I. THE CURRENT ECONOMIC CRISIS AND MACRO-ECONOMIC IDEAS
The current economic crisis has a number of aspects, not least an intellectual one occurring at the interface between today’s dominant approach to macro-economic theory and empirical evidence. It is not just that most economists failed to foresee what was coming and that their models now require a little readjustment in order to catch up with the evidence. Rather, it is that such re-adjustments are impossible if they persist in adhering to the fundamental principles upon which these models are grounded. That is why renewed interest in John Maynard Keynes’s ideas is so welcome, even if so far this seems to be having little effect on what departments of economics are actually teaching.Footnote 1
On my reading of its history, what is nowadays called macro-economics is driven by a rather distinctive internal dynamic that differs significantly from the story of unidirectional technical progress which, in the eyes of so many of its current practitioners, makes the study of that history unnecessary. Though the sub-discipline ceased to be a series of essentially ad hoc responses to current events at some time in the mid-nineteenth century, contemporary happenings have nevertheless continued to exert a systematic influence on its subsequent development. That is because they have provided not just policy challenges, but also ongoing empirical tests of the evolving theoretical content of macro-economics.
Rather than presenting a picture of unrelenting technical progress within something resembling a single Kuhnian paradigm that has been in place since the eighteenth century—the prevailing view of most of its modern practitioners, including Robert E. Lucas, as we shall see below—the evolution of macro-economics has been characterized by a strong element of essentially Popperian “conjecture and refutation,” and it has also displayed a tendency to double back on itself from time to time to pick up still useful but temporarily mislaid ideas as starting points for subsequent development.Footnote 2 Two factors seem to have played a role here. First, as Harry Johnson (Reference Johnson1971) noted, the ability to cope with a currently important policy issue can be an important determinant of a macro-economic doctrine’s success. As particular policy problems come and go, therefore, so do ideas that can address them, eventually falling into neglect—sometimes temporarily—not always because they have been refuted by empirical happenings but because they have been rendered temporarily irrelevant by them. Thus, the problem of unemployment dominated inter-war discourse, but in the late 1960s the need to cope with inflation gave a strong impetus to what Johnson called “the monetarist counter-revolution,” in the process generating renewed interest in the ideas of Henry Thornton (e.g. 1802), Irving Fisher (e.g. 1911), and Knut Wicksell (e.g. 1898) among others, and giving a considerable head start to those who had already read these authors. And second, sometimes lines of investigation are abandoned for want of the analytic means needed to carry them further, only to be resurrected later when advances in technique remove these barriers. A striking example of this tendency is in the changing treatment of expectations in macro-economics, first during the 1930s when, at the hands of Keynes (Reference Keynes1936), exogenous expectations replaced technically unmanageable ideas about their endogeneity, and later from the 1950s onward, as the arts of modeling error learning and then the formation of rational expectations were mastered.Footnote 3
The interaction over time of macro-economic ideas with the facts that either refute them or render them temporarily irrelevant is also a two-way affair. As I noted in David Laidler (Reference Laidler and Mizen2003), forward-looking rational behavior on the part of policy makers and private agents alike must always be based on models—formal or informal—of the economic environment in which they are operating and will therefore be conditioned by those models. Ideas inconsistent with the way the economy actually works nevertheless affect its performance, therefore, eventually in ways that produce unexpected results that reveal the inconsistency in question. It is therefore hard to understand economic events, and hence the very empirical basis of macro-economics itself, without also understanding how the ideas that helped generate them evolved.
The current economic crisis is a prime example of a series of events grossly inconsistent with the very macro-economic ideas that helped produce them, but these happenings seem to bear a strong resemblance to those that three quarters of a century ago provoked what is generally called the “Keynesian Revolution.” Prompted by these considerations, this paper examines some of the relationships between macro-economic ideas and macro-economic crises, both recent and not so recent. It first examines what Robert E. Lucas, who did as much as anyone to push the frontier of macro-economics to its current location, has had to say about that frontier’s relationship to real-world economic phenomena, and also about Keynes’s place in the history of macro-economics. Then it suggests that Lucas’s understanding of how economics has evolved is flawed, and that this accounts for his equally flawed interpretation of Keynes’s place in that history. It also considers some implications for understanding today’s awkward economic facts of taking another look at aspects of Keynes’s General Theory, not so much its speculations about the role of psychology and social conventions in the economic decisions of individual agents that George Akerlof and Robert Shiller (2009) have recently highlighted under the label Animal Spirits, however, as its insights into the influence of the monetary and financial systems on the coordination of these decisions.
II. LUCAS ON MACRO-ECONOMICS AND KEYNES
Talking about “My Keynesian Education” at the 2003 HOPE conference on The IS-LM Model: Its Rise, Fall, and Strange Persistence (Michel De Vroey and Kevin Hoover Reference De Vroey and Hoover2004) Lucas pointed, almost as an aside, to
the problem that the new theories, the theories embedded in general equilibrium dynamics of the sort that we know how to use pretty well now—there’s a residue of things they don’t let us think about. They don’t let us think about the U.S. experience in the 1930s or about financial crises and their real consequences in Asia and Latin America, they don’t let us think very well about Japan in the 1990s (2004, p. 23).
This remark dates from 2003, when the final “Greenspan boom” was in full swing, and the “residue” of problems to which Lucas referred did indeed seem rather remote from the immediate but apparently well-established economic environment in which it was made, and we should judge its offhand tone in this context. Viewed with hindsight, though, the remark was ominous, because those same models now, in 2009, do not help us to think about problems that are dominating the current evolution of the entire world economy—some residue!Footnote 4
Lucas (Reference Lucas, De Vroey and Hoover2004) also argued that Keynesian economics was equally unhelpful with such issues, but here it is important to note that the “Keynesian economics” that formed the basis of his education in the early 1960s can be summarized as IS-LM (Investment equals Saving–Liquidity preference equals Money supply)—albeit the rather sophisticated version of it set out in Martin Bailey’s (Reference Bailey1962) textbook National Income and the Price Level—plus large-scale econometric models, of which the then emerging Brookings Model was something of a paradigm. This “Keynesian economics” had evolved from The General Theory—whether legitimately or not—as a means of coming to grips with, and designing policy to influence, the behavior of real income and employment, but in the early 1960s it had been extended to deal with inflation by the addition of an analysis of exogenous “cost-push” factors, and/or a simple and apparently permanent inverse relationship between inflation and unemployment—a Phillips curve. When, at the decade’s end, these modifications began to come to empirical grief in the face of failed policy experiments that they themselves had helped inspire, a further patch-up ensued. Endogenously determined, albeit adaptive, inflation expectations were introduced into the Phillips curve, with Lucas himself, in co-operation with the Leonard Rapping, being a notable contributor to these early efforts—see Lucas and Rapping (Reference Lucas and Rapping1969). Although this simple modification, inspired by the work of Milton Friedman (Reference Friedman1968) and Edmund Phelps (Reference Phelps1967), was more empirically helpful than is now commonly believed,Footnote 5 its deployment nevertheless came too late to save the day for the macro-economics in which Lucas had been trained, because it did nothing to protect it from deeper theoretical probing along two inter-related lines.
First, the fact that endogenous expectations had been invoked to make macro-economic models work in an inflationary environment made it hard to avoid asking whether simple adaptive formulae were the best that either economic agents or those modeling their behavior could do. Second, the Phillips curve had been launched by A. W. (Bill) Phillips (Reference Phillips1958) as an empirical relationship with only scanty theoretical justification, but one important strand of the late 1960s literature, of which Lucas and Rapping (Reference Lucas and Rapping1969) were also pioneers, grounded it in an analysis of the economy’s supply side that could be combined with the IS-LM system to form a highly aggregated model of general economic equilibrium. Such a model clearly had to have links to micro-economic general equilibrium analysis, and the exploration of these links fitted naturally into a then already on-going search for the micro-foundations of macro-economics.
As everyone knows, what was soon labeled new-classical economics emerged largely from Lucas’s own efforts to come to grips with these two issues. Initially, it was his replacement of adaptive by rational expectations that attracted most of the attention, but the closely related explicit application to traditionally macro questions of micro general equilibrium analysis marked a much more fundamental change in the then dominant approach to macro-economics, because it broke the area’s last remaining intellectual links to Keynes’s General Theory. Moreover, though Lucas’s contribution launched a radically new vision of how a market economy functions, he himself thought of it more modestly as involving the exploitation of newly available analytic techniques to deal with age-old problems that the macro-economic theory of the 1960s and the macro-econometrics that went with it had, like everything that had preceded it since David Hume, proved unable to address. Viewing the General Theory through the prism created by this macro-economics, he saw it as having created an unhelpful detour in the discipline’s otherwise orderly history, and interpreted its temporary success as a consequence of the historical situation that had prompted its writing.
In Lucas’s view, Keynes had not advanced economics but had merely offered an ad hoc political response to the circumstances of the Great Depression, a response which, seen in relation to what Lucas believed to have been the already long-established internal dynamic of economics, was of no lasting scientific value. In 2004, he made this point as follows:
Keynes’s real contribution is … not Einstein-level theory, new paradigm, all this … that’s just so much hot air. … [I]n writing the General Theory, Keynes was viewing himself as a spokesman for a discredited profession … in a situation where people are ready to throw in the towel on capitalism and liberal democracy and go with fascism or corporatism, protectionism, socialist planning. … What he hits on is that the government should take on new responsibilities … for stabilizing overall spending flows. … And … for everybody in the post-war period—I’m talking about Keynesians and monetarists both—that’s the agreed upon view… .
So I think this was a great political achievement. It gave us a lasting image of what we need economists for. I’ve been talking about the internal mainstream of economics, that’s what we researchers live on, but as a group we have to earn our living by helping people diagnose situations that arise and helping them understand what is going on and what we can do about it. That was Keynes’s whole life. He was a political activist from beginning to end (2004, pp. 23–24).
I shall return below to the view of macro-economics’ internal dynamic implicit in this last paragraph. For the moment, though, note that Lucas is hardly the first commentator on the General Theory to draw attention to its author’s talents as a polemicist, and to be led to dismiss that book’s claims to scientific importance in the process. The opening sentences of Arthur C. Pigou’s (Reference Pigou1936) review of the same book are justly famous, at least among historians of economic thought:
When, in 1919, he wrote The Economic Consequences of the Peace, Mr. Keynes did a good day’s work for the world, in helping it back towards sanity. But he did a bad day’s work for himself as an economist. For he discovered then, and his sub-conscious mind has not been able to forget since, that the best way to win attention for one’s own ideas is to present them in a matrix of sarcastic comment upon other people (p. 115).
And in 1936, as we all know, Pigou was no more appreciative of the General Theory as a substantive contribution to economic analysis than Lucas would later be. Indeed, like Lucas, he drew an explicit and unflattering comparison between Keynes and Einstein: "Einstein actually did for Physics what Mr. Keynes believes himself to have done for Economics" (1936, p. 115).Footnote 6
But Elizabeth Johnson (Reference Johnson, Johnson and Johnson1978a and b), who was every bit as sensitive as Lucas to Keynes’s political agenda, without, however, at the same time dismissing his scientific contributions, argued that we should pay attention to Keynes’s skills as a writer if we are fully to appreciate his economics, and nowhere did he make more careful use of these than when he chose the titles of his books. Economic Consequences was indeed written to influence current policy. So were A Tract on Monetary Reform (1923), The Means to Prosperity (1933), and How to Pay for the War (1940), not to mention those occasional pieces collected together as Essays in Persuasion (1931). Their author’s purposes are clearly expressed in the titles of these works. But A Treatise on Money (1930) and The General Theory of Employment, Interest and Money (1936) signal altogether more serious academic intentions. Pigou missed the point of the General Theory initially, perhaps because he was hurt, and justifiably so, by its unfair attacks upon his own work, but, with some acknowledged help from David Champernowne, he would later change his mind about its scientific importance (see Pigou Reference Pigou1938). Lucas’s views on these matters, however, have not changed for three decades, most likely for two closely related reasons, namely the logical structure of his own macro-economics, and his understanding of the place of that macro-economics in the history of the subject.
III. LUCAS ON PROGRESS IN MACRO-ECONOMICS
The development of the expectations-augmented Phillips curve in the late 1960s forced increased attention to be paid to the properties of the supply side of macro-economic models, and one strand in the resulting literature treated the relationship itself as an aggregate supply curve, along which, as Lucas (Reference Lucas, De Vroey and Hoover2004) points out with specific reference to Lucas and Rapping’s (Reference Lucas and Rapping1969) version of the analysis, “we have a cleared labor market at every point in time” (p. 26). But as a simple matter of logic, the labor market can be cleared only if the demanders of labor both expect to sell what labor produces and are able to do so. Furthermore, the aggregate demand side of the economy, to which Lucas and Rapping thought their analysis was complementary, was at that time invariably modeled in IS-LM terms and had a postulate about supply side behavior embedded in it, namely, that what was demanded was also being produced, for how else could the economy be on an IS-curve along which desired investment equaled desired saving?Footnote 7
The aggregate demand and supply curves of the typical macro-economic model of around 1970s, that is to say, were not behavior relationships, but equilibrium loci, and to be on either, the economy had to be on both, that is in full equilibrium. However, the then popular large scale quantitative versions of such models that had figured prominently in Lucas’s Keynesian education were made up of difference equations that not only tried to deal with out-of-equilibrium adjustment processes, but were also estimated one at a time and then brought together in systems which Lucas disparagingly but with some justification compared to a “church supper”—“a completely crazy way to put together a general equilibrium model of the whole economy. Nobody’s thinking about the whole thing” (2004, p. 21).
Lucas seems to have understood from a very early stage, however, that to bring coherence to these models required more than merely an effort to think self-consciously about the “whole thing” when constructing them. He was familiar with Don Patinkin’s (Reference Patinkin1957) search for Walrasian foundations for Keynesian macro-economics, and had noticed that in the resulting systems, movements between equilibria occurred in real time as prices that adjusted in accordance with what he refers to as the “mechanical auctioneer dynamics that Samuelson introduced.” As he pointed out in (2004), because “the rate of change of price in any one market ought to depend on excess demand and supply in all markets in the system” (p. 15) anything could happen in this process.Footnote 8 Crucially also, Lucas recognized that this characteristic had a counterpart in those abovementioned difference-equation-based macro-econometric models into which “Keynesian theory … [had breathed] some economic life.” To match the data they relied on inter-equilibria adjustment processes characterized by parameters whose values were left to be determined freely by those data, and as a result they could explain (almost—sometimes there were sign or magnitude restrictions) anything and therefore (almost) nothing. As he put it in his Nobel Lecture (Lucas Reference Lucas1996, p. 252), “The dynamics had a kind of patched in quality, fitting the facts, but only in a manner that suggests they could equally well fit any facts.” And to complete this unsatisfactory picture, the expectations that implicitly or explicitly entered into determining the behavior described by the individual equations of these models were routinely unrelated to the outcomes generated by the models themselves.
In (2004), Lucas described the development of macro-economics once these problems were recognized as involving the replacement of this Keynesian economics, in which he had initially been educated,
by the Arrow-Debreu model, which shows how you can take what seems to be a static general equilibrium model and talk about markets for contingent claims, talk about any kind of dynamics you’d like, coming right out of the economics. No auctioneer, or the auctioneer worked very quickly … we didn’t know this theory existed back in 1960 although it did. But now its potential is getting realized. It has completely succeeded in taking over growth theory, most of public finance, financial economics. Now it’s coming in use in macroeconomics with real business cycle theory; certain kinds of monetary variations have been introduced with success. … (p. 23)
This of course is the approach that doesn’t let us think about that “residue of things” which includes financial crises and the depression; but it is even more limiting than that. It is also makes it hard for its exponents to double back to the insights of an earlier era in economics for help with this problem, because the most important of those insights were into the workings of economies in whose description that phrase deployed by Lucas—“no auctioneer”—is to be taken literally, so that the agents operating within them must themselves set the prices at which they then trade. An economy with “no auctioneer” is thus fundamentally different from one presided over by a “very quick” auctioneer, because when the latter sets the prices of future goods and/or state contingent claims on them alongside those of current goods, he reduces the logic of “any kind of dynamics you like” to that of a static general equilibrium model and precludes the possibility of trade and production happening at non-market-clearing prices, something that is all too likely to happen in his absence.Footnote 9
As some extremely distinguished exponents of the Arrow-Debreu model—for example Frank H. Hahn (e.g. 1982)—have argued, this quick auctioneer’s activities render such an economy crucially different from any that we might encounter in the real world, not least because they eliminate any essential role for money in its processes of price formation or exchange.Footnote 10 The market mechanisms embedded in the Arrow-Debreu model are thus at best a metaphor for the monetary and financial systems through which exchange among agents both at a moment in time and over time is mediated in the real world. It is because financial crises involve failures of the latter systems that models which analyze only situations in which they are working cannot be of any help in understanding these events.
This does not in and of itself make dynamic general equilibrium models bad economics, because to resort to a metaphor is merely to say that one thing behaves “as if” it were another. Many of us are quite comfortable with empirically testing “as if” statements about how the economy functions, tentatively accepting them if they seem to be useful and otherwise rejecting them. Some of us are even willing to deploy particular already well-tested “as if” hypotheses in contexts where they seem likely to work, though we also know that in others they don’t. Dynamic general equilibrium modeling in macro-economics, of the type that evolved from Lucas’s (Reference Lucas1972) “money-supply surprise” model, could, and still can, therefore, be defended on an “as if” basis as one potentially useful approach among others, even if limited in its applicability. Indeed in some places from the early 1990s until recently, such an approach—essentially that codified by Woodford (Reference Woodford2003)—provided a very useful framework for monetary policy making, But, as a matter of fact, since the mid-1970s, the market-clearing postulate has often been treated by its exponents, not as a refutable conjecture about how economies might helpfully be modeled in some circumstances, but as an axiom, alongside that of rational maximizing behavior, upon whose acceptance hinges the admissibility of any model as a source of predictions that are then worth testing.Footnote 11
One can see why well articulated micro-economic foundations began to take strong precedence over empirical evidence in macro-economics in the early 1970s. After all, a permanent inflation–unemployment trade-off had been presented as a well established fact in the preceding decade and had quickly found a central place not just in the textbooks, but in the policy debates of the period too, but Phelps and Friedman had challenged its authority with a priori micro-economic reasoning before the empirical counter-examples that confirmed their skepticism had been generated. Once inflation began to generate these counter-examples in the early 1970s, however, it became patently obvious that, in this important instance, the careful deductive analysis of rational behavior had been a better guide to assessing propositions about real world economies than had empirical generalizations with no obvious basis in such reasoning.
Even so, the temptation to draw general conclusions about how to proceed with macro-economic modeling from examples such as this has its dangers. The rational maximization postulate is most easily deployed in macro-economics by adopting the “representative agent” simplification to dispose of all the many complications that disparities and interactions among multiple agents can create for the analysis of aggregate behavior. When such models are extended to multiple-agent formulations, the assumption of continuously clearing markets plays the apparently primarily technical role of enabling the analysis to continue to focus on rational maximizing behavior, and in particular on the formulation and use of rational expectations to guide it.Footnote 12
But in fact, this assumption of clearing markets, technically useful though it certainly is, is also of immense substantive significance. To treat it as an axiom that defines the borderline between what is and is not admissible as science, rather than an empirical hypothesis that ought itself to be tested, is to do nothing less than resolve by assumption, and hence place beyond debate, two of the oldest and most contentious questions in the history of economics, namely whether, and if so how, a decentralized market economy is capable of coordinating the individual consumption and production decisions of those who participate in it. Only the answers “yes” to the first and “as if by agents who respond to the market clearing prices posted by an auctioneer who works very quickly” to the second are admissible. Other responses, such as “sometimes” and “by indirect exchange mediated by monetary and financial systems that are subject to failure from time to time,” let alone any more radical suggestions that might boil down to “never,” are ruled out.
But, in Lucas’s view the deployment of Arrow-Debreu analysis represents nothing more (nor less) than the fulfillment of what was always the goal of a long line of theorists, stretching from David Hume to Don Patinkin who, wanting “… to think in general equilibrium terms … resort to disequilibrium dynamics only because the analytic equipment available to them offers no alternative” (1996, p. 253). Or, as he put it (2004)
I see the progressive … element in economics as entirely technical: better mathematics, better mathematical formulation, better data, better data-processing methods, better statistical methods, better computational methods, I think of all progress in economic thinking, in the kind of basic core of economic theory as developing entirely as learning how to do what Hume, and Smith and Ricardo wanted to do, only better (2004, p. 22).
However, as we have seen, no theoretical system based on the assumption of continuously clearing markets—and this is true even of those mislabeled “new Keynesians” models deployed for example by Woodford (Reference Woodford2003) in which money-wage and nominal price adjustment is slowed down by arbitrarily introduced overlapping contracts—can deal with some of the critical monetary and financial features of any real world economy.Footnote 13 It is not just that such macro-economic models cannot address the policy issues that the recent convulsions in financial markets have created, or that it is difficult for anyone brought up under their influence to conceive of such events occurring in the first place, but that those who treat their market-clearing properties as defining the boundaries of scientific macro-economics place such questions outside of the sub-discipline’s subject matter. That is why a crisis in macro-economics is an integral part of the current economic situation.
IV. KEYNES AS AN ALLEGED SCIENTIFIC OUTLIER
The extent to which modern economic theory has contributed to the upheavals that began to shake economies in 2007 is a problem for future economic historians and historians of economic thought to worry about. For today’s economists, or at least those who take the subject’s past seriously, a more immediate question is where in that past we might look for guidance as we try to reconfigure its future.
One answer already widely on offer, from Akerlof and Shiller (Reference Akerlof and Shiller2009) among others, is the Keynesian economics of the late 1930s. Somewhat paradoxically, although these authors have a high opinion of Keynes as an economist, their assessment of the facts of his place in the history of macro-economics is very similar to Lucas’s. They too treat him as an outlier who tried and failed to divert economics from a continuously developing orthodoxy that could be traced back to Adam Smith and the other founders of Classical economics. As they describe that orthodoxy,
According to traditional economics, free market capitalism will be essentially perfect and stable. … This line of reasoning goes back to Adam Smith. … If people rationally pursue their own economic interests in such markets, they will exhaust all mutually beneficial opportunities to produce goods and exchange with one another [and this] results in full employment" (2009, p. 2).
And of course at least one other authority took a similar view of Keynes’s place in the history of economic thought even before the publication of the General Theory, namely its author. Recall that famous letter to George Bernard Shaw, later quoted on the back cover of the “Papermac” edition of the book in question: “… you have to know that I believe myself to be writing a book on economic theory which will largely revolutionise [sic]—not, I suppose at once but in the course of the next ten years—the way the world thinks about economic problems”; but note also that the General Theory’s own account of “Classical economics” was castigated by Pigou (Reference Pigou1936) as “a macedoine of misrepresentation,” and that another reviewer, Frank H. Knight (Reference Knight1937) described Keynes’s treatment of it in the following similar, if more even-tempered, terms:
the references under this phrase are the sort of caricatures, which are typically set up as straw men for purposes of attack in controversial writing … the reader of Mr. Keynes’s book would do well to keep in mind that references to “classical economics” are to be interpreted as relating to economic analysis at the stage at which uncertainty and monetary disturbances are assumed absent (1937, p. 101).
This judgment is surely right, for though Keynes did have interesting and sometimes novel things to say about “uncertainty and monetary disturbances,” it is by now a commonplace of the history of macro-economic thought that he systematically downplayed the fact that so did many of his then academically respectable contemporaries and predecessors. It is worrisome therefore that, at a time when the sub-discipline is having so much difficulty coping with empirical puzzles that seem to be associated with these very phenomena, important theorists like Akerlof and Shiller, and of course Lucas, remain content to take Keynes’s own account of his place in that history at face value.
V. SOME PSYCHOLOGICAL ELEMENTS OF KEYNES’S MACRO-ECONOMICS
In Animal Spirits Akerlof and Shiller invoke Keynes’s name and vocabulary mainly as rhetorical devices in support of the view that modern macro-economics should take a little more notice of psychology and rely somewhat less on utility-maximizing hypotheses. The ultimate test of their claims here must be whether they help us to understand the world we live in better than the alternative, and on this question I am willing to suspend judgment pending the outcome of further empirical investigations. But, though it is certainly true that there is much in the General Theory that derives from Keynes’s own keen interest in psychology, a point often stressed by Gilles Dostaler—most recently in (2007, Ch. 6)—not all of the precedents for such an approach on display in the General Theory are encouraging.
To begin with, Keynes invoked not maximizing behavior but psychology to get the stable marginal propensity to consume that he needed to generate a stable multiplier, which was both a useful simplification to deploy in his explanation of unemployment as a consequence of deficient private investment and a crucial lynchpin for his policy recommendation that this gap could be filled by public expenditure. He told his readers that though the “propensity to consume” was influenced by both subjective and objective factors, the former, namely “those psychological characteristics of human nature … social practices and institutions … though not unalterable … are unlikely to undergo a material change over a short period of time, except in abnormal or revolutionary circumstances” so that, ultimately
The fundamental psychological law, upon which we are entitled to depend with great confidence both a priori from our knowledge of human nature and from the detailed facts of experience, is that men are disposed, as a rule and on average, to increase their consumption as their income increases, but not by as much as the increase in their income (1936, p. 98).
As Allan Hynes (Reference Hynes1998) has documented, however, this building block of The General Theory, which Bertil Ohlin had criticized as an over-simplification as early as (1937), began to crack under the weight of empirical evidence in the early 1940s and was ultimately replaced by the utility-maximizing models of Franco Modigliani and Richard Brumberg (Reference Modigliani, Brumberg and Kurihara1954) and Friedman (Reference Friedman1957).
Or to cite another example, Allan Meltzer (Reference Meltzer1988, p. 146) has drawn attention to Keynes’s way of switching between extrapolative and regressive hypotheses about interest-rate expectations. Perhaps the most striking example of this occurred as the needs of his case changed between the Treatise on Money (1930) and the General Theory. In the Treatise, when discussing the use of monetary policy to influence the long-term rate of interest in order to offset the effects on investment of swings in “the spirit of enterprise,” he argued that the short interest rates that the central bank could undoubtedly influence “affect long rates more than one might expect” because “mob psychology”—at first sight an unreliable foundation for monetary policy—in fact provided the basis for “a homeopathic cure. …The real prospects do not suffer such large and quick changes as does the spirit of enterprise” so “it is not unreasonable to depend on short-period influences for counteracting a violent, and perhaps unreasoning change in sentiment.” In the General Theory, on the other hand, though Keynes still presented that same long rate of interest as “a highly psychological phenomenon” he quickly qualified this characterization by suggesting that it was “more accurately … a highly conventional … phenomenon …” which would sometimes be impervious to monetary policy because
Any level of interest which is accepted with sufficient conviction as likely to be durable will be durable; subject of course, in a changing society to fluctuations for all kinds of reasons round the expected normal. … it may fluctuate for decades about a level which is chronically too high for full employment. … (italics in original)
As to swings in the above-mentioned “spirit of enterprise,” which had already appeared, though not under that label, in the work of Frederick Lavington (Reference Lavington1922) and Pigou (e.g. 1927) as driven by cumulative and contagious “errors of optimism and pessimism,” these would reappear in the General Theory as consequences of exogenous changes in “animal spirits” that are there presented as more important determinants of the marginal efficiency of capital than any “weighted average of quantitative benefits multiplied by quantitative probabilities” (p. 161). Now Roger Backhouse and Laidler (Reference Backhouse, Laidler, De Vroey and Hoover2004) among others have followed Jan Kregel (Reference Kregel1976) in pointing out that Keynes’s treatment of long-term expectations about the profitability of investment as exogenous enabled him to develop an essentially comparative static and hence analytically manageable framework to expound the central theme of the General Theory, namely that it is changes in expectations about the future which largely cause variations in the present level of employment, so this was surely a fruitful and productive simplification.Footnote 14 But though Keynes had many amusing things to say about the gyrations of the “spirit of enterprise” and its interactions with what he called “speculation” in his (1936) Chapter 12 on “The State of Long-term Expectation,” the fact remains that in accounting for swings in investment by attributing them to exogenous changes in “animal spirits” (or to exogenous changes in anything else for that matter) he left them entirely unexplained.Footnote 15
Now Keynes was not the first economist to appeal to psychology when lacking an explanation for seemingly important facts: There is, for example, more psychology and less rational maximization at the heart of the Wealth of Nations (1776) than either Lucas or Akerlof and Shiller seem willing to concede. Adam Smith even invoked an inherent “propensity to truck and barter” to explain the phenomenon of exchange itself. Nor was Keynes the last to follow such a line, as Akerlof and Shiller’s own example shows. But even given the advances that the latter tell us have been made since 1936 in the course of a further seventy years work on the idea of “animal spirits” (see p. xi), one wonders whether their optimistic claim that the modern version of this concept provides “easy answers” to the eight undoubtedly important economic questions which their book addresses is justified.Footnote 16 Perhaps such answers are just a bit too easy when it is permissible to cast them in terms of behavioral hypotheses that, not being firmly grounded in rational maximization and constrained by its mechanics, can be custom-tailored to the relevant facts.
VI. KEYNES ON MONEY AND COORDINATION
Keynes’s deployment of exogenous long-term expectations to permit the analytic simplifications needed to generate a coherent theory of employment provides compelling support for Lucas’s views on how the availability of analytic techniques can sometimes limit the scope of economic theorizing, and subsequent developments in the treatment of expectations just as surely illustrate his contentions about the importance of purely technical advances in the permitting theory to move forward. As we have seen, however, Lucas goes beyond such claims. He explicitly treats technical advances as in and of themselves defining theoretical progress, and he implicitly treats that progress as coming at no cost. If he is right here, any doubling back of economic thought to the 1930s, or any other era, is unnecessary, now or at any other time.
Akerlof and Shiller clearly reject this viewpoint, but it is not necessary to share their enthusiasm for psychology in general or “animal spirits” in particular, to agree with their broader judgment. As Backhouse and Laidler (Reference Backhouse, Laidler, De Vroey and Hoover2004) pointed out, much else besides Keynes’s deployment of psychological ideas became hidden from view as IS-LM cast its ever deepening shadow over macro-economics in the 1940s and 50s. This is not to argue that such economics was not legitimately derived from the General Theory.Footnote 17 But it is to argue that there was much else both in that book, and in the literature in whose context it should be read, that did not find its way into the so-called Keynesian education of Lucas and of most of his contemporaries.
In particular Keynes attributed an utterly central role to money in his new theory of how the economy functioned. Readers of his Preface were told that
… whilst … money enters into the economic scheme in an essential and peculiar manner, technical monetary detail falls into the background. A monetary economy, we shall find, is one in which changing views about the future are capable of influencing the quantity of employment and not merely its direction (p. vii).
Even so, had the General Theory said no more than that the key to macro-economic instability was to be found in the workings of the monetary system, this would not have set it apart from a host of other writings of the period. John Stuart Mill’s insights into the way an excess supply of output as a whole could arise as a counterpart to an excess demand for money in times of crisis first appeared in a paper written in the late 1820s, though not published until (1844), but they were also set out in his Principles (1848), and were taken up in due course by Alfred Marshall and Mary Marshall (1879). In the inter-war years, much developed—they were on prominent display in Ralph Hawtrey’s (e.g. 1919, 1932) work. Furthermore, as Axel Leijonhufvud (Reference Leijonhufvud1981) has documented so persuasively, Knut Wicksell’s (Reference Wicksell1898) analysis of how an interest rate determined within the monetary system might disrupt capital markets’ ability to maintain equilibrium between saving and investment inspired a lively and diverse subsequent literature from which Keynes’s own Treatise on Money drew much of its inspiration. By 1936, that is to say, the literature of macro-economics was dominated, not by the view that the economy could be analyzed “as if” functioning by barter, but by a bewildering variety of efforts, each in its own way unsatisfactory, to articulate just what it was about money that made an economy that used it different in general, and crisis prone in particular.
As I argued in Laidler (Reference Laidler1999), there was therefore nothing original about this question when the General Theory posed it, nor were any of the concepts Keynes deployed in formulating his particular response new. The multiplier was borrowed from Richard Kahn (Reference Kahn1931) and Jens Warming (Reference Warming1932), the “marginal efficiency of capital” from Irving Fisher (Reference Fisher1907)—where it was called the “rate of return over cost”—while “liquidity preference”—also a new name for an older idea—came from Lavington (Reference Lavington1921) by way of Keynes’s own earlier Treatise on Money and Hicks (Reference Hicks1935).
What was new in (1936) was a uniquely powerful and coherent synthesis of these ideas. The point of liquidity preference theory was that money, whose use as a means of exchange and unit of account made coordinated economic activity feasible in the first place, also could and did function as a store of value alongside claims to the income streams generated by capital goods.Footnote 18 Thus
The psychological time-preferences of an individual require two distinct sets of decisions to carry them out completely. The first is concerned with that aspect of time-preference which … determines how much of his income he will consume and how much he will reserve in some form of command over future consumption.
But this decision having been made, there is a further decision … namely in what form he will hold the command over future consumption which he has reserved, whether out of his current income or from previous savings. Does he want to hold it in the form of immediate, liquid command (i.e. in money or its equivalent)? Or is he prepared to part with immediate command for a specified or indefinite period, leaving it to future market conditions to determine on what terms he can, if necessary, convert deferred command over specific goods into immediate command over goods in general? (p. 166)
In the “as if” model of a barter economy that Keynes misleadingly identified as representing the sum total of classical economics, the rate of interest has only one function, to equilibrate the first of the abovementioned decisions with the investment choices of firms. But, he argued, a monetary economy is not an “as if barter” economy precisely because, when money can be held as a store of value, the rate of interest also acquires a crucial role in portfolio allocation decisions, which may undermine its capacity simultaneously to coordinate the allocation of resources over time.
Keynes posited the logical possibility of coordination failures in a monetary economy, not their logical necessity, however, and he went to considerable trouble to describe the circumstances under which a monetary economy would converge on an equilibrium in which what he called the “neutral” rate of interest—that which would equate saving and investment at a full employment level of income—would prevail. However, writing as he was in the mid 1930s (and recall that in his native Britain, mass unemployment had been chronic since 1921) it was nevertheless reasonable for him to conclude that at least some of the many factors to which he could point as making this outcome unlikely had in fact prevailed, and to go on to suggest both a diagnosis of the depression, and policy remedies for it.
The General Theory’s treatment of these issues will seem to many nowadays both dated and politically naive:Footnote 19 It attributes the depression of its times to a chronic lack of investment opportunities that are privately profitable even at a low positive value of the long-term rate of interest, this state of affairs being caused in turn by a secular slowdown in technical progress which was expected to persist into the foreseeable future; it then suggests remedies that require some permanent changes in economic and social organization: “a much lower rate of interest than has ruled hitherto,” a “state of affairs … quite compatible with some measure of individualism, yet it would mean the euthanasia of the rentier” (pp. 375–6), not to mention a “somewhat comprehensive socialization of investment” (p. 378).
We might conjecture that the clearly political content of these recommendations combined with the dramatic flair with which they were advanced as seeming “… to a nineteenth-century publicist or to a contemporary American financier to be a terrific encroachment on individualism” (p. 380) were what helped to convince Lucas that their author’s theoretical ideas could be of no account for the development of scientific economics. But, as Patinkin (Reference Patinkin1983) pointed out in the course of his critique of Meltzer’s (Reference Meltzer1981) “different perspective” on the General Theory, that book’s final Chapter 24, in which these ideas appear, is not a conventional closing summary of a scientific monograph dedicated to expounding a new theory, but merely “Concluding notes on the Social Philosophy towards which the General Theory might lead” (italics added).
Keynes himself thus knew that the political appendages of the General Theory were not logically implied by its theoretical economic content, but by certain empirical judgments with which he supplemented the book’s analytic framework. But claims on behalf of the importance and originality of that theoretical economic content are not, as Lucas would have them, “hot air.” That “general theory” to which the book’s title alludes is, after all, a systematic and successful effort to show why the logic of a monetary economy cannot in general be reduced to that of an economy in which, to borrow Lucas’s terminology, a very quick auctioneer is always at work, and how that economy’s distinguishing monetary characteristics render it prone to coordination failures. Furthermore, this analytic barrier is operative in both directions because, as Peter Howitt has argued in a number of places (e.g. 1997, 2006), the logic of macro-economic systems that rely on the fiction of an auctioneer to support the interlinked hypotheses of clearing markets and rational expectations cannot be extended to encompass coordination failures either. If therefore, as Lucas would have us do, we treat the differences between these two classes of models as reflecting not the presence of different substantive and testable “as-if” hypotheses about the way in which economies actually function, but simply the technical superiority of one of them, which we are then bound to choose on these grounds, we not only find ourselves unable to think about coordination failures, but we also classify any efforts to do so as scientifically retrogressive, thus ruling them out.
VII. THE GENERAL THEORY’S CURRENT RELEVANCE
The General Theory of Employment, Interest and Money is not a treatise on the overall anatomy of economic crises. Though it does have things to say about their origins, it is primarily about why they involve unemployment, and what might be done about this, and it is with these same issues that it is most likely to be helpful today. Keynes argued that an understanding of monetary matters was essential to grasping how the level of employment could sometimes be deficient, that this very understanding implied that wage and price flexibility, which operated through its effects on the monetary system, could not and should not be relied on to restore it to a satisfactory level, and also that expansionary monetary policy might sometimes be of limited usefulness for this purpose.
It is convenient to take up the role of price flexibility, or the lack thereof, in the story first of all. Pace Akerlof and Shiller and a host of others, the significance of price stickiness for a monetary economy does not rest in any essential way on money illusion, contractually determined money wage rigidities, or any other such dei ex machinis. Rather, as became entirely clear only in the 1960s from the work of Robert Clower (Reference Clower, Hahn and Brechling1965) and Leijonhufvud (Reference Leijonhufvud1968), it derives from the more fundamental fact that, though in such an economy variations in the general price level are required to keep the supply and demand for money in equilibrium at full employment, there is no auctioneer present to set this variable’s value.Footnote 20 Its behavior is therefore the aggregate byproduct of the activities of a host of individual agents, each of whom sets the nominal price of whatever is traded in his or her own market with no regard to that aggregate outcome. These agents can bring as much (or little) flexibility and rationality to these activities as we care to attribute to them, but so long as they do not always make exactly the right array of decisions to keep the overall price level at its market-clearing value, then trading at false prices will take place in some markets, with consequences for subsequent decisions about quantities whose aggregate outcomes may sometimes look like multiplier processes.
Not that Keynes regarded a degree of price level stickiness as a drawback in a monetary economy. On the contrary, he clearly believed that the predictability in nominal values that it brought with it helped to make rational decisions possible, and from the Tract on Monetary Reform (1923) to How to Pay for the War (1940) he was a firm advocate of price level stability as a policy goal. This was his position in the General Theory as well, even in the face of employment problems:
… the chief result of …[money wage flexibility] would be to cause a great instability of prices, so violent perhaps as to make business calculations futile in an economic society functioning after the manner of that in which we live. To suppose that a flexible wage policy is a right and proper adjunct of a system … which on the whole is one of laissez-faire, is the opposite of the truth (p. 269).
But when it came to explaining the possibility of unemployment arising, and warning about limits on the efficacy of monetary cures for this state of affairs, Keynes emphasized not price stickiness per se but money’s capacity to satisfy liquidity preference, though of course this did derive from its role in the processes of price formation and exchange that price stability helped support.
According to Keynes, the basic reason that unemployment could arise was that, to use a more modern terminology than he deployed himself, liquidity preference interfered with the rate of interest’s ability to keep the allocation of resources over time coordinated in the face of changing investment and saving decisions, thus forcing it to abdicate this task to movements in income and employment; and monetary cures for unemployment were unreliable because the sensitivity of the demand for money to the rate of interest might place limits on the expansionary impact of increases in real balances, even when these were brought about directly by policy-induced increases in the supply of nominal balances, rather than price level variations.
Keynes’s claims on behalf of the novelty and importance of these implications of his monetary theory of the rate of interest generated much immediate criticism, and the ensuing debate established a point that is crucially relevant to today’s policy debates: namely that, though his explanation of how variations in employment could occur and his speculations about the weakness of monetary stabilizing measures are linked to one another by his insights into the unique nature of a monetary economy, one can accept the former without also assenting to the latter. Dennis Robertson (Reference Robertson1936), Jacob Viner (Reference Viner1936), and Hawtrey (Reference Hawtrey1937) each in his own way took this position, and there is a lot to be said for it.
The key point here is that the interest sensitivity of the demand for money must be nontrivial for a monetary economy to generate lapses from high employment in the face of shocks to saving and investment, but it needs to be very high to support doubts about the effectiveness of monetary stimulus in the face of such lapses, and there is much less empirical evidence to support this latter position than the former. That is why Keynes’s claims about the importance of liquidity preference as the key to coordination failures that can generate unemployment are still worth taking seriously, while his doubts about monetary policy as a cure for them are more questionable. Though he was himself ambivalent about the empirical relevance of that extreme case which Dennis Robertson later called the liquidity trap, there is ample textual evidence that he believed liquidity preference usually to be highly interest sensitive and unstable as well, a matter that Friedman (Reference Friedman and Gordon1974) in particular emphasized when he denied that his demand for money function was not so much a restated quantity theory as a development of Keynes’s monetary theory.Footnote 21
If one can accept Keynes’s view of the monetary roots of unemployment without being obliged to accept his skepticism about the efficacy of monetary remedies for it, however, this still leaves open questions about the adequacy of the explanations on offer in the General Theory for the onset of a crisis like today's. There is much that is attractive here: not least the parallel between Keynes’s paradoxical proposition, already explored at length, that the very institution of monetary exchange that makes a market economy possible also renders it vulnerable to failure, and his equally paradoxical argument that the modern financial markets, which do so much to mobilize savings and make them available for investment, also undermine the rationality of the latter activity.Footnote 22 As he put it: “The social object of skilled investment should be to defeat the dark forces of time and ignorance which envelop our future. The actual private object of the most skilled investment today is to “beat the gun” as the Americans so well express it, to outwit the crowd, and to pass the bad, or depreciating, half crown to the other fellow” (p. 155). And his recognition that “these tendencies are a scarcely avoidable outcome of having successfully organized ‘liquid’ investment markets” did not soften his judgment that “when the capital development of a country becomes the by-product of the activities of a casino, the job is likely to be ill done” (p. 159).
It is hard to deny that all this has considerable contemporary resonance in 2009, or to fail to recognize that Hyman Minsky’s (e.g. 1982) nowadays finally popular analysis of financial markets’ tendencies to develop into gigantic Ponzi schemes has some of its roots here. But Keynes’s analysis is nevertheless incomplete, as has already been noted above.Footnote 23 His insights into the role of financial markets in detaching investment decisions from fundamentals notwithstanding, a spontaneous collapse in animal spirits and hence in the marginal efficiency of capital and in investment expenditure is more an ex post rationalization of an economic crisis than an explanation for it, and it also leaves too many facts unaccounted for. That is why, as I have explained at greater length in Laidler (Reference Laidler2007), I find analyses of booms and their collapse into financial crisis along Austrian — see e.g. Friedrich von Hayek (1929), and Roger Garrison (Reference Garrison2001)—or Robertsonian—see e.g. Robertson (Reference Robertson1928)—lines more satisfactory, in particular because these find a place for the incomplete capital expenditure projects that, as an empirical matter, always seem to be left behind by such events; and if forced to refine this choice further, I would opt for Robertson, who avoided Hayek’s tendency towards extreme policy pessimism about how to deal with the subsequent slump because he had a better appreciation of the fact that bubbles are primarily sectoral in nature, even though the consequences of their collapse may often be economy-wide.
Even so, my skepticism about the adequacy of Keynes’s views on monetary policy, and on the causes of crises notwithstanding, I hope that it is clear from the foregoing discussion that his interpretation of falling income and employment as a monetary economy’s way of moving saving and investment back towards equilibrium when its mechanisms prevent the rate of interest from doing the trick is as relevant today as it was in 1936, and an idea that modern macro-economics would do well to revive.
VIII. KEYNES’S CONTEMPORY SIGNIFICANCE
Lucas (Reference Lucas, De Vroey and Hoover2004) described Keynes’s response to the depression as that of a political activist whose claims to have contributed to economic theory were unsustainable. And yet Lucas was not willing to argue that Keynes’s influence was over, because further progress along “the internal mainstream of economics … that … we researchers live on,” whose flow Keynes had temporarily interrupted, had not yet produced an adequate treatment of matters related to monetary instability. “Some people just deny that there are real effects of monetary instability, but I think that it is just a mistake” and in any event, researchers “as a group … have to earn our living by helping people diagnose situations that arise and helping them understand what is going on and what we can do about it.”
While waiting for the above-mentioned “internal mainstream” to produce relevant results, therefore, Lucas seems willing, when discussing current policy as he does in a recent lecture (2008), to offer analysis in the spirit he attributes to Keynes, and also to Friedman, whose approach he doesn’t find too different (see 2004, p. 24).Footnote 24 But, as I argued at the outset of this paper, the mainstream of macro-economics does not follow a straight line characterized by purely technical advances. It is much more wayward, and sometimes encounters empirical obstacles sufficiently strong to divert its course and even turn it temporarily back on itself in search of theoretical ideas that have been prematurely mislaid in the face of changing policy problems or inadequacies of available analytic techniques.
Today’s economic crisis is just such an obstacle, and my main purpose in this paper has been to argue that it calls for a reconsideration of Keynes’s insights into the mechanics of inter-temporal coordination in a monetary economy and their proneness to occasional failure. More generally, because I have also argued that this exercise will not even be attempted by exponents of a macro-economics that insists on beginning everything from the assumption that markets clear continuously, I have also implied that it is not just Keynes’s specific answer that nowadays needs reconsideration, but perhaps even more the basic question that he shared with his contemporaries: namely, what is it about a monetary economy that sometimes causes its coordination mechanisms to break down? Perhaps Lucas’s recent ventures into “helping people diagnose … and understand what is going on” in today’s economy will lead him, not to mention the many other researchers he has influenced, to restore this question to its rightful place at the center of macro-economic theory.
