The rate of change of debt (with respect to GDP) is like your speed of travel the faster you drive, the sooner you'll get there but there's a twist. On the way out, increasing debt makes the journey more pleasant the additional spending increases aggregate demand and this experience is what fooled neocla.s.sical economists, who ignore the role of debt in macroeconomics, into believing that the economy was experiencing a 'Great Moderation.' But rising debt increases the distance you have to travel backwards when you want to reduce debt, which is what the USA is now doing. So rising debt feels great on the outward drive from LA east (increasing debt), but lousy when you want to head home again (and reduce debt).

The Credit Impulse is like acceleration it's a measure of the g-forces, so to speak, generated by either rapid acceleration or rapid deceleration. Acceleration in the debt level felt great on the way up: it was the real source of the booms in the Ponzi economy that the USA has become. Equally, acceleration in the opposite direction in effect going backwards at an accelerating speed is terrifying: as the rate of decline in debt increases, the fall in aggregate demand increases and unemployment explodes.

The interactions of the level of debt, rate of growth of debt and the Credit Impulse are akin to those between distance, speed and acceleration as well and here I'll limit my a.n.a.logy to the last few years, when America went from increasing debt the drive from LA to New York and then abruptly changed direction into deleveraging.

The reversal of direction necessarily involves your acceleration changing from zero or positive to negative, and it feels dreadful: imagine the feeling of slamming on the brakes, putting the car in reverse, and then driving backwards at an accelerating speed.

At some point, however, you will reach the maximum reverse speed of the car, and at that point the terrifying feeling of driving backwards more rapidly will give way to merely the unpleasant feeling of driving backwards at high speed. If you then start driving backwards less rapidly, you will actually feel a positive acceleration even though you are still driving backwards. However, if you keep slowing down your reverse speed, then at some point you will reverse direction, and start heading back towards New York again. You can't maintain positive acceleration indefinitely without at some point changing from a negative to a positive velocity, and thus resuming your journey towards a place that you were initially trying to leave.

We can now get a handle on why this recession has been so extreme compared to its post-World War II predecessors, and why I believe that the crisis has many years to run.

First, all three debt indicators reached levels that are unprecedented in the post-World War II period. The debt-to-GDP ratio, which began the post-war period at barely 50 percent, increased by a factor of 6 in the subsequent five decades to reach a peak of 298 percent of GDP in early 2009.

Secondly, while private debt itself grew at a relatively constant if volatile 10 percent per annum between 1955 and 2008, the debt-financed proportion of aggregate demand rose from 5 percent in the 1950s to 28 percent in 2008.

13.15 Relatively constant growth in debt This occurred because the rate of growth of nominal debt was about 3 percent higher than that of nominal GDP from 1945 till 2008. The impact of rising debt on aggregate demand therefore doubled every twenty-three years.12 It then plunged to minus 19 percent in early 2010 an unprecedented event in post-World War II economic history. This debt level is still falling, though the rate of fall has slowed in recent times, from a peak rate of minus 19 percent of GDP in early 2010 to minus 12 percent in September 2010 (the last date at which debt data were available at the time of writing).

13.16 The biggest collapse in the Credit Impulse ever recorded 13.17 Growing level of debt-financed demand as debt grew faster than GDP Thirdly, the Credit Impulse averaged plus 1.2 percent from 1955 till 2008, and then hit at an unprecedented minus 27 percent in 2009 at the depths of the downturn. It is now returning toward zero which in part reflects its inevitable return toward zero as deleveraging becomes entrenched.13 This puts far less drag on aggregate demand, but also removes the 'turbo boost' that a positive Credit Impulse gave to growth in the previous half-century. The Credit Impulse will also tend to be negative while deleveraging continues, just as it tended to be positive when rising debt was boosting aggregate demand. This means the economy will have a tendency toward recessions rather than booms until the debt-to-GDP ratio stabilizes at some future date.

The interaction of these three factors will determine the economic future of the United States (and many other OECD nations, which are in a similar predicament).

The Credit Impulse, as the most volatile factor, will set the immediate economic environment. While it remains negative, the rate at which the USA is deleveraging accelerates, so it therefore had to rise again at some stage as it has since mid-2009. This will accelerate aggregate demand, but it can't lead to a sustained rise in aggregate demand without causing the debt-to-GDP ratio to rise. That is extremely unlikely to happen, since even after the deleveraging of the last two years, the aggregate level of private debt is 100 percent of GDP higher than it was at the start of the Great Depression.

These dynamics of debt were the key cause of both the Great Moderation and the Great Recession, yet they were completely ignored by neocla.s.sical economists because of their fallacious belief that changes in private debt have no macroeconomic effects (Bernanke 2000: 24). Therefore, far from making sure that 'It' won't happen again, as Bernanke a.s.serted in 2002, by ignoring and in fact abetting the rise in private debt, neocla.s.sical economists have allowed the conditions for another Great Depression to develop. Worse, a comparison of today's debt data to those from 192040 shows that the debt-deflationary forces that have been unleashed in the Great Recession are far larger than those that caused the Great Depression see Figure 13.18.

13.18 The two great debt bubbles Debt deflation then and now Comparing the 1920s1940s to now the Roaring Twenties and the Great Recession to the 'Noughty Nineties'14 and the Great Recession is feasible, but complicated both by differences in the economic circ.u.mstances at the time, and differences in the quality of the statistics.

A major complication is the extreme volatility in economic performance over the 1920s no one was writing about 'the Great Moderation' back then. The decade began and ended with a depression, and recorded output fluctuated wildly. The average increase in nominal GDP over 192129 was 4.5 percent, but it fluctuated wildly from 2 to +13 percent, with a standard deviation of 4.4 percent. In contrast, the Noughty Nineties recorded a higher rate of nominal growth of 5.3 percent, and this was very stable, ranging between 2.6 and 6.6 percent with a standard deviation of only 1.4 percent.

13.19 Change in nominal GDP growth then and now However, as well as being a decade of stock market speculation, the 1920s also saw serious Schumpeterian investment and 'creative destruction.' It was the decade of the Charleston and The Great Gatsby, but it was the decade of the production line, technological innovation in manufacturing and transportation, and the continuing transformation of American employment from agriculture to industry. The average rate of real economic growth was therefore higher in the 1920s than in the period from 1999 to 2009 though disentangling this from the gyrations in the price level is extremely difficult.

13.20 Real GDP growth then and now For example, the nominal rate of growth in 1922/23 was 13 percent, but the 'real' rate was an even higher 20 percent. This impossible level reflected the simultaneous recovery from deflation of over 10 percent to inflation of 3 percent, and unemployment falling and hence output rising from 12 percent to 2.5 percent as the economy recovered from the depression of January 1920 to June 1921.

13.21 Inflation then and now Overall, the rate of unemployment is the best means to compare the two periods, but here we run into the distortions caused by politically motivated redefinitions of the unemployment rate since the late 1970s (see Box 13.1). The U-3 measure for 19992009 averages 5 percent, only marginally higher than the average of 4.7 percent for 192029; but the U-6 measure for 19992009 averages 8.8 percent, and I regard this as a fairer comparison of the two periods.

13.22 Unemployment then and now The upshot of all this is that the Roaring Twenties saw more real growth than the Noughty Nineties, and this masked the importance of debt at the time. But categorically, the fundamental cause of both the Great Depression and the Great Recession was the bursting of a debt-financed speculative bubble that had fueled the false but seductive prosperity of the previous decade.

The Great Depression remains the greatest economic crisis that capitalism has ever experienced, but on every debt metric, the forces that caused the Great Recession are bigger.15 Private debt rose 50 percent over the 1920s, from $106 billion (yes, billion) in 1920 to $161 billion by 1930; it rose from $17 trillion to $42 trillion between 1999 and 2009 a 140 percent increase.

13.23 Nominal private debt then and now In inflation-adjusted terms, the increase was very similar a 72 percent increase over the Roaring Twenties versus an 85 percent increase from 1999 to 2009. Remarkably, the real level of debt grew at almost precisely the same rate for the first eight years in both periods a rate of about 7 percent per year. This chimes with one implication of the monetary model of capitalism I outline in the next chapter: banks increase their profits by increasing debt, and they therefore have an incentive to increase debt as fast as is possible. The easiest way to do this is to fund Ponzi schemes, which were the hallmark of both the Roaring Twenties and the 'Noughty Nineties.'

13.24 Real debt then and now Though the rate of growth of debt was similar, the level of debt compared to GDP is far higher now than in the 1930s. The debt-to-GDP ratio was 175 percent when the Great Depression began; it is over 100 percent higher today, and hit 298 percent before it began to reverse in 2009. The degree of deleveraging needed to eliminate the Ponzi overhang is therefore much higher today than it was in 1930.

13.25 Debt to GDP then and now Rising debt fueled the Roaring Twenties, just as rising debt fueled the false prosperity of the Internet and Subprime Bubbles in the 'Noughty Nineties.' Since the rate of real economic growth was higher back in the 1920s than today, the debt ratio itself remained roughly constant prior to the bursting of the Ponzi Scheme in the 1920s; however, debt grew as rapidly in real terms in the 1920s as it did in the noughties, and the collapse of debt in real terms when the crisis. .h.i.t was also remarkably similar.

But from there they diverge, because the second scourge of the 1930s deflation has yet to occur in a sustained manner during the Great Recession. Consequently, while the real burden of debt rose during the early 1930s even as the nominal level of debt was falling, so far the Great Recession has involved falling debt in both real and nominal terms.

13.26 Real debt growth then and now One possible reason for the marked difference in inflationary dynamics between the two periods is the composition of private debt. In the 1920s, the vast bulk of the debt was owed by business. Business debt was three times that of household debt, and four times that of the financial sector. Therefore, when the Roaring Twenties boom collapsed as debt-financing fell, businesses were the ones in serious financial difficulties. As Fisher surmised, individual businesses responded by cutting their markups to try to entice customers into their stores and not their compet.i.tors', leading to a general fall in the price level that actually increased the debt-to-GDP ratio, even as nominal debt levels fell.

Today the ranking is reversed in the insolvency stakes: the financial sector carried the highest level of debt leading into the Great Recession virtually 125 percent of GDP, five times the level of debt it had in 1930. Households come second now, with a debt level of almost 100 percent of GDP, two and a half times the level they had in 1930. The business sector carried a modest debt level of 80 percent of GDP, when compared to its 1930s level of 110 percent though even this is more than twice its debt level during the 'Golden Age' of the 1950s and 1960s.

This composition difference may have implications for how the debt-deflationary dynamics of the Great Recession will play out. The prospects of a 1930s-style deflationary collapse are low, since businesses do not face the direct pressure of insolvency that they faced back then. However, their retail customers, the consumer sector, have never been this debt-enc.u.mbered, and it is far harder for households to reduce debt than it is for businesses: to put it colloquially, businesses can get out of debt by going bankrupt, ceasing investment, and sacking the workers. Bankruptcy is far more painful for individuals than companies; it is much harder to stop consuming than it is to stop investing, and households can't 'sack the kids.'

This implies a far less severe tendency to deflation, but a more intractable one at the same time since consumer demand will remain muted while debt levels remain high.

13.27 The collapse of debt-financed demand then and now 13.28 Debt by sector business debt then, household debt now Finally, though the Roaring Twenties became a reference period for frivolous speculation in popular culture, they have nothing on the Noughty Nineties. Debt-financed spending never exceeded 10 percent of GDP in the 1920s. In the noughties, it rarely fell below 20 percent of GDP. The popular culture of the twenty-first century may ignore the Roaring Twenties and see the Noughty Nineties as the hallmark of delusional economic behavior.

Given this much higher level of debt-financed speculation, the plunge into negative territory was far faster in 2008/09 than it was in 192931 but the reversal of direction has also been far more sudden. The change in debt went from adding 28 percent of GDP to aggregate demand in 2008 to subtracting 19 percent from it in 2010, but the rate of decline turned around merely a year after the crisis began, compared to the three years that elapsed before the debt-financed contribution started to rise from the depths in the 1930s (see Figure 13.27) (a large part of this may be the product of the huge intervention by both the federal government and the Federal Reserve).

The Credit Impulse was also far more dramatic in the noughties than in the twenties: it was higher during the boom, and plunged far more rapidly and deeply during the slump. The Credit Impulse took four years to go from its positive peak of 2.5 percent before the Great Depression to 16 percent in 1931. It began from the much higher level of 5 percent in late 2007 and fell to a staggering 26 percent in late 2009 a plunge of over 30 percent in just two years versus an 18 percent fall over four years in the Great Recession.

13.29 The Credit Impulse then and now The collapse in debt-financed aggregate demand was the key factor behind both the Great Depression and the Great Recession. Though debt-financed demand played less of a role in the 1920s than it did in the noughties, the collapse in the Great Depression was as deep as today's, and far more prolonged, which caused unemployment to hit the unprecedented level of 25 percent in 1932. When the Credit Impulse finally rose again in 1933, so did employment, and unemployment fell to just over 11 percent in mid-1937 leading to hopes that the depression was finally over.

However, debt-financed demand turned negative once again in 1938, and unemployment rose with it to 20 percent. Only with the onset of the war with j.a.pan did unemployment fall back to the average experienced during the 1920s.

13.30 Debt-financed demand and unemployment, 192040 The same pattern has played out during the Great Moderation and Great Recession. When debt-financed demand collapsed, unemployment exploded to 10 percent on the U-3 measure, and 17 percent on the more comparable U-6 measure. Just as significantly, the unemployment rate stabilized when the decline in debt-financed demand turned around. Though the huge fiscal and monetary stimulus packages also played a role, changes in debt-financed demand dominate economic performance.

One statistical indicator of the importance of debt dynamics in causing both the Great Depression and the Great Recession and the booms that preceded them is the correlation coefficient between changes in debt and the level of unemployment. Over the whole period from 1921 till 1940, the correlation coefficient was minus 0.83, while over the period from 1990 till 2011, it was minus 0.91 (versus the maximum value it could have taken of minus one). A correlation of that scale, over time periods of that length, when economic circ.u.mstances varied from bust to boom and back again, is staggering.

13.31 Debt-financed demand and unemployment, 19902011 The Credit Impulse confirms the dominant role of private debt. The correlation between the Credit Impulse and the rate of change of unemployment was minus 0.53 in 192240, and minus 0.75 between 1990 and 2011.

13.32 Credit Impulse and change in unemployment, 192040 13.33 Credit Impulse and change in unemployment, 19902010 Changes in the rate of change of credit also lead changes in unemployment. When the Credit Impulse is lagged by four months, the correlation rises to minus 0.85.

13.34 The Credit Impulse leads change in unemployment This correlation is, if anything, even more staggering than that between debt-financed demand and the level of unemployment. The correlation between change in unemployment and the Credit Impulse is one between a rate of change and the rate of change of a rate of change. There are so many other factors buffeting the economy in addition to debt that finding any correlation between a first-order and second-order effect is remarkable, let alone one so large, and spanning such different economic circ.u.mstances from the recession of the early 1990s, through the 'Great Moderation,' into the Great Recession and even the apparent beginnings of a recovery from it.

Fighting the Great Recession The global economy won't return to sustained growth until debt levels are substantially reduced. With debt at its current level, the general tendency of the private sector will be to delever, so that the change in credit will deduct from economic growth rather than contributing to it. Any short-term boost to demand from the Credit Impulse such as that occurring in early 2011 will ultimately dissipate, since if it were sustained then ultimately debt levels would have to rise again. Since the household sector in particular is debt-saturated, credit growth will hit a debt ceiling and give way to deleveraging again. The US economy in particular is likely to be trapped in a never-ending sequence of 'double dips,' just as j.a.pan has been for the last two decades.

There is a simple, but confrontational, way to stop this process: a unilateral write-off of debt.

This policy which occurred regularly in ancient societies, where it was known as a Jubilee (Hudson 2000: 347) goes strongly against the grain of a modern capitalist society, where paying your debts is seen as a social obligation. But the ancient and biblical practice addressed a weakness in those societies the tendency for debtors to become hopelessly indebted given the enormous interest rates that were common then: Mesopotamian economic thought c. 2000 BC rested on a more realistic mathematical foundation than does today's orthodoxy. At least the Babylonians appear to have recognized that over time the debt overhead came to exceed the ability to pay, culminating in a concentration of property ownership in the hands of creditors. While debts grew exponentially, the economy grew less rapidly. The earning capacity of Babylonian rural producers hardly could be reconciled with creditor claims mounting up at the typical 33.333 percent rate of interest for agricultural loans, or even at the commercial 20 percent rate. Such charges were unsustainable for economies as a whole. (Ibid.: 348) It would be foolish to deny that we have a similar weakness in modern capitalist society: our tendency to be sucked into Ponzi schemes by a banking sector that profits from rising debt.

As I explain in the next chapter, when lending is undertaken for investment or consumption, debt tends not to get out of hand. But when borrowing is undertaken to speculate on a.s.set prices, debt tends to grow more rapidly than income. This growth causes a false boom while it is happening, but results in a collapse once debt growth terminates as it has done now.

Though borrowers can be blamed for having euphoric expectations of unsustainable capital gains, in reality the real blame for Ponzi schemes lies with their financiers the banks and the finance sector in general rather than the borrowers. That was blindingly obvious during the Subprime Bubble in the USA, where many firms willfully wrote loans when they knew or should have known that borrowers could not repay them.

Such loans should not be honored. But that is what we are doing now, by maintaining the debt and expecting that debtors should repay debts that should never have been issued in the first place.

The consequences of our current behavior are twofold. First, the economy will be enc.u.mbered by a debt burden that should never have been generated, and will limp along for a decade or more, as has j.a.pan. Secondly, the financial sector will continue to believe that 'the Greenspan Put' will absolve them from the consequences of irresponsible lending.

A debt jubilee would address both those consequences. First, debt repayments that are hobbling consumer spending and industrial investment would be abolished; secondly, this would impose the pain of bankruptcy and capital loss on the financial sector a pain it has avoided in general thus far through all the rescues since Greenspan's first back in 1987.

Needless to say, this would not be an easy policy to implement.

Its biggest hurdle would be political: it is obvious that the major political force in the USA today and much of the OECD is the financial sector itself. Since widespread debt abolition would bankrupt much of this sector, and eliminate individual fortunes (those that have not already been salted away), it will be opposed ferociously by that sector.

The same was the case though on a smaller scale than today during the Great Depression. It took a Ferdinand Pecora (Perino 2010) to turn the tide against the bankers then, and a Franklin Roosevelt (Roosevelt 1933) to convert that tide into political power and policies that included debt moratoria.

The recent Financial Crisis Inquiry Commission (Financial Crisis Inquiry Commission 2011) was a farce compared to Pecora's work, and Obama's administration to date has focused more on returning the financial sector to its old ways than on bringing it to account.

The policy would also need to re-establish the practice of banking providing working capital and investment funds for industrial capitalism. This should be the primary role of banking, but it virtually died out as the financial sector became more and more an engine for speculation, so that most companies today raise their funds on the commercial paper market.16 A debt jubilee would bankrupt many banks, and put them into receivership; though it would be painful, the receivers could also be required to re-establish this key but neglected banking practice.

It would also be necessary to compensate to some extent those not in debt as well though they would also benefit from the sudden increase in spending power that such a policy would cause.

Such a policy would have to be accompanied by inst.i.tutional reforms to finance that prevented a travesty like the Subprime Bubble from recurring; I discuss some possible reforms at the end of Chapter 14. It would also be far from a panacea for America's woes on its own, since it would also expose the extent to which the gutting of American industry in the last three decades has been disguised by the growth of the financial sector on the back of the Ponzi schemes of the stock and housing markets. The finance sector would shrink dramatically, and unlike in the 1930s, there would not be potential factory jobs awaiting unemployed financial advisors.

A debt jubilee, and the reforms I suggest in Chapter 14, is politically improbable now. But the alternative I believe is a decade or more of economic stagnation. At some stage we are going to have to accept the wisdom in Michael Hudson's simple phrase that 'Debts that can't be repaid, won't be repaid.'

Conclusion The data on debt confirm the conclusions that can be reached from a.s.sessing the logical coherence or lack of it in neocla.s.sical theory: every methodological choice neocla.s.sical economics made was wrong. The belief that economics can be reduced to microeconomics is false; money and credit cannot be ignored, capitalism cannot be modeled as a single 'representative agent,' finance destabilizes the economy, and the economy is permanently in disequilibrium.

If we are to develop an economics that is relevant to capitalism, then it must be a strictly monetary, dynamic theory in which finance plays a fundamentally destabilizing role. In the next chapter, I show how such an economic theory can be developed, by building on the work of both the great non-neocla.s.sical economists and recent empirical work by economists from the 'post-Keynesian' school of thought.

14 | A MONETARY MODEL OF CAPITALISM.

Many of the foundations on which neocla.s.sical macroeconomics is built arose from persevering with methodological choices that the nineteenth-century founding fathers of neocla.s.sicism made out of expediency rather than preference. They a.s.sumed that all economic processes occurred in equilibrium, so that they could model the economy using comparative statics rather than using more difficult dynamic differential equations; they avoided thinking about money and modeled the simpler process of barter instead; they ignored uncertainty about the future and, as Keynes put it, tried to 'deal with the present by abstracting from the fact that we know very little about the future' (Keynes 1937: 215) and so on.

Though these choices made it easy to concoct simple parables about supply and demand, they actually made mathematical modeling of the economy harder, not easier. The absurdities that later neocla.s.sicals added from the fallacy of the horizontal demand curve to the intellectual travesty of the 'representative agent' were products of clinging to these simple parables, despite the deep research that contradicted them.

Economists trained on these methods are now scrambling to make ad hoc modifications to the core neocla.s.sical parable to produce hybrid models that mimic the real-world phenomenon of the Great Recession which, according to the parables, cannot occur.1 Though such models will superficially ape reality, they will do so for the reasons that Solow gave, that the addition of various 'imperfections' results in a model that 'sounds better and fits the data better' simply because 'these imperfections were chosen by intelligent economists to make the models work better [...]' (Solow 2001: 26).

This is the difficult road to relevance take a theoretical framework in which the real-world phenomenon you are trying to describe cannot happen, and tinker with it until something resembling reality emerges. It will not last. Once the global economy emerges from this crisis, if this approach still dominates economics, then within decades these 'imperfections' will go the way of the dodo. Economists will return to the core parable, and the crisis we are now in will be seen as the result of bad Federal Reserve policy,2 rather than a manifestation of capitalism's innate instability amplified by a finance sector that is almost designed to generate Ponzi schemes.

We have to do better than that. We have to start with foundations from which the phenomena of reality emerge naturally by constructing monetary models of capitalism built on the melded visions of Marx, Schumpeter, Keynes and Minsky.

Methodological precepts An essential first step towards a meaningful macroeconomics is to acknowledge the one profound lesson from the failure of the neocla.s.sical experiment: that strong reductionism is a fallacy. Macroeconomic phenomena and even phenomena within one market are emergent properties of the dynamic, disequilibrium interactions of individuals and social groups in a rich inst.i.tutional environment, constrained by the physical, temporal and environmental realities of production. These phenomena will not be predictable from the behavior of isolated individuals. Instead, macroeconomics is a self-contained field of a.n.a.lysis, and it must be reconstructed as such. The reductionist route must be abandoned.

There are basically two routes by which models of a new 'emergent phenomena' macroeconomics could be built: the 'bottoms-up' approach that has always dominated economics, but modified in the light of the modern knowledge of complex systems; or the 'tops-down' approach that typified the work of Marx, Schumpeter, Keynes and Minsky, in which the economy is described at the level of aggregates evolutionary change, social cla.s.ses, aggregate production, aggregate debt levels and so on.

The former approach takes the macroeconomic phenomena as given, and attempts to build computer-based multi-agent models in which those macroeconomic phenomena arise as emergent properties of the models. The latter works at the level of aggregates, and puts the verbal models of the great non-neocla.s.sical thinkers into the form of dynamic equations.

Most economists who are trying to build macroeconomic models that transcend the neocla.s.sical dead end are taking the former approach (Barr, Ta.s.sier et al. 2008; Seppecher 2010).3 This approach is worthwhile, though there are inherent difficulties in it that I discuss briefly later. I have taken the latter approach of trying to put the Marx-Schumpeter-Keynes-Minsky vision directly into mathematical form.

Doing this turned out to be far easier than I expected, once I made money the starting point of my a.n.a.lysis of capitalism.

Endogenous money One of the many issues on which Keynes failed to convince his fellow economists was the importance of money in modeling the economy. One reason for this was that money's explicit role in the General Theory itself was restricted largely to the impact of expectations about an uncertain future, and the difference between real and nominal wages. Keynes acknowledged that money did not feature heavily in his technical a.n.a.lysis, and that he saw a substantial continuity between monetary a.n.a.lysis and the Marshallian model of supply and demand: whilst it is found that money enters into the economic scheme in an essential and peculiar manner, technical monetary detail falls into the background. A monetary economy, we shall find, is essentially one in which changing views about the future are capable of influencing the quant.i.ty of employment and not merely its direction. But our method of a.n.a.lyzing the economic behavior of the present under the influence of changing ideas about the future is one which depends on the interaction of supply and demand, and is in this way linked up with our fundamental theory of value. We are thus led to a more general theory, which includes the cla.s.sical theory with which we are familiar, as a special case. (Keynes 1936: xxiixxiii; emphases added) It is therefore difficult to attack neocla.s.sical 'supply and demand'-oriented models of money as misrepresentations of Keynes. Nonetheless, the post-Keynesian school of thought has made the fundamental importance of money a byword of its a.n.a.lysis. An essential aspect of this has been the empirically based a.n.a.lysis of how money is created (detailed in the previous chapter), which contradicts the conventional fractional reserve banking, 'Money Multiplier' model of money formation.

Having empirically eliminated one model of money creation, another was needed but the initial attempts to create one were clumsy. Rather than the 'vertical money supply curve' of Hicks's IS-LM model, some post-Keynesian economists proposed a 'horizontal money supply curve' in which banks simply pa.s.sively supplied whatever quant.i.ty of credit money firms wanted, at the prevailing interest rate. This model, known as 'Horizontalism' (Moore 1988b), led to a lengthy dispute within post-Keynesian economics over whether the money supply curve was horizontal, or sloped upwards (Dow 1997).

This dispute put the empirically accurate findings of post-Keynesian researchers into the same methodological straitjacket that neocla.s.sical economics itself employed: the equilibrium a.n.a.lysis of intersecting supply and demand curves. Though this was hardly the intention of the originators of endogenous money a.n.a.lysis, it effectively made monetary a.n.a.lysis an extension of supply and demand a.n.a.lysis.

Partic.i.p.ants in this debate were aware of the limitations of this approach as Sheila Dow observed, '[T]he limitations of a diagrammatic representation of a non-deterministic organic process become very clear. This framework is being offered here as an aid to thought, but it can only cope with one phase of the process, not with the feedbacks' (ibid.: 74). But one of the great ironies of economics is that, because critics of neocla.s.sical economics were themselves trained by neocla.s.sical economists, most critics weren't trained in suitable alternative modeling methods, such as differential equations or multi-agent simulation.

For real a.n.a.lytic progress to be made, a watertight basis for Keynes's a.s.sertion that money 'enters into the economic scheme in an essential and peculiar manner' was required, as well as a methodological approach that captured the feedback effects that diagrams and equilibrium a.n.a.lysis could not.

The former was supplied by the 'Monetary Circuit' school in Europe, and specifically the Italian economist Augusto Graziani. Graziani argued that, if money is treated as just another commodity subject to the 'laws' of supply and demand, then the economy is effectively still a barter system: all that has happened is that one more commodity has been added to the mix, or singled out as the commodity through which all barter must occur. This is quant.i.tative change, not qualitative, and yet something qualitative must change if a monetary economy is to be distinguished from a barter system.

Graziani's brilliant insight was that, for a monetary economy to be clearly distinguished from a barter economy, the monetary economy could not use a commodity as money. Therefore money had to be a non-commodity something that was intrinsically worthless, and which could not be simply produced as commodities themselves can: 'a commodity money is by definition a kind of money that any producer can produce for himself. But an economy using as money a commodity coming out of a regular process of production, cannot be distinguished from a barter economy' (Graziani 1989: 3). This then led to a simple but profound principle: 'A true monetary economy must therefore be using a token money, which is nowadays a paper currency' (ibid.: 3).

The fact that a monetary economy uses a token something that is intrinsically worthless as a means of exchange implies two further key conditions 'In order for money to exist': b) money has to be accepted as a means of final settlement of the transaction (otherwise it would be credit and not money); c) money must not grant privileges of seigniorage to any agent making a payment. (Ibid.: 3) From this Graziani derived the insight that 'any monetary payment must therefore be a triangular transaction, involving at least three agents, the payer, the payee, and the bank': The only way to satisfy those three conditions is to have payments made by means of promises of a third agent, the typical third agent being nowadays a bank [...] Once the payment is made, no debt and credit relationships are left between the two agents. But one of them is now a creditor of the bank, while the second is a debtor of the same bank. (Ibid.: 3; all emphases in original) 14.1 The neocla.s.sical model of exchange as barter This perspective clearly delineates a monetary vision of capitalism from the neocla.s.sical barter paradigm. As shown in Figure 14.1, in the neocla.s.sical world, transactions are two-sided, two-commodity, barter exchanges: person A gives person B one unit of commodity X in return for some number of units of commodity Y. Calling one of these 'the money commodity' does not alter the essentially barter personality of the transaction.

But in our monetary world, transactions are three-sided, single-commodity, financial exchanges, as portrayed in Figure 14.2: person B instructs bank Z to debit Y units of currency from B's account, and credit A's account with the same amount, in return for which person A gives person B one unit of commodity X.

Banks are thus an essential component of capitalism, and are inherently different to industrial firms. Firms produce goods (and services) for sale by combining labor and other commodities in a production process that takes both time and effort. Banks generate and honor promises to pay that are used by third parties to facilitate the sale of goods.4 Therefore firms and banks must be clearly distinguished in any model of capitalism: 'Since in a monetary economy money payments necessarily go through a third agent, the third agent being one that specializes in the activity of producing means of payment (in modern times a bank), banks and firms must be considered as two distinct kinds of agents [...] In any model of a monetary economy, banks and firms cannot be aggregated into one single sector' (ibid.: 4; emphasis in original).

14.2 The nature of exchange in the real world This simple but profound perspective on what is the essence of a monetary capitalist economy yielded two essential requirements for a model of capitalism: * all transactions involve transfer of funds between bank accounts; * the minimum number of cla.s.ses5 in a model of capitalism is three: capitalists, workers and bankers.

It also implied that the best structure for modeling the financial side of capitalism is a double-entry system of bank accounts. This led me to develop a means to derive dynamic monetary models of capitalism from a system of double-entry bookkeeping accounts (Keen 2008, 2009b, 2010, 2011), and a remarkable amount of the Marx-Schumpeter-Keynes-Minsky perspective on capitalism arose naturally out of this approach.

I'll outline the simplest possible version of this model before expanding it to provide a monetary version of the Minsky model outlined in Chapter 13.

A 'pure credit' economy Our modern monetary economy is a system of such complexity that it makes the outrageous contraptions of Rube Goldberg, Heath Robinson and Bruce Petty appear trite by comparison: the Bank of International Settlements, central banks, commercial banks; merchant banks, hedge funds, superannuation funds, building societies; fiat money, credit money, multiple measures of money (base money, M0, M1, M2, M3, broad money); reserve ratios, Taylor Rules, Basel Rules ...

Many of these components were inst.i.tuted to try to control bank lending after the catastrophe of the Great Depression; many others were responses by the financial system to evade the intentions of these controls. To my cynical eye, the evasive maneuvers of the financial system have been far more effective than the regulatory structures themselves, and in essence our financial system approximates the behavior of the almost completely unregulated private banks of the 'free banking' period in the nineteenth century.

14.3 A nineteenth-century private banknote For that reason, my base monetary model is a pure credit economy with no government or central bank, in which the private bank prints its own paper notes, and where transactions involve transferring paper notes from the accounts of the buyers to that of the sellers. There are three cla.s.ses workers, capitalists and bankers and, in the simplest possible model with no Ponzi lending behavior, firms are the only borrowers, and they borrow in order to be able pay the wages needed to hire workers.

Five accounts are needed to describe the basic monetary flows in this system: 1 a vault, in which the bank stores its notes prior to lending; 2 a 'bank safe,' into and out of which interest payments are made; 3 deposit accounts for firms, into which money lent by the banks is put and through which all the firm sector's transactions occur; 4 deposit accounts for workers, into which their wages are paid; and 5 a loan register, which is not an account as such, but a ledger that records the amounts that have been lent by the banks to firms, and on which loan interest is charged.

The basic monetary operations that occur in this simple model are:6 1 the banking sector makes loans to the firm sector; 2 the banks charge interest on outstanding loans; 3 firms pay the interest; 4 firms hire workers; 5 workers and bankers consume the output of the firms; and 6 firms repay their loans.

These operations are shown in Table 14.1, which (based on the standard accounting practice of showing 'a.s.sets minus liabilities equals equity') shows the economy from the point of view of the banks, with the banking sector's a.s.sets on the left-hand side of the ledger and its liabilities and residual equity on the right-hand side.7 Actual transfers of money are shown in normal text, while operations that are not money transfers but accounting operations such as the bank recording that interest due on loans has been paid are shown in italics.

TABLE 14.1 A pure credit economy with paper money Since all the entries in this table indicate flows into and out of accounts (or additions and subtractions from the loan ledger), a remarkable thing is possible: a dynamic model of this monetary model can be derived just by 'adding up' the entries in the columns, as in Table 14.2.

TABLE 14.2 The dynamics of a pure credit economy with no growth This model can be simulated if we put values on these flows. Some of these are obvious: the interest charged, for example, will equal the rate of interest on loans times the amount currently recorded on the loan ledger; interest paid is the rate of interest on deposits times the amount currently in the firms' deposit accounts.8 Others lending from the vault, payment of wages, consumption by workers and bankers and loan repayment will in the real world depend on a whole host of factors, but to model the simplest possible system, I relate them here to the balances in these other accounts, and use constants rather than variables simply to see whether the model is viable: obviously, if it's impossible to find a set of constants that makes this model viable, then no set of variables is likely to do it either.

Thus lending from the vault is modeled occurring at some constant rate times the amount of money in the vault; the flow of wages is some constant times the balance in firms' deposit accounts; workers' and bankers' consumption depend on the balances in the workers' deposit accounts and the safe respectively; while the flow of loan repayments is some constant times the amount of loans outstanding.

The constants (known as 'time constants' in dynamic modeling)9 used tell us how many times in a year the given account will turn over so a value of , for example, indicates that the balance in the relevant account will be turned over every two years. One obvious value here is that for workers' consumption: since workers' wages are paid on a weekly basis, and most of workers' incomes is expended on consumption, the constant for workers' consumption will be 26 indicating that the balance in the workers' accounts turns over twenty-six times a year. For the sake of ill.u.s.tration, I use for lending money (so that the vault turns over every two years), 3 for wages, 1 for bankers' consumption, 26 for workers' consumption, 1/10 for loan repayment, and I set the rate of interest on loans to 5 percent and the rate of interest on deposits to 2 percent.

If the model starts with $100 million initially in the vault and no money in any other account, then after ten years, the amount in the vault falls to $16.9 million, with $83.1 million in outstanding loans, $2.7 million in the safe, $72.1 million in firm deposit accounts, and $8.3 million in the workers' deposit accounts see Figure 14.4.10 It is also possible to calculate the annual wages bill and bank earnings. The annual wages bill is the time constant for wage payments times the balance in the firms' deposit account, which is three times $72.1 million or $216.3 million, while bank gross earnings are the rate of interest on loans times the outstanding loan balance (5 percent times $83.1 million or $4.16 million) minus the rate of interest on deposits times the firms' deposit balance (2 percent times $72 million or $1.44 million), for a net bankers' income of $2.7 million per annum.

Capitalists' income isn't as obvious in this simple model, and to explain it properly will require incorporating production and pricing as well. But we can imply what profits are by realizing that net annual income in this simple model equals the sum of wages plus profits the income of bankers cancels out and adds nothing to aggregate income (see Table 14.3).

TABLE 14.3 Net incomes Since wages represent part of the net surplus generated in production, profits must represent the remainder. If workers' wages represent, say, 75 percent of net income, then profits represent 25 percent so in this numerical example they equal $72.1 million.11 14.4 Bank accounts Annual income in this example is thus $288.4 million almost three times the amount of money in the model, and precisely four times the amount of money in firms' deposit accounts. How can this be? Marx's insight into why Say's Law is invalid in a capitalist economy holds the key. Remember that Say's Law holds under simple commodity production (CommodityMoneyCommodity), but not in capitalism, because that also has the circuit MoneyCommodityMore Money. Marx also pointed out that this 'Circuit of Capital' takes time: it involves getting money in the first place, using it to hire workers and buy inputs, combine them in a production process, ship the finished goods and finally sell them to customers. There is thus a time lag between outlaying M and earning M+, which Marx called the 'period of turnover.' This can be significantly shorter than a year, though it's highly unlikely to be as short as the example Marx himself gave: 'Let the period of turnover be 5 weeks, the working period 4 weeks [...] In a year of 50 weeks [...] Capital I of 2,000, constantly employed in the working period, is therefore turned over 12 times. 12 times 2,000 makes 25,000' (Marx 1885: ch. 16).

Expressed as a fraction of a year, Marx's example gives a value of 1/12.5 for the period of turnover and in general, the smaller the number, the faster a given amount of money turns over, and the more profit (and wages) that can be generated. Marx's numerical example was extreme, but the basic insight is correct, that a given sum of money can finance several times as much turnover in a given year.

The period of turnover can also be derived for our example, using the facts that the value of the time constant for wages is 3, and 75 percent of national income goes to workers as wages. Total income wages plus profits is thus four times the amount of money in the firms' deposit accounts. The turnover period is therefore one year divided by 4: it takes three months, in this toy economy, to go from M to M+.

Though the turnover period is an unfamiliar concept, it's related to the well-known if less well-defined concept of the velocity of money. The turnover period tells us how often the money in firms' deposit accounts turns over; the velocity of money in this model is the value of wages plus profits (GDP, which is $288.4 million in this example) divided by either the total money supply ($100 million) or the money in active circulation, which is the sum of the amounts in the deposit accounts plus the safe ($83.1 million). Measured the former way, the velocity of money is 2.88; measured the latter way, it's 3.47.