|
|
|
|
CIAO DATE: 03/02
Critical Review
Winter–Spring 1998 (Vol.12 Nos.1–2)
Boettke’s Austrian Critique of Mainstream Economics: An Empiricist’s Response
By Thomas Mayer *
Abstract
Many of Boettke’s criticisms of formalist economics are justified. However, he defines formalism so broadly that it becomes practically synonymous with mainstream economics, while his criticisms primarily target the sins of formalist economics more narrowly defined. And since he treats Austrian economics as the only viable alternative to mainstream economics, he incorrectly awards victory to Austrian economics. While Austrian economics has some valuable ideas to contribute to mainstream economics, it has serious deficiencies of its own.
Starting in the late 1940s, economics underwent a major revolution, a revolution at least as much in methodology and “scientific taste” as in substantive doctrine. Previously economists, like other scientists and serious scholars, treated rigor as an important virtue. But in the last half-century the value attached to rigor, relative to the other desirable characteristics of a theory—such as empirical confirmation and relevance to practical problems—rose sharply. The leaders of this formalist revolution now seem to classify all economic analyses that are not mathematically formalized modeling as “mere talk.” One might be excused for gaining the impression that formalists consider this dichotomy to be more important than the contrast between analyses that are empirically confirmed and those that are empirically disconfirmed.
Despite its insistence on explicit statements and rigor, the formalist revolution was not itself founded on a detailed and rigorous analysis of why formalism is superior to the older way of doing economics. Typically, formalists recite the advantages of rigorous mathematical treatment, but fail to compare them to the advantages of less formal treatment (cf. Backhouse 1997; Montgomery 1997). The great success of formalism should therefore be attributed not entirely, or necessarily even in large part, to superior logic, but, at least in part, to its advocates being brilliant scholars, highly respected for their substantive work. Another explanation is that—at least to those who confuse mathematics and empirical science—formalism fits the image of economics as a science, and that is the image economists want to project. Moreover, because—within the confines of mathematical models—formalist analysis provides a high degree of certitude, many economists find formalism comforting. That it sharply differentiates the professional economist’s product from cocktail-party chatter is also satisfying. In addition, it has allowed the young, who generally had better mathematical training than their elders, to achieve dominance.
But although the formalist revolution succeeded in establishing rules to which economists must adhere if they want to publish on mainstream topics in the “respectable” journals, it did not capture the hearts and minds of all economists. It is not certain even that it commands the adherence of half of all the academic economists, and it is likely that outside of academia only a distinct minority of economists accept it. But the tone of economics is set by those who publish in the leading journals and teach in the major research universities. There, formalism dominates.
But by now counterrevolutionaries are active. Peter Boettke is one of them. Being a counterrevolutionary myself, I applaud his attack on formalism, though I define it much more narrowly than he does (see Mayer 1993 and 1995). Counterrevolutionaries, like other revolutionaries, are a quarrelsome lot, so rather than endorse Boettke’s criticisms by repeating them, or repeating my own previous criticisms (Mayer 1993), I will criticize some aspects of Boettke’s case against what he calls formalism, while omitting the many points about which I can only say “right on.”
There are two themes running through Boettke’s essay, a powerful criticism of contemporary economics for ignoring important older insights that do not fit into the Procrustean bed of formalized analysis, and an acclaim of Austrian economics as the best alternative. I will deal primarily with the second theme. In summary, I criticize Boettke mainly for confounding formalism with a broader version of mainstream economics, thus posing a false dichotomy between formalist and Austrian economics and overstating the contribution of the latter.
Before coming to these criticisms one should note two of the strong points of Boettke’s essay. One is his distinction between the idealizations used in economic theory and the criteria that should be used to judge the functioning of an economic system, a point discussed below. Another is his insistence that when discussing broad issues of economic policy, we have to look beyond what economic theory on its own can tell us. Institutions do matter.
Since Boettke gives a prominent role to Abba Lerner as a formalist and proponent of market socialism who got that subject wrong because he ignored institutions, it may be appropriate to digress with a story Lerner once told me. After he wrote The Economics of Control, a book that pays scant regard to institutions, he believed that he had demonstrated that it matters little whether a country is capitalist or socialist, as long as its managers follow the correct optimization rules he had set out. But he added that he preferred socialism because (if I remember correctly) it allows a better distribution of income. However, he said that subsequently he changed his mind—because of false teeth. On a visit to Israel he noticed that it did a thriving export business in false teeth. This, said Lerner, was logical because making false teeth is a business that requires much skilled labor and little capital. But, Lerner added, no government planning agency intending to increase exports would ever think of false teeth. Since Boettke stresses the importance of innovation under capitalism, there is therefore much less disagreement between him and Lerner than he suggests.
Boettke’s Critique of Formalism
Boettke writes that in describing economic behavior in mathematical language, formalists underplay the complexity of the real world. Hence formalism swept away “historical work on the complex web of institutions that undergird capitalist dynamics” (1997b, 21). Formalists, such as Samuelson, have “drained economic theory of institutional context. . . . Parsimony won out over thoroughness. Formalism denied scientific status to realistic theory. . . . Ideas that defied the techniques of formal analysis came to be considered unworthy of serious consideration” (ibid., 21, 22, emphasis in original.)
Boettke’s delineation of what he calls formalism thus focuses on two separate characteristics. One is formalism’s extensive abstraction from institutional context. He is right in believing that this is required for—and fostered by—mathematical modeling. But not only by mathematical modeling. As he himself points out, all thinking requires abstraction from a wealth of detail. When Austrian economists talk about entrepreneurship, they, too, are abstracting from the real world, ignoring, for example, such “irrelevant details” as whether the firm (entrepreneur) is a closely held corporation, or one with widely dispersed stockholders who have little power over management.
What matters is whether it is the important or the peripheral characteristics that are being abstracted from. But how do we know which are the important ones? When Austrians complain that formalist economics abstracts from the fact that historical time is irreversible, formalists can reply that time being irreversible is an irrelevant characteristic that does not affect the insights and predictions generated by their theories.
Milton Friedman (1953, ch. 1) has argued that until we know whether a theory’s predictions are accurate we cannot say whether the unrealistic nature of its assumptions—that is, its abstractions—matter. An alternative criterion stresses explanation in place of prediction. According to this criterion, the right abstractions yield a theory that makes us understand the phenomenon, providing what Fritz Machlup (1978, 145) called a sense of “Ahaness,” because it elucidates the mechanism by which the supposed cause generates the claimed effect. Though philosophers of science and methodologists argue about the choice between these criteria, most of us pay some attention to both, though we differ in the relative weight we give them. Even if a theory seems to explain well, in the sense of linking a particular phenomenon smoothly to our prior beliefs, we do not accept it if its predictions are consistently falsified. Conversely, we usually reject as a mere spurious correlation a hypothesis that predicts well, but “makes no sense.” But neither approach rejects a theory merely because it abstracts.
One might perhaps respond that some abstractions are so obviously wrong that even without knowing how well a theory based on them predicts or explains, we can say with confidence that this theory must be incorrect. But even though there are instances where this is true (e.g., abstracting away self-interested behavior, so that we assume that firms are driven only by altruistic motives), are there many such cases in economics? Boettke is able to say Yes only because he takes it as a given that Austrian theory is correct. Hence, if formalist economics ignores some variable that plays a significant role in Austrian theory, the formalist theory must be wrong. But why assume, a priori, that Austrian theory is correct?
A more nuanced version of Boettke’s criticism of formalist abstraction is possible, however. This is the view that often mathematical modeling not only requires more abstraction than verbal analysis, but also that it is harder to be mindful of some of the abstractions that have been made when we read a mathematical analysis than when we read a verbal analysis (see Keynes 1936, 297–98). Moreover, it is tempting to decide what part of reality to model and what part to abstract from by the criterion of what is mathematically tractable, rather than by the criterion of what is important for the question at hand. What makes this temptation worse is that modelers usually do not point out that, since they are abstracting from some characteristics that may be salient, their conclusions are of limited value. To be sure, they may seem to guard against this possibility by showing that their model gives a good fit to the data, but that is a necessary and not a sufficient condition for acceptance. The good fit may well be the result of extensive data mining. Moreover, even though the model gives a good fit, previously published rival models may give a better fit. 1
What is important here (and incidentally is also basic to Friedman’s 1953 essay) is that we use a theory and its abstractions to deal with a particular problem or question. A valid abstraction when addressing one question can be an invalid one when addressing another. 2 When trying to determine whether there is a stable relationship between changes in bank reserves and the money supply, we can safely abstract away the fact that the actions of government officials are influenced by their self-interest. But we cannot abstract away the fact that the ratio of the public’s demand for currency relative to deposits varies. Conversely, when we ask, as Boettke does, whether market socialism could work effectively, we can abstract away the stability of the public’s demand for currency, but not the motives of government officials.
Boettke’s criticism of formalism for its heroic abstractions is therefore too general. He would have to show that these abstractions result in theories that neither predict well nor foster understanding. To be sure, he does discuss why abstracting away the motives that would drive government officials under market socialism invalidates any formalistic case for such a system. And although he does not provide any hard evidence, I find his argument plausible. But market socialism is hardly a central topic in the work of formalists, and it was originally presented without formal models by Oskar Lange (1939) and developed further by Abba Lerner (1944), whose use of mathematics generally consisted of simple geometry. Formalists might therefore readily concede that Boettke’s criticism of market socialism is correct, and yet go about most of their business undisturbed.
The second part of Boettke’s criticism of formalism—that the medium tends to become the message—is valid. Formalists tend to evaluate ideas by their suitability for modeling, and to judge models much too much by their technical sophistication and elegance, and much too little by the insights they provide into economic behavior or by their predictive success. Deirdre McCloskey (1985) is right in objecting that economics departments have appropriated the criteria that are proper for a mathematics department: rigor, generality, and elegance. Thus, a leading mathematical economist and mathematician, Gerard Debreu (1991, 5), writes:
| In the past two decades, economic theory has been carried away further by a seemingly irresistible current that can be explained only partly by the intellectual successes of its mathematization. Essential to an attempt at a fuller explanation are the values imprinted on an economist by his study of mathematics. When a theorist who has been so typed judges his scholarly work, those values do not play a silent role: they may play a decisive role. The very choice of the questions to which he tries to find answers is influenced by his mathematical background. Thus, the danger is ever present that the part of economics will become secondary, if not marginal, in that judgment. 3 |
The issue is therefore not the fact that formalist economists use mathematics, but what they use it for, and the role mathematical techniques play in the criteria by which they evaluate work. Someone might use a mathematically sophisticated and complex model to solve a problem that is central to the “study of mankind in the ordinary business of life”—Alfred Marshall’s (1947, 1) definition of economics—without abstracting away those institutional details that are relevant for the applicability of the conclusion. He may thereby produce good economics. Someone else who uses no mathematics might tackle a trivial problem, or use too restrictive abstractions, and thus produce bad economics.
There is another kind of formalism on which Boettke does not touch (see Mayer 1993, ch. 3). As an ideal type, this formalism insists on deriving conclusions to every problem explicitly from first principles, which in economics means the assumptions of utility maximization (in practice often narrowed to mean income maximization) and rational behavior, along with a minimum of other assumptions. This type of formalism takes as its model not the natural sciences, but mathematics and logic, with their greater stress on demonstrative reasoning. By contrast, there is what can be called empirical-science economics, again an ideal type. It is more concerned with predicting or explaining empirically observable characteristics of the economy, and less with rigor, parsimony, and elegance.
In practice, formalist economists also want their models to tell us something that is applicable to the real world: they do not, for example, work with models that assume that agents are completely altruistic. On the other side, empirical economists may sometimes start with the same propositions that formalists use. The difference is that formalists are much more likely to treat these propositions as axioms, while empirical economists are more likely to treat them as working assumptions that are usually, but not necessarily always, correct.
Such a distinction is alien to Boettke’s framework because, though explicitly contrasting the older neoclassical economics with formalist economics, in much of his essay he divides economists into Austrians, (old) institutionalists, Marxists, and formalists. Since he quickly dismisses institutionalists and Marxists, he has only to show the fallacy of formalism to award victory to the Austrians. But that is too facile. It tars the substantial majority of economists with the same brush, despite major methodological differences among them. Those who take an empirical approach—and there are many—are thereby made to share responsibility for the extreme degree to which abstraction is carried by those who are formalists by my much narrower definition. To accuse the likes of Milton Friedman, Charles Goodhart, Joseph Stiglitz, or Paul Krugman, to name only a few leading mainstream economists, of ignoring “the complex web of institutions,” or of putting the medium above the message, is surely wrong.
Boettke’s discussion of the Chicago school illustrates this tendency to place too many economists into the formalist camp. He strongly approves of the early Chicago school led by Frank Knight, who used the equilibrium concept as an ideal type. He is then critical to some extent of Milton Friedman, and to a much greater extent of Robert Lucas, for treating equilibrium, not as an ideal type, but as a description of the economy. But even Lucas, who is much more formalist than Friedman and Knight, did not always see the economy as in almost permanent equilibrium, at least not in his early work on the aggregate supply function (Lucas 1973). There he dealt with an economy in which it takes time for economic agents to discover when the quantity of money changes and adjust their prices accordingly. Similarly, Friedman attributes fluctuations in output and employment primarily to changes in the quantity of money to which prices have not yet adjusted. In general, it is a mistake to see Friedman as a formalist (my definition) on the basis of his famous methodological essay; far from advocating formalism, Friedman rejects it (Friedman 1953, 11–12, 24–25, 277–300; Hirsch and de Machi 1990; Hammond 1996, ch. 2). It is only real-business-cycle theorists who see the economy in almost continuous equilibrium.
But even if one were to interpret Boettke’s attack on what he calls formalist economics only as a criticism of formalism on my narrower definition, it would still go too far. Like most methodologists, Boettke is a monotheist—there is only one true methodology. No tradeoff at the margin is allowed. This might be appropriate if economics had only one purpose. Now in a sense it does: to provide an explanatory and predictive apparatus that ranks high on the criteria of rigor, elegance, generality, and parsimony, as well as on the criteria of accuracy and applicability to many real-world situations. But a theory that fully satisfies all of these criteria would be a rare find indeed. So there is usually a need for tradeoffs, or for multiple theories operating at different levels of abstraction.
One can justify some formalist economics on l’art pour l’art grounds because, like any other piece of rigorous and elegant reasoning, it provides intellectual satisfaction. If we support research in pure mathematics and in art history “for the glory of mankind,” we should also support some research in formalist economics. 4 Moreover, some formalist models can be justified, not as the end product of economic analysis, but as an intermediate product. Allan Gibbard and Hal Varian (1978) have argued that many economic models are “caricatures” intended to highlight some particular feature of the economic process, even though they give a distorted picture of the economy. Other economists can then combine the lessons learned from caricature models to build a more balanced model of the economy. The trouble is that putting these caricature models together into a realistic description of the economy seems less attractive or more difficult than building additional caricature models, so that too little of the former and too much of the latter gets done.
The question is therefore not whether some formalist economics should be done, but how much. The glib answer that some economists give, “Let the academic market decide,” will not do. The academic economist’s “market” consists of other academics, not consumers who pay with their own money for what they demand, or institutions that are held closely accountable by the general public or by students (the ultimate customers of this research.) Academic economists tend to treat as interesting problems those that are technically difficult, even if they do not have much bearing on how the economy functions or on policy choices. Thus, in academia (or more precisely in those fields of academia where the influence of academics predominates over those of practitioners), the interests of producers tend to outweigh those of consumers, a clear example of market failure. I therefore agree with Boettke that there is far too much formalist research (on my narrow definition), but I would not like to see all of it eliminated.
Markets versus Planning
Another problem with Boettke’s analysis is that he focuses so much on the problem of free markets versus planning, an issue on which the Austrians have centered much of their attention. This is obviously an important issue for economics. But it is only one of many. Most papers in economics journals either deal with completely different issues, or, if they do deal directly with the question of free markets versus planning, they do so in a specific and narrow context, such as flexible versus fixed exchange rates. Mainstream economics devotes much of its effort to small questions rather than basic systemic questions. For example, the first paper in the September, 1997 issue of the American Economic Review deals with an empirical model of international specialization, and the next three papers have the following titles: “The International Transmission of Financial Shocks: The Case of Japan,” “A Political-Economic Analysis of Free-Trade Agreements,” and “An Empirical Assessment of the Proximity-Concentration Trade-off between Multinational Sales and Trade.” Even the two papers in that issue that do compare economic systems (“Privatization in Eastern Germany: Management Selection and Economic Transition” and “Competition or Compensation: Supplier Incentives under the American and Japanese Subcontracting Systems”) do so in narrowly circumscribed ways. This narrow focus has proved fruitful. Normal-science research is how a mature field, which economics has become by now, makes its day-to-day advances.
Boettke and David L. Prychitko (1994) tell us that in recent years Austrians have moved away from their traditional emphasis on ideology. But their research agenda is still too strongly influenced by the “big” issue of markets versus planning.
In addition, the issue of planning versus free markets is a problem that belongs only partly within economics. As Boettke rightly points out, it does little good to contrast an idealized picture of market socialism with capitalism as it works in practice. But determining how market socialism would work in practice raises some exceedingly difficult issues for which an economist’s expertise is not sufficient. For example, would pressure groups induce the state to prop up firms that should be allowed to fail? This is a question for political science, and not for the unreflective application of economic assumptions to political phenomena. Sociological questions also arise: what elites would replace the capitalist elites, and what effect would that change have? We know by now that social engineering is fraught with uncertainty and danger; the law of unintended consequences rules. There is a saying in the military: no plan survives contact with the enemy. Something similar applies to economic policy.
What economists can do much better than evaluate how market socialism would work in practice is to compare the actual workings of a market system with the workings of an idealized market system. And what one can do well is what one does, particularly if one is an academic. This creates a natural bias against the free market (though one that may perhaps be fully or more than fully offset by biases in the other direction.) Boettke is fully justified in complaining about this bias. Where Boettke is less justified is in rejecting the work of those who, like Stiglitz, George Akerlof, or Herbert Simon, among others, have analyzed the implications of factors such as asymmetric information and limited rationality. To be sure, such analyses can be used to demonstrate that the actual functioning of a market system falls short of its ideals. Stiglitz et al. generally do not quantify the losses from these shortfalls, and perhaps they are minor. All the same, they exist. Students of market processes, such as the Austrians, should welcome, rather than deplore, any work that tells us more about how markets operate, even if it creates a bias in favor of planning on the part of those who fail to reflect sufficiently on the shortfalls of actual, as opposed to ideal, socialism. In general, we do not object to increased knowledge because some people may misuse it.
Ideology in Economics
Another problem is Boettke’s treatment of rival schools as driven by ideology. In one sense ideology is the metaphysical core of a research program, and therefore unobjectionable. 5 In another sense, used by those who complain about the ideology of their opponents, ideology is a tendency to accept or reject evidence on the basis of whether it fits one’s preconceptions, or a tendency to select theories on the basis of their policy implications, instead of the other way round. As such, ideology demonstrates obtuseness or lack of intellectual honesty, at least with oneself, if not with others. To accuse someone or some school of being ideological is therefore to make a serious charge, albeit in relatively polite language. Unless one has evidence to back up that type of charge it is better not to make it. To be sure, if an intelligent and well-informed person rejects my excellent arguments, I am tempted to attribute this to willfulness rather than to the unconvincing nature of my arguments—which, after all, I find utterly convincing. But this temptation should be resisted.
However, there is still another way of looking at ideology: to treat it as undesirable in many cases, but also as natural and sometimes as justified. It is natural to give more credence to evidence that supports one’s position than to evidence that rebuts it, if only because this eliminates an unpleasant feeling of cognitive dissonance. Moreover, in some cases this tendency may be justified. Suppose I have much evidence that a proposition is correct. I am now presented with evidence that it is false. Hence, I must usually reject either all the old evidence that supports it, or else the new evidence. 6 Suppose that, upon careful reflection, I cannot find any errors either in the supportive evidence or in the new contradictory evidence. One possibility is to say that I do not know whether the proposition is true. But suppose I have to make some decision, or that the supportive evidence is very strong. It may then be reasonable for me to adhere to my former belief, and to treat the new, contradictory evidence as an anomaly that will sooner or later be somehow resolved, even though this makes me seem ideological. 7 For all of these reasons, it is better not to attribute ideology to one’s opponents.
Moreover, it is far from clear that one should blame antimarket ideology on the rise of formalism, as Boettke does. Institutionalism, too, can be used to justify interventionism. Indeed, institutionalism has been the traditional source of interventionist arguments in economics, while neoclassical theorists have usually defended the market system.
New Keynesian Theory
Boettke’s treatment of the New Keynesians does not do them justice. New classical economists had attacked Keynesian theory for lacking microfoundations for its basic assumption of wage or price inflexibility. In response, New Keynesians have shown that due to a number of factors (among which the efficiency wages that Boettke stresses are not necessarily the most important), wage and price inflexibility are consistent with rational income maximization, thus refuting the new classical criticism. To be sure, New Keynesians have not succeeded in measuring the absolute or even the relative importance of the various factors they analyze, so their demonstration of the mere existence of these factors does not allow us to claim that we necessarily live in a Keynesian (or monetarist) rather than a new classical world. However, their primary task was not to do that, but merely to demonstrate that models with wage or price inflexibility are consistent with a belief in rational utility maximization, and do not require any implausible ad hoc assumptions. That wages and prices actually are not very flexible is an empirical claim (for which there is substantial empirical evidence) that needs to be added to New Keynesian theoretical models to refute new classical theory.
Austrians and Post Keynesians
The criticisms of mainstream economics that Boettke and other Austrians make from a right-wing perspective have much in common (both substantively and in style of argument) with the criticisms of the Post Keynesians, who represent the left wing of Keynesianism. Post Keynesians are, like Boettke, critical of formalism and its focus on equilibrium, and like the Austrians, they want to reintroduce historical time into economics. They also stress the prevalence of uncertainty, and its distinction from mere risk where the probability distribution is unknown. Thus, suppose one accepts Boettke’s arguments about the inadequacy of mainstream economics in these respects. One does not then have to draw the same conclusions as he does; instead one could become a Post Keynesian. Again, Boettke’s tendency to see methodological disagreements in economics as a dichotomy between the Austrians on one side and everyone else on the other muddies the waters.
The Role of Austrian Economics
Boettke’s paper is at least as much a pro-Austrian tract as it is a criticism of formalism. It is useful to distinguish two roles that Austrian economics could play. One is to supplement mainstream economics, and the other is to replace it.
As a substitute for mainstream economics, Austrian economics is not likely to be successful, and for good reason. Mainstream economics does have many shortcomings—particularly the version I have called formalist economics. But it also has many successes, especially when accompanied by an empirical mindset, as it is in the work of economists such as Anthony Atkinson, George Akerlof, Friedman, Franco Modigliani, Robert Solow, and James Tobin, to mention just a few modern masters. It has produced testable—and confirmed—hypotheses about important aspects of economic behavior (see Mayer 1995, ch. 11). We would be the poorer if this work were to cease, or if it were to become only a minor tributary to the stream of economics.
Moreover, Austrian economics suffers from several weaknesses. One is the Austrians’ tendency to address their discussions primarily to each other, to spend too much time questioning the purity of each other’s doctrine, and to debate minor deviations among the elect; in other words, to behave like a stereotypical sect.
Another weakness is the Austrians’ concentration on “big” problems instead of on the day-to-day problems that constitute the work of a normal science. In Lakatosian terms, too much of Austrian research deals with the metaphysical core, and too little with the protective belt. Austrian economics needs fewer generals making grand plans, and more privates fighting the war one hill at a time. (But Boettke should not be blamed for this failing of Austrian economics. On the contrary, he has been a leader among Austrians in emphasizing “small,” manageable problems.)
What is discouraging is that when Austrians have dealt with small problems, often the results have not been fruitful. Boettke’s The Elgar Companion to Austrian Economics (1994) has a section of short papers on applied economics, presumably intended to illustrate the fruitfulness of Austrian economics when applied to specific problems. I do not think that the section succeeds in this effort. Many of these essays do make valid points, but too often these are points originally made by economists who are more closely identified with mainstream than with Austrian economics. Showing that they are also consistent with Austrian economics is not enough to show that Austrian economics should replace mainstream economics. When these papers do cite specific Austrian contributions, these are usually insights that originated a long time ago—suggesting, again, that Austrian economics is not a progressive research program.
All in all, despite its intellectual ferment in the 1980s (see Lavoie, 1994a), modern Austrian economics has so far, at least, not been very productive with respect to those problems that interest mainstream economists, and also, I would argue, those that interest the general public. Part of this is due, of course, to Austrian theory not providing the rationalizations for government intervention that much of the public and many economists welcome. But that is not the whole story. While in the hands of an intellectual giant like Hayek, Austrian economics is a powerful tool, in the hands of lesser economists it is not as powerful a tool as mainstream theory is. And that is a serious drawback. Inevitably, most researchers are not giants, and a productive methodology is one that provides ways in which they, too, can make useful contributions. We should evaluate methodologies not only in the context of verification, but also in the context of discovery.
The lesser fruitfulness of Austrian economics in the hands of most economists is due not only to its previously discussed focus on “big” problems, but also to the much smaller role that Austrians accord to that great reservoir of tasks for many economists, empirical work. Many Austrians consider econometrics to be useless. Leading Austrians reject on methodological grounds the validity of aggregates, such as the price level. Thus Hayek (1935, 5) wrote: “From the very nature of economic theory, averages can never form a link in its reasoning.” Austrians therefore reject as inapplicable to the study of society the instrumentalism that is used so much in the natural sciences. This position strikes me as mistaken, but not being a philosopher, I will not attempt to discuss it further.
If they reject econometrics, Austrians could instead use qualitative economic history as a testing ground, particularly if they are willing to use averages. Although some have done so—for example, by arguing that an unregulated banking system has worked well in the past (for a summary see Schuler 1994)—on the whole, Austrians do little historical work, because of their strong abhorrence of historicism with its disdain of economic theory. But economic history can be used both to support and to make use of economic theory, instead of as a substitute for it. So it is not surprising that some Austrians—including Boettke—are now urging their colleagues to make more use of economic history, and to test both the correctness and the explanatory power of their theories (see Boettke, Horwitz, and Prychitko 1994). Indeed, Boettke (1994, 5) has argued that the claim that Austrians reject empirical research is “a misreading of the Austrian tradition,” though he adds that Austrians also believe that “the significant debates in the social sciences are theoretical debates.”
All the same, Austrians do less empirical work than mainstream economists do, perhaps because they want to engage in the more significant debates. Another reason might be that many Austrians adhere to a mentalist subjectivism that suggests that empirical work is not likely to be successful (see Lavoie 1994b). By rejecting econometrics and making only very limited use of economic history, Austrians have denied themselves the opportunity to test their theories, though this, too, has changed somewhat in recent years (see Lavoie 1994a).
Austrians might respond, along with some mainstream economists (for instance, Robbins 1932), that empirical testing is not needed, that it suffices to construct a valid logical chain beginning with axioms we know from introspection and casual observation to be true, and ending with whatever conclusion follows deductively. But while that may allow us to argue that a certain effect does occur, it usually cannot tell us whether it is significant or not. For example, Hayek (1935) attributed business cycles to the changes in the relative prices of various goods induced by changes in the money supply. Skeptics might respond that, while such relative price changes contribute to business cycles, it is possible that they explain only a trivial proportion of the fluctuations we experience as business cycles. Similarly, for a long time Austrians have talked about the importance of complementarity among various types of capital. But how much complementarity is there? Only recently has an economist (and one who has only tenuous links with the Austrians) provided empirical evidence that such complementarity does matter empirically (Montgomery 1996).
Austrian Economics as a Complement to Mainstream Economics
Viewed as a supplement to, and not a replacement for, mainstream economics, Austrian economics has something to contribute. Mainstream economics has narrowed its vision to a limited set of ideas that modelers find tractable. The Austrians therefore have something valuable to contribute when they insist that the entrepreneur is more than some nondescript graduate of an MBA program who mechanically grinds out the profit-maximizing solution to a standardized problem. Similarly, their insistence that price competition is only one aspect—and not necessarily the most important aspect—of the competitive process adds a valuable insight that game-theoretic models of oligopoly are likely to miss. Insistence on entrepreneurs being innovators operating in a fog of uncertainty, as well as insistence on irreversible historical time, are other Austrian ideas that may deserve much more attention than the mainstream gives them.
One might liken the difference between mainstream and Austrian economics to the difference between a flashlight and a lantern. The former illuminates sharply along a narrow beam. The latter provides all-round but less sharp illumination. Which one is preferable depends on one’s purpose, and neither should necessarily be rejected because the other is better at solving some problems. Consider, for example, Boettke’s criticism of the equilibrium concept used in mainstream economics. He is right in saying that its beam misses much of what we should see, such as the entrepreneur’s creation of new goods and new markets, so that some of the standard criticisms of product differentiation are simplistic. On the other hand, the insight that markets tend toward equilibrium provides us with a powerful tool for predicting how they will respond to certain shocks. If we were to relinquish standard equilibrium analysis we would lose much. But if we fail to see that there is much more to market behavior than a tendency toward narrowly defined equilibrium, we also lose much. 8 Mainstream economists and Austrians have something to learn from each other.
Improving Communication
Unfortunately, the lines of communication between Austrian and mainstream economics are feeble. I doubt that many mainstream economists read any Austrian economics, though this situation seems to have improved recently. Some economists who are not identified as Austrians now take Austrian ideas seriously (see Caldwell 1982, Hoover 1988, Montgomery 1996). But communication still needs much improvement. One factor hindering it is the somewhat arrogant belief of many mainstream economists that the writings of heterodox economists, or for that matter just about all economics written prior to the 1970s, has nothing to teach them.
On the other side, it is still the case, though perhaps to a lesser extent now than before, that an Austrian discussion of a mainstream proposition often consists in large part of showing that it is inconsistent with Austrian economics, perhaps even that it has been explicitly rejected by one of the Founding Fathers. Given the disdain with which most mainstream economists treat Austrian economics, such introversion is understandable—nor is it necessarily worse than the introversion of the formalists, who tend to treat economics as the study, not of the economy, but of other economists’ models. But introversion on either side hinders progress.
Even when Austrians go beyond condemning some work as being un-Austrian, their criticisms usually deal in generalities, such as the mainstream’s use of illegitimate aggregate concepts, instead of getting down into the trenches. It is easy for mainstream economists to ignore such broad criticism and say: “Yes, right, but our theories nonetheless work.” It would be more difficult to ignore the Austrians’ criticisms if these would more often consist of evidence (or, more specifically, what mainstream economists consider evidence) that a particular mainstream theory or model fails to predict or explain because it ignores a specific point made by Austrians.
While few mainstream economists are likely to become converts to Austrian economics in its entirety, more might take specific Austrian ideas seriously if these were shown to be relevant to their specific concerns, and were reformulated in a way that is compatible with mainstream methodological criteria. Granted that Austrian economics is built on philosophical foundations that differ sharply from those of mainstream economics, it would “sell” better if the methodology and the contents were marketed separately. Similarly, a more politically neutral version of Austrian economics might find additional markets.
Austrians may, of course, object that they want to do more than bring about a few changes in mainstream economics. But aiming at an attainable goal is better than aiming at one that is out of reach. Moreover, a series of small changes may add up to a large change. One should not exaggerate the incompatibility of paradigms.
This is not to deny that over the years Austrian economics has already made some major contributions to mainstream economics. Its insight into the problem of rational resource allocation under socialism is one, and so is Hayek’s work on the role of decentralized information in general. But more interaction is needed. Since Austrians are in the minority, as a practical matter it is up to them to build the needed bridges.
In summary, then, Boettke’s criticism of what he calls “formalism” goes too far. He defines formalism too broadly because he sees methodological disputes as pitting Austrians plus a few neoclassical economists against everybody else, thus ignoring some basic distinctions among the “everybody else.” In particular, he overlooks the important empiricist tradition in modern economics.
All the same, though I have not discussed these points, there is much in Boetke’s criticisms of contemporary mainstream economics that is correct and well argued. Moreover, his essay does serve an important and useful purpose in drawing attention to Austrian economics, which has some important ideas to contribute. But substituting Austrian economics for what Boettke calls formalism, or awarding it the main role in such a synthesis, would be a mistake
References
Backhouse, Roger. 1997. “The Rhetoric and Methodology of Modern Macroeconomics.” In Reflections on the Development of Modern Macroeconomics, ed. Brian Snowdon and Howard Vane. Cheltenham, England: Edward Elgar.
Boettke, Peter J. 1994. The Elgar Companion to Austrian Economics. Cheltenham, England: Edward Elgar.
Boettke, Peter J. 1997a. Private communication.
Boettke, Peter J. 1997b. “Where Did Economics Go Wrong? Modern Economics as a Flight from Reality.” Critical Review 11(1): 11–64.
Boettke, Peter, and David Prychitko. 1994a. “Introduction: The Present Status of Austrian Economics: Some (Perhaps Biased) Institutional History behind Market Process Theory.” In Boettke and Prychitko 1994b.
Boettke, Peter, and David Prychitko. 1994b. The Market Process. Cheltenham, England: Edward Elgar.
Boettke, Peter, Steven Horwitz, and David Prychitko. 1994. “Beyond Equilibrium Economics: Reflections on the Uniqueness of the Austrian Tradition.” In Boettke and Prychitko 1994b.
Cagan, Phillip. 1965. The Effect of Pension Plans on Aggregate Savings. New York: National Bureau of Economic Research.
Caldwell, Bruce. 1987. Beyond Positivism. London: George Allen and Unwin.
Debreu, Gerard. 1984. “Economic Theory in a Mathematical Mode.” American Economic Review 74(3): 267–78.
Debreu, Gerard. 1986. “Theoretical Models: Mathematical Form and Economic Content.” Econometrica 54(6): 1259–70.
Debreu, Gerard. 1991. “The Mathematization of Economic Theory.” American Economic Review 81(1): 1–7.
Foley, Richard. 1979. “Justified Inconsistent Beliefs.” American Philosophical Quarterly 16.
Friedman, Milton. 1953. Essays in Positive Economics. Chicago: University of Chicago Press.
Gibbard, Allan, and Hal Varian. 1978. “Economic Models.” Journal of Philosophy 75: 665–67.
Hammond, Daniel. 1996. Theory and Measurement. New York: Cambridge University Press.
Hayek, Friedrich von. 1935. Prices and Production. London: Routledge.
Hirsch, Abraham, and Neil de Marchi. 1990. Milton Friedman: Economics in Practice. Hemel Hemstead, England: Harvester Wheatsheaf.
Hoover, Kevin. 1988. The New Classical Macroeconomics. Oxford: Blackwell.
Keynes, John Maynard. 1936. The General Theory of Employment, Interest and Money. New York: Harcourt Brace.
Lange, Oskar. [1939] 1970. “On the Economic Theory of Socialism.” In On the Economic Theory of Socialism, ed. Benjamin Lippincot. New York: Augustus Kelley.
Lavoie, Don. 1994a. “Preface: The Meaning of Market Processs.” In Boettke and Prychitko 1994b.
Lavoie, Don. 1994b. “The Interpretative Turn.” In Boettke and Prychitko 1994b.
Lerner, Abba. 1994. The Economics of Control. London: Macmillan.
Lucas, Robert. 1973. “Some International Evidence on Output-Inflation Trade-Offs.” American Economic Review 63(3).
Machlup, Fritz. 1978. Methodology of Economics and Other Social Sciences. New York: Academy Press.
Marshall, Alfred. 1947. Principles of Economics. London: Macmillan.
Mayer, Thomas. 1993. Truth versus Precision in Economics. Cheltenham, England: Edward Elgar.
Mayer, Thomas. 1995. Doing Economic Research. Cheltenham, England: Edward Elgar.
Mayer, Thomas. 1996. “The Dark Side of Economic Modeling.” In Foundations of Research in Economics, ed. Steven Medema and Warren Samuels. Cheltenham, England: Edward Elgar.
McCloskey, D. N. 1985. The Rhetoric of Economics. Madison: University of Wisconsin Press.
Montgomery, Michael. 1996. “‘Austrian’ Persistence? Time to Build Survey Evidence on the Time Structure of Investment Spending.” Unpublished manuscript.
Montgomery, Michael. 1997. “On the Philosophical Foundations of ‘Rigor’: Macroeconomic Modeling and Ceteris Paribus.” Unpublished manuscript.
Robbins, Lionel. 1932. The Nature and Significance of Economic Science. London: Macmillan.
Schuler, Kurt. 1994. “The History of Free Banking.” In Boettke 1994.
Endotes
*: Thomas Mayer, Department of Economics, University of California, Davis, telephone (510) 549-0504, telefax (510) 549-9472, wishes to thank Peter Boettke, Kevin Hoover, Michael Montgomery, Nancy Wulwick, and Andrea Salanti for helpful comments. Back.
Note 1: For other problems with modeling see Mayer 1996. Back.
Note 2: Physicists on a practical level also use different theories when dealing with large objects and with subatomic particles, though, of course, their theories fit together. In economics, too, we can work with general enough premises to cover the entire field, but then our analysis tends to be vacuous. Whether this reflects a fundamental difference between the natural and the social sciences, or is merely a difference in the stages of their development, is another question. Back.
Note 3: The last sentence should be read only as a qualification and warning against overvaluing the beneficial role of mathematics in economics, since elsewhere Debreu (1984 and 1986) has stressed the virtues of mathematical economics. Back.
Note 4: In Bergamo, Italy, I saw just below the dome of a church a fresco that can be seen—and seen only faintly—from the top of a certain tower. Presumably it was painted “for the glory of God,” not for people’s enjoyment or instruction. One can justify some papers in Econometrica, which almost nobody can read, in the same way, given how few readers there are even for most of the less-technical papers in economic journals, and how unlikely it is that they will have any influence on policy or on economists’ thinking. Back.
Note 5: One might, however, object that in the social sciences, knowledge is so precarious that one should limit as much as one reasonably can those propositions that are privileged as uncontroversial. Many ideological propositions would then not qualify for the core. Back.
Note 6: But that is not always so. In some situations it is reasonable to hold two conflicting views (see Foley 1979). Back.
Note 7: Here is a concrete example. Phillip Cagan (1965) found that being covered by a corporate pension scheme induces households to save more on their own. This implies an increasing marginal utility of wealth, and therefore conflicts with the well-established belief that marginal utility decreases as one obtains more. I therefore did not accept Cagan’s findings, though I could find no fault with his analysis. Subsequently, when Cagan’s data were reanalyzed, it turned out that his conclusion had been wrong. Back.
Note 8: Boettke (1997b, 30) recognizes this when he writes that “equilibrium theorizing is not to be rejected, according to Hayek, but its real purpose must be constantly kept in mind. Formal modeling can be a very good servant, but a poor master.” However, the general impression that Boettke gives is of a strong condemnation of equilibrium theory. Back.