An Introduction to Post Keynesian Economics: Involuntary Unemployment with Perfectly Flexible Wages and Prices1
Abstract Post Keynesian economics is based on the economics of John Maynard Keynes. Unlike Keynesianism, it does not rely on rigidities or imperfections to explain less-than-full employment, a condition that they argue is the rule rather than the exception. One key implication of this is that Post Keynesians do not view economics as being the science of scarcity. In the real world, our most difficult challenge is related to the pervasive and chronic abundance of one particular resource: labor. The consequences of this abundance are significant, widespread, and avoidable. This article explains how they come to such a conclusion.
Keywords: Keynes, Post Keynesian, unemployment, General Theory, heterodox JEL Codes: E12, E24, E32,
John T. Harvey Professor of Economics TCU Box 298510 Texas Christian University Fort Worth, TX 76129 817-257-7230 [email protected]
FINAL VERSION PUBLISHED AS: Harvey, J.T., 2016. An Introduction to Post Keynesian Economics: Involuntary Unemployment With Perfectly Flexible Wages and Prices. The American Economist, 61(2), pp.140-156.
Post Keynesian economics is based on the economics of John Maynard Keynes. According to Post Keynesians, what has become known as “Keynesianism” is a poor substitute for what was originally argued in the General Theory (Keynes, 1936). The model developed there never depended on rigidities or imperfections to explain economic downturns, but instead described an economy that tended toward crisis and less-than-full employment even with perfectly flexible wages and prices. Thus, while Keynesians argue that we could stay in a perpetual state of expansion were it not for external shocks and policy errors, Keynes’ model saw involuntary unemployment as the rule rather than the exception and explained business cycles as inherent in the system rather than exogenous. One of the key implications of all this is that Post Keynesians do not see economics as being the science of scarcity. In the real world, our most difficult challenge is related to the pervasive and chronic abundance of one particular resource: labor. The consequences of this abundance are significant, widespread, and avoidable. The goal of this paper is to introduce the reader to the main elements of that argument. Though Post Keynesianism actually goes beyond Keynes to incorporate the unique contributions of scholars like Michael Kalecki, Piero Sraffa, Nicholas Kaldor, and Joan Robinson, this essay will focus primarily on the school of thought’s namesake. I will begin by explaining what Keynes actually said about the labor market and full employment, then discuss the importance of investment spending in achieving the latter, and following that offer an overview of Post Keynesian policy recommendations. Along the way, I will touch on the critical role played by the financial sector, the meaning and consequence of fundamental uncertainty, and the specifics of the endogenously-generated business cycle that is characteristic of market economies. The concluding comments include a brief overview of contemporary Post Keynesian activities. 2
I. Labor Market Contrary to what you will read in almost any economics textbook, Keynes’ explanation of involuntary unemployment absolutely did not depend on the existence of sticky wages. This is one of the most pervasive misconceptions in all of economics and one that, were it true, would reduce Keynes’ contribution to tacking a horizontal line onto the standard labor-market diagram. Some of the problem lies with the fact that the General Theory is undoubtedly a very difficult read (Chick, 2006). But, it is nevertheless very clear on this point: the existence of involuntary unemployment does not depend on any labor, social, or policy obstinance or “interference.” To be fair, Keynes does discuss the possibility of trade union opposition to reductions in the money wage and he agrees as an empirical matter that while workers will not object to a general fall in real wages (as created by price inflation, for example), they will resist a fall in pay relative to other workers. However, he counted any resulting unemployment as voluntary (Keynes 1936, pp. 8 and 14). Besides this, even were one to insist that those observations revealed a belief that wages may be downwardly rigid in the real world, there is absolutely no question that his theoretical analysis assumed perfectly flexible wages and prices. Keynes’ argument with respect to involuntary unemployment goes as follows (this section is based on Chick 1983, which I highly recommend to those wanting to decipher the General Theory). His labor demand curve is the locus of points showing the profit maximizing levels of employment at each real wage (i.e., where the real wage equals the marginal product of labor). This is labeled Nd on Figure 1. Meanwhile, the labor supply curve shows where the real wage equals the marginal disutility of work (Ns on Figure 1). None of this is the least bit controversial and Keynes does not take issue with either. The usual argument is that the interaction of these 3
functions yields W0/P0 and N0 as the full-employment equilibrium. This is so because at points below that real wage there is excess demand for labor. Firms are disappointed but possess a mechanism that will put them back on Nd: they can offer a higher nominal wage. They do so until the market comes to rest at W0/P0 and N0. Meanwhile, above the intersection we have an excess supply of labor and involuntary unemployment. This time it is the workers who are disappointed, but they, too, have a means of returning to the curve that represents their desired outcome: they can offer to work for less. Once again, the market clears.
Figure 1: Traditional labor-market diagram. All well and good, says Keynes, regarding the excess demand for labor. But where the conventional analysis falls short is in considering excess supply. No such adjustment toward the intersection actually takes place. The critical error is the treatment of Ns as a set of equilibrium points when, in fact, it only shows labor’s preference. Workers, or rather the involuntarily unemployed, have no means of enforcing their wishes. Hence, while the labor supply curve is an indispensable reference that shows us the maximum number of willing workers who would be 4
forthcoming at each real wage, points of actual employment lie only on the labor demand curve. The level of employment must be a point on Nd; it need not also be a point on Ns. Consider such a situation. Above the intersection, the short side of the market will prevail so that we would be on Nd somewhere to the left of Ns. Firms are perfectly happy as the real wage equals the marginal product of labor and they are therefore maximizing profits. Those lucky enough to have jobs are also happy, especially so since they are working at a real wage that lies above the marginal disutility of labor at that volume of employment. They have no reason to offer to work for less. That leaves the involuntarily unemployed as the only unhappy party, many of whom would most certainly be willing to work at less than the prevailing real wage–but no one asked them. Because they are already profit maximizing, firms are not hiring. If there are no help wanted ads, there is no one with whom the involuntarily unemployed can negotiate. Nor are they in any realistic sense in direct competition with those who do have jobs. They can remain involuntarily unemployed indefinitely. Post Keynesians view this as a much more realistic scenario. During the Great Depression, for example, there is no reason to believe that firms were not hiring the profit maximizing quantity of labor. In addition, it is also very likely that those who were employed counted themselves as extremely lucky and, while willing to work for less, they had no reason to make such a request. That leaves the millions of unemployed. Would they have been willing to work for less, particularly given that modern public social safety nets did not exist? Of course, and yet from 1930 through 1941, unemployment in the US fell below double digits only twice (the endpoints). This is perfectly consistent with Keynes’ labor market analysis. Firms and workers were content and while the involuntarily unemployed were not, they had no means of translating their wishes 5
into reality. And although it seems reasonable to believe that firms would slowly replace more expensive workers with cheaper ones as retirements and terminations took place, 1) it is also possible that they might choose not to replace open positions at all until the economy shows signs of recovery and 2) the intransigence of the unemployment rate suggests that even if this force exists, it is extremely weak. By contrast, it is difficult to explain the 1930s with the traditional Keynesian inflexiblewage story. Any forces that might have served as an impediment to wage adjustment would have most certainly have fallen away by the early 1930s, when the number of unemployed was eight times what it had been in 1929. Keynes raised this very issue in the General Theory: Moreover, the contention that the unemployment which characterises a depression is due to a refusal by labour to accept a reduction of money-wages is not clearly supported by the facts. It is not very plausible to assert that unemployment in the United States in 1932 was due either to labour obstinately refusing to accept a reduction of money-wages or to its obstinately demanding a real wage beyond what the productivity of the economic machine was capable of furnishing...Labour is not more truculent in the depression than in the boom–far from it. (Keynes, 1936, p. 9). The labor markets of the Great Depression and the Great Recession were not being prevented from reaching equilibrium, they were already there. The fact is that there is no reason to expect real wages fall when they are above the intersection. Nor does Keynes suggest that the solution is to find some means of forcing them to do so, for wages are simultaneously firms’ primary cost and their major source of demand. Lowering wages is not a guarantee of higher 6
employment when it may also lower spending. Besides which, the initial increase in unemployment was not caused by a spike in wages. A proper understanding of the labor market requires that we reframe the issue entirely by shifting our attention from price variables to income variables.
II. Reframing Involuntary Unemployment Real wages are not the key to fluctuations in employment. As Keynes wrote in the General Theory, “Wide variations are experienced in the volume of employment without any apparent change either in the minimum real demands of labour or in its productivity” (Keynes, 1936, p. 9). While it would be foolish to argue that they are irrelevant, far more significant in causing changes in the hiring decision are changes in sales forecasts. This is the premise underlying Keynes’ Z-D or aggregate supply-aggregate demand diagram shown in Figure 2.2 Just as in the traditional labor market, the horizontal axis is the quantity of labor. The vertical, however, is expected nominal sales or planned spending. The Z curve is essentially labor demand and is derived directly from the equation that defined Nd.3 Consistent with the traditional analysis, it is a set of profit maximizing points for the firm. Firms making this determination in the Keynes/Post Keynesian world, however, focus not on wages but expected nominal sales: Py. This is not to say that wages are ignored, they are just held constant and reflected in the slope of Z (which, assuming no region of increasing returns, takes the given shape and begins at the origin). If wages rise, Z pivots inward; if they fall, it pivots outward. Note, however, that given Keynes’ framework, the impact of the change in wages on demand would also need to be considered.
Figure 2: Keynes’ Z-D diagram.
The sales expectation is in money terms because that is the world in which firms live. Their contracts are in money terms, their debts are in money terms, the wages they pay are in money terms, etc. Were we doing a study over a twenty-year period, then it would be instructive to deflate values (not to mention the many other adjustments that may be necessary). But, over the short run–and the long run is simply a series of short runs–entrepreneurs are worried about earning dollars. The positive slope of Z is a function of the fact that as firms predict higher sales on the Py axis, so they demand a larger quantity of workers, N. Z becomes increasingly steep due to the diminishing marginal productivity of labor when combined with a given amount of capital. It takes a greater jump in expected sales to convince the entrepreneur to go from 1000 to 1001 workers than from 100 to 101 since in the former case the last worker is adding less to output (we are assuming the same wage for every worker). At this point, one may be tempted to guess that the Post Keynesians are going to attribute 8
demand deficiencies to coordination failures whereby pessimistic sales forecasts lead to lay offs and thereby self-fulfilling prophecies. Though the analysis certainly allows for that, the simplifying assumption that firms’ forecasts are always correct is often made so that we can get to the more fundamental problem. It will be found on the D curve. The D curve shows the planned level of spending (Py) at each level of employment (N), where the latter is measured such that an increase of x units of N always yields the same increase in income.4 It is positively sloped since as employment rises, so does total income and therefore planned spending. The slope of D reflects the fact that as N rises, consumption does as well but at a decreasing rate. This is due to Keynes’ fundamental psychological law:5 The fundamental psychological law, upon which we are entitled to depend with great confidence both a priori from our knowledge of human nature and from the detailed facts of experience, is that men are disposed, as a rule and on the average, to increase their consumption as their income increases, but not by as much as the increase in their income (Keynes, 1936, p. 96). Last, the vertical intercept of D represents autonomous planned spending. Because he is assuming a closed economy with no government, this reduces to investment spending.6 The intersection of Z and D yields the level of employment. To the right of N0 on Figure 2, the profit-maximizing level of sales required to justify those levels of employment will not, according to D, be forthcoming. To the left of N0, profits are sufficient to convince firms to expand employment. Where, one may ask, is full employment? Because it is not in any direct manner related to the process by which the economy reaches equilibrium, it is simply a reference point (Nf) on the horizontal axis. It can lie can lie anywhere from the current equilibrium level of 9
N to the right. For reasons discussed below, it is usually to the right. Note that while Nf would move in response to changes in the wage (presumably to the right as wages rise and the left when they fall), it is important to bear in mind that any changes will also affect the Z and D curves. In addition, realistically speaking, very few people can actually choose to withdraw their labor services in the event of a fall in wages. In fact, they will almost certainly be forced to try to take on extra work because so much consumer spending is related to either basic necessities or contractual payments. The point here is that wages are much more than just another price. They simultaneously affect both supply and demand and those who earn them are the raison d’être for economic activity. Wages are further complicated by the fact that they are dependent not only on traditional economic variables, but social and political factors. Economic agents actively fight for larger shares of the income pool and power is an indispensable part of that story.
III. Full employment: does saving cause investment or investment cause saving? Figure 2 makes it clear that the key to reaching full employment is the vertical intercept of D. This is especially so given that, as D gets flatter, Z gets steeper. This means that at the very moment that firms need progressively larger jumps in sales in order to justify additional hiring, the consumption expenditure forthcoming from a given level of income is declining. Just as in real life, as unemployment declines so it becomes increasingly difficult to squeeze one more job out of the economy. Figure 3 shows a Z-D diagram at full employment. Aggregate expenditures and income are Pyf, consumption expenditures are D1, and savings are D2. In a closed economy with no 10
government sector, this implies that we can reach Nf if and only if investment is large enough to absorb the saving that would occur at the full-employment level of income. There is nothing terribly controversial about this statement. However, say Keynes and the Post Keynesians, there is absolutely no reason to believe that this will occur except by coincidence. There is not, as in some macro models, an equilibrating force pushing us towards such a point. We can therefore come to rest anywhere from Nf to the left.
Figure 3: Keynes’ Z-D diagram at full employment.
The story could now go in a number of different directions since discussing investment requires explaining expectations formation, financial markets, interest rates, household saving, and the factors driving the profitability of physical capital. However, it might be most useful to start by examining something Post Keynesians actually reject: the loanable funds theory of interest. The latter, as is well known, assumes both saving and investment to be a function of the interest rate, 11
the former a positive one and the latter negative. It further assumes that household saving is the sole source of loanable funds, that firms require those funds to finance investment, and that banks pay interest on saving and earn it on the money that is borrowed. Banks, therefore, have an incentive to loan out every penny in their vaults. Supporters of loanable funds believe that the interaction of these forces moves the rate of interest to the level that causes the quantity of investment to be exactly equal to the quantity of saving and, significantly, either of these can be the driving7 force. For example, in the event of an increase in the propensity to invest, this creates an excess demand for loanable funds that leads banks to increase interest rates until sufficient saving is brought forth to finance the planned projects. If the propensity to save increases, banks find themselves with unborrowed funds in their vaults. They address this by lowering interest rates until investment rises sufficiently to absorb total saving. Note that this modeling of the financial sector leads to the happy conclusion that we have nothing to fear with respect to investment being sufficient to offset the full-employment level of savings. Returning to Figure 3, if households wish to save D2, then interest rates will adjust until firms borrow however much households opt not to spend–problem solved. Hence, even if we allow for the Keynes/Post Keynesian characterization of the determination of employment up to this point, tacking on the loanable funds theory of interest brings us right back to a world in which there is an automatic tendency toward full employment. Post Keynesians, however, raise two objections to the loanable funds theory. First, saving, and therefore consumption, is not a function of interest rates, but of income. In a model like loanable funds where there is a systemic tendency toward full employment and, therefore, the full12
employment level of income, the latter is not going to be a major factor. We know how much income will be, the only question is how it will be split between consumption and saving. Interest rates determine this. In Keynes, however, wide fluctuations in the level of economic activity are expected and thus play a significant role in spending decisions. Furthermore, Keynes assumes that agents operate in a world of uncertainty (more on this below), which not only creates an incentive to save even without the offer of interest, but it leads people to be willing to hold the most barren of all assets: cash. In the Post Keynesian specification, interest is not the reward for abstaining from consumption, but for parting with liquidity. But this alone is insufficient to break the loanable funds theory, for it only means that households do not respond to changes in interest rates. So long as firms still do and household saving continues to represent the sole source of loanable funds, the outcome is the same: interest rates will adjust until any level of saving can be absorbed, meaning that there exists no obstacle to full employment. Post Keynesians object to another premise of the theory, however, and that is that savings represent loanable funds. Such an assumption is based on a very simplistic view of the financial sector. Banks do not loan depositors’ savings, they extend credit, and so long as agents are willing to accept the resulting pseudo-IOUs as payment then loans can be made without a single penny of prior savings. Indeed, one does not even need a bank. If a merchant allows a farmer to buy a $20 bag of seed with an IOU, that merchant can then spend the IOU, as can every other subsequent recipient of that debt in the economy. Money is created out of thin air and all that is required is faith that the farmer will eventually bring forth a harvest sufficient to repay both the $20 (plus any interest). Even in the event that the farmer eventually defaults, the IOU nevertheless served as money in the interim.8 13
While a modern financial system with a central bank is more complex, the outcome is the same. When an entrepreneur approaches a bank for a $1 million loan, the bank evaluates the proposal and, if sufficiently impressed, adds $1 million to their assets under “loans” and $1 million to their liabilities under “demand deposits.” The books once again balance and the entrepreneur walks out with a check book connected to the newly-created $1 million deposit. The only potential problem for the banker is meeting the reserve requirement, which, even if the system is loaned up, can be satisfied by selling assets to the central bank. The latter, if targeting interest rates as modern central banks do, must oblige or those rates will rise. More than that, if they are convinced that these demands for funding are legitimate, they will want to accommodate them because that is their job (Furey, 2013). The upshot of all this is that in a modern financial system, savings in no realistic way represent a constraint on investment funding. This is hardly a new idea. As Keynes wrote, The investment market can become congested through the shortage of cash. It can never become congested through the shortage of saving. This is the most fundamental of my conclusions within this field (Keynes, 1937, p. 669). If savings are neither a function of interest rates nor the source of loans, then the loanable funds theory does not hold. As a consequence, with respect to the forces determining the levels of aggregate saving and investment in a macroeconomy and assuming that bank finance is forthcoming, it is investment that is in the driver’s seat. No amount of change in the desire to save will have any direct impact on investment whatsoever. But investment will always create the corresponding level of saving. An increase in investment funded by credit extended by the financial sector shifts D upward, and as D moves it raises employment, income, and saving. The 14
last will increase by precisely the same amount as the investment. Therefore 1) we can reach full employment if and only if investment is sufficiently large to absorb the saving that would occur at that level, but 2) there is no guarantee that this will happen. In fact, as seen below, there are many reasons why it may not.
IV. Investment, the business cycle, and stagnation Physical investment is an extremely expensive, long-term, and largely irreversible commitment. A reasonably well-founded forecast of future profitability is therefore an absolute necessity if we expect a rational individual to pull the trigger. Unfortunately, it is impossible to meet this condition in the real world because the future is inherently unknown. Sufficient data simply do not exist. This was a question that fascinated Keynes: why do people continually make important decisions based on arguments that, while no doubt well-researched and carefully considered, are not and cannot be conclusive? He pursued this in his Treatise on Probability (Keynes, 1921) and though the extent of its connection to the General Theory is debated, there is no question that his study of this phenomenon led him to reject classical probability theory as a meaningful representation of entrpreneurial decision making. The key sticking point is that, according to Keynes, investors operate in an environment of fundamental uncertainty, meaning that they cannot know all the possible future outcomes nor their likelihoods. This information is available at a roulette table. Indeed, the house knows it very well and takes it into account in determining the winners’ pay outs. But, realistically speaking, nothing approaching this exists for the entrepreneur. Nor is assuming that investors act as if they had sufficient information to create expected 15
value calculations warranted (Keynes, 1937, p. 215). Such an assumption leaves us with a characterization of the investment decision that differs too significantly from what entrepreneurs actually face. Theirs is a far more dangerous and unpredictable world than that found around a roulette wheel and, as a consequence, they exhibit a much more fundamental kind of instability. For, while the gambler may be disappointed when the ball comes to rest on red rather than black, she is not then led to question the entire foundation of the forecast that led to her choice. Rational individuals at the gaming table know the odds and that those odds do not change. The next bet, even if she picks a different color, will be placed with the same probability distribution in mind. This is not so when profits fall below expectations; indeed it cannot be since the probability distribution was unknowable in the first place. As a consequence, an entrepreneurial forecast, ...being based on so flimsy a foundation...is subject to sudden and violent changes. The practice of calmness and immobility, of certainty and security, suddenly breaks down. New fears and hopes will, without warning, take charge of human conduct. The forces of disillusion may suddenly impose a new conventional basis of valuation (Keynes, 1937, pp. 214-5). We do not and cannot generate mathematically objective forecasts, nor do we act as if we do. At the heart of every investment decision is the realization that we have “only the vaguest idea of any but the most direct consequences of our acts (Keynes, 1937, p. 213). To reiterate, disappointment in this world can create a much more fundamental sort of panic than that which may occur around a roulette table. Given this, one may wonder why any coldly rational individual would ever undertake an 16
investment project more significant than a vendor’s cart. They would not, but then we are not coldly rational. Rather, “a large proportion of our positive activities depend on spontaneous optimism rather than on a mathematical expectation” (Keynes, 1936, p. 161). When our spontaneous optimism (also known as animal spirits) outweighs the reluctance created by the fundamental uncertainty of the real world, we act; when it does not, we do not. Though this is a very basic version of what Keynes developed in the Treatise on Probability and elsewhere, it captures the essence. It is impossible to understand entrepreneurial decision making as strictly rational because sufficient information does not exist for entrepreneurs to develop conclusive arguments. At some level, they must rely on emotion. They have no choice. That said, they are not wildly irrational, either. Investors do their very best to collect and analyze data and weigh various options. The problem is that, in the end, such calculations will inevitably prove insufficient as the final basis for a decision. Keynes summarized the situation as follows: We should not conclude from this that everything depends on waves of irrational psychology. On the contrary, the state of long-term expectation is often steady, and, even when it is not, the other factors exert their compensating effects. We are merely reminding ourselves that human decisions affecting the future, whether personal or political or economic, cannot depend on strict mathematical expectation, since the basis for making such calculations does not exist; and that it is our innate urge to activity which makes the wheels go round, our rational selves choosing between the alternatives as best we are able, calculating where we can, but often falling back for our motive on whim or sentiment or chance (Keynes, 1936, pp. 162-3). 17
One may fear that modeling human economic behavior in this way renders it unpredictable and therefore impossible to model. This is not true. Not only does a clear pattern emerge, but it explains both the business cycle and the general tendency for the economy to come to rest at less than full employment. Dealing with the former first, say for example that we are just emerging from the depths of recession. Animal spirits are likely to be dampened and, since it is safer to wait than act, rates of investment are low. Some capital formation will be occurring, however, especially if repair and maintenance ignored during the downturn can no longer be delayed. Similarly, consumer durable spending may be recovering as it becomes less feasible to simply repair aging cars, refrigerators, and washers. For these reasons and others, realized rates of profit from any investment that does happen to occur are likely to exceed expectations. This and the fading memory of the downturn fuel optimism, causing entrepreneurs to revise their forecasts of upward–and rightly so, for prospects are, indeed, brighter than they had been earlier. The expansion begins. Otherwise idle entrepreneurs will now join in, encouraged by the apparent success of those who acted first (Keen, 1995, p. 611). The necessary credit to fund rising investment will be forthcoming because bankers’ forecasts are guided by these same principles and will therefore be similarly enthusiastic. Unemployment falls, incomes rise, and realized sales tend to justify investors’ newly-optimistic forecasts. The intercept of D shifts up and the economy moves toward, perhaps even to, Nf. But the seeds of the downturn are already being sown because realized rates of return will inevitably decline and fall below the buoyed expectations. For one, everything else being equal, the rise in the demand for physical capital increases its cost. In addition, financing may be getting 18
more expensive, in part because central banks tend to raise their interest rate target during expansions (an admittedly exogenous factor) and also because rising debt/income levels make individual agents less creditworthy (an endogenous one). But these are, according to Post Keynesians, minor compared to the effect of the rising stock of physical capital. Two factors are relevant here. First, capital that comes online in late expansion faces more competition than that which became operational in the early to middle period. The new restaurant in town is more likely to prosper when it is the only one than when it is one of five. Second, because it does not take long to complete an investment project, the rate of increase of investment spending will begin to wane and perhaps even become negative as some entrepreneurs reach target levels of capacity.9 This has the potential to lower aggregate demand just at the point when entrepreneurs are hoping to recoup the high cost of investment and begin reducing their now-heavy debt loads. There therefore comes a point in every expansion when, regardless of monetary or fiscal policy or exogenous shocks (all of which can certainly aggravate matters), business owners’ animal-spirits fueled expectations clash with reality. That reality need not be particularly dismal, just less pleasing than expected. ...it is an essential characteristic of the boom that investments which will in fact yield, say, 2 per cent in conditions of full employment are made in the expectation of a yield of; say, 6 per cent, and are valued accordingly. When the disillusion comes, this expectation is replaced by a contrary “error of pessimism”, with the result that the investments, which would in fact yield 2 per cent. in conditions of full employment, are expected to yield less than nothing; and the resulting collapse of new investment then leads to a state of unemployment in which the investments, 19
which would have yielded 2 per cent. in conditions of full employment, in fact yield less than nothing (Keynes, 1936, pp. 321-2). This is not, incidentally, a model of the business cycle premised on the idea that overinvestment is the culprit. It is not that entrepreneurs built more capital than would be profitable at full employment, but that they undertook projects based on expectations that were bound to be disappointed. The inevitable disillusion triggers the recession before we have reached a socially-optimal stock of physical capital, not after. For that reason, slowing the expansion is not a reasonable means of addressing this problem. Indeed, Keynes discusses at length the folly of raising interest rates during an upturn. Thus, time and again, the deceleration in investment characteristic of late expansion leads to lagging sales which disappoints expectations. This disappointment prompts agents to revise their forecasts. As the latter were generated in an environment of uncertainty and therefore lacked a firm foundation, the new estimates may reflect a considerable level of panic. Investment collapses, bringing down with it incomes and employment. The cycle starts again once bad memories fade and firms and consumers find it absolutely necessary to replace aging durables. This–not monetary policy or changes in productivity–creates the business cycle. These factors are consistent not just with the existence of a business cycle, but with the general tendency for investment to fall short of the level necessary to reach Nf. Furthermore, the more productive our society becomes, the more difficult it is for a pure free-market economy to generate levels of investment spending consistent with full employment. It is in this context that Keynes’ makes his well-known comment on the supposed folly of public investment: It is a curious thing, worthy of mention, that the popular mind seems only to be 20
aware of this ultimate perplexity where public investment is concerned, as in the case of road-building and house-building and the like. It is commonly urged as an objection to schemes for raising employment by investment under the auspices of public authority that it is laying up trouble for the future. 'What will you do,' it is asked, 'when you have built all the houses and roads and town halls and electric grids and water supplies and so forth which the stationary population of the future can be expected to require?' But it is not so easily understood that the same difficulty applies to private investment and to industrial expansion; particularly to the latter, since it is much easier to see an early satiation of the demand for new factories and plant which absorb individually but little money, than of the demand for dwelling-houses (Keynes, 1936, pp. 106). The developed economies of the world solved the core scarcity problem decades ago. No one is going hungry or naked for reasons that have anything to do with opportunity costs. Though it may be true that we must make choices when it comes to the distribution of XBoxes, cuttingedge cell phones, or theater tickets, we can easily produce more than enough of the things necessary for life and several levels above that. What has vexed us for many years is finding a means of generating a job for all those willing to work. That the economics discipline has almost completely missed this fact has led to many meaningless models and disastrous policies. Austerity programs, for example, are based on a fundamental misunderstanding of the operation of macroeconomies. They are akin to bleeding the patient. Data are, incidentally, consistent with the claim that investment spending and profits tend to decline in late expansion. Table 1 shows averages for nine of the ten US expansions since 1950 21
divided into two categories: the early to mid period and the final four quarters (the excluded expansion was 1980:IV to 1981:II, which was too short to allow such a comparison). Note first that the average rate of growth of real investment spending is, indeed, significantly higher in the early to mid period (15.9%) than in the late (4.5%).10 Second, quarters of negative investment growth are more than twice as likely in the final year of an expansion than earlier. Last, corporate profits begin to shrink as the upturn weakens, creating a situation where disappointment is likely.11 These profits are not necessarily poor, particularly in light of the fact that the decline is rather small relative to the long period of increase. The problem occurs when they are less than what was forecast.
Table 1: Averages for all US expansions since 1950 lasting > four quarters. Expansion Stage Early to Mid
% Negative Inv
Last Four Quarters 4 4.5% 53% -1.1% KEY (all data from Federal Reserve Bank of St. Louis) %Ä Inv: Percentage change in real gross private domestic investment spending since previous quarter, annual rates. Length: Length of stage in quarters. % Negative Inv: Percentage of quarters during stage that real investment spending growth was negative. %Ä Corp Profits: Percentage change in after-tax corporate profits (deflated by the GDP deflator) since previous quarter, annual rates.
V. Davidson’s Capital Market Diagram In terms of formal models of the above, Paul Davidson’s is perhaps the most useful (Davidson, 1978 and 2011). Taking first the demand for physical capital, he specifies it as: (1)
Dk = f(pk, i, Ö, E) - - + + where Dk is the quantity of capital demanded, pk is the price of capital goods, i is the rate of discount used by entrepreneurs in considering the present value of expected future profits, Ö is the expected growth in demand for the products produced by the capital in question, and E is the number of investors able to obtain finance for their projects. Note that the financial sector is immediately linked to the real in the form of E. This is extremely important in Post Keynesian economics. Note further that Ö, as a forecast, is subject to the factors associated with fundamental uncertainty and animal spirits discussed above (as is E but in a less direct way). This function, when placed in pk and quantity of capital space, yields the negatively-sloped line labeled Dk in Figure 4. Changes in i, Ö, or E will cause shifts in the directions implied by the signs of their partial derivatives. The vertical line Sk marks the current stock of capital and its intersection with Dk yields ps or the price entrepreneurs would be willing to pay for an existing unit (Davidson’s stock demand price). The current level of investment is a function of the comparison of this subjective price with what must actually be paid to build new capital. For example, suppose agents, taking into account the existing quantity of capital and all the factors in equation (1), were willing to pay $1 million for a representative factory. Say at the same time that it costs only $500,000 to build one. This means that new factories will be constructed and net investment will be positive. But, if entrepreneurs valued existing factories at $250,000 while building cost $500,000, then there is a glut of factories and net investment will be negative. 23
Figure 4: Davidson’s capital market diagram, stock supply price determination. This begs the question of how current costs are determined. This is shown in Figure 5. The function Sk+sk shows the sum of the existing stock of capital (Sk) plus a supply function for the building of new capital (sk). The latter is positively sloped on the assumption that the greater the volume of current investment, the higher the cost of building materials, labor, architectural and contracting services, and so on. Price pm is the minimum at which one could possibly build a new unit of capital. To determine the price at which new capital can be produced, pf (Davidson’s flow supply price), Sk+sk must be compared to the total demand for capital. This is not Dk, however, as the capital-goods producing industry will be busy with more than just new construction. Depreciation must be addressed, as well. It was important to Keynes, incidentally, that this be understood as a function of diminishing profitability of capital rather than a decline in actual physical productivity. In other words, an entrepreneur is induced to undertake replacement investment not simply because a machine breaks down (which plays a role, of course), but also when what it produces–even if it does so at a high rate of technical efficiency–is no longer judged 24
to be saleable at the expected profit. Returning to the point at hand, Figure 5 shows the total demand for capital as Dk+dk, where dk is replacement investment. How much depreciation is actually replaced turns out to be a function of the difference between ps and pf, as will be seen below.12
Figure 5: Davidson’s capital market diagram. The intersection of the total demand for capital, Dk+dk, and the total supply, Sk+sk, yields the price entrepreneur’s must pay to have a new unit of capital produced: pf. When this lies below ps, positive net investment occurs; when it does not, net investment is negative. The former is illustrated in Figure 5. There, gross investment is k3 - k1, that just to replace depreciation is k3 - k2 (equal to the distance between Dk and dk), and net investment is k2 - k1. Note what this implies about the next time period for, while Figure 5 is an equilibrium model and in that sense shows an economy at rest, it is just a snapshot of a dynamic process. Not only is Davidson careful to stress the possibility of interactions among functions (which is a general theme in Post Keynesian 25
economics), but the graph shows that next period the stock of capital will be at k2. Ceteris paribus (leaving the demand curves in place and shifting the supply curves to the right), this means that ps and pf will move closer together, thereby reducing the levels of both gross and net investment. As the former is what we would transfer to the Z-D diagram as the vertical intercept of D, it illustrates the difficulty of reaching Nf. Not only is the existence of a level of gross investment sufficient to absorb the full-employment level of savings a mere coincidence, but there is no reason to believe that it is sustainable because as the stock of capital increases, so the incentive to continue adding to it decreases. Were this the entire story, the implication would be that we eventually reach a stationary state where ps = pf, net investment is zero and entrepreneurs, presumably at their target level of capacity, simply replace depreciation period after period.13 But this ignores the dynamics of Dk, which is driven in no small part by entrepreneurs’ and financiers’ expectations (recall Ö and E in equation (1)). Consistent with the above discussion, while fundamental uncertainty makes it impossible for them to generate mathematically objective forecasts, there is nevertheless a clear and predictable pattern to their behavior. It is best understood by considering the stages of the business cycle.
Table 2: Impact of business cycle stage on expectations Stage
The key is really the same as has been already described above, but this time with changes in expectations being reflected by shifts of Dk on the capital market diagram. Just as before, surprises, both pleasant and unpleasant, can lead to rapid and significant forecast revisions. In early expansion, agents are relatively pessimistic. The pick up in investment and, therefore, sales and profits come as a pleasant surprise. This fuels the animal spirits of entrepreneurs and of those who lend to them and the consequent rise in Ö and E from equation (1) causes a rightward shift in Dk (and Dk+dk). This shift is assumed to be more than sufficient to offset the rightward movement of Sk+sk that occurs any time there is positive net investment, leading to the period of strong investment that characterizes the mid-expansion phase. Since those realized outcomes are consistent with expectations, no significant revisions of forecasts take place. The demand curves stay in place. But, the steady rightward shifts of the supply curves eventually take their toll. By late expansion, entrepreneurs find that their predictions were overly optimistic. They may discount this at first, but they will eventually come to realize that at least some of projects are not going to earn what they had hoped. Entrepreneurs are rudely reminded that they never had a firm basis for their expectations in the first place and their error of optimism “is replaced by a contrary ‘error of pessimism’” (Keynes, 1936, p. 322). This manifests itself in a collapse in Ö in equation (1), causing a sudden leftward shift in (Dk+dk). What had been slowing but positive net investment 27
now becomes negative net investment and the financial sector, stung by defaults and frightened by the sudden downturn, will likely ration credit more carefully, lowering E. In addition, rising liquidity preference may lead to an increase in the rate of discount (i). It may be some time before the negative net investment creates sufficient scarcity in the capital market for investment to rise again, which is why Keynes feared that slumps would be more protracted than booms.
VI. Economics without full employment There are, of course, many other aspects of Post Keynesianism. There is, for example, an entire alternate micro theory based on the importance of oligopoly, market power, internal financing, and non-price competition (Eichner, 1976). There is a trade theory focused on supply chains and absolute advantage (Milberg and Winkler, 2013). Post Keynesians emphasize the study of financial markets and explain how asset prices are often more heavily influenced by short-term psychology than factors truly associated with the long-term viability of the asset issuer–especially the more we “liberalize” such markets (Grabel, 2003). Hyman Minsky hypothesized that economic agents tend to take on debt to the point where it creates systemic financial fragility (Minsky, 1982). With respect to inflation, Post Keynesians maintain that the conventional wisdom that central banks can cause it by creating an excess supply of money is based on a primitive understanding of the financial sector. Every tool available to the central bank requires the conscious and voluntary cooperation of the recipient of the new cash: you cannot force someone to sell their Treasury Bill and you cannot force someone to take out a loan (Davidson and Weintruab, 1973). If they chose to do so, then they are not left with excess money balances, but just what they wanted. Money is like a haircut: it is impossible to create without there being a 28
corresponding and simultaneous demand. There is also an exchange rate theory premised on the idea that in today’s world it is financial capital, not trade, that drives currency prices in both the short and long run and that trade imbalances can exist indefinitely (Harvey, 2009). What all of these examples have in common is that they do not assume, over any time horizon, a tendency for the economy to be drawn toward full employment. Quite the contrary, for all the reasons suggested earlier, Post Keynesians believe that it is very difficult to get to full employment and that even if we do, it is unsustainable. Post Keynesian analysis therefore does not focus on opportunity costs since this is a useful concept only when current capacity is being fully utilized. The real question is what determines our current level of capacity utilization in the first place. To reiterate a point made earlier, Post Keynesians are not denying that scarcity is never an issue, it is just not the issue. Involuntary unemployment is. A popular (though not universally endorsed) policy prescription is the creation of a jobs program (Tchnernva, 2014). This differs from traditional Neoclassical Keynesian deficit spending. The goal of the latter is to indirectly induce the private sector to hire all those who are willing to work but unable to find jobs. Tax cuts, infrastructure expenditures, defense spending, and the like are supposed to create an income and employment multiplier that brings us back to the perimeter of the production possibilities frontier. But realistically speaking, “How many missiles would the government have to order before a job trickles down to Harlem?” (Wray, 2007, p. 17). Post Keynesians argue instead that if it is the government’s intention to create jobs, then they should do so directly. This is a much more efficient method and will still stimulate the private sector. We are not short of useful things for people to do in our society, only of those that will earn a profit. Even the most unskilled person can do something in the public sector that can 29
contribute to social welfare and the public good, while simultaneously beginning to learn new skills, gain new work experience, and enhance their own human capital which will make them more employable in the eyes of private employers” (Tcherneva, 2012, p. 32). This policy not only addresses chronic unemployment in a way that traditional deficit spending cannot, but the expenditures can be driven by formulae and might therefore be more difficult to abuse politically. If your state has 750,000 unemployed, then you are allotted 750,000 times the established wage plus operating expenses. You do not affect your state’s share via political favors, nor does this require an act of Congress. And since, as emphasized by the Modern Monetary Theorists in Post Keynesian economics, it is impossible for the US to default on debt denominated in its own currency, there is no limit to the government’s ability to fulfill this role (Fawley and Juvenal, 2011). Last, remuneration would be set somewhat below the private sector’s equivalent so that as the latter grew workers would migrate to it. Post Keynesians are not arguing that the market is bad, only that it is insufficient.
VII. Conclusions Continuing the last point, Post Keynesians neither pro- nor anti-markets. Markets are tools and as such it makes no more sense to be for or against them than it would to be for or against hammers. One tends to be very “pro-hammer” when the job at hand involves driving a nail into a piece of wood. On other hand, most people are generally “anti-hammer” when it comes to cutting something neatly in half. It is not a question of politics or ideology, but of pragmatics. Policy choices are to be determined on a case-by-case basis and we will likely have to experiment 30
with and change policies over time. Post Keynesian economists can be found in a number of universities and research centers.14 Though they publish in a variety of journals, the primary ones remain the Journal of Post Keynesian Economics (co-founded in 1978 by Sidney Weintraub and Paul Davidson) and the Cambridge Journal of Economics (founded in 1977 by the Cambridge Political Economy Society).15 Their research is frequently presented in sessions at the annual meetings of the Association for Evolutionary Economics, the Association for Heterodox Economics, the Association for Institutional Thought, and the Eastern Economics Association. Post Keynesian economics has also been in the news quite a bit since the Financial Crisis of 2007-8. Hyman Minsky’s name appears frequently even in the popular press, the Washington Post featured an article on Modern Monetary Theory (Matthews, 2012), the graduate student who exposed the spreadsheet errors in Carmen Reinhart and Kenneth Rogoff’s paper on the impact of government debt was enrolled in a Post Keynesian program (University of Massachusetts at Amherst; Coy, 2013), and the current chief economist for the US Senate Budget Committee and for US Presidential Candidate Bernie Sanders, Stephanie Kelton, is a Post Keynesian. Post Keynesians are a very active group who continue to apply, modify, and extend the work of economists like Michael Kalecki, Piero Sraffa, Nicholas Kaldor, Joan Robinson, and, of course, John Maynard Keynes. While they tend to be particularly interested in macro theory and policy, their interests can be found under each of the Journal of Economic Literature classification categories. Some of what they do looks quite similar to what one finds in the mainstream in that you might see systems of equations, graphical analyses, and econometrics. Other aspects, like open-systems modeling, fundamental uncertainty, and their focus on the 31
institutional specifics of the monetary sector, look more foreign to those trained traditionally. One thing is for certain, however: you will never read “assume full employment” in a Post Keynesian study.
Endnotes 1.The author is indebted to Victoria Chick, Paul Davidson, and Kiril Tochkov for their insightful comments. Remaining errors are, of course, the sole responsibility of the author. 2.Note that interpreting Keynes’ Z-D analysis has not been without controversy (King 1994). I employ that used by Davidson (1972) and Chick (1983). Davidson and Smolenky (1964) is an early development. 3.Starting with W/P=MPN, multiply both sides by P and divide by MPN, yielding W/MPN=P. Now multiply both sides by y and the left by (N/N) to give (W/MPN)y(N/N)=Py. Rearrange and express y/N as APN, or the average product of labor, and we are left with the final equation for Z: [W(APN/MPN)]N=Py. 4.Simply put, he is measuring N in units of unskilled labor and at wage units equal to what an unskilled laborer would earn (see Chick, 1983, pp. 68-70). For example, if unskilled labor earns $1000 per month and an accountant earns $3000, then hiring one more accountant moves us three units on the N axis. 5.My personal preference is to give D decreasing slope as I believe this to be more consistent with the fundamental psychological law, but not all Post Keynesians do. 6.This is not strictly true, as he acknowledged that some investment spending is a function of N while some consumption spending is autonomous. However, it simplifies the analysis and does not change the conclusion. 7.In both cases, of course, the initial change in investment or saving that sets the forces in motion is partly dampened by the interest rate movement it causes. 8.Such an arrangement would obviously not be viable in today’s more complex and impersonal world, but the principle is the same. 9.Expansions have averaged just over 58 months since World War Two. Meanwhile, while this can obviously vary, a restaurant (for example) should take well under a year to complete (see for example http://www.martinkovicmilford.com/q_and_a_restaurant.php). 10.As suggested earlier, while the basic model is laid out in nominal terms, Post Keynesians nevertheless typically deflate values when they are being compared over a long period of time. 11. That these profits are not the same as the rate of return on investment suggests that we use some caution in making inferences, but it nevertheless seems likely that there would be a strong correlation. 12.The functions are drawn parallel on the assumption that their price sensitivities are identical.
13.This would be the case whether we started with ps > pf or ps < pf. 14.A helpful guide can be found on the Heterodox Economics Newsletter website (http://www.heterodoxnews.com/HEN/home.html). Also useful are the sites for the Post Keynesian Study Group (https://www.postkeynesian.net/), the Association for Heterodox Economics (http://hetecon.net/), and the Real-World Economics Review (http://www.paecon.net/PAEReview/). 15.Note that the latter is less exclusively Post Keynesian.
References Chick, Victoria. (1983). Macroeconomics After Keynes: A Reconsideration of the General Theory, Cambridge, MA: MIT Press. Chick, V. (2006). The General Theory is difficult: whose fault is that?. Revista de Economia, 32(2), 135-151. Coy, Peter. (2013, April 18). Economists' Spreadsheet Error Upends the Debt Debate, Bloomberg Business. Retrieved from http://www.bloomberg.com/bw/articles/2013-04-18/economists-spreadsheet-error-upends -the-debt-debate Davidson, P. (1978). Money and the real world. London: Macmillan. Davidson, P. (Ed.). (2011). Post Keynesian macroeconomic theory. Edward Elgar Publishing. Davidson, P., & Smolensky, E. (1964). Aggregate demand and supply analysis. New York: Harper and Row. Weintraub, S., & Davidson, P. (1973). Money as Cause and Effect. Economic Journal, 83(332), 1117-32. Eichner, A. (1976). The megacorp and oligopoly (Vol. 1979). Cambridge: Cambridge University Press. Fawley, B. W., & Juvenal, L. (2011). Why health care matters and the current debt does not. The Regional Economist, 19(4), 4-5. Retrieved from https://www.stlouisfed.org/publications/regional-economist/october-2011/why-health-care -matters-and-the-current-debt-does-not
Furey, K. (2013, March). A reading on money and money creation. In Forum for Social Economics (Vol. 42, No. 1, pp. 38-58). TF. Grabel, I. (2003). Averting crisis? Assessing measures to manage financial integration in emerging economies. Cambridge Journal of Economics, 27(3), 317-336. Harvey, J. T. (2009). Currencies, capital flows and crises: A post Keynesian analysis of exchange rate determination. Routledge. Keen, S. (1995). Finance and economic breakdown: modeling Minsky's" financial instability hypothesis". Journal of Post Keynesian Economics, 607-635. Keynes, John Maynard. (1921). A Treatise on Probability, London: Macmillan. Keynes, John Maynard. (1936). The General Theory of Employment, Interest and Money, London: Macmillan. Keynes, J. M. (1937). The ex-ante theory of the rate of interest. The Economic Journal 47(168), 663-669. King, J. E. (1994). Aggregate supply and demand analysis since Keynes: a partial history. Journal of Post Keynesian Economics 17(1), 3-31. Matthews, Dylan. (2012, February 18). Modern Monetary Theory is an unconventional take on economic strategy. The Washington Post. Retrieved from http://www.washingtonpost.com/business/modern-monetary-theory-is-an-unconventionaltake-on-economic-strategy/2012/02/15/gIQAR8uPMR_story.html Milberg, W., & Winkler, D. (2013). Outsourcing economics: Global value chains in capitalist development. Cambridge University Press.
Minsky, H. P. (1982). Can it happen again? Essay on instability and finance. New York: ME Sharpe. Tcherneva, P. R. (2012). Inflationary and distributional effects of alternative fiscal policies: An augmented Minskyan-Kaleckian model. Levy Economics Institute, Working Paper No. 706. Retrieved from http://www.levyinstitute.org/pubs/wp_706.pdf Tcherneva, P. R. (2014). Reorienting fiscal policy: a bottom-up approach. Journal of Post Keynesian Economics, 37(1), 43-66. Wray, L. R. (2007). The employer of last resort programme: could it work for developing countries?. Internat. Labour Organization. Retrieved from http://www.cfeps.org/elm07-5.pdf