundaries between economics and psychology. Clearly dissatisfied with the version of this chapter in the first edition, Robbins made significant alterations that suppressed his almost naïve reliance on simple “deductions” from standard supply and demand schemes (1932, pp. 75, 77) in favor of a discussion that focused on how individuals value economic (that is scarce) goods. The reference to the Fisher equation was suppressed completely, in favor of a sustained discussion of “simple and indisputable facts of experience” that all related to “the way in which the scarcity of goods … shows itself in the world of reality” (1984, p. 78).

All in all, these changes moved Robbins's discussion of the nature and validity of economic postulates from the realm of a priori, deductive principles that were more-or-less valid on a logical rather than experiential basis, to the realm of day-to-day experiences of an economizing individual. With this shift in emphasis there was an increased favorable in-text discussion of Austrian school economists, which he set in opposition to the historical schools in England and Germany. Even though we see less footnoting of German and Austrian authors than in the first edition, their importance was raised by incorporating them more extensively into the body of the text.

This shift in emphasis also brought Robbins's discussion more in line with section 4, which discussed the “borderlands” of economics and psychology (p. 83). After all, in the first edition it was not so clear how a discussion of the “purely formal” tautology of MV = PT (1932, p. 82) might bring one to the importance of the allegedly psychological foundations of economics. But from Robbins's sustained contrast between the historical schools in England and Germany, which emphasized the historical situatedness of their inferences, and the Austrian school, which based the “general applicability” of their assumptions on the “indisputable fact” that individuals are able to scale their preferences (pp. 78–81), the natural question to ask was, what sustained such a claim? And here, the reference science clearly was psychology. Robbins's claim that the “relative valuations” of individuals were at the heart of the subject matter of economics (p. 83) became more tied in with Austrian economics vis-à-vis the historical school, but also made it necessary to take a stance on the precise boundaries between economics and psychology.

In both editions Robbins went to some pain to argue that if there was any psychology involved in economics at all, this had nothing to do with Fachpsychologie. Robbins argued as follows: He challenged those who argued that relative valuations were dependent upon “the validity of particular psychological doctrines” (p. 83). He straightforwardly dismissed such arguments as resting on inexact thought. Unfortunately, these “borderlands of Economics are the happy hunting-ground” of all those who advanced the one or the other psychological theory for providing the foundations for the economic “assumption of relative valuations” (p. 83). Those susceptible to this were very willing to make the foundations of economics dependent upon current fancies in psychology that shift “every five years or so” and were not willing to confront the subject proper of economics—that is, “the implications of choice in a world of scarcity” (pp. 83–84). The result was to be “bamboozled into believing” that “fashionable psychology” matters, while economics really had nothing to do with it (p. 84).

It was the economists’ own fault, unfortunately. Robbins gave instances of economists who illegitimately crossed the border: Jevons and Edgeworth, whose names were “a sufficient reminder of a line of really competent economists who did make pretensions of this sort.” The Austrians (Menger, Böhm-Bawerk) carefully avoided “this kind of misconception” (pp. 84–85).

What was this misconception precisely? Robbins referred to it in two different ways, as “psychological hedonism” and as “a theory of pleasure and pain” (pp. 84–85). These are by no means the same, since almost any psychological theory from Locke onwards (if not before) in one way or the other was about pleasures and pains, without necessarily being “hedonistic.” Robbins's interchangeable use of both labels was understandable enough, however, because Edgeworth persistently referred to his own theory of pleasures and pains as “hedonic psychology.” One may thus guess that it is Edgeworth in particular Robbins had in view, which is confirmed by Robbins's mostly critical references to Edgeworth's work elsewhere in the Essay.

Robbins quotes from Edgeworth's Mathematical Psychics (1881) in which Edgeworth compared the human agent to a “pleasure machine.” And he made note of attempts to “exhibit the law of diminishing marginal utility as a special case of the Weber-Fechner Law” (p. 85). These attempts were made by Edgeworth in his intriguing New and Old Method of Ethics of 1877.Footnote ³ In a footnote Robbins referred to Max Weber's Die Grenznutzlehre und das psychophysische Grundgesetz that was first printed in the Archiv für Sozialwissenschaft und Sozialpolitik for 1908. The first edition of the Essay considered Weber's argument “conclusive”—a word Robbins dropped in the second edition. Robbins also dropped the addition that Weber's own views on the basis of political economy rested “upon a certain degree of misconception” of what economics was about (1932, p. 85). What this misconception consisted of was left unspecified, but it will be of our concern in this paper.

From this point onwards, the first and the second editions of the Essay are very different. In the first edition Robbins expands on the economists’ assumption of “scales of relative valuations,” an assumption that serves as a “datum” to the economist, even though he granted “psychologists or perhaps even physiologists” that the question why “the human animal attaches particular values … to particular things” could be sensibly studied (1932, p. 86), but it was not the question of interest to the economist. As emphasized by Hands (Reference Hands2007), the flavor of Robbins's discussion of human valuations opened the door to behaviorism, and it was quite understandable that Souter in his criticism of Robbins accused him of just that: a behaviorist theory of human valuation.Footnote ⁴ Thus, Robbins wrote that economists only needed to assume “the obvious fact that different possibilities offer different stimuli to behavior, and that these stimuli can be arranged in order of their intensity” (1932, p. 86).Footnote ⁵

In the first edition, the main reason for Robbins to distance himself from the “hedonism” of Jevons and Edgeworth was discussed in section 5 of chapter IV; the alleged focus of economists on self-interest. He forcefully rejected the view of man as an “egotist” or “pleasure machine, invariably driven by self-interest,” which he considered to be implied in the “hedonistic trimmings of the work of Jevons and his followers” (1932, pp. 86–87). The absurdity of this belief was “exasperating … to any competent economist” and simply led to a “failure” to understand their business (1932, p. 87). It was clear though that Robbins was aware his argument was driven rather by rhetoric than by substance, and so it is exactly this part of the Essay that went into thorough reconstruction. The second edition, then, heavily relied on Max Weber's analysis of means–ends rationality, in particular Weber's essays Die Objektivität sozialwissenschaftlichen und sozialpolitischer Erkenntnis and Der Sinn der Wertfreiheit der soziologischen und ökonomischen Wissenschaft, as much as it relied on the essay on Fechner and marginalism reprinted in his Wissenschaftslehre (p. 90).

To bring out the importance of this shift in Robbins's argument, I will first show how psychophysics made its way into economics at the end of the nineteenth century. Though Jevons and Edgeworth are the most important instances, particularly in Germany there was a broadly sympathetic audience to attempts to use psychophysics to solve socio-economic problems. This audience included economists generally associated with the German Historical School such as Lujo Brentano. Though Max Weber is commonly associated with the Historical School as well, it is well known that Weber's attitude towards it was ambivalent, to say the least. On one hand Weber rejected what he considered its fundamental irrationalism; on the other hand, he held on to explanations in the social realm that were historically situated. This is where Weber's criticism of psychophysics and his elaboration of the notion of means–ends rationality came in. I will then return to Robbins's Essay to see how Robbins used Weber's criticism, but departed from it as well.

III. THE “PHYSICAL GROUNDWORK” OF ECONOMICS

In the second half of the nineteenth century, attempts to search for what Stanley Jevons called the “physical groundwork” of political economy became increasingly popular after a period in which political economists, largely following John Stuart Mill, maintained strict boundaries between the natural sciences and the science of mind, including political economy, though this did not mean that they denied political economy the lawlike character they so much admired of the natural sciences, rational mechanics in particular. Well-known is Mill's resolute negative answer in his Logic on one of the vexatae quaestiones in the science of mind whether mental states are reducible to physiological states (Mill Reference Mill and Robson1843 [1974], p. 850). This answer enabled him to ascribe causal power to mental motives and to maintain freedom of the will as well. Thus Mill could uphold the lawlike character of political economy—a science of motives—and freedom of the will because an individual could always refrain from acting upon a motive.

Mill's delicate squaring of the circle in moral philosophy explains the puzzled, if not straightforwardly dismissive, reactions to sparse attempts to ground the principles of political economy in “the afferent trunks of nerve-fibre” as pursued, for example, by an outsider to political economy like Richard Jennings (Reference Jennings1855 [1969]). According to the Irish political economist and Mill adept John Elliot Cairnes, such attempts would turn “political economy into a wholly different subject than the world has hitherto known it”—reason enough to think this alien to the business of the political economist (Maas Reference Maas2005a,Reference Maasb; White Reference White and Mirowski1994).

In the second half of the century, two developments gave increasing credibility to attempts to ground economics in human physiology. The first were developments within psychophysiology itself. (Psycho-)physiologists like Thomas Laycock, William Carpenter, and Henry Maudsley questioned with increasing confidence the existing boundaries between the phenomena of mind and matter that lay at root of Mill's solution to the scientific character of political economy, though not all of them went as far as Laycock in claiming that all mental states were, after all, just emanations of brain states, turning consciousness into an epiphenomenon.Footnote ⁶

The second development was the emergence of thermodynamics that, rightly or wrongly, gave credibility to the idea that mental phenomena were a particular manifestation of energy, just like electricity or mechanical force. These last ideas can be found widely in the nineteenth century, not only in very different intellectual groups in Britain, but all through Europe.Footnote ⁷ The idea that the human body was no more than a heat engine was expressed in France, Germany, and Britain alike and physiologists ardently attempted to discover the precise mechanism of how the human body converted food into useful effect, and it is no coincidence that many of the frontrunners in thermodynamics were frontrunners in physiology and psychophysics as well (Rabinbach Reference Rabinbach1992).

Even John Stuart Mill considered the idea of a conversion of energy into different forms, including mental states, in an exchange of letters with his close friend, the psychologist Alexander Bain, though Bain (a former adherent of phrenology) was more sanguine about this than Mill. William Carpenter's “correlation of forces” was a particular instance of this idea that aimed to salvage a non-reducible status for mental states, and so would save freedom of the will as well. Thomas Huxley by contrast was much more sympathetic to the “shibboleth of materialism” that “thought is a secretion of the brain” and famously expressed this in his essay in the Fortnightly Review in 1874 in which he compared man to a machine (Maas Reference Maas2005a, p. 162).

The unifying concept was the concept of work: Just as heat could be measured by its mechanical equivalent in units of work, so we could measure the mechanical equivalent of the work of the mind.Footnote ⁸ The surprising consequence was a shift from the labor theory of value of the classical economists to a theory of value that histories of economics refer to as “subjective,” but that encompassed mental and physical work. The best examples of political economists in Britain who were clearly inspired by these developments in psychophysiology and thermodynamics are William Stanley Jevons and Francis Ysidro Edgeworth, just the two economists Robbins explicitly referred to as mistakenly relying on “fancy psychological theories.”Footnote ⁹ Jevons dismissed Mill's “convenient” approach to political economy as methodologically naïve in that it apparently only needed the “private laboratory of the mind,” rather than genuine “instruments” and “apparatus,” to measure its subject of study (Maas Reference Maas2005b). For Jevons, by contrast, precision measurement by means of experiments turned mere speculations into matters of fact that could be observed.

To show how mathematical, numerical precision could be given to economics Jevons published an experimental study into the “natural laws of fatigue” in Nature (1870) and it is informative that Fechner at the time made similar experiments. In his New and Old Methods of Ethics, or “Physical Ethics” and “Methods of Ethics” (1877), a highly intriguing work that he published on his own expenses, Edgeworth made a term-for-term translation of the Weber–Fechner “law” into the hedonic calculus.Footnote ¹⁰ Edgeworth consistently translated German psychophysics into the “hedonic” language of pleasure and pain.

According to Edgeworth, economics searched for explanations of the “axiom” of self-interest, and such explanations pointed to a theory that was based in man's physiology. Because “the necessary cerebral investigations” were impossible (Edgeworth Reference Edgeworth1877, p. 20), we needed to take recourse to other, indirect methods of measurement that could provide evidence on the relation of man's physiology to his behavior. Edgeworth considered the new experimental practices of psychophysics of Helmholtz, Fechner, Wundt, and Delboeuf as indirect methods of investigation that were to be preferred over the “introspective marks of brain activity” that were favored by John Stuart Mill and others (Newman 2004, p. 197). Approvingly Edgeworth quoted Alfred Barratt, the author of Physical Ethics (1869), who had argued that bodily states could be used to indicate mental states, just as the thermometer was used to measure temperature (1877, p. 6).Footnote ¹¹

From this brief overview, it should be clear that Jevons and Edgeworth looked at the economic agent as an energetic system that optimizes pleasures and pains. Anson Rabinbach (Reference Rabinbach1992) aptly refers to this system as “the human motor”; man was some sort of engine converting food into useful effect. The lack of direct methods to investigate this “human motor” made Jevons and Edgeworth search for indirect, experimental methods of research. Edgeworth found such an indirect method precisely in the German discourse of (experimental) psychophysics. There was a clear aim to mathematize, to think about agency in functional form and in terms of optimization. For both Jevons and Edgeworth, to link the principles of political economy to man's biological frame entailed the introduction of new tools of research that had been considered alien to the field before then: mathematics, diagrammatic expositions, and, eventually, experiments. To think about the relation of mind and matter in terms of “energy” or “correlation” was present in neurophysiological research at least up to the fifties, as is witnessed from the following quote of the 1950s from the neuroscientist Sir Charles Scott Sherrington: “… we find that the energy-system with which we correlate the mind has of course extension and parts” (quoted from Bennett and Hacker Reference Bennett and Hacker2002, p. 36).

It was from Marshall Hall's early nineteenth century theory of the motor system that speculations about conversions and/or reductions of mental to physiological states all started. I know of no references of Robbins to Sherrington, but it should be clear that Sherrington was building on similar sets of ideas as Jevons and Edgeworth. These were not just “psychological fancies,” but ideas that could get you the Nobel Prize (as was awarded to Sherrington in 1932, the same year Robbins's Essay appeared). But reading Sherrington's words, it was and is not so clear what these have to do with economics; indeed, was the search for the “physical groundwork” of economics not a red herring from the very start?

IV. WEBER, THE GERMAN HISTORICAL SCHOOL, AND PSYCHOPHYSICS

That psychophysics had nothing to do with economics is certainly the message to be drawn from Max Weber's famous essay on the relevance of Fechner's work for marginalist economics. As I mentioned before, Weber published this essay originally in 1908, at a time when his engagement with psychophysics was particularly intense. Around this year he seriously engaged with the work of industrial psychologists, who used their experimental research to intervene in policy debates about the efficiency of work, and the Social Question in general (Rabinbach Reference Rabinbach1992; Brain Reference Brain2001). The major publication that resulted from this engagement was his Zur Psychophysik der industriellen Arbeit that appeared over 1908–1909 in four parts in the Archiv für Sozialwissenschaft und Sozialpolitik, but he remained engaged with psychophysics at least until the end of 1911 (Probleme der Arbeiterpsychologie).Footnote ¹² As a matter of fact, his engagement with psychophysics dates back much earlier and is closely related with the distance he took from his old teachers Wilhelm Roscher and Karl Knies and his rapprochement to the Austrian School.

Weber succeeded Knies in 1896 as professor of economics at the University of Heidelberg. His Roscher und Knies und die logischen Probleme der historischen Nationalökonomie (Reference Weber and Oakes1975a) emerged from a planned Festschrift to honor the university (published in 1902).Footnote ¹³ The end result was a collection of essays that contained a crushing criticism of the approach of the German Historical School to economics. In this book, Weber used Carl Menger's notion of economizing behavior as the relevant benchmark for understanding economic agency in contrast with then current research from psychophysicists such as Wundt, Münsterberg, and others.

Weber agreed with his former teacher Wilhelm Roscher that the “fundamental and substantive problem of economics” was the unintended coordination of self-interested actions: “How are the origins and persistence of the institutions of economic life to be explained, institutions which were not purposefully created by collective means, but which nevertheless—from our point of view—function purposefully?” (Weber Reference Weber and Oakes1975a, p. 80). This was the problem posed by the Dutch physician Bernard Mandeville, in his notorious Fable of the Bees, and “many of his successors, consciously or unconsciously agreed with [his] view: economic self-interest is that power which … ‘always wills evil but does good’” (Weber Reference Weber and Oakes1975a, p. 83).Footnote ¹⁴

But Weber did not agree with the Mandevillian irrationalist psychology Wilhelm Roscher and Karl Knies used in support of their historical approach to economics. Although Roscher generally disagreed with “Mandeville and the Enlightenment,” he followed Mandeville in considering self-interest as an “instinct” on which man invariably acted. Thus, the actions of an individual were conceived as inherently irrational, because instinctive. Building on this psychology Roscher and Knies “hypostatized the concept of an essentially irrational and unique ‘Volksgeist’,” and they even made this “Volksgeist” the individual that transformed through history (Weber Reference Weber and Oakes1975a, p. 61).

Weber contrasted this approach with that of Wilhelm Dilthey and his former teachers Wilhelm Windelbandt and Heinrich Rickert, who searched for “meaning” in history; that is, who aimed to “understand” the actions of individual agents as “intelligible” (Weber Reference Weber and Oakes1975a, p. 65). Thus, Weber introduced his main argument for thinking about human behavior as categorically different from a natural, instinctive process. The crux was the intentional character of human agency—that is to say, its inherent possibility to be interpreted rationally; human actions were susceptible to “a meaningful interpretation.” For that reason, Weber considered “individual human conduct … intrinsically less irrational” than “the individual natural event” (Weber Reference Weber and Oakes1975a, p. 125).

Hence, the possibility of understanding human actions rationally was the distinguishing feature between events in the natural and in the social realm. In the realm of nature we can search for causal explanations, but we cannot ask for a reason. To ask for a reason was to ask for the intentions of the individual agent. To understand individual human conduct was to ascribe “a ‘rational’ interpretation in terms of intentions and beliefs” (Weber Reference Weber and Oakes1975a, p. 127). This did not exclude causal explanations. Rather, as the German sociologist Georg Simmel had argued, to ascribe rationality to an action was to causally explain this action from a reason (Weber Reference Weber and Oakes1975a, pp. 152–154).Footnote ¹⁵

Weber exemplified his argument with one important type of rational ascription that was of particular use in economics, namely the “‘rational’ interpretation which employs the categories of ‘ends’ and ‘means’” (Weber Reference Weber and Oakes1975a, p. 186). From the sequel, it was clear that Weber had Carl Menger's theory of economizing behavior in view. Weber gave the following explanation of the “logic” of this behavior. Given a certain intention or goal x, an actor needs to select a means y, rather than y′ or y′′, as the best means (on the basis of existing empirical evidence) to obtain that goal. Though such judgments were evaluative, they were empirical in the sense that, for example, y rather than y′ was more appropriate in terms of sacrifices. Weber emphasized there was nothing “subjective” in such estimations; everybody would come to the same conclusion and so nothing hinged on one's psychological dispositions (Reference Weber and Oakes1975a, pp. 186–191).

Weber thus disentangled the iron chains between self-interest and economics that had been such a dominant feature of nineteenth century political economy. By contrast, for explanations in terms of means and ends, no assessment needed to be made of the psychology of the individual. The only assumption to be made was that an individual was able to rank the objectively given means in regard to a given end. According to Weber, that was trivial in terms of logic, not in terms of psychology. Though evaluative, such an analysis was completely “objective” in the sense that the different possible courses of actions could be assessed by anyone. No recourse needed to be made to the notion of preferences or to any similar notion. In the sense of being the “best” choice, such an evaluation could be called “rational” (Weber Reference Weber and Oakes1975a, p. 191).

Weber emphasized the “self-evident” character of such calculations. As an outcome they produced a causal account of action, though not nomological as in the natural sciences. There might easily be deviations from the “best,” that is “rational” course of action, and such deviations might be explained as the “nonrational elements of actual economic action” (Weber Reference Weber and Oakes1975a, p. 189). For that reason, Weber called such explanations “idealtypical,” to distinguish them from the causal nomological accounts of the natural sciences; one could not infer actual actions from idealtypical evaluations, but only the “objectively possible” courses of action (Reference Weber and Oakes1975a, pp. 189–190).

This was how rational explanations were used in economics and the social sciences more broadly. In economics, the assumption of “pure rationality” served to “theoretically” deduce the consequences of “economic situations” (Weber Reference Weber and Oakes1975a, p. 189). Similar to the use of rational explanations in history, these deduced consequences could be compared to their real world counterparts to distinguish the rational from the nonrational aspects of concrete actions. Though lawlike, rational explanations differed from natural law explanations in that they provided only interpretations of events, as opposed to a “law of nature” that “must be true in concrete cases” (Weber Reference Weber and Oakes1975a, p. 190). Means–ends interpretations thus implied the rationalization of empirical reality, but did not imply the rationality of reality itself. They served to circumvent the concrete complexity of empirical, real-world situations by proceeding on the assumption that these situations were the result of intentional optimizing actions.

Thus, Carl Menger's theory of human economizing behavior was for Weber a particular instance where the understanding of economic agency in terms of rationality produced causal explanations. On this notion, Weber argued, there was no need to explain man's economic behavior from his biological or psychological frame. Even though Weber noted the “extraordinarily importance” (Reference Weber and Oakes1975a, p. 111) of the psychophysical experiments of Wundt, Münsterberg, and others in his essays on Roscher and Knies, they did not illuminate the problem of choosing the best means for a purpose.

V. WEBER'S 1908 ESSAY ON FECHNER AND MARGINALISM

Weber reached similar conclusions in his famous 1908 essay on the relation of marginalism to psychophysics. Weber was prompted to write the essay by Lujo Brentano's enthusiastic account of recent developments in economics, in which Brentano stressed the importance of psychophysics for economics (Brentano Reference Brentano1908).Footnote ¹⁶ From the preceding it will come as no surprise that this claim was in Weber's eyes untenable. But now Weber was more explicit about the economists’ engagement with explanations in terms of rankings of needs. Weber was persistent that such rankings could be made quite independent of the experimental investigations of the psychophysicists.

According to Weber “common experience” taught us that “men … are motivated by ‘needs’.” Common experience taught us also that people made a ranking of needs according to their “urgency.” And common experience taught us that men were able to act “expediently,” that is on the basis of “prior calculation” (Weber Reference Weber and Schneider1975b, p. 29). Weber claimed that these assumptions did not ask for any detailed investigation into the psychological complexity of these needs. Weber was well aware that from the viewpoint of a psychologist, concepts like “purposive action,” “experiencing,” or “prior calculation” were all highly complex. But this was simply irrelevant to the economist (Reference Weber and Schneider1975b, p. 30). On the basis of “entirely trivial, but undisputable facts of everyday experience,” the economist could “theoretically conceive of a relatively large number of people” each of whom used his available resources “for the sole and exclusive purpose of peaceably achieving an ‘optimum’ of satisfaction of his various competing needs” (Weber Reference Weber and Schneider1975b, 29).

To think of such trivialities as the “foundations” of economics was perhaps dazzling, but “yet this is the situation” (Weber Reference Weber and Schneider1975b, p. 30). Weber emphasized that the economist did not need to recast these everyday facts “to make them susceptible to the psychologists’ usual work with revolving drums or other laboratory apparatus!” He could even obtain “mathematical formulations for his theoretically conceived course of economically relevant action” without taking recourse to any refined psychological theories, because his concern was not an investigation into man's psychology, but an evaluation of given means to an end (Weber Reference Weber and Schneider1975b, p. 30).

Interestingly, Weber added an historical note in which he claimed that the notion of means–ends rationality had gained increasing empirical validity in the “capitalist epoch.” To think about the agent as optimizing the use of limited means for a given end was exactly the situation of “an agent who constantly carries on ‘economic enterprise,’ and it treats his life as the object of his ‘enterprise’ controlled according to calculation” (Weber Reference Weber and Schneider1975b, p. 32). In so far then as the economist assumed anything about the individual's “psyche,” he assumed a “merchant's soul” (Reference Weber and Schneider1975b, p. 32). Just as a merchant was able to numerically rank the “intensity” of his needs, and the means to fulfill them, so did the economist theorize on the “increasingly true assumption” that “everyone were to shape his conduct towards his environment exclusively according to the principles of commercial bookkeeping—and, in this sense, ‘rationally.’” (Reference Weber and Schneider1975b, pp. 32–33).

For Weber, rationality was “an approximation to reality that has implicated the destiny of ever-wider layers of humanity. And it will hold more and more broadly, as far as our horizons allow us to see” (Weber Reference Weber and Schneider1975b, p. 33). The truth of economics rested on the worldwide spread of capitalist accounting techniques, not on the truth of the psychophysics of Fechner and others. Indeed, the economist's approach was the very “opposite” of any psychology (Weber Reference Weber and Schneider1975b, p. 32).Footnote ¹⁷ As Weber remarked in his criticism of industrial psychology, his methods were “far off the methods of measurement that we find in the laboratory of the experimental psychologists” (1995, p. 128).

We have seen that Weber reframed topics that had been of concern to John Stuart Mill in a new setting. He replaced Mill's recourse to a categorical split between the realms of matter and mind by a categorical split between understanding natural causal relations and understanding the causal structure of human actions as fundamentally intentional and goal-oriented. For Weber, intentionality and rationality were different sides of the same coin. Means–end rationality enabled him to steer away from a discourse in which an economic agent acted mechanically out of self-interest. On the macro level rationality served to distance his own “interpretative” approach in the social sciences from the irrationalism of the Historical School; on the micro level rationality served as the foil against which to understand individual human conduct in causal, yet non-mechanical terms.

As a consequence, Weber could deny the relevance of physiological (and psychological) research to the concerns of economists, and he could substitute means–ends rationality as its simple and effective alternative. In economics, no recourse needed to be made to the physiology of mankind, or any other psychological theory; means–ends rationality was the panacea to all. Historically, this kind of rationality was linked to the rise of capitalist accounting techniques—the “merchant's soul.” Though Weber criticized the Historical School for its irrationalism, his own conception of economics was still fundamentally historically oriented. His historical orientation and his rejection of psychology made the experimental method of no interest to economics.

VI. ROBBINS, WEBER, AND THE ECONOMIC PROBLEM

From the foregoing, it will be immediately apparent how important Weber was for Robbins in restructuring his own discussion of the relation of economics and psychology. In the first edition, we have seen, Robbins referred to Weber's essay on psychophysics and marginalism, which he considered “conclusive” (1932, p. 85n3). Robbins also noted that Weber's analysis of the “logical basis” of economics rested “upon a certain degree of misconception” (1932, p. 85n3). We have also seen that Robbins's own hand in distinguishing economics from psychology in the first edition was not completely felicitous, because he got jammed into a discussion of the idea of “scales of valuation” that, as Hands (Reference Hands2009) observes, was itself too behaviorist to convincingly separate the problem of economics from an approach in which economics was considered to be about selfish conduct that more or less mechanically played out in the market.

In this thoroughly rewritten part of his argument, Robbins expanded on the meaning of means–ends explanations, and how these differed from explanations in the natural sciences. While Robbins's discussion in the first edition still was framed in the behaviorist terms of stimulus and response, now his discussion was far more aligned with Weber's notion of intentionality. At first Robbins followed Weber's analysis of means–ends rationality. The concept of purposive conduct had a structure that could not be fully captured by the methods of the natural sciences, especially because ends were unobservables. This did not mean that means–ends explanations were unscientific. Rather, this very notion was sufficient to serve as a causally explanatory term—to enter “in the chain of causal explanation” (1984, p. 90).

Given Weber's strong rejection of the relevance of psychology to economics, it comes as a surprise that Robbins also argues that if we were to do our “job as economists” right, we “must include psychological elements” in our explanations (p. 89). In contrast with Weber, Robbins conceived of purposive conduct as a psychological notion, though not related to one of those “fashionable” theories that change once every five years (p. 90).Footnote ¹⁸ Though Weber would have agreed that there was a difference with natural science explanations, Robbins argued that the structure of means–ends rationality was “psychical, not physical.” There was a point then where economics did depend on psychology, but not in terms of a naturalistic kind of psychology; that is not by means of psychological theories that try to explain mental states in terms of physiological states.

Robbins's rejection of the relevance of specific psychological theories was therefore to embrace another theory that hinged on an explanation of rationality in terms of means and ends, which was rooted in the daily experiences of individuals (pp. 91–92). For Robbins this meant that economics, one way or another, included “psychological elements” in its explanations (p. 89). Weber, we have seen, did not consider means–end rationality so much as a psychological theory, but as an aspect of human behavior in a specific historical setting. Its objective structure could be assessed by everyone, independent of one's psychological dispositions.

We can explain this different assessment of the relation of means–ends explanations to psychology if we pay attention to the differences in emphasis on the variety of needs an individual aimed to satisfy. Even though Weber made mention of this, his whole discussion of means–ends rationality was intended to show its structural characteristics and its relation to a specific historical epoch and a particular exemplary agent (the merchant) for whom there basically was only one goal: profit. This enabled Weber to downplay the importance of psychology altogether. Things look very different for Robbins. For Robbins economics was centrally about choice in an economizing context. Weber could disentangle economics from psychology by emphasizing the objective structure of means–ends rationality. Robbins, by contrast, did not want to follow Weber's argument, because the mere formal structure of means to ends obfuscated the way this structure worked out in economics.

To see this we need to go back to Robbins's discussion of means and ends in the second chapter of the Essay, a discussion that remained largely unaltered in the second edition. This chapter was about the crucial difference between a technical and an economic problem. Both had the structure of means to ends, but in very different ways. Whereas a technique was about means and ends, it was only about choosing the best means to a specific end, for example making fire. An economic problem emerged only when there were various ends that could not all be satisfied at the same time—a situation of scarcity. It was only here that one could speak of economizing behavior. The problem then became, as Robbins explained on the economists’ favorite example of Robinson Crusoe, “how much wood to use for fires and how much for fencing” (1984, p. 35). That was “no longer a purely technical problem” (p. 35). Robbins referred to “Professor Mayer's very elegant way of putting the distinction, the problem of technique arises when there is one end and a multiplicity of means, the problem of economy when both the end and means are multiple” (p. 35). That was the problem of choice studied by the economist; the economist studied choice-rationality, not instrumental rationality, where instrumental rationality dealt only with choosing the best means to reach a goal. In the second edition Robbins emphasized that the economist solved this problem of choice by means of indifference curves. The optimal use of resources did not just hinge on “technical possibilities,” but could be solved only when the “consumption indifference curves are known” (p. 35n1).

I already briefly mentioned Robbins's remark in the first edition of the Essay about “a certain degree of misconception” in Weber's view of economics, a remark he dropped in the second. I guess that this misconception is exactly situated in Weber's ambiguity on the distinction between instrumental and choice rationality. And I also think Robbins was right.Footnote ¹⁹ If we follow for example the German reception of Weber, we find that economics and technology by and large have been regarded as dealing with the very same problem, even though German sociologists were aware of the difference between instrumental and choice rationality.Footnote ²⁰

Focusing on the formal structure of means to ends enabled Weber to consider this structure, as we have seen, as an “objective” fact and to discard psychology. For Robbins, the variety of an individual's needs was central and so it was impossible to merely focus on the formal structure of means to ends; once an individual's needs were important, psychology was back in the picture. Robbins minimized the importance of psychology to the “simple and indisputable facts of experience”—that is the assumption of transitivity of preferences, something that could be verified on a day-to-day basis. But this was psychology nevertheless, even though it explicitly was not the Fachpsychologie that was embraced by Jevons and his followers. In that sense, we might well argue that Giocoli's discussion of the “escape from psychology” appertains more to Weber than to Robbins (Giocoli Reference Giocoli2003, pp. 79–90).

As a consequence Robbins cut the boundaries between economics, psychology, and history very differently from Weber. While Weber situated economics historically, Robbins pushed the economic problem outside of history. Competition was to be found in any problem of choice. As such, it was not a historical problem, but a problem that would arise in any situation where a variety of needs competed for scarce means. Rather than being a social phenomenon, competition became a phenomenon that was inherent to the economic agent; it was what economizing behavior was all about. Historical or sociological accounts could thicken the description of how these conditions played out in specific circumstances, but they could not alter the general problem-setting the economist faced. Against Marx and materialist social theory Robbins argued that the Marxian dictum that “the history of tools is the history of mankind” ignored the “autonomous changes on the demand side” that were expressed in the variety of needs that simultaneously needed to be satisfied (1984, p. 44). These needs formed the “‘psychological’ (or, for that matter ‘physiological’) side to scarcity” (p. 44). Without them, there was no economic problem separate from technology.

VII. IN CONCLUSION

If we compare the very different stances on the relevance of psychophysics to economics of Jevons and Edgeworth on the one hand, and Robbins and Weber on the other, we see that they entail very different views of the economic agent. For Jevons and Edgeworth, economics was concerned with an analysis of how individuals, considered as pleasure machines, behave in the market. Their explanatory strategy was to turn to man's biological frame and to consider how this behavior followed from an alleged optimization process of man's energetic dispositions. Balancing pleasures and pains, Jevons considered economic man (a “trading body,” not a self-conscious individual) as an energetic system following the principle of least action. This was not a social, but a natural fact—pleasures and pains governed the individual by necessity. For Edgeworth something similar can be said.

Things look very different for Max Weber. His focus was the business man, the “merchant” who used “double book-keeping” to decide on the best course of action. There was no assumption needed about his psychological or biological frame, whether such assumptions were about “instincts” or “self-interest” or whatever. Rather, the economic agent acted in a specific historical (and social) context that put one specific goal, profit maximization, up front. Given this goal, everyone could assess the best means to obtain this end. That was not a natural law; it was a rational assessment that nevertheless provided a causal explanation for economic action. For Weber, it was no different for other economic agents, and for that reason no assumptions were needed about mankind's biological or psychological constitution. For Weber the emergence of a market economy was a social, not a biological fact.

Weber's criticism of psychophysics served Robbins to stake out a neutral playing ground for the economist. Weber's understanding of means–ends rationality helped Robbins to steer away from convoluted associations with the “universal bogey-man” of economic man—solely motivated by self-interest—and utilitarianism and other programs that crossed the borders between the objective statements of science in which ends were taken as given, to policy prescriptions where ends were set. In contrast with Weber, Robbins disentangled economics from history and re-aligned it with psychology, though on a limited basis. Weber mentioned, but did not acknowledge, that the assumption of ordered preferences, or needs, at the end of the day was a psychological assumption. Robbins grounded this assumption on the daily experiences of economizing individuals. To economize should be distinguished from solving a technical problem. The “misconception” Robbins may have perceived in Weber's work was that both were not sufficiently distinguished.

I think it is undeniable that twentieth-century economists by and large subscribed to the basic tenet of Robbins’s reasoning. With all changes in the concept of rationality over the twentieth century the basic instinct of economists has been to take rationality as the decisive causal factor in social explanations. Well into the 1970s, means–ends rationality was what rationality meant to an economist. In his Fels lectures Arrow wrote that an economist “by training thinks of himself as the guardian of rationality, the ascriber of rationality to others, and the prescriber of rationality to the social world.” At the end of the day, rationality was about means and ends: “Rationality, after all, has to do with means and ends and their relation” (Arrow Reference Arrow1974, pp. 16–17). Arrow's phrasing conceals however the distinction Robbins made between a technical and an economic problem, and it was just this distinction that reconnected economics to psychology.

After Robbins, most economists fell into a kind of Kantian dogmatic slumber before being awakened by systematic violations of the trivially easy ranking of A over B over C somewhere in the seventies (Bruni and Sugden Reference Bruni and Sugden2007). It was only once it became increasingly indisputable from psychologists’ experiments that even very simple choice situations are really not that simple, that economists gradually became convinced that the almost tacit assumption that individuals can rank their preferences, should, perhaps, be exchanged for a reinvestigation of how individuals do or do not optimize in concrete situations. That does mean, for a growing community of economists, that one needs to reinvestigate the “physical groundwork of economics.”

It is somewhat ironic, I think, that for some contemporary economists, such a reinvestigation leads them to use Robbins's Essay in support of their endeavors. If the crucial difference between an economic and a technical problem is, in modern terms, the existence of opportunity costs, there is no reason per se to stop at the level of the individual agent: any situation in which various needs compete over scarce means is then open to the toolbox of the economist.Footnote ²¹ The irony is that while Robbins steered away from attempts to vivisect the economic agent, nowadays his definition is used to do just that.