Archive for the 'Keynes' Category

Hicks on IS-LM and Temporary Equilibrium

Jan, commenting on my recent post about Krugman, Minsky and IS-LM, quoted the penultimate paragraph of J. R. Hicks’s 1980 paper on IS-LM in the Journal of Post-Keynesian Economics, a brand of economics not particularly sympathetic to Hicks’s invention. Hicks explained that in the mid-1930s he had been thinking along lines similar to Keynes’s even before the General Theory was published, and had the basic idea of IS-LM in his mind even before he had read the General Theory, while also acknowledging that his enthusiasm for the IS-LM construct had waned considerably over the years.

Hicks discussed both the similarities and the differences between his model and IS-LM. But as the discussion proceeds, it becomes clear that what he is thinking of as his model is what became his model of temporary equilibrium in Value and Capital. So it really is important to understand what Hicks felt were the similarities as well as the key differences between the temporary- equilibrium model, and the IS-LM model. Here is how Hicks put it:

I recognized immediately, as soon as I read The General Theory, that my model and Keynes’ had some things in common. Both of us fixed our attention on the behavior of an economy during a period—a period that had a past, which nothing that was done during the period could alter, and a future, which during the period was unknown. Expectations of the future would nevertheless affect what happened during the period. Neither of us made any assumption about “rational expectations” ; expectations, in our models, were strictly exogenous.3 (Keynes made much more fuss over that than I did, but there is the same implication in my model also.) Subject to these data— the given equipment carried over from the past, the production possibilities within the period, the preference schedules, and the given expectations— the actual performance of the economy within the period was supposed to be determined, or determinable. It would be determined as an equilibrium performance, with respect to these data.

There was all this in common between my model and Keynes'; it was enough to make me recognize, as soon as I saw The General Theory, that his model was a relation of mine and, as such, one which I could warmly welcome. There were, however, two differences, on which (as we shall see) much depends. The more obvious difference was that mine was a flexprice model, a perfect competition model, in which all prices were flexible, while in Keynes’ the level of money wages (at least) was exogenously determined. So Keynes’ was a model that was consistent with unemployment, while mine, in his terms, was a full employment model. I shall have much to say about this difference, but I may as well note, at the start, that I do not think it matters much. I did not think, even in 1936, that it mattered much. IS-LM was in fact a translation of Keynes’ nonflexprice model into my terms. It seemed to me already that that could be done; but how it is done requires explanation.

The other difference is more fundamental; it concerns the length of the period. Keynes’ (he said) was a “short-period,” a term with connotations derived from Marshall; we shall not go far wrong if we think of it as a year. Mine was an “ultra-short-period” ; I called it a week. Much more can happen in a year than in a week; Keynes has to allow for quite a lot of things to happen. I wanted to avoid so much happening, so that my (flexprice) markets could reflect propensities (and expectations) as they are at a moment. So it was that I made my markets open only on a Monday; what actually happened during the ensuing week was not to affect them. This was a very artificial device, not (I would think now) much to be recommended. But the point of it was to exclude the things which might happen, and must disturb the markets, during a period of finite length; and this, as we shall see, is a very real trouble in Keynes. (pp. 139-40)

Hicks then explained how the specific idea of the IS-LM model came to him as a result of working on a three-good Walrasian system in which the solution could be described in terms of equilibrium in two markets, the third market necessarily being in equilibrium if the other two were in equilibrium. That’s an interesting historical tidbit, but the point that I want to discuss is what I think is Hicks’s failure to fully understand the significance of his own model, whose importance, regrettably, he consistently underestimated in later work (e.g., in Capital and Growth and in this paper).

The point that I want to focus on is in the second paragraph quoted above where Hicks says “mine [i.e. temporary equilibrium] was a flexprice model, a perfect competition model, in which all prices were flexible, while in Keynes’ the level of money wages (at least) was exogenously determined. So Keynes’ was a model that was consistent with unemployment, while mine, in his terms, was a full employment model.” This, it seems to me, is all wrong, because Hicks, is taking a very naïve and misguided view of what perfect competition and flexible prices mean. Those terms are often mistakenly assumed to meant that if prices are simply allowed to adjust freely, all  markets will clear and all resources will be utilized.

I think that is a total misconception, and the significance of the temporary-equilibrium construct is in helping us understand why an economy can operate sub-optimally with idle resources even when there is perfect competition and markets “clear.” What prevents optimality and allows resources to remain idle despite freely adjustming prices and perfect competition is that the expectations held by agents are not consistent. If expectations are not consistent, the plans based on those expectations are not consistent. If plans are not consistent, then how can one expect resources to be used optimally or even at all? Thus, for Hicks to assert, casually without explicit qualification, that his temporary-equilibrium model was a full-employment model, indicates to me that Hicks was unaware of the deeper significance of his own model.

If we take a full equilibrium as our benchmark, and look at how one of the markets in that full equilibrium clears, we can imagine the equilibrium as the intersection of a supply curve and a demand curve, whose positions in the standard price/quantity space depend on the price expectations of suppliers and of demanders. Different, i.e, inconsistent, price expectations would imply shifts in both the demand and supply curves from those corresponding to full intertemporal equilibrium. Overall, the price expectations consistent with a full intertemporal equilibrium will in some sense maximize total output and employment, so when price expectations are inconsistent with full intertemporal equilibrium, the shifts of the demand and supply curves will be such that they will intersect at points corresponding to less output and less employment than would have been the case in full intertemporal equilibrium. In fact, it is possible to imagine that expectations on the supply side and the demand side are so inconsistent that the point of intersection between the demand and supply curves corresponds to an output (and hence employment) that is way less than it would have been in full intertemporal equilibrium. The problem is not that the price in the market doesn’t allow the market to clear. Rather, given the positions of the demand and supply curves, their point of intersection implies a low output, because inconsistent price expectations are such that potentially advantageous trading opportunities are not being recognized.

So for Hicks to assert that his flexprice temporary-equilibrium model was (in Keynes’s terms) a full-employment model without noting the possibility of a significant contraction of output (and employment) in a perfectly competitive flexprice temporary-equilibrium model when there are significant inconsistencies in expectations suggests strongly that Hicks somehow did not fully comprehend what his own creation was all about. His failure to comprehend his own model also explains why he felt the need to abandon the flexprice temporary-equilibrium model in his later work for a fixprice model.

There is, of course, a lot more to be said about all this, and Hicks’s comments concerning the choice of a length of the period are also of interest, but the clear (or so it seems to me) misunderstanding by Hicks of what is entailed by a flexprice temporary equilibrium is an important point to recognize in evaluating both Hicks’s work and his commentary on that work and its relation to Keynes.

Aggregate Demand and Coordination Failures

Regular readers of this blog may have noticed that I have been writing less about monetary policy and more about theory and methodology than when I started blogging a little over three years ago. Now one reason for that is that I’ve already said what I want to say about policy, and, since I get bored easily, I look for new things to write about. Another reason is that, at least in the US, the economy seems to have reached a sustainable growth path that seems likely to continue for the near to intermediate term. I think that monetary policy could be doing more to promote recovery, and I wish that it would, but unfortunately, the policy is what it is, and it will continue more or less in the way that Janet Yellen has been saying it will. Falling oil prices, because of increasing US oil output, suggest that growth may speed up slightly even as inflation stays low, possibly even falling to one percent or less. At least in the short-term, the fall in inflation does not seem like a cause for concern. A third reason for writing less about monetary policy is that I have been giving a lot of thought to what it is that I dislike about the current state of macroeconomics, and as I have been thinking about it, I have been writing about it.

In thinking about what I think is wrong with modern macroeconomics, I have been coming back again and again, though usually without explicit attribution, to an idea that was impressed upon me as an undergrad and grad student by Axel Leijonhufvud: that the main concern of macroeconomics ought to be with failures of coordination. A Swede, trained in the tradition of the Wicksellian Stockholm School, Leijonhufvud immersed himself in the study of the economics of Keynes and Keynesian economics, while also mastering the Austrian literature, and becoming an admirer of Hayek, especially Hayek’s seminal 1937 paper, “Economics and Knowledge.”

In discussing Keynes, Leijonhufvud focused on two kinds of coordination failures.

First, there is a problem in the labor market. If there is unemployment because the real wage is too high, an individual worker can’t solve the problem by offering to accept a reduced nominal wage. Suppose the price of output is $1 a unit and the wage is $10 a day, but the real wage consistent with full employment is $9 a day, meaning that producers choose to produce less output than they would produce if the real wage were lower, thus hiring fewer workers than they would if the real wage were lower than it is. If an individual worker offers to accept a wage of $9 a day, but other workers continue to hold out for $10 a day, it’s not clear that an employer would want to hire the worker who offers to work for $9 a day. If employers are not hiring additional workers because they can’t cover the cost of the additional output produced with the incremental revenue generated by the added output, the willingness of one worker to work for $9 a day is not likely to make a difference to the employer’s output and hiring decisions. It is not obvious what sequence of transactions would result in an increase in output and employment when the real wage is above the equilibrium level. There are complex feedback effects from a change, so that the net effect of making those changes in a piecemeal fashion is unpredictable, even though there is a possible full-employment equilibrium with a real wage of $9 a day. If the problem is that real wages in general are too high for full employment, the willingness of an individual worker to accept a reduced wage from a single employer does not fix the problem.

In the standard competitive model, there is a perfect market for every commodity in which every transactor is assumed to be able to buy and sell as much as he wants. But the standard competitive model has very little to say about the process by which those equilibrium prices are arrived at. And a typical worker is never faced with that kind of choice posited in the competitive model: an impersonal uniform wage at which he can decide how many hours a day or week or year he wants to work at that uniform wage. Under those circumstances, Keynes argued that the willingness of some workers to accept wage cuts in order to gain employment would not significantly increase employment, and might actually have destabilizing side-effects. Keynes tried to make this argument in the framework of an equilibrium model, though the nature of the argument, as Don Patinkin among others observed, was really better suited to a disequilibrium framework. Unfortunately, Keynes’s argument was subsequently dumbed down to a simple assertion that wages and prices are sticky (especially downward).

Second, there is an intertemporal problem, because the interest rate may be stuck at a rate too high to allow enough current investment to generate the full-employment level of spending given the current level of the money wage. In this scenario, unemployment isn’t caused by a real wage that is too high, so trying to fix it by wage adjustment would be a mistake. Since the source of the problem is the rate of interest, the way to fix the problem would be to reduce the rate of interest. But depending on the circumstances, there may be a coordination failure: bear speculators, expecting the rate of interest to rise when it falls to abnormally low levels, prevent the rate of interest from falling enough to induce enough investment to support full employment. Keynes put too much weight on bear speculators as the source of the intertemporal problem; Hawtrey’s notion of a credit deadlock would actually have been a better way to go, and nowadays, when people speak about a Keynesian liquidity trap, what they really have in mind is something closer to Hawtreyan credit deadlock than to the Keynesian liquidity trap.

Keynes surely deserves credit for identifying and explaining two possible sources of coordination failures, failures affecting the macroeconomy, because interest rates and wages, though they actually come in many different shapes and sizes, affect all markets and are true macroeconomic variables. But Keynes’s analysis of those coordination failures was far from being fully satisfactory, which is not surprising; a theoretical pioneer rarely provides a fully satisfactory analysis, leaving lots of work for successors.

But I think that Keynes’s theoretical paradigm actually did lead macroeconomics in the wrong direction, in the direction of a highly aggregated model with a single output, a bond, a medium of exchange, and a labor market, with no explicit characterization of the production technology. (I.e., is there one factor or two, and if two how is the price of the second factor determined? See, here, here, here, and here my discussion of Earl Thompson’s “A Reformulation of Macroeconomic Theory,” which I hope at some point to revisit and continue.)

Why was it the wrong direction? Because, the Keynesian model (both Keynes’s own version and the Hicksian IS-LM version of his model) ruled out the sort of coordination problems that might arise in a multi-product, multi-factor, intertemporal model in which total output depends in a meaningful way on the meshing of the interdependent plans, independently formulated by decentralized decision-makers, contingent on possibly inconsistent expectations of the future. In the over-simplified and over-aggregated Keynesian model, the essence of the coordination problem has been assumed away, leaving only a residue of the actual problem to be addressed by the model. The focus of the model is on aggregate expenditure, income, and output flows, with no attention paid to the truly daunting task of achieving sufficient coordination among the independent decision makers to allow total output and income to closely approximate the maximum sustainable output and income that the system could generate in a perfectly coordinated state, aka full intertemporal equilibrium.

This way of thinking about macroeconomics led to the merging of macroeconomics with neoclassical growth theory and to the routine and unthinking incorporation of aggregate production functions in macroeconomic models, a practice that is strictly justified only in a single-output, two-factor model in which the value of capital is independent of the rate of interest, so that the havoc-producing effects of reswitching and capital-reversal can be avoided. Eventually, these models were taken over by modern real-business-cycle theorists, who dogmatically rule out any consideration of coordination problems, while attributing all observed output and employment fluctuations to random productivity shocks. If one thinks of macroeconomics as an attempt to understand coordination failures, the RBC explanation of output and employment fluctuations is totally backwards; productivity fluctuations, like fluctuations in output and employment, are the not the results of unexplained random disturbances, they are the symptoms of coordination failures. That’s it, eureka! Solve the problem by assuming that it does not exist.

If you are thinking that this seems like an Austrian critique of the Keynesian model or the Keynesian approach, you are right; it is an Austrian critique. But it has nothing to do with stereotypical Austrian policy negativism; it is a critique of the oversimplified structure of the Keynesian model, which foreshadowed the reduction ad absurdum or modern real-business-cycle theory, which has nearly banished the idea of coordination failures from modern macroeconomics. The critique is not about the lack of a roundabout capital structure; it is about the narrow scope for inconsistencies in production and consumption plans.

I think that Leijonhufvud almost 40 years ago was getting at this point when he wrote the following paragraph near toward end of his book on Keynes.

The unclear mix of statics and dynamics [in the General Theory] would seem to be main reason for later muddles. One cannot assume that what went wrong was simply that Keynes slipped up here and there in his adaptation of standard tools, and that consequently, if we go back and tinker a little more with the Marshallian toolbox his purposes will be realized. What is required, I believe, is a systematic investigation from the standpoint of the information problems stressed in this study, of what elements of the static theory of resource allocation can without further ado be utilized in the analysis of dynamic and historical systems. This, of course, would be merely a first step: the gap yawns very wide between the systematic and rigorous modern analysis of the stability of simple, “featureless,” pure exchange systems and Keynes’ inspired sketch of the income-constrained process in a monetary exchange-cum production system. But even for such a first step, the prescription cannot be to “go back to Keynes.” If one must retrace some step of past developments in order to get on the right track – and that is probably advisable – my own preference is to go back to Hayek. Hayek’s Gestalt-conception of what happens during business cycles, it has been generally agreed, was much less sound that Keynes’. As an unhappy consequence, his far superior work on the fundamentals of the problem has not received the attention it deserves. (pp. 401-02)

I don’t think that we actually need to go back to Hayek, though “Economics and Knowledge” should certainly be read by every macroeconomist, but we do need to get a clearer understanding of the potential for breakdowns in economic activity to be caused by inconsistent expectations, especially when expectations are themselves mutually dependent and reinforcing. Because expectations are mutually interdependent, they are highly susceptible to network effects. Network effects produce tipping points, tipping points can lead to catastrophic outcomes. Just wanted to share that with you. Have a nice day.

Explaining the Hegemony of New Classical Economics

Simon Wren-Lewis, Robert Waldmann, and Paul Krugman have all recently devoted additional space to explaining – ruefully, for the most part – how it came about that New Classical Economics took over mainstream macroeconomics just about half a century after the Keynesian Revolution. And Mark Thoma got them all started by a complaint about the sorry state of modern macroeconomics and its failure to prevent or to cure the Little Depression.

Wren-Lewis believes that the main problem with modern macro is too much of a good thing, the good thing being microfoundations. Those microfoundations, in Wren-Lewis’s rendering, filled certain gaps in the ad hoc Keynesian expenditure functions. Although the gaps were not as serious as the New Classical School believed, adding an explicit model of intertemporal expenditure plans derived from optimization conditions and rational expectations, was, in Wren-Lewis’s estimation, an improvement on the old Keynesian theory. The improvements could have been easily assimilated into the old Keynesian theory, but weren’t because New Classicals wanted to junk, not improve, the received Keynesian theory.

Wren-Lewis believes that it is actually possible for the progeny of Keynes and the progeny of Fisher to coexist harmoniously, and despite his discomfort with the anti-Keynesian bias of modern macroeconomics, he views the current macroeconomic research program as progressive. By progressive, I interpret him to mean that macroeconomics is still generating new theoretical problems to investigate, and that attempts to solve those problems are producing a stream of interesting and useful publications – interesting and useful, that is, to other economists doing macroeconomic research. Whether the problems and their solutions are useful to anyone else is perhaps not quite so clear. But even if interest in modern macroeconomics is largely confined to practitioners of modern macroeconomics, that fact alone would not conclusively show that the research program in which they are engaged is not progressive, the progressiveness of the research program requiring no more than a sufficient number of self-selecting econ grad students, and a willingness of university departments and sources of research funding to cater to the idiosyncratic tastes of modern macroeconomists.

Robert Waldmann, unsurprisingly, takes a rather less charitable view of modern macroeconomics, focusing on its failure to discover any new, previously unknown, empirical facts about macroeconomic, or to better explain known facts than do alternative models, e.g., by more accurately predicting observed macro time-series data. By that, admittedly, demanding criterion, Waldmann finds nothing progressive in the modern macroeconomics research program.

Paul Krugman weighed in by emphasizing not only the ideological agenda behind the New Classical Revolution, but the self-interest of those involved:

Well, while the explicit message of such manifestos is intellectual – this is the only valid way to do macroeconomics – there’s also an implicit message: from now on, only my students and disciples will get jobs at good schools and publish in major journals/ And that, to an important extent, is exactly what happened; Ken Rogoff wrote about the “scars of not being able to publish stick-price papers during the years of new classical repression.” As time went on and members of the clique made up an ever-growing share of senior faculty and journal editors, the clique’s dominance became self-perpetuating – and impervious to intellectual failure.

I don’t disagree that there has been intellectual repression, and that this has made professional advancement difficult for those who don’t subscribe to the reigning macroeconomic orthodoxy, but I think that the story is more complicated than Krugman suggests. The reason I say that is because I cannot believe that the top-ranking economics departments at schools like MIT, Harvard, UC Berkeley, Princeton, and Penn, and other supposed bastions of saltwater thinking have bought into the underlying New Classical ideology. Nevertheless, microfounded DSGE models have become de rigueur for any serious academic macroeconomic theorizing, not only in the Journal of Political Economy (Chicago), but in the Quarterly Journal of Economics (Harvard), the Review of Economics and Statistics (MIT), and the American Economic Review. New Keynesians, like Simon Wren-Lewis, have made their peace with the new order, and old Keynesians have been relegated to the periphery, unable to publish in the journals that matter without observing the generally accepted (even by those who don’t subscribe to New Classical ideology) conventions of proper macroeconomic discourse.

So I don’t think that Krugman’s ideology plus self-interest story fully explains how the New Classical hegemony was achieved. What I think is missing from his story is the spurious methodological requirement of microfoundations foisted on macroeconomists in the course of the 1970s. I have discussed microfoundations in a number of earlier posts (here, here, here, here, and here) so I will try, possibly in vain, not to repeat myself too much.

The importance and desirability of microfoundations were never questioned. What, after all, was the neoclassical synthesis, if not an attempt, partly successful and partly unsuccessful, to integrate monetary theory with value theory, or macroeconomics with microeconomics? But in the early 1970s the focus of attempts, notably in the 1970 Phelps volume, to provide microfoundations changed from embedding the Keynesian system in a general-equilibrium framework, as Patinkin had done, to providing an explicit microeconomic rationale for the Keynesian idea that the labor market could not be cleared via wage adjustments.

In chapter 19 of the General Theory, Keynes struggled to come up with a convincing general explanation for the failure of nominal-wage reductions to clear the labor market. Instead, he offered an assortment of seemingly ad hoc arguments about why nominal-wage adjustments would not succeed in reducing unemployment, enabling all workers willing to work at the prevailing wage to find employment at that wage. This forced Keynesians into the awkward position of relying on an argument — wages tend to be sticky, especially in the downward direction — that was not really different from one used by the “Classical Economists” excoriated by Keynes to explain high unemployment: that rigidities in the price system – often politically imposed rigidities – prevented wage and price adjustments from equilibrating demand with supply in the textbook fashion.

These early attempts at providing microfoundations were largely exercises in applied price theory, explaining why self-interested behavior by rational workers and employers lacking perfect information about all potential jobs and all potential workers would not result in immediate price adjustments that would enable all workers to find employment at a uniform market-clearing wage. Although these largely search-theoretic models led to a more sophisticated and nuanced understanding of labor-market dynamics than economists had previously had, the models ultimately did not provide a fully satisfactory account of cyclical unemployment. But the goal of microfoundations was to explain a certain set of phenomena in the labor market that had not been seriously investigated, in the hope that price and wage stickiness could be analyzed as an economic phenomenon rather than being arbitrarily introduced into models as an ad hoc, albeit seemingly plausible, assumption.

But instead of pursuing microfoundations as an explanatory strategy, the New Classicals chose to impose it as a methodological prerequisite. A macroeconomic model was inadmissible unless it could be explicitly and formally derived from the optimizing choices of fully rational agents. Instead of trying to enrich and potentially transform the Keynesian model with a deeper analysis and understanding of the incentives and constraints under which workers and employers make decisions, the New Classicals used microfoundations as a methodological tool by which to delegitimize Keynesian models, those models being insufficiently or improperly microfounded. Instead of using microfoundations as a method by which to make macroeconomic models conform more closely to the imperfect and limited informational resources available to actual employers deciding to hire or fire employees, and actual workers deciding to accept or reject employment opportunities, the New Classicals chose to use microfoundations as a methodological justification for the extreme unrealism of the rational-expectations assumption, portraying it as nothing more than the consistent application of the rationality postulate underlying standard neoclassical price theory.

For the New Classicals, microfoundations became a reductionist crusade. There is only one kind of economics, and it is not macroeconomics. Even the idea that there could be a conceptual distinction between micro and macroeconomics was unacceptable to Robert Lucas, just as the idea that there is, or could be, a mind not reducible to the brain is unacceptable to some deranged neuroscientists. No science, not even chemistry, has been reduced to physics. Were it ever to be accomplished, the reduction of chemistry to physics would be a great scientific achievement. Some parts of chemistry have been reduced to physics, which is a good thing, especially when doing so actually enhances our understanding of the chemical process and results in an improved, or more exact, restatement of the relevant chemical laws. But it would be absurd and preposterous simply to reject, on supposed methodological principle, those parts of chemistry that have not been reduced to physics. And how much more absurd would it be to reject higher-level sciences, like biology and ecology, for no other reason than that they have not been reduced to physics.

But reductionism is what modern macroeconomics, under the New Classical hegemony, insists on. No exceptions allowed; don’t even ask. Meekly and unreflectively, modern macroeconomics has succumbed to the absurd and arrogant methodological authoritarianism of the New Classical Revolution. What an embarrassment.

UPDATE (11:43 AM EDST): I made some minor editorial revisions to eliminate some grammatical errors and misplaced or superfluous words.

Krugman on Minsky, IS-LM and Temporary Equilibrium

Catching up on my blog reading, I found this one from Paul Krugman from almost two weeks ago defending the IS-LM model against Hyman Minsky’s criticism (channeled by his student Lars Syll) that IS-LM misrepresented the message of Keynes’s General Theory. That is an old debate, and it’s a debate that will never be resolved because IS-LM is a nice way of incorporating monetary effects into the pure income-expenditure model that was the basis of Keynes’s multiplier analysis and his policy prescriptions. On the other hand, the model leaves out much of what most interesting and insightful in the General Theory – precisely the stuff that could not easily be distilled into a simple analytic model.

Here’s Krugman:

Lars Syll approvingly quotes Hyman Minsky denouncing IS-LM analysis as an “obfuscation” of Keynes; Brad DeLong disagrees. As you might guess, so do I.

There are really two questions here. The less important is whether something like IS-LM — a static, equilibrium analysis of output and employment that takes expectations and financial conditions as given — does violence to the spirit of Keynes. Why isn’t this all that important? Because Keynes was a smart guy, not a prophet. The General Theory is interesting and inspiring, but not holy writ.

It’s also a protean work that contains a lot of different ideas, not necessarily consistent with each other. Still, when I read Minsky putting into Keynes’s mouth the claim that

Only a theory that was explicitly cyclical and overtly financial was capable of being useful

I have to wonder whether he really read the book! As I read the General Theory — and I’ve read it carefully — one of Keynes’s central insights was precisely that you wanted to step back from thinking about the business cycle. Previous thinkers had focused all their energy on trying to explain booms and busts; Keynes argued that the real thing that needed explanation was the way the economy seemed to spend prolonged periods in a state of underemployment:

[I]t is an outstanding characteristic of the economic system in which we live that, whilst it is subject to severe fluctuations in respect of output and employment, it is not violently unstable. Indeed it seems capable of remaining in a chronic condition of subnormal activity for a considerable period without any marked tendency either towards recovery or towards complete collapse.

So Keynes started with a, yes, equilibrium model of a depressed economy. He then went on to offer thoughts about how changes in animal spirits could alter this equilibrium; but he waited until Chapter 22 (!) to sketch out a story about the business cycle, and made it clear that this was not the centerpiece of his theory. Yes, I know that he later wrote an article claiming that it was all about the instability of expectations, but the book is what changed economics, and that’s not what it says.

This all seems pretty sensible to me. Nevertheless, there is so much in the General Theory — both good and bad – that isn’t reflected in IS-LM, that to reduce the General Theory to IS-LM is a kind of misrepresentation. And to be fair, Hicks himself acknowledged that IS-LM was merely a way of representing one critical difference in the assumptions underlying the Keynesian and the “Classical” analyses of macroeconomic equilibrium.

But I would take issue with the following assertion by Krugman.

The point is that Keynes very much made use of the method of temporary equilibrium — interpreting the state of the economy in the short run as if it were a static equilibrium with a lot of stuff taken provisionally as given — as a way to clarify thought. And the larger point is that he was right to do this.

When people like me use something like IS-LM, we’re not imagining that the IS curve is fixed in position for ever after. It’s a ceteris paribus thing, just like supply and demand. Assuming short-run equilibrium in some things — in this case interest rates and output — doesn’t mean that you’ve forgotten that things change, it’s just a way to clarify your thought. And the truth is that people who try to think in terms of everything being dynamic all at once almost always end up either confused or engaging in a lot of implicit theorizing they don’t even realize they’re doing.

When I think of a temporary equilibrium, the most important – indeed the defining — characteristic of that temporary equilibrium is that expectations of at least some agents have been disappointed. The disappointment of expectations is likely to, but does not strictly require, a revision of disappointed expectations and of the plans conditioned on those expectations. The revision of expectations and plans as a result of expectations being disappointed is what gives rise to a dynamic adjustment process. But that is precisely what is excluded from – or at least not explicitly taken into account by – the IS-LM model. There is nothing in the IS-LM model that provides any direct insight into the process by which expectations are revised as a result of being disappointed. That Keynes could so easily think in terms of a depressed economy being in equilibrium suggests to me that he was missing what I regard as the key insight of the temporary-equilibrium method.

Of course, there are those who argue, perhaps most notably Roger Farmer, that economies have multiple equilibria, each with different levels of output and employment corresponding to different expectational parameters. That seems to me a more Keynesian approach, an approach recognizing that expectations can be self-fulfilling, than the temporary-equilibrium approach in which the focus is on mistaken and conflicting expectations, not their self-fulfillment.

Now to be fair, I have to admit that Hicks, himself, who introduced the temporary-equilibrium approach in Value and Capital (1939) later (1965) suggested in Capital and Growth (p. 65) that both the Keynes in the General Theory and the temporary-equilibrium approach of Value and Capital were “quasi-static.” The analysis of the General Theory “is not the analysis of a process; no means has been provided by which we can pass from one Keynesian period to the next. . . . The Temporary Equilibrium model of Value and Capital, also, is quasi-static in just the same sense. The reason why I was contented with such a model was because I had my eyes fixed on Keynes.

Despite Hicks’s identification of the temporary-equilibrium method with Keynes’s method in the General Theory, I think that Hicks was overly modest in assessing his own contribution in Value and Capital, failing to appreciate the full significance of the method he had introduced. Which, I suppose, just goes to show that you can’t assume that the person who invents a concept or an idea is necessarily the one who has the best, or most comprehensive, understanding of what the concept means of what its significance is.

The Trouble with IS-LM (and its Successors)

Lately, I have been reading a paper by Roger Backhouse and David Laidler, “What Was Lost with IS-LM” (an earlier version is available here) which was part of a very interesting symposium of 11 papers on the IS-LM model published as a supplement to the 2004 volume of History of Political Economy. The main thesis of the paper is that the IS-LM model, like the General Theory of which it is a partial and imperfect distillation, aborted a number of promising developments in the rapidly developing, but still nascent, field of macroeconomics in the 1920 and 1930s, developments that just might, had they not been elbowed aside by the IS-LM model, have evolved into a more useful and relevant theory of macroeconomic fluctuations and policy than we now possess. Even though I have occasionally sparred with Scott Sumner about IS-LM – with me pushing back a bit at Scott’s attacks on IS-LM — I have a lot of sympathy for the Backhouse-Laidler thesis.

The Backhouse-Laidler paper is too long to summarize, but I will just note that there are four types of loss that they attribute to IS-LM, which are all, more or less, derivative of the static equilibrium character of Keynes’s analytic method in both the General Theory and the IS-LM construction.

1 The loss of dynamic analysis. IS-LM is a single-period model.

2 The loss of intertemporal choice and expectations. Intertemporal choice and expectations are excluded a priori in a single-period model.

3 The loss of policy regimes. In a single-period model, policy is a one-time affair. The problem of setting up a regime that leads to optimal results over time doesn’t arise.

4 The loss of intertemporal coordination failures. Another concept that is irrelevant in a one-period model.

There was one particular passage that I found especially impressive. Commenting on the lack of any systematic dynamic analysis in the GT, Backhouse and Laidler observe,

[A]lthough [Keynes] made many remarks that could be (and in some cases were later) turned into dynamic models, the emphasis of the General Theory was nevertheless on unemployment as an equilibrium phenomenon.

Dynamic accounts of how money wages might affect employment were only a little more integrated into Keynes’s formal analysis than they were later into IS-LM. Far more significant for the development in Keynes’s thought is how Keynes himself systematically neglected dynamic factors that had been discussed in previous explanations of unemployment. This was a feature of the General Theory remarked on by Bertil Ohlin (1937, 235-36):

Keynes’s theoretical system . . . is equally “old-fashioned” in the second respect which characterizes recent economic theory – namely, the attempt to break away from an explanation of economic events by means of orthodox equilibrium constructions. No other analysis of trade fluctuations in recent years – with the possible exception of the Mises-Hayek school – follows such conservative lines in this respect. In fact, Keynes is much more of an “equilibrium theorist” than such economists as Cassel and, I think, Marshall.

Backhouse and Laidler go on to cite the Stockholm School (of which Ohlin was a leading figure) as an example of explicitly dynamic analysis.

As Bjorn Hansson (1982) has shown, this group developed an explicit method, using the idea of a succession of “unit periods,” in which each period began with agents having plans based on newly formed expectations about the outcome of executing them, and ended with the economy in some new situation that was the outcome of executing them, and ended with the economy in some new situation that was the outcome of market processes set in motion by the incompatibility of those plans, and in which expectations had been reformulated, too, in the light of experience. They applied this method to the construction of a wide variety of what they called “model sequences,” many of which involved downward spirals in economic activity at whose very heart lay rising unemployment. This is not the place to discuss the vexed question of the extent to which some of this work anticipated the Keynesian multiplier process, but it should be noted that, in IS-LM, it is the limit to which such processes move, rather than the time path they follow to get there, that is emphasized.

The Stockholm method seems to me exactly the right way to explain business-cycle downturns. In normal times, there is a rough – certainly not perfect, but good enough — correspondence of expectations among agents. That correspondence of expectations implies that the individual plans contingent on those expectations will be more or less compatible with one another. Surprises happen; here and there people are disappointed and regret past decisions, but, on the whole, they are able to adjust as needed to muddle through. There is usually enough flexibility in a system to allow most people to adjust their plans in response to unforeseen circumstances, so that the disappointment of some expectations doesn’t become contagious, causing a systemic crisis.

But when there is some sort of major shock – and it can only be a shock if it is unforeseen – the system may not be able to adjust. Instead, the disappointment of expectations becomes contagious. If my customers aren’t able to sell their products, I may not be able to sell mine. Expectations are like networks. If there is a breakdown at some point in the network, the whole network may collapse or malfunction. Because expectations and plans fit together in interlocking networks, it is possible that even a disturbance at one point in the network can cascade over an increasingly wide group of agents, leading to something like a system-wide breakdown, a financial crisis or a depression.

But the “problem” with the Stockholm method was that it was open-ended. It could offer only “a wide variety” of “model sequences,” without specifying a determinate solution. It was just this gap in the Stockholm approach that Keynes was able to fill. He provided a determinate equilibrium, “the limit to which the Stockholm model sequences would move, rather than the time path they follow to get there.” A messy, but insightful, approach to explaining the phenomenon of downward spirals in economic activity coupled with rising unemployment was cast aside in favor of the neater, simpler approach of Keynes. No wonder Ohlin sounds annoyed in his comment, quoted by Backhouse and Laidler, about Keynes. Tractability trumped insight.

Unfortunately, that is still the case today. Open-ended models of the sort that the Stockholm School tried to develop still cannot compete with the RBC and DSGE models that have displaced IS-LM and now dominate modern macroeconomics. The basic idea that modern economies form networks, and that networks have properties that are not reducible to just the nodes forming them has yet to penetrate the trained intuition of modern macroeconomists. Otherwise, how would it have been possible to imagine that a macroeconomic model could consist of a single representative agent? And just because modern macroeconomists have expanded their models to include more than a single representative agent doesn’t mean that the intellectual gap evidenced by the introduction of representative-agent models into macroeconomic discourse has been closed.

How to Think about Own Rates of Interest, Version 2.0

In my previous post, I tried to explain how to think about own rates of interest. Unfortunately, I made a careless error in calculating the own rate of interest in the simple example I constructed to capture the essence of Sraffa’s own-rate argument against Hayek’s notion of the natural rate of interest. But sometimes these little slip-ups can be educational, so I am going to try to turn my conceptual misstep to advantage in working through and amplifying the example I presented last time.

But before I reproduce the passage from Sraffa’s review that will serve as our basic text in this post as it did in the previous post, I want to clarify another point. The own rate of interest for a commodity may be calculated in terms of any standard of value. If I borrow wheat and promise to repay in wheat, the wheat own rate of interest may be calculated in terms of wheat or in terms of any other standard; all of those rates are own rates, but each is expressed in terms of a different standard.

Lend me 100 bushels of wheat today, and I will pay you back 102 bushels next year. The own rate of interest for wheat in terms of wheat would be 2%. Alternatively, I could borrow $100 of wheat today and promise to pay back $102 of wheat next year. The own rate of interest for wheat in terms of wheat and the own rate of interest for wheat in terms of dollars would be equal if and only if the forward dollar price of wheat is the same as the current dollar price of wheat. The commodity or asset in terms of which a price is quoted or in terms of which we measure the own rate is known as the numeraire. (If all that Sraffa was trying to say in criticizing Hayek was that there are many equivalent ways of expressing own interest rates, he was making a trivial point. Perhaps Hayek didn’t understand that trivial point, in which case the rough treatment he got from Sraffa was not undeserved. But it seems clear that Sraffa was trying — unsuccessfully — to make a more substantive point than that.)

In principle, there is a separate own rate of interest for every commodity and for every numeraire. If there are n commodities, there are n potential numeraires, and n own rates can be expressed in terms of each numeraire. So there are n-squared own rates. Each own rate can be thought of as equilibrating the demand for loans made in terms of a given commodity and a given numeraire. But arbitrage constraints tightly link all these separate own rates together. If it were cheaper to borrow in terms of one commodity than another, or in terms of one numeraire than another, borrowers would switch to the commodity and numeraire with the lowest cost of borrowing, and if it were more profitable to lend in terms of one commodity, or in terms of one numeraire, than another, lenders would switch to lending in terms of the commodity or numeraire with the highest return.

Thus, competition tends to equalize own rates across all commodities and across all numeraires. Of course, perfect arbitrage requires the existence of forward markets in which to contract today for the purchase or sale of a commodity at a future date. When forward markets don’t exist, some traders may anticipate advantages to borrowing or lending in terms of particular commodities based on their expectations of future prices for those commodities. The arbitrage constraint on the variation of interest rates was discovered and explained by Irving Fisher in his great work Appreciation and Interest.

It is clear that if the unit of length were changed and its change were foreknown, contracts would be modified accordingly. Suppose a yard were defined (as once it probably was) to be the length of the king’s girdle, and suppose the king to be a child. Everybody would then know that the “yard” would increase with age and a merchant who should agree to deliver 1000 “yards” ten years hence, would make his terms correspond to his expectations. To alter the mode of measurement does not alter the actual quantities involved but merely the numbers by which they are represented. (p. 1)

We thus see that the farmer who contracts a mortgage in gold is, if the interest is properly adjusted, no worse and no better off than if his contract were in a “wheat” standard or a “multiple” standard. (p. 16)

I pause to make a subtle, but, I think, an important, point. Although the relationship between the spot and the forward price of any commodity tightly constrains the own rate for that commodity, the spot/forward relationship does not determine the own rate of interest for that commodity. There is always some “real” rate reflecting a rate of intertemporal exchange that is consistent with intertemporal equilibrium. Given such an intertemporal rate of exchange — a real rate of interest — the spot/forward relationship for a commodity in terms of a numeraire pins down the own rate for that commodity in terms of that numeraire.

OK with that introduction out of the way, let’s go back to my previous post in which I wrote the following:

Sraffa correctly noted that arbitrage would force the terms of such a loan (i.e., the own rate of interest) to equal the ratio of the current forward price of the commodity to its current spot price, buying spot and selling forward being essentially equivalent to borrowing and repaying.

That statement now seems quite wrong to me. Sraffa did not assert that arbitrage would force the own rate of interest to equal the ratio of the spot and forward prices. He merely noted that in a stationary equilibrium with equality between all spot and forward prices, all own interest rates would be equal. I criticized him for failing to note that in a stationary equilibrium all own rates would be zero. The conclusion that all own rates would be zero in a stationary equilibrium might in fact be valid, but if it is, it is not as obviously valid as I suggested, and my criticism of Sraffa and Ludwig von Mises for not drawing what seemed to me an obvious inference was not justified. To conclude that own rates are zero in a stationary equilibrium, you would, at a minimum, have to show that there is at least one commodity which could be carried from one period to the next at a non-negative profit. Sraffa may have come close to suggesting such an assumption in the passage in which he explains how borrowing to buy cotton spot and immediately selling cotton forward can be viewed as the equivalent of contracting a loan in terms of cotton, but he did not make that assumption explicitly. In any event, I mistakenly interpreted him to be saying that the ratio of the spot and forward prices is the same as the own interest rate, which is neither true nor what Sraffa meant.

And now let’s finally go back to the key quotation of Sraffa’s that I tried unsuccessfully to parse in my previous post.

Suppose there is a change in the distribution of demand between various commodities; immediately some will rise in price, and others will fall; the market will expect that, after a certain time, the supply of the former will increase, and the supply of the latter fall, and accordingly the forward price, for the date on which equilibrium is expected to be restored, will be below the spot price in the case of the former and above it in the case of the latter; in other words, the rate of interest on the former will be higher than on the latter. (“Dr. Hayek on Money and Capital,” p. 50)

In my previous post I tried to flesh out Sraffa’s example by supposing that, in the stationary equilibrium before the demand shift, tomatoes and cucumbers were both selling for a dollar each. In a stationary equilibrium, tomato and cucumber prices would remain, indefinitely into the future, at a dollar each. A shift in demand from tomatoes to cucumbers upsets the equilibrium, causing the price of tomatoes to fall to, say, $.90 and the price of cucumbers to rise to, say, $1.10. But Sraffa also argued that the prices of tomatoes and cucumbers would diverge only temporarily from their equilibrium values, implicitly assuming that the long-run supply curves of both tomatoes and cucumbers are horizontal at a price of $1 per unit.

I misunderstood Sraffa to be saying that the ratio of the future price and the spot price of tomatoes equals one plus the own rate on tomatoes. I therefore incorrectly calculated the own rate on tomatoes as 1/.9 minus one or 11.1%. There were two mistakes. First, I incorrectly inferred that equality of all spot and forward prices implies that the real rate must be zero, and second, as Nick Edmunds pointed out in his comment, a forward price exceeding the spot price would actually be reflected in an own rate less than the zero real rate that I had been posited. To calculate the own rate on tomatoes, I ought to have taken the ratio of spot price to the forward price — (.9/1) — and subtracted one plus the real rate. If the real rate is zero, then the implied own rate is .9 minus 1, or -10%.

To see where this comes from, we can take the simple algebra from Fisher (pp. 8-9). Let i be the interest rate calculated in terms of one commodity and one numeraire, and j be the rate of interest calculated in terms of a different commodity in that numeraire. Further, let a be the rate at which the second commodity appreciates relative to the first commodity. We have the following relationship derived from the arbitrage condition.

(1 + i) = (1 + j)(1 + a)

Now in our case, we are trying to calculate the own rate on tomatoes given that tomatoes are expected (an expectation reflected in the forward price of tomatoes) to appreciate by 10% from $.90 to $1.00 over the term of the loan. To keep the analysis simple, assume that i is zero. Although I concede that a positive real rate may be consistent with the stationary equilibrium that I, following Sraffa, have assumed, a zero real rate is certainly not an implausible assumption, and no important conclusions of this discussion hinge on assuming that i is zero.

To apply Fisher’s framework to Sraffa’s example, we need only substitute the ratio of the forward price of tomatoes to the spot price — [p(fwd)/p(spot)] — for the appreciation factor (1 + a).

So, in place of the previous equation, I can now substitute the following equivalent equation:

(1 + i) = (1 + j) [p(fwd)/p(spot)].

Rearranging, we get:

[p(spot)/p(fwd)] (1 + i) = (1 + j).

If i = 0, the following equation results:

[p(spot)/p(fwd)] = (1 + j).

In other words:

j = [p(spot)/p(fwd)] – 1.

If the ratio of the spot to the forward price is .9, then the own rate on tomatoes, j, equals -10%.

My assertion in the previous post that the own rate on cucumbers would be negative by the amount of expected depreciation (from $1.10 to $1) in the next period was also backwards. The own rate on cucumbers would have to exceed the zero equilibrium real rate by as much as cucumbers would depreciate at the time of repayment. So, for cucumbers, j would equal 11%.

Just to elaborate further, let’s assume that there is a third commodity, onions, and that, in the initial equilibrium, the unit prices of onions, tomatoes and cucumbers are equal. If the demand shift from tomatoes to cucumbers does not affect the demand for onions, then, even after the shift in demand, the price of onions will remain one dollar per onion.

The table below shows prices and own rates for tomatoes, cucumbers and onions for each possible choice of numeraire. If prices are quoted in tomatoes, the price of tomatoes is fixed at 1. Given a zero real rate, the own rate on tomatoes in period is zero. What about the own rate on cucumbers? In period 0, with no change in prices expected, the own rate on cucumbers is also zero. However in period 1, after the price of cucumbers has risen to 1.22 tomatoes, the own rate on cucumbers must reflect the expected reduction in the price of a cucumber in terms of tomatoes from 1.22 tomatoes in period 1 to 1 tomato in period 2, a price reduction of 22% percent in terms of tomatoes, implying a cucumber own rate of 22% in terms of tomatoes. Similarly, the onion own rate in terms of tomatoes would be 11% percent reflecting a forward price for onions in terms of tomatoes 11% below the spot price for onions in terms of tomatoes. If prices were quoted in terms of cucumbers, the cucumber own rate would be zero, and because the prices of tomatoes and onions would be expected to rise in terms of cucumbers, the tomato and onion own rates would be negative (-18.2% for tomatoes and -10% for onions). And if prices were quoted in terms of onions, the onion own rate would be zero, while the tomato own rate, given the expected appreciation of tomatoes in terms of onions, would be negative (-10%), and the cucumber own rate, given the expected depreciation of cucumbers in terms of onions, would be positive (10%).

own_rates_in_terms_of_tomatoes_cucumbers_onions

The next table, summarizing the first one, is a 3 by 3 matrix showing each of the nine possible combinations of numeraires and corresponding own rates.

own_rates_in_terms_of_tomatoes_cucumbers_onions_2

Thus, although the own rates of the different commodities differ, and although the commodity own rates differ depending on the choice of numeraire, the cost of borrowing (and the return to lending) is equal regardless of which commodity and which numeraire is chosen. As I stated in my previous post, Sraffa believed that, by showing that own rates can diverge, he showed that Hayek’s concept of a natural rate of interest was a nonsense notion. However, the differences in own rates, as Fisher had already showed 36 years earlier, are purely nominal. The underlying real rate, under Sraffa’s own analysis, is independent of the own rates.

Moreover, as I pointed out in my previous post, though the point was made in the context of a confused exposition of own rates,  whenever the own rate for a commodity is negative, there is an incentive to hold it now for sale in the next period at a higher price it would fetch in the current period. It is therefore only possible to observe negative own rates on commodities that are costly to store. Only if the cost of holding a commodity is greater than its expected appreciation would it not be profitable to withhold the commodity from sale this period and to sell instead in the following period. The rate of appreciation of a commodity cannot exceed the cost of storing it (as a percentage of its price).

What do I conclude from all this? That neither Sraffa nor Hayek adequately understood Fisher. Sraffa seems to have argued that there would be multiple real own rates of interest in disequilibrium — or at least his discussion of own rates seem to suggest that that is what he thought — while Hayek failed to see that there could be multiple nominal own rates. Fisher provided a definitive exposition of the distinction between real and nominal rates that encompasses both own rates and money rates of interest.

A. C. Pigou, the great and devoted student of Alfred Marshall, and ultimately his successor at Cambridge, is supposed to have said “It’s all in Marshall.” Well, one could also say “it’s all in Fisher.” Keynes, despite going out of his way in Chapter 12 of the General Theory to criticize Fisher’s distinction between the real and nominal rates of interest, actually vindicated Fisher’s distinction in his exposition of own rates in Chapter 17 of the GT, providing a valuable extension of Fisher’s analysis, but apparently failing to see the connection between his discussion and Fisher’s, and instead crediting Sraffa for introducing the own-rate analysis, even as he undermined Sraffa’s ambiguous suggestion that real own rates could differ. Go figure.

How to Think about Own Rates of Interest

Phil Pilkington has responded to my post about the latest version of my paper (co-authored by Paul Zimmerman) on the Sraffa-Hayek debate about the natural rate of interest. For those of you who haven’t been following my posts on the subject, here’s a quick review. Almost three years ago I wrote a post refuting Sraffa’s argument that Hayek’s concept of the natural rate of interest is incoherent, there being a multiplicity of own rates of interest in a barter economy (Hayek’s benchmark for the rate of interest undisturbed by monetary influences), which makes it impossible to identify any particular own rate as the natural rate of interest.

Sraffa maintained that if there are many own rates of interest in a barter economy, none of them having a claim to priority over the others, then Hayek had no basis for singling out any particular one of them as the natural rate and holding it up as the benchmark rate to guide monetary policy. I pointed out that Ludwig Lachmann had answered Sraffa’s attack (about 20 years too late) by explaining that even though there could be many own rates for individual commodities, all own rates are related by the condition that the cost of borrowing in terms of all commodities would be equalized, differences in own rates reflecting merely differences in expected appreciation or depreciation of the different commodities. Different own rates are simply different nominal rates; there is a unique real own rate, a point demonstrated by Irving Fisher in 1896 in Appreciation and Interest.

Let me pause here for a moment to explain what is meant by an own rate of interest. It is simply the name for the rate of interest corresponding to a loan contracted in terms of a particular commodity, the borrower receiving the commodity now and repaying the lender with the same commodity when the term of the loan expires. Sraffa correctly noted that in equilibrium arbitrage would force the terms of such a loan (i.e., the own rate of interest) to equal the ratio of the current forward price of the commodity to its current spot price, buying spot and selling forward being essentially equivalent to borrowing and repaying.

Now what is tricky about Sraffa’s argument against Hayek is that he actually acknowledges at the beginning of his argument that in a stationary equilibrium, presumably meaning that prices remain at their current equilibrium levels over time, all own rates would be equal. In fact if prices remain (and are expected to remain) constant period after period, the ratio of forward to spot prices would equal unity for all commodities implying that the natural rate of interest would be zero. Sraffa did not make that point explicitly, but it seems to be a necessary implication of his analysis. (This implication seems to bear on an old controversy in the theory of capital and interest, which is whether the rate of interest would be positive in a stationary equilibrium with constant real income). Schumpeter argued that the equilibrium rate of interest would be zero, and von Mises argued that it would be positive, because time preference implying that the rate of interest is necessarily always positive is a kind of a priori praxeological law of nature, the sort of apodictic gibberish to which von Mises was regrettably predisposed. The own-rate analysis supports Schumpeter against Mises.

So to make the case against Hayek, Sraffa had to posit a change, a shift in demand from one product to another, that disrupts the pre-existing equilibrium. Here is the key passage from Sraffa:

Suppose there is a change in the distribution of demand between various commodities; immediately some will rise in price, and others will fall; the market will expect that, after a certain time, the supply of the former will increase, and the supply of the latter fall, and accordingly the forward price, for the date on which equilibrium is expected to be restored, will be below the spot price in the case of the former and above it in the case of the latter; in other words, the rate of interest on the former will be higher than on the latter. (p. 50)

This is a difficult passage, and in previous posts, and in my paper with Zimmerman, I did not try to parse this passage. But I am going to parse it now. Assume that demand shifts from tomatoes to cucumbers. In the original equilibrium, let the prices of both be $1 a pound. With a zero own rate of interest in terms of both tomatoes and cucumbers, you could borrow a pound of tomatoes today and discharge your debt by repaying the lender a pound of tomatoes at the expiration of the loan. However, after the demand shift, the price of tomatoes falls to, say, $0.90 a pound, and the price of cucumbers rises to, say, $1.10 a pound. Sraffa posits that the price changes are temporary, not because the demand shift is temporary, but because the supply curves of tomatoes and cucumbers are perfectly elastic at $1 a pound. However, supply does not adjust immediately, so Sraffa believes that there can be a temporary deviation from the long-run equilibrium prices of tomatoes and cucumbers.

The ratio of the forward prices to the spot prices tells you what the own rates are for tomatoes and cucumbers. For tomatoes, the ratio is 1/.9, implying an own rate of 11.1%. For cucumbers the ratio is 1/1.1, implying an own rate of -9.1%. Other prices have not changed, so all other own rates remain at 0. Having shown that own rates can diverge, Sraffa thinks that he has proven Hayek’s concept of a natural rate of interest to be a nonsense notion. He was mistaken.

There are at least two mistakes. First, the negative own rate on cucumbers simply means that no one will lend in terms of cucumbers for negative interest when other commodities allow lending at zero interest. It also means that no one will hold cucumbers in this period to sell at a lower price in the next period than the cucumbers would fetch in the current period. Cucumbers are a bad investment, promising a negative return; any lending and investing will be conducted in terms of some other commodity. The negative own rate on cucumbers signifies a kind of corner solution, reflecting the impossibility of transporting next period’s cucumbers into the present. If that were possible cucumber prices would be equal in the present and the future, and the cucumber own rate would be equal to all other own rates at zero. But the point is that if any lending takes place, it will be at a zero own rate.

Second, the positive own rate on tomatoes means that there is an incentive to lend in terms of tomatoes rather than lend in terms of other commodities. But as long as it is possible to borrow in terms of other commodities at a zero own rate, no one borrows in terms of tomatoes. Thus, if anyone wanted to lend in terms of tomatoes, he would have to reduce the rate on tomatoes to make borrowers indifferent between borrowing in terms of tomatoes and borrowing in terms of some other commodity. However, if tomatoes today can be held at zero cost to be sold at the higher price prevailing next period, currently produced tomatoes would be sold in the next period rather than sold today. So if there were no costs of holding tomatoes until the next period, the price of tomatoes in the next period would be no higher than the price in the current period. In other words, the forward price of tomatoes cannot exceed the current spot price by more than the cost of holding tomatoes until the next period. If the difference between the spot and the forward price reflects no more than the cost of holding tomatoes till the next period, then, as Keynes showed in chapter 17 of the General Theory, the own rates are indeed effectively equalized after appropriate adjustment for storage costs and expected appreciation.

Thus, it was Keynes, who having selected Sraffa to review Hayek’s Prices and Production in the Economic Journal, of which Keynes was then the editor, adapted Sraffa’s own rate analysis in the General Theory, but did so in a fashion that, at least partially, rehabilitated the very natural-rate analysis that had been the object of Sraffa’s scorn in his review of Prices and Production. Keynes also rejected the natural-rate analysis, but he did so not because it is nonsensical, but because the natural rate is not independent of the level of employment. Keynes’s argument that the natural rate depends on the level of employment seems to me to be inconsistent with the idea that the IS curve is downward sloping. But I will have to think about that a bit and reread the relevant passage in the General Theory and perhaps revisit the point in a future post.

 UPDATE (07/28/14 13:02 EDT): Thanks to my commenters for pointing out that my own thinking about the own rate of interest was not quite right. I should have defined the own rate in terms of a real numeraire instead of $, which was a bit of awkwardness that I should have fixed before posting. I will try to publish a corrected version of this post later today or tomorrow. Sorry for posting without sufficient review and revision.

UPDATE (08/04/14 11:38 EDT): I hope to post the long-delayed sequel to this post later today. A number of personal issues took precedence over posting, but I also found it difficult to get clear on several minor points, which I hope that I have now resolved adequately, for example I found that defining the own rate in terms of a real numeraire was not really the source of my problem with this post, though it was a useful exercise to work through. Anyway, stay tuned.


About Me

David Glasner
Washington, DC

I am an economist at the Federal Trade Commission. Nothing that you read on this blog necessarily reflects the views of the FTC or the individual commissioners. Although I work at the FTC as an antitrust economist, most of my research and writing has been on monetary economics and policy and the history of monetary theory. In my book Free Banking and Monetary Reform, I argued for a non-Monetarist non-Keynesian approach to monetary policy, based on a theory of a competitive supply of money. Over the years, I have become increasingly impressed by the similarities between my approach and that of R. G. Hawtrey and hope to bring Hawtrey's unduly neglected contributions to the attention of a wider audience.

Enter your email address to follow this blog and receive notifications of new posts by email.

Join 279 other followers


Follow

Get every new post delivered to your Inbox.

Join 279 other followers