This was that the market process had to include a production stage where the quant.i.ty and value of output exceeded the value of inputs in Marx's terms and in Sraffa's (discussed in Chapter 6), a surplus is produced. The capitalist pays a fair price for his raw materials, and a fair wage to his employees. They are then combined in a production process which generates commodities for sale where the physical quant.i.ty of commodities and their monetary value exceed the quant.i.ty and value of inputs. The commodities are then sold for more than the cost of the raw materials and workers' wages, yielding a profit. The profit allows the capitalist to fulfill his desire to acc.u.mulate wealth, without robbing any other market partic.i.p.ants, and without having to buy commodities below their value and sell them above it.7 Say's Law and Walras's Law, on the other hand, begin from the abstraction of an exchange-only economy: an economy in which goods exist at the outset, but where no production takes place (production is shoehorned into the a.n.a.lysis at a later point, but unsatisfactorily, as I outline below). The market simply enables the exchange of pre-existing goods. In such an economy, surplus in Marx's sense would be impossible. Equally, if one agent desired to and did acc.u.mulate wealth, that would necessarily involve theft in the Say's Principle sense. However, this condition does not hold when we move from the fiction of an exchange-only economy to the reality of a production and exchange economy. With production, it is possible for agents to desire to acc.u.mulate wealth without therefore aspiring to be thieves.

Marx formalized this a.n.a.lysis in terms of two 'circuits,' the 'Circuit of Commodities' and the 'Circuit of Capital.'

In the Circuit of Commodities, people come to market with commodities, which they exchange for money in order to buy other commodities. Marx stylized this as C M C: CommodityMoneyCommodity.

Though Marx discussed various ways in which this circuit could fail owing primarily to delays between the sale of one commodity and the purchase of the next generally speaking it obeys Walras's Law. Each 'agent' desires to convert commodities of a given value into different commodities of equivalent value.

However, in the Circuit of Capital, people came to market with money, with the intention of turning this money into more money. These agents buy commodities specifically, labor and raw materials with money, put these to work in a factory to produce other commodities, and then sell these commodities for (hopefully) more money, thus making a profit. Marx stylized this as M C M+: MoneyCommodityMore money.

The complete circuit, and the one which emphasizes the fallacy behind Walras's Law, was M C(L, MP) ... P ... C+c M+m: MoneyLabor and means of production ... Production ... Different commodities, of greater value than paid for the labor and means of productionSale of commodities to generate more money This circuit specifically violates Say's Principle and Walras's Law. Rather than simply wanting to exchange one set of commodities for another of equivalent value, the agents in this circuit wish to complete it with more wealth than they started with. If we focus upon the commodity stages of this circuit, then, as Marx says, these agents wish to supply more than they demand, and to acc.u.mulate the difference as profit which adds to their wealth. Their supply is the commodities they produce for sale. Their demand is the inputs to production they purchase the labor and raw materials. In Say's Principle's terms, the sum of these, their excess demand, is negative. When the two circuits are added together, the sum of all excess demands in a capitalist economy is likewise negative (prior to the introduction of credit, which we consider below).

This explanation of why Say's Law and Walras's Law don't apply to a market economy is far clearer than Keynes's, and the great pity is that Keynes didn't use it in the General Theory, because it was in his 1933 draft. In this draft, Keynes observes that Marx made the 'pregnant observation' that: [T]he nature of production in the actual world is not C M C', i.e. of exchanging commodity (or effort) for money in order to obtain another commodity (or effort). That may be the standpoint of the private consumer. But it is not the att.i.tude of business, which is a case of M C M', i.e., of parting with money for commodity (or effort) in order to obtain more money. (Dillard 1984: 424, citing Keynes's Collected Works, vol. 29, p. 81) Keynes continued in a footnote that this vision of capitalism as having two circuits, one of which was motivated solely by the desire to acc.u.mulate wealth, in turn implied the likelihood of periodic crises when expectations of profit were not met: Marx, however, was approaching the intermediate truth when he added that the continuous excess of M' [over M] would be inevitably interrupted by a series of crises, gradually increasing in intensity, or entrepreneur bankruptcy and underemployment, during which, presumably M must be in excess. My own argument, if it is accepted, should at least serve to effect a reconciliation between the followers of Marx and those of Major Douglas, leaving the cla.s.sical economics still high and dry in the belief that M and M' are always equal. (Ibid.: 424, citing Keynes's Collected Works, vol. 29, p. 82n.) Unfortunately, Keynes later subst.i.tuted his own convoluted reasoning for Marx's, I expect for two reasons. First, his argument was an attempt to put Marx's logic into the Marshallian framework in which Keynes was educated; secondly, he probably made a political judgment, at a time when Stalin's power was rising and communism had great political appeal, not to acknowledge the 'father of communism' in his critique of conventional economics.

Had Marx's clear logic been brought to center stage by Keynes, it is feasible that the 'neocla.s.sical counter-revolution' initiated by Hicks might not have even commenced, because the fact that Keynes rejected Walras's Law, and his sound reasons for doing so, would have been so much clearer. So although Keynes's decision can be understood in the context of his times, with the benefit of hindsight, it was a serious mistake. Keynes's obscure and confusing argument allowed economists to continue believing that Walras's Law was an irrefutable truth. Only those working outside the neocla.s.sical mainstream realized otherwise.

Credit and the fallacy of Walras's Law.

Minsky, like Keynes before him, also omitted any reference to Marx in his own work, but his reasons for doing so are far easier to accept: given that he was an American academic during the McCarthyist period, any acknowledgment of Marx would have seriously impeded his academic career, if not ended it altogether.8 However, he was strongly influenced by Marx's a.n.a.lysis, and took Marx's logic one step farther. He pointed out that since there is a buyer for every seller, and since accounting demands that expenditure must equal receipts, and yet growth also occurs over time, then credit and debt must make up the gap. Credit and debt are therefore fundamental to capitalism: If income is to grow, the financial markets, where the various plans to save and invest are reconciled, must generate an aggregate demand that, aside from brief intervals, is ever rising. For real aggregate demand to be increasing, [...] it is necessary that current spending plans, summed over all sectors, be greater than current received income and that some market technique exist by which aggregate spending in excess of aggregate antic.i.p.ated income can be financed. It follows that over a period during which economic growth takes place, at least some sectors finance a part of their spending by emitting debt or selling a.s.sets. (Minsky 1982 [1963]: 6; emphasis added) Minsky's insight here points out the pivotal blind spot in thinking which leads neocla.s.sical and Austrian economists to believe respectively in Walras's Law and Say's Law: they fail to consider the role of credit in a capitalist economy.

Say, banker, can you spare a dime? Say's Law and Walras's Law envisage a world in which commodities are purchased only from the proceeds of selling other commodities, and in which commodities are the only things that are bought and sold. As Kates put it: According to the law of markets [Say's Law], aggregate demand was a conception unnecessary for a proper understanding of the cyclical behavior of economies. There were, of course, purchases and sales, and one could add together in some way everything bought during a period of time and describe this as aggregate demand [...] [but] demand was not thought of as independent of supply. Instead, demand was const.i.tuted by supply; one could not demand without first having produced. Or to be more precise, demand in aggregate was made up of supplies in aggregate. (Kates 2003: 734) In contrast to the position put by Kates, the world in which we live is one in which goods are purchased using both the proceeds of selling other goods and credit, while what is bought and sold includes existing a.s.sets as well as newly produced goods.

Aggregate demand is therefore aggregate supply plus the change in debt, while aggregate demand is expended on both commodities and a.s.sets (shares and property).9 This guarantees the overall accounting balance that is an integral part of both Say's Law and Walras's Law, but it includes both the role of credit and the role of a.s.set sales in a capitalist economy, which both of those 'laws' omit. Those 'laws' are thus relevant only to a world of either pure exchange or simple commodity production the world that Marx characterizes as CMC but are not relevant to the (normally) growing capitalist world in which we actually live.

The Say's Law/Walras's Law fallacy of ignoring the role of credit is the foundation of the neocla.s.sical (and Austrian) argument that 'general gluts' and depressions are impossible, and that all crises are really sectoral imbalances which can be corrected by price adjustments alone. Once this fallacy is removed, depressions or 'general gluts' (and general booms) are possible, and the contraction of credit plays a key role in them. But credit which is not backed by existing goods is also an essential feature of an expanding economy as well, as Schumpeter explains more clearly than either Minsky or Marx.

Schumpeter focused upon the role of entrepreneurs in capitalism, and made the point that an entrepreneur is someone with an idea but not necessarily the finance needed to put that idea into motion.10 The entrepreneur therefore must borrow money to be able to purchase the goods and labor needed to turn his idea into a final product. This money, borrowed from a bank, adds to the demand for existing goods and services generated by the sale of those existing goods and services.

The fundamental notion that the essence of economic development consists in a different employment of existing services of labor and land leads us to the statement that the carrying out of new combinations takes place through the withdrawal of services of labor and land from their previous employments [...] this again leads us to two heresies: first to the heresy that money, and then to the second heresy that also other means of payment, perform an essential function, hence that processes in terms of means of payment are not merely reflexes of processes in terms of goods. In every possible strain, with rare unanimity, even with impatience and moral and intellectual indignation, a very long line of theorists have a.s.sured us of the opposite [...]

From this it follows, therefore, that in real life total credit must be greater than it could be if there were only fully covered credit. The credit structure projects not only beyond the existing gold basis, but also beyond the existing commodity basis. (Schumpeter 1934: 95, 101; emphasis added) This Marx-Schumpeter-Minsky perspective thus integrates production, exchange and credit as holistic aspects of a capitalist economy, and therefore as essential elements of any theory of capitalism. Neocla.s.sical economics, in contrast, can only a.n.a.lyze an exchange or simple commodity production economy in which money is simply a means to make barter easier.

Say's Principle, which insists that the sum of all notional excess demands is zero, is a model of a capitalist economy without production and, most importantly, without capitalists.

Walrasian rejoinders?

There are a number of objections which economists could make to this Marx-Schumpeter-Minsky model of a monetary production economy.

First, Marx's circuits clearly cover not one market, but two: one when the capitalist buys his inputs, the other when he sells his outputs. Since these are two distinct markets in time, there is no reason, even under Walras's Law, why demands in one should equal supplies in the other. However, in each market, Walras's Law will apply.

Secondly, it is incorrect to conflate the exchange process with the production process. It is quite possible that agents could purchase inputs to production in one market, then combine them in production, produce a larger value of commodities and subsequently bring those commodities to sale at a subsequent market.

Thirdly, Marx's notion of a surplus implies that there are some commodities which can be purchased and, through production, turned into a larger value of commodities. This implies a 'free lunch.' If such a possibility ever existed, it would have long ago been 'arbitraged' away by the price of these commodities rising, or the price of the outputs falling.

Fourthly, Marx neglects the concept of a rate of time discount. Though some agents may appear to want to acc.u.mulate over time, if we discount future incomes to reflect the fact that the commodities that income will enable you to buy will be consumed in the future, then overall these agents are simply maintaining their level of satisfaction over time.

Taking the first and second hypothetical objections together, one of the strengths of Marx's approach is that his model covers a process through time, rather than merely considering an instant in time. In reality, at the aggregate level, exchange and production occur simultaneously. Factories are continuously producing commodities, sales rooms continually moving recently produced stock, workers are being paid wages, and spending them on consumer goods. Marx's circuits a.n.a.lysis captures the organic nature of the production and exchange processes of a market economy, whereas the neocla.s.sical approach artificially separates these into distinct stages.

This organic approach therefore enables Marx to consider the economy as a dynamic process, in which growth is an integral aspect of a capitalist economy. As part of this process, there are some agents who are continually acc.u.mulating wealth (when economic conditions are favorable), and others who are continually simply maintaining their level of economic well-being (though they can also gain in wealth if real wages are increasing, and if the wage exceeds subsistence).

Walras's Law, on the other hand, is best suited to the economic irrelevance of an exchange-only economy, or a production economy in which growth does not occur (which Marx called simple commodity production). If production and growth do occur, then they take place outside the market, when ironically the market is the main intellectual focus of neocla.s.sical economics. Conventional economics is thus a theory which suits a static economy and which can only be adapted to a dynamic economy with great difficulty, if at all when what is needed are theories to a.n.a.lyze dynamic economies. Marx's 'through time' model of circuits is thus better suited to the a.n.a.lysis of a market economy than the 'moment in time' model of Walras's Law.

Marx's model of capitalist expectations is also far more valid than Walras's. A capitalist might well have his purchases and supplies balanced in any one market, as Walras's Law requires. However, purchases in this period are undertaken with the intention of selling a greater quant.i.ty in the next market. Marx's notion of the circuit of capital thus provides a link between one 'market period' and the next, which Walras's Law does not.

Since Say's Law and Walras's Law are in fact founded upon the hypothesized state of mind of each market partic.i.p.ant at one instant in time, and since at any instant in time we can presume that a capitalist will desire to acc.u.mulate, then the very starting point of Say's/Walras's Law is invalid. In a capitalist economy, the sum of the intended excess demands at any one point in time will be negative, not zero. Marx's circuit thus more accurately states the intention of capitalists by its focus on the growth in wealth over time, than does Walras's Law's dynamically irrelevant and factually incorrect instantaneous static snapshot.

The arbitrage argument highlights the difference between neocla.s.sical theory and Marx's theory of value,11 which I discuss in more detail in Chapter 17 (where, I had better point out, I reject the 'labor theory of value' which conventional Marxists regard as integral to Marx's a.n.a.lysis). Neocla.s.sical theory basically argues that the average rate of profit is driven down to the marginal productivity of capital, so that profit simply reflects the contribution which capital makes to output. This rate of profit is then called 'normal profit,' treated as a cost of production, and notionally set as the zero mark. Only profit above this level, called super-normal profit, is formally acknowledged in the theory, and in the pervasive theory of perfectly compet.i.tive equilibrium, super-normal profit is zero, so that profit fails to appear as a variable in economic theory.

The notion that profit is determined by the marginal product of capital was debunked in Chapter 7. Marx's theory of value, on the other hand, sees profit as a surplus of sales over the cost of production, allows for a positive rate of profit, and makes the rate of profit an integral part of the theory of production and exchange.

The time discount argument, that people are simply maintaining their level of satisfaction over time, has problems on at least two fronts. First, it is very hard to believe, for example, that Warren Buffett would feel that his level of wealth in 2011 was equivalent to his wealth in 1970. Successful capitalists would clearly feel that they have gained in wealth over time and unsuccessful capitalists would definitely know that they have lost or at least failed to gain. Secondly, all this argument does is move the zero position when calculating whether someone is acc.u.mulating, staying the same, or losing out. If the normal rate of time discount is, say, 2 percent, then anyone who is acc.u.mulating wealth at more than 2 percent per annum is increasing their wealth the sum of their time-discounted excess demands is negative.

So what?

The Walrasian argument that the sum of all excess demands is zero provides an apparent center of gravity for the economy. Rather like a seesaw, if one sector of the economy is down, then another sector is necessarily up. Furthermore, economists postulate that there are countervailing forces at work: a 'down' sector will have the price of its output driven down, thus increasing demand and restoring balance, and vice versa for an 'up' sector. The seesaw will ultimately return to balance.

A negative sum for aggregate excess demand and the requirement that this be made up for by borrowing money from banks moves that center of gravity. Instead of the economy behaving like a seesaw where the pivot is carefully placed at the center of gravity, it behaves like one where the pivot is instead off-center, and can move abruptly one way or the other. A down sector is not necessarily offset by an up sector, so that, contrary to Walras's Law, the entire economy can remain down or up for an indefinite period.

In particular, a general slump is feasible. As Keynes argued, a decline in spending on consumption by consumers could lead investors to also reduce their demand for investment goods, so that the economy could remain in a situation of inadequate excess demand.

The key destabilizing force is investment. As both Keynes and Marx emphasize, investment is undertaken not for its own sake, but to yield a profit. If expectations of profit evaporate, then so too will investment spending, and the economy will be thrown into a general slump. Equally, if expectations of profit become too euphoric, investment can be overdone, and the economy can be thrown into an unsustainable boom in that the profits expected by the investors will not be realized, and the boom will give way to bust. The non-Say's/Walras's Law vision of the economy shared by Marx, Schumpeter, Keynes and Minsky thus accords with the manifest instability of the macroeconomy, whereas Walras's Law a.s.serts that, despite appearances to the contrary, the macroeconomy really is stable.

At the same time, this potential for instability is also a necessary aspect of the potential for growth. Instability, in and of itself, is not a bad thing, but in fact is fundamental to any dynamic, growing system. To extend the seesaw a.n.a.logy, the fact that the real-world economic seesaw is not in equilibrium means that the only way to stop it tipping over is to keep the seesaw itself moving in the direction of its imbalance.12 The neocla.s.sical obsession with equilibrium is therefore a hindrance to understanding the forces that enable the economy to grow, when growth has always been a fundamental aspect of capitalism.13 Unfortunately, this perspective on Keynes's General Theory was buried beneath economists' mistaken belief that Walras's Law was incontrovertible. By forging a reconciliation between Keynes and Walras, the resulting 'Keynesian economics' was not Keynes's economics at all. It is little wonder that this Keynesian 'straw man' was so easily deconstructed by its conservative critics.

Say no more! Though Keynes unintentionally obscured Marx's critique of Say's Law, he also provided an eloquent explanation of why this shallow, simplistic notion held, and continues to hold, such a strong grip upon the minds of economists: That it reached conclusions quite different from what the ordinary uninstructed person would expect, added, I suppose, to its intellectual prestige. That its teaching, translated into practice, was austere and often unpalatable, lent it virtue. That it was adapted to carry a vast and consistent logical superstructure, gave it beauty. That it could explain much social injustice and apparent cruelty as an inevitable incident in the scheme of progress, and the attempt to change such things as likely on the whole to do more harm than good, commended it to authority. That it afforded a measure of justification to the free activities of the individual capitalist, attracted to it the support of the dominant social force behind authority. (Keynes 1936) However, Keynes continued, this contrariness had, by the time of the Great Depression, led to a diminution of economics in the eyes of the public: But although the doctrine itself has remained unquestioned by orthodox economists up to a late date, its signal failure for purposes of scientific prediction has greatly impaired, in the course of time, the prestige of its pract.i.tioners. For professional economists, after Malthus, were apparently unmoved by the lack of correspondence between the results of their theory and the facts of observation a discrepancy which the ordinary man has not failed to observe, with the result of his growing unwillingness to accord to economists that measure of respect which he gives to other groups of scientists whose theoretical results are confirmed by observation when they are applied to the facts. (Ibid.) Despite Marx's and Keynes's critiques, Say's Law and Walras's Law lived on, and still dominate economic thinking today. The attempts by well-meaning economists like Clower and Leijonhufvud to reconcile Keynes with Walras's Law thus robbed Keynes of a vital component of his argument, making 'Keynesian economics' a severely emasculated version of Keynes's thought. But this was far from the only way in which Keynesian economics became a travesty of Keynes's original vision.

Hamlet without the prince.

Rather as the Bible is for many Christians, the General Theory is the essential economics reference which few economists have ever read including the vast majority of those who call themselves Keynesian economists.

There are many reasons for this.

One is that the General Theory is a difficult book. There are at least two roots to this difficulty. The good root is that Keynes was so much more insightful than most other economists that the concepts in the General Theory are difficult for more ordinary mortals to grasp; the bad root is that, as Keynes himself acknowledged, the book was replete with concepts from the very school of economics which he was hoping to overthrow. As cited previously, in the Preface Keynes observed that The composition of this book has been for the author a long struggle of escape, and so must the reading of it be for most readers if the author's a.s.sault upon them is to be successful a struggle of escape from habitual modes of thought and expression. The ideas which are here expressed so laboriously are extremely simple and should be obvious. The difficulty lies, not in the new ideas, but in escaping from the old ones, which ramify, for those brought up as most of us have been, into every corner of our minds. (Ibid.: xxiii) The second and most important reason is this transcendental truth: neocla.s.sical economists don't believe that macroeconomics should exist. The att.i.tude of strong reductionism is so strong in neocla.s.sical economics that the very existence of macroeconomics as an independent field of research within economics was an affront to them. Neocla.s.sical economists could read the General Theory and find it incomprehensible, because the concepts they expect utility-maximizing consumers, profit-maximizing producers, equilibrium and so on are not the foundations of Keynes's thought.14 A final reason for not reading it is laziness: it is much easier to read the 'Reader's Digest' version given in a textbook than it is to slog through the unabridged original. As a result, many economists were inclined to rely upon summaries, rather than reading the original. Keynes obliged by providing his own summary, of just fifteen pages, in 1937.

Keynes and uncertainty The key concept in Keynes's summary was the impact of expectations upon investment, when those expectations were about what might happen in an uncertain future.

Investment is undertaken to augment wealth, and yet the outcome of any investment depends upon economic circ.u.mstances in the relatively distant future. Since the future cannot be known, investment is necessarily undertaken on the basis of expectations formed under uncertainty. Keynes was at pains to distinguish the concept of uncertainty from the simpler concept of risk.

Risk occurs when some future event can only be one of a number of already known alternatives, and when there is a known history of previous outcomes which enables us to a.s.sign a reliable and definite probability to each possible outcome. A dice roll is an example of risk. The dice can land only on one of six sides, and therefore only one of six numbers will turn up. If they are fair dice, each number has a 1 in 6 chance of turning up. The theory of probability can then be used to help predict the chances of various patterns of numbers occurring in future rolls of the dice.

Uncertainty is fundamentally different, and it has proved to be a difficult concept for economists before and after Keynes to grasp. Keynes gave several examples. Neither roulette, nor life expectancy, nor even the weather qualified. Instead, uncertainty referred to such things as the chance that war might break out (this was in 1937, not long before Chamberlain's 'peace in our time' deal with Hitler), the rate of interest twenty years in the future, or when some invention would become obsolete. I gave a more positive and I hope evocative example of uncertainty in Chapter 8.

Probability theory cannot be used to help guide us in these circ.u.mstances because there is no prior history to go on, and because the outcomes are not constrained to any known finite set of possibilities. As Keynes put it, 'About these matters there is no scientific basis on which to form any calculable probability whatever. We simply do not know' (Keynes 1937: 214).

Faced with this uncertainty, and yet compelled to act in spite of it, we develop conventions to help us cope. In marked contrast to his clumsy critique of Say's Law in the General Theory, Keynes, in explaining these conventions, was at his eloquent best: How do we manage in such circ.u.mstances to behave in a manner which saves our faces as rational, economic men? We have devised for the purpose a variety of techniques, of which much the most important are the three following: 1 We a.s.sume that the present is a much more serviceable guide to the future than a candid examination of past experience would show it to have been hitherto. In other words we largely ignore the prospect of future changes about the actual character of which we know nothing.

2 We a.s.sume that the existing state of opinion as expressed in prices and the character of existing output is based on a correct summing up of future prospects, so that we can accept it as such unless and until something new and relevant comes into the picture.

3 Knowing that our own individual judgment is worthless, we endeavor to fall back on the judgment of the rest of the world which is perhaps better informed. That is, we endeavor to conform with the behavior of the majority or the average. The psychology of a society of individuals each of whom is endeavoring to copy the others leads to what we may strictly term a conventional judgment. (Ibid.: 214) Keynes notes that expectations formed in this manner are certain to be disappointed, but there is no other way in which to form them. Expectations are therefore bound to be fragile, since future circ.u.mstances almost inevitably turn out to be different from what we expected. This volatility in expectations will mean sudden shifts in investor (and speculator) sentiment, which will suddenly change the values placed on a.s.sets, to the detriment of anyone whose a.s.sets are held in non-liquid form.

As a consequence, money plays an essential role in a market economy because of its instant liquidity. The extent to which we desire to hold our wealth in the form of non-income-earning money, rather than income-earning but illiquid a.s.sets, 'is a barometer of the degree of our distrust of our own calculations and conventions concerning the future' (ibid.: 216).

This 'liquidity preference,' Keynes argued, determines the rate of interest: the less we trust our fragile expectations of the future, the higher the rate of interest has to be to entice us to sacrifice unprofitable but safe cash for potentially profitable but volatile a.s.sets.

In a.s.sets themselves investors face two broad alternatives: lending money at the prevailing rate of interest (effectively purchasing bonds), or buying shares which confer part-ownership of capital a.s.sets. Both these activities are effectively 'placement,' as Blatt (1983) put it, rather than investment proper, however, which is the building of new capital a.s.sets (Boyd and Blatt 1988).15 New capital a.s.sets are produced not for their own sake, but in expectation of profits, and profits will come in the form of capital gain if their market prices (the result of placement activity) exceed their costs of construction. Physical investment is, therefore, also extremely volatile because 'it depends on two sets of judgments about the future, neither of which rests on an adequate or secure foundation on the propensity to h.o.a.rd [the flip side of liquidity preference] and on opinions of the future yield of capital-a.s.sets' (Keynes 1937: 218). These two factors, which play a key role in determining how much investment takes place, are likely to feed upon and destabilize each other: if we become more pessimistic about the future prospects of investments, we are likely to want to h.o.a.rd more, not less.

Having explained why expectations are so important in economic practice and economic theory, why uncertainty makes expectations so fragile and volatile, and how these factors affect the rate of interest and the level of investment, Keynes returned once more to an attack on Say's Law. He divided expenditure into consumption which is relatively stable and investment which is highly volatile and emphasized that investment is the key determinant of the level and rate of change of output (and hence employment). His theory was, therefore, a theory 'of why output and employment are so liable to fluctuation' (ibid.: 221). In contrast to the unintelligible summary in the General Theory itself, Keynes gave a relatively pithy summary in which expectations, investment and uncertainty had pivotal roles: The theory can be summed up by saying that, given the psychology of the public, the level of output and employment as a whole depends on the amount of investment. I put it in this way, not because this is the only factor on which aggregate output depends, but because it is usual in a complex system to regard as the causa causans that factor which is most p.r.o.ne to sudden and wide fluctuation.

More comprehensively, aggregate output depends on the propensity to h.o.a.rd, on the policy of the monetary authority as it affects the quant.i.ty of money, on the state of confidence concerning the prospective yield of capital-a.s.sets, on the propensity to spend and on the social factors which influence the level of the money-wage. But of these several factors it is those which determine the rate of investment which are most unreliable, since it is they which are influenced by our views of the future about which we know so little. (Ibid.: 221; emphasis added) Keynes peppered this paper with observations about how conventional economics ignored the issue of uncertainty, and how expectations are formed under uncertainty, by simply a.s.suming the problem away: 'I accuse the cla.s.sical economic theory of being itself one of these pretty, polite techniques which tries to deal with the present by abstracting from the fact that we know very little about the future' (ibid.: 215).

Finally, in a departure from the General Theory of just a year earlier, Keynes criticized the concept of the 'marginal efficiency of capital' the ratio of the yield of newly produced capital a.s.sets to their price. Whereas he used this concept extensively in the General Theory, here he argued that it is indeterminate, since the price of capital a.s.sets is so volatile, and there will be a different 'marginal efficiency of capital' for every different level of a.s.set prices. Rather than being a determinant of investment, the 'marginal efficiency of capital' might simply be a by-product of the level of investment and current expectations.

These are all difficult concepts, especially for economists who were bred in the neocla.s.sical tradition in which 'at any given time facts and expectations were a.s.sumed to be given in a definite and calculable form; and risks, of which, though admitted, not much notice was taken, were supposed to be capable of an exact actuarial computation' (ibid.: 213).

But if Keynes had truly unleashed a revolution in economic thought, then economists should have attempted to get their minds around these difficult concepts, and fought to escape from the 'habitual modes of thought and expression' which had gripped them prior to the calamity of the Great Depression.

Did economists do so? Some did, but the majority did not and for that reason the profession bifurcated into two camps: a minority which swore fealty to Keynes's revolutionary vision (who generally call themselves 'post-Keynesian'), and a majority which paid lip-service to some of Keynes's words, but which rapidly fell back into old, familiar ways. These economists ignored Keynes's General Theory, and even his pithy summary, instead clutching at another alleged summary by one J. R. Hicks.

Slimming Keynes down to size: the IS-LM model Hicks's 'Mr. Keynes and the Cla.s.sics' purported to be a book review of the General Theory. Hicks began by disputing that neocla.s.sical economists held quite the views Keynes alleged that they held, and therefore tried to construct a more typical cla.s.sical theory 'in a form similar to that in which Mr. Keynes sets out his own theory [...] Thus I a.s.sume that I am dealing with a short period in which the quant.i.ty of physical equipment of all kinds available can be taken as fixed' (Hicks 1937: 148).

Was this really the manner in which Keynes set out his own theory? Not according to Keynes, who criticized 'the cla.s.sics' (by which he meant what we today call neocla.s.sical economists) for working with a model in which 'the amount of factors employed was given' (Keynes 1937: 212). Nonetheless, Hicks continued. He summarized the 'typical Cla.s.sical theory' in three equations, which argued that: the amount of money determined total output (output was some constant times the money stock); the rate of interest determined the level of investment; and the rate of interest determined the level of savings (and savings equaled investment).16 The first equation determines total output and total employment,17 while the second two simply determine how much of output is devoted to investment, and how much to current consumption. If the savings rate increases, then so does investment. Increasing money wages 'will necessarily diminish employment and raise real wages' (Hicks 1937), while the obverse policy cutting money wages will necessarily increase employment and reduce the real wage. Decreasing the money supply directly decreases income and employment, and is the main explanation for economic downturns (an argument which Milton Friedman later revived).

Clearly, Keynes's theory was substantially different from this. But how did Hicks summarize Keynes? In three more equations, where: the demand for money depends upon the rate of interest (in place of the 'cla.s.sical' fixed relationship between money and output); investment is a function of the rate of interest; and savings is a function of income.

h.e.l.lo? What happened to uncertainty, expectations, liquidity preference determining the rate of interest, speculative capital a.s.set prices, and so on? They are nowhere to be seen. Sometime later, Hyman Minsky commented that 'Keynes without uncertainty is rather like Hamlet without the Prince' (Minsky 1975: 75), but this is what Hicks served up as Keynes. Even the Reader's Digest would draw the line at this level of abridging, but this was not the end of Hicks's rephrasing of Keynes's Shakespearean sonnets into schoolyard doggerel.

He next argued that 'Keynes's' first equation omitted the impact of income on demand for money. This was the traditional 'transactions demand for money' argument that some level of money was needed to finance everyday transactions, so that an increase in income would generate an increase in the demand for money. To be truly general, said Hicks, the 'general theory' should include the impact of income on the demand for money, as well as the impact of the rate of interest.

Keynes had omitted discussion of the transactions demand for money because this demand was relatively stable, and therefore less important than the more important demand set by liquidity preference. But Hicks believed that 'however much stress we lay upon the "speculative motive," the "transactions motive" must always come in as well' (Hicks 1937: 153). So he proposed a revised set of equations in which the demand for money depends upon two variables the rate of interest and the level of income though not, as Keynes had it, on 'the degree of our distrust of our own calculations and conventions concerning the future' (Keynes 1937: 216).

With this revision, Keynes, who was at such pains to distinguish himself from his predecessors primarily though not exclusively on the basis of the importance he attached to uncertainty and expectations is pushed back into the camp from which he desired to escape. As Hicks put it: 'With this revision, Mr. Keynes takes a big step back to Marshallian orthodoxy, and his theory becomes hard to distinguish from the revised and qualified Marshallian theories, which, as we have seen, are not new. Is there really any difference between them, or is the whole thing a sham fight? Let us have recourse to a diagram' (Hicks 1937: 153).

Hicks's diagram explains why his rendition of Keynes was so readily accepted by economists, while Keynes's own summary was ignored (see Figure 10.1). It was the old familiar totem of two intersecting curves, though now relabeled to reflect its somewhat different derivation: in place of 'S' and 'D' for supply and demand, we now had 'IS' and 'LM.' The 'Totem of the Micro,' as Leijonhufvud satirized the supply and demand diagrams of Marshallian microeconomics, now had a bigger sibling for macroeconomics though it was not derived in a way that microeconomists would accept, nor did it reach conclusions about the macro economy with which they would agree, as we shall see later.

The downward-sloping curve, the equivalent of the microeconomic demand curve, was derived from the investment and savings relations in Hicks's model. The upward-sloping curve, the equivalent of the microeconomic supply curve, was derived from the money demand relation (on the a.s.sumption that the money supply was controlled by the monetary authorities, and was therefore determined outside the model).

10.1 Hicks's model of keynes.

The IS curve showed all those combinations of the rate of interest (i) and the level of income (I) which yielded equilibrium in the goods market. The LL curve (which economists today call the LM curve) showed all those combinations of the rate of interest and the level of income which gave equilibrium in the money market.

10.2 Derivation of the downward-sloping IS curve.

10.3 Derivation of the upward-sloping LM curve.

Here, at last, in comparison to the strange concepts of Keynes, economists were back on familiar ground. As Hicks put it: Income and the rate of interest are now determined together at P, the point of intersection of the curves LL and IS. They are determined together; just as price and output are determined together in the modern theory of demand and supply. Indeed, Mr. Keynes's innovation is closely parallel, in this respect, to the innovation of the marginalists. (Ibid.) One problem with this 'general theory,' however, was that many of Keynes's conclusions could not be derived from it something which would not have surprised Keynes a great deal, since this model omitted his key concepts of uncertainty and expectations. But Hicks had an apparent dilemma: But if this is the real 'General Theory,' how does Mr. Keynes come to make his remarks about an increase in the inducement to invest not raising the rate of interest? It would appear from our diagram that a rise in the marginal-efficiency-of-capital schedule must raise the curve IS; and, therefore, although it will raise income and employment, it will also raise the rate of interest. (Ibid.: 154) To Keynes, the reason why an increased desire to invest would not necessarily raise the rate of interest is because the latter was determined by liquidity preference, which 'is a barometer of the degree of our distrust of our own calculations and conventions concerning the future.' In a depressed economy, an increase in investment could well reduce the 'degree of distrust,' leading to a fall in the rate of interest rather than a rise. But with Hicks's picture of Keynes shorn of uncertainty, conventions and expectations, there were no such mechanisms to draw upon. Fortunately, Hicks's model provided a simple and far more conventional solution: simply bend the curves: This brings us to what, from many points of view, is the most important thing in Mr. Keynes's book. It is not only possible to show that a given supply of money determines a certain relation between Income and interest (which we have expressed by the curve LL); it is also possible to say something about the shape of the curve. It will probably tend to be nearly horizontal on the left, and nearly vertical on the right. This is because there is (1) some minimum below which the rate of interest is unlikely to go, and (though Mr. Keynes does not stress this) there is (2) a maximum to the level of income which can possibly be financed with a given amount of money. If we like we can think of the curve as approaching these limits asymptotically. (Ibid.) This 'liquidity trap' enabled Hicks to provide an explanation for the Great Depression, and simultaneously reconcile Keynes with 'the Cla.s.sics.' Keynes was consigned to one end of the LM curve, where the liquidity trap applied, and 'the Cla.s.sics' to the other, where full employment was the rule (see Figure 3.1). In the 'cla.s.sical' range of the LM curve, conventional economics reigned supreme: there was a maximal, full employment level of income, where any attempts to increase output would simply cause a rising interest rate (or inflation, in extensions of the IS-LM model). In the 'Keynesian' region, monetary policy (which moved the LM curve) was ineffective, because the LM curve was effectively horizontal, but fiscal policy (which moved the IS curve) could generate greater output and hence employment without increasing interest rates. A higher level of government expenditure could shift the IS curve to the right, thus moving the point of intersection of the IS and LM curves to the right and raising the equilibrium level of output.

10.4 'Reconciling' Keynes with 'the Cla.s.sics'

Hicks put the position pithily. In the 'Keynesian region' of his model, a depression can ensue because traditional monetary policy is ineffective but Keynes's prescription of fiscal policy can save the day: 'So the General Theory of Employment is the economics of Depression' (ibid.: 155).

Hicks next proposed that, for reasons of mathematical elegance rather than economic relevance, all three variables (demand for money, investment and savings) should be made functions of both income and the rate of interest (though not uncertainty or expectations): In order to elucidate the relation between Mr. Keynes and the 'Cla.s.sics,' we have invented a little apparatus. It does not appear that we have exhausted the uses of that apparatus, so let us conclude by giving it a little run on its own.

With that apparatus at our disposal, we are no longer obliged to make certain simplifications which Mr. Keynes makes in his exposition. We can reinsert the missing i in the third equation, and allow for any possible effect of the rate of interest upon saving; and, what is much more important, we can call in question the sole dependence of investment upon the rate of interest, which looks rather suspicious in the second equation. Mathematical elegance would suggest that we ought to have I and i in all three equations, if the theory is to be really General. (Ibid.: 156) Economists, having been threatened by Keynes with the need to completely retrain themselves, could now engage in their favorite game of tobogganing up and down one curve, moving another to the left or right, just as they did in microeconomics. It is little wonder that this Hicksian IS-LM model was adopted as the basis for 'Keynesian' economics, and equally little wonder that, many years later, macroeconomics was converted to a subset of microeconomics.

The true origins of IS-LM Though 'cutting-edge' economic a.n.a.lysis has left Hicks's model behind, most macroeconomists still think in IS-LM terms, and this model is still the common fodder served up to undergraduate students as Keynesian economics. It therefore still has pedagogic and disciplinary relevance. So the question arises: from where did this model emanate? It clearly was not derived from Keynes's General Theory, apart from the adoption of some of Keynes's terminology. The mystery of its origins was finally solved by one Sir John Hicks an older, be-nighted and somewhat wiser J. R. Hicks.

ISLM: an apology Hicks's detective work was published in a paper ent.i.tled 'IS-LM: an explanation,'18 but in many ways it was an apology. Published in the non-orthodox Journal of Post Keynesian Economics in 1980, the paper's opening sentence was: 'The IS-LM diagram, which is widely, though not universally, accepted as a convenient synopsis of Keynesian theory, is a thing for which I cannot deny that I have some responsibility' (Hicks 1981: 141).

Even after this rueful opening, Hicks clung to a very Walrasian vision of Keynes, and elsewhere he described the IS-LM diagram as 'a product of my Walrasianism' (Hicks 1979: 990). But he conceded that his rendition had erroneously omitted any discussion of uncertainty or expectations. His explanation as to how he could have missed so fundamental an aspect of Keynes's thought was that, shortly before the General Theory was published, he had published a paper which, he believed, had strong similarities to Keynes's argument (Hicks 1935).19 What he then published as a review of Keynes was actually a restatement of his own model, using some of Keynes's terminology.

Hicks saw two key problems in cross-dressing as Keynes. The first was that his model was 'a flexprice model [...] while in Keynes's the level of money wages (at least) was exogenously determined' (Hicks 1981: 141);20 the second 'more fundamental' problem was that Hicks's model used a period of a single week, while Keynes used 'a "short-period," a term with connotations derived from Marshall; we shall not go far wrong if we think of it as a year' (Hicks 1980).

Discussing the second problem, Hicks argued that the difference in period length had a drastic impact upon the relevance of expectations. With a time period of just a week, it is not unreasonable to keep expectations constant and therefore to ignore them. But keeping expectations constant over a year in an IS-LM model does not make sense, because 'for the purpose of generating an LM curve, which is to represent liquidity preference, it will not do without amendment. For there is no sense in liquidity, unless expectations are uncertain' (Hicks 1981: 152).

This was precisely the point Keynes himself made, in ironic form, in 1937: Money [...] is a store of wealth. So we are told, without a smile on the face. But in the world of the cla.s.sical economy, what an insane use to which to put it! For it is a recognized characteristic of money as a store of wealth that it is barren; whereas practically every other form of storing wealth yields some interest or profit. Why should anyone outside a lunatic asylum wish to use money as a store of wealth?

Because, partly on reasonable and partly on instinctive grounds, our desire to hold money as a store of wealth is a barometer of the degree of our distrust of our own calculations and conventions concerning the future [...] The possession of actual money lulls our disquietude; and the premium which we require to make us part with money is the measure of the degree of our disquietude. (Keynes 1937: 21516) Thus, without uncertain expectations, there is no sense in liquidity preference, and Hicks cannot justify the LM half of his IS-LM model. But with uncertain expectations, there is no sense in equilibrium a.n.a.lysis either, since equilibrium can be maintained only if expectations are continually being fulfilled. Hicks concluded that the equilibrium/constant expectations framework of the IS-LM model was theoretically unsound, and practically irrelevant to the problems of the macroeconomy: I accordingly conclude that the only way in which IS-LM a.n.a.lysis usefully survives as anything more than a cla.s.sroom gadget, to be superseded, later on, by something better is in application to a particular kind of causal a.n.a.lysis, where the use of equilibrium methods, even a drastic use of equilibrium methods, is not inappropriate [...]

When one turns to questions of policy, looking towards the future instead of the past, the use of equilibrium methods is still more suspect. For one cannot prescribe policy without considering at least the possibility that policy may be changed. There can be no change of policy if everything is to go on as expected if the economy is to remain in what (however approximately) may be regarded as its existing equilibrium. (Hicks 1981) There is one more, crucial weakness in Hicks's model that he touched upon but did not consider properly, and which would invalidate his model even if an LM curve could be derived: his use of 'Walras's Law' to reduce the number of markets in the model from three to two: 'Keynes had three elements in his theory: the marginal efficiency of capital, the consumption function, and liquidity preference. The market for goods, the market for bonds, and the market for money [...]' (ibid.: 142).

He then explained that he dropped the second of these markets by applying Walras's Law: 'One did not have to bother about the market for "loanable funds," since it appeared, on the Walras a.n.a.logy, that if these two "markets" were in equilibrium, the third must be also. So I concluded that the intersection of IS and LM determined the equilibrium of the system as a whole' (ibid.: 142).

Next he noted that there was in fact one other market that should be part of the model: the labor market which was, of course, an integral part of Keynes's a.n.a.lysis in the General Theory itself: 'In strictness, we now need four markets, since labor and goods will have to be distinguished [...]' (ibid.: 1423).

He went on to argue that its omission was justified, but here his neocla.s.sical fixation with equilibrium a.n.a.lysis led him astray, because ignoring for the moment that Walras's Law is false in a capitalist economy Walras's Law allows you to drop one market only when all other markets are in equilibrium.21 In the IS-LM model, this applies only where the two curves cross: where the combination of GDP and the interest rate is such that both the goods market (the IS curve) and the money market (the LM curve) are in equilibrium. Then, in a three-market model goods, money, and labor if the money and goods markets are in equilibrium, then so too must be the labor market.

However, if the combination of the interest rate and the GDP are such that one of these two markets is out of equilibrium, then so too must be the labor market. Therefore only in equilibrium can the labor market be ignored. At any other location, the labor market must also be considered and therefore the IS-LM model is incomplete. Everywhere except at the point of intersection of IS and LM, it needs to be the IS-LM-'LSLD' model (where 'LS' and 'LD' refer to labor supply and labor demand respectively).

Furthermore, since at anywhere except the intersection of IS and LM at least one of those two markets is in disequilibrium, the third, ignored 'LSLD' market must also be in disequilibrium: wages must be higher (or lower) than the level that will clear the labor market. Therefore price-setting in this market and the other one that is in disequilibrium must be a dynamic, disequilibrium process, not a simple calculation of the equilibrium wage. Even Hicks's emasculated version of Keynes's macroeconomics must employ dynamic, disequilibrium a.n.a.lysis, in contrast to the comparative static mode in which the IS-LM model is normally applied.

This IS-LM model is thus invalid, even on its own terms, if it is pushed anywhere beyond working out what rate of interest and GDP combination represent equilibrium in the economy. To be used as a model of economic dynamics, it must become a three-equation model, and these must all be disequilibrium equations. This is not how IS-LM is taught, or used.22 But in its heyday, the IS-LM model gave economists something they had never really had previously: a framework on which to build models that were not merely drawings, or symbolic equations, but numerical equations that they could use to predict the future course of the economy.

The age of large-scale econometric models.

Hicks's model and the later development of the 'Aggregate Supply-Aggregate Demand' model set off the heyday of attempts by economists to turn these models into numerical simulations of the economy, using the newly developed tool of the computer.

With a careful choice of parameter values, these models could generate a reasonable fit between the inputs ('exogenous variables') and the variables like future output and employment levels ('endogenous variables'). If a model's fit to the data wasn't too good, it could be improved by fine-tuning the parameters, or adding more variables, and as a result most of these models 'grew like Topsy.' One of the earliest such model, developed by Lawrence Klein (Klein 1950; Renfro 2009), had just six equations; eventually models with thousands of equations were developed and many are still in use.

There were five aspects of these models that made them easy to simulate, but which also made them fundamentally unsuited for economic a.n.a.lysis.

First, the models were frequently linear variables in the equations were multiplied by constants, and added together to produce predictions for other variables in contrast to the nonlinear models outlined in Chapter 9, so they couldn't develop interactions between variables that caused cyclical behavior, let alone complex behavior as in Lorenz's weather model.

Secondly, even when nonlinearities existed when employment was divided by population to calculate an employment rate, for example, or when logarithms of variables were used rather than the raw variables the model was solved as if these nonlinearities did not affect the system's tendency towards equilibrium (McCullough and Renfro 2000). Simulations therefore worked out what the equilibrium of the model would be, and their predictions had the economy converging to this point over time (see, for example, Renfro 2009: 46).23 Thirdly, the models effectively a.s.sumed that the economy's dynamics involved movements from one equilibrium position to another, with movement being caused by 'exogenous shocks' events external to the economy (such as damaging floods or unexpected bountiful harvests). This continued the convention in econometrics of seeing fluctuations in the economy having non-economic causes: 'The majority of the economic oscillations which we encounter [...] seem to be explained by the fact that certain exterior impulses. .h.i.t the economic system and thereby initiate more or less regular oscillations' (Frisch 1933: 171 [1]).24 Even though this argument was made during the Great Depression, where no 'external impulse' could be blamed for the crisis, and when economists like Schumpeter, Keynes and Fisher were arguing that cycles and possibly breakdowns were endemic to capitalism, this belief became the standard view in numerical simulations of the economy.25 Fourthly, they were based on a neocla.s.sical vision of the economy, and therefore omitted the credit and debt variables that we now know are crucial to macroeconomics.

Finally, they omitted any consideration of how expectations are formed under pervasive uncertainty, a key aspect of Keynes's vision of the macroeconomy that was lacking in the parent IS-LM model.

There were therefore many good grounds on which these models could have been criticized. However, the one focused on by economists was something entirely different: they objected to these numerical models simply because, as with Hicks's stylized IS-LM model from which they were derived, they argued that there could be involuntary unemployment, and that the level of unemployment could be affected by government demand-management policies conclusions which neocla.s.sical economists mistakenly believed contradicted neocla.s.sical microeconomics.

From IS-LM to the representative agent.