07.10.2021

Marxist monetary theory

It is profit, not government spending, that drives capitalist investment, writes Michael Roberts

Bill Mitchell - a professor in economics and director of the Centre of Full Employment and Equity, at the University of Newcastle, NSW, Australia - is one of the world’s leading exponents of what is called Modern Monetary Theory.

MMT has gained traction in the labour movement in recent years on the grounds that it provides powerful new arguments to refute the claims of mainstream economics that governments need to balance their books: ie, keep spending in line with tax revenues and not allow government debt to spiral. In order to balance the books, so the mainstream argument goes, government spending must be cut, taxes raised and debt levels reduced, even if that means more poverty and worse public services.

However, MMT is supposed to offer the answers to the Austerians, both theoretically and in alternative policies and Mitchell, among other MMT supporters, has tirelessly campaigned for the adoption of MMT measures in the labour movement as the key answer to ending unemployment and austerity.¹

I have spent much ink on my blog and in papers discussing and debating the merits of MMT as the answer to capitalist policies of austerity and unemployment.² In my humble opinion, it falls short of achieving its claims and objectives because it ignores the social structure of a capitalist economy and argues that the recognition and manipulation of the monetary system can solve the problems without ending capitalism itself.

At a recent fringe meeting at the Labour Party conference, I was invited to debate the merits of MMT with professor Mitchell, who was contributing via Zoom from Australia. Shortly after the debate, he posted the following on his blog:

I gave a talk at the Resist event in Brighton UK last Sunday evening. On the panel was a person who dismissed Modern Monetary Theory (MMT) as irrelevant to the real challenges that arose under capitalism and he invoked Marx a lot. It was not a very illuminating interchange, because not only did he misrepresent what MMT was, but, in my view, he also seemed to think that we could extrapolate Marx’s scant ideas of money directly into the situation faced by nations today.³

In his posts, professor Mitchell seeks to dismiss my arguments that MMT does not take us very far in explaining the nature of a capitalist economy or providing decisive economic policies that will help labour. He thinks that I have not only “misunderstood MMT”, but also have “a scant understanding of Marx’s monetary theory”. He says Marx would never have agreed that government spending could not end unemployment or that capitalists would not increase production and employment if government spending orders came in.

In the debate I said that Marx attacked the idea that, just by manipulating money, governments could solve unemployment and poverty. This was the view of Pierre Proudhon, the leading socialist of the mid-19th century. Marx responded:

Can the existing relations of production and the relations of distribution which correspond to them be revolutionised by a change in the instrument of circulation, in the organisation of circulation? Further question: Can such a transformation of circulation be undertaken without touching the existing relations of production and the social relations which rest on them?⁴

Out of date?

In his post, Mitchell refers to professor Duncan Foley (at least he gets named!), who asserts that, for Marx,

... the movement of commodities is largely determined outside the monetary sphere, and that movements of money in most cases are determined by those commodity movements ... The theoretical question then arises as to which is the determining factor. Does the movement of money determine the movement of commodities or the movement of commodities determine the movement of money?⁵

Mitchell rejects Foley’s approach to money and reckons that Marx is out of date with this critique of Proudhon because back then money was gold or backed by gold, which meant its value was anchored to the production cost of gold mining. Now, in the modern world of ‘fiat currencies’, the state can create money that is not tied to the value or price of gold and therefore the state can increase the amount of money at will.

I do not agree that this makes Marx out of date. Fiat money does not change the role or nature of money in a capitalist economy - its value is still tied to the labour time taken in capitalist accumulation. Commodity money (gold) contains value, while non-commodity money represents/reflects value, and because of this both can measure the value of any other commodities and express it in price form. Modern states are clearly crucial to the reproduction of money and the system in which it circulates. But their power over money is quite limited - the state mint can print any numbers on its bills and coins, or the central bank add any number of digits to the government’s bank account, but that cannot decide what those numbers refer to. That is determined in countless price-setting decisions by mainly private firms, reacting strategically to the structure of costs and demand they face, in competition with other firms.

MMT supporters in effect reckon that any reduction in private-sector investment can be replaced or added to by government investment ‘paid for’ by the creation of money out of thin air. But this money will lose its value (ie, purchasing power) if it does not bear any relation to value created by the productive sectors of the capitalist economy, which still dominate the economy. Instead, the result will be rising prices and/or falling profitability that will eventually choke off production in the private sector. Unless MMT proponents are then prepared to move to a Marxist policy conclusion - namely the appropriation of the finance sector and the ‘commanding heights’ of the productive sector through public ownership and a plan of production, thus curbing or ending the law of value in the economy - the policy of government spending through unlimited money creation will fail.

As far as I can tell, MMT exponents studiously avoid and ignore such a policy conclusion - perhaps because like Proudhon they misunderstand the reality of capitalism, preferring ‘tricks of circulation’; or perhaps because they actually oppose the abolition of the capitalist mode of production (indeed, that is the view of most MMT exponents, although not professor Mitchell).

Mitchell goes on to highlight something in one of my blog posts on MMT. I had a plotted figure that shows there is no inverse correlation between increased government spending and unemployment.⁶ On the contrary, among OECD countries, the higher the level of government spending, the higher unemployment!

Mitchell now applied his superior understanding of statistics to this result: “In Statistics 101 or Econometrics 101, one of the first things students learn is to be careful in attributing causality. Correlation is not causality.” Yes, this correlation does not prove causation. Indeed, that is what I conclude in the post: namely that there must be other reasons than the amount of state spending to explain unemployment rates. What drives unemployment rates in a capitalist economy is the rate of investment and employment made by the capitalist sector. When there is a slump in investment and people are laid off to join what Marx called “the reserve army of labour”, government spending rises, as benefit spending increases. Having said that, there is no evidence that unemployment falls when government spending rises, because there are much more compelling causes.

In his second post, Mitchell claims that he can show “how nonsensical it is to claim that capitalist firms will not expand production if they have idle capacity and can increase profits by responding to increased sales orders”.⁷ And that “the government sector is not bound by the so-called dynamics of private capital accumulation and under certain conditions can typically command productive resources from the non-government sector through increased spending without introducing inflationary pressures”.

He takes us through a short course on the history of crisis theory, from the views of Say and Ricardo that general overproduction was impossible, to the underconsumption theories of Sismondi, Luxemburg and others. And then he says: “Marx considered the accumulation process would lead to an excessive build-up of capital which would suppress the rate of profit and it was this dynamic that generated the crisis.” Well, yes, that was Marx’s theory of crises, more or less in a nutshell.

Having said this, however, Mitchell quickly takes a step back:

But we need to be careful in unpicking the logic here. Yes, the owners of capital control production and employment and their expectations of future returns dictate the rate at which the capital stock accumulates over time. But equally, when considering the causes of crises, we cannot avoid focusing on the realisation issue, because it was through market exchange that capitalists were able to realise the surplus value they had expropriated in the production process by exploiting their workers into the monetary form of profit.

Mitchell then claims Marx for his own theory of crises: “As we move through history, the scholars that followed Marx clearly understood that effective demand was a causal factor in determining unemployment and recession.” So there we have it. Forget Marx’s profitability theory as the driver of accumulation and thus investment demand. Let’s revert to the same position as orthodox Keynesians: that crises are due to a lack of ‘effective demand’.

Contradictions

Now I have spent much time showing that the Keynesian-Mitchell theory of lack of effective demand is an inadequate explanation of regular and recurring crises in capitalist production.⁸ But here let me mention just one point. Why does capitalist production seem to have enough effective demand for years or even a decade, and then suddenly investment and production collapse, unemployment rockets and there is then a ‘lack of effective demand’? MMT has no answer to this question. The answer lies in the contradictions within the capitalist mode of production: specifically in the tendency for the profitability of capital to fall over time and eventually lead to a fall in total profits and value creation. Then investment demand collapses, leading to a lack of aggregate demand, so that capitalist production cannot be ‘realised’. This is the causal sequence in crises.

Mitchell invokes the post-Keynesian arguments of Michał Kalecki to justify his view that it is aggregate demand that drives sales, production and profits. He states that “to offset any tendency for the rate of profit on the expanded capital stock to fall and thus offset the possibility of a crisis”, Kalecki wrote: “… if effective demand adequate to secure full employment is created by stimulating private investment, the devices which we use for it must cumulatively increase to offset the influence of the falling rate of profit.”

Talking again of correlation, identities and causation, there is the Kalecki identity (investment = profits); and there is the correlation (investment moves with profits); but which is the causal direction?⁹ Kalecki argues that investment drives profits, but this is back to front. In my view, Marx correctly argued that in a capitalist economy profits drive investment, not vice versa, as Kalecki argues. Profits call the tune, not investment or consumption. That is why boosting government spending, either by traditional borrowing methods or by ‘printing money’ will not guarantee expansion of employment or faster growth.

With government investment averaging just 2%-3% of GDP in most major economies and capitalist investment averaging 15%-20%, it is going to take a huge jump in government spending to replace capitalist investment - indeed to the point of ending the dominance of capitalist investment entirely.

But is there any empirical proof that profits drive investment and it is not the ‘lack of effective demand’ that leads to slumps? Yes, there is. Let me just cite one study by José Tapia: Investment, profits and crises: theories and evidence, chapter 3.¹⁰ Is there any empirical evidence showing that increased government spending has little or no effect on boosting ‘aggregate demand’? Again, José Tapia has done the stats.¹¹ These are not correlations, but applied causal analysis.

Tapia finds that “overall these results seem quite inconsistent with the hypothesis that an increase in government spending will pump-prime the economy by raising private investment”. So that “the Keynesian view ... is also inconsistent with the finding that the net effect of lagged government expending on private investment is rather null or even significantly negative in recent decades”.

Professor Mitchell is convinced that “capitalist firms will respond to increased sales demand by producing goods and services. If they think the demand is stable, they will invest and build productive capacity if they are currently at full capacity.” But, just in case, “if they decide for any reason not to respond, then the government can always employ and produce itself”. So we have a sort of two-stage policy: first we must break with the policies of austerity coming from the mainstream by adopting MMT policies; and afterwards, because it is a separate issue, we must look at changing the social structure as good socialists.

I take this to mean that in applying MMT to government policies, we can ignore (for now) the continuance of fossil fuel production, big pharma control on health, the reckless greed of capitalist banking, the dominance of the tech and media monopolies. These niceties of capitalism are irrelevant to the aims of MMT, which has nothing to say on these economic formations. But precisely because MMT ignores that very social structure, its pursuit of achieving full employment through the ‘tricks of circulation’ of money will fail.

Michael Roberts blogs at thenextrecession.wordpress.com