A Marxist theory of inflation

In my previous post on inflation, I spelt out why mainstream theories of inflation have been proved wrong empirically; leaving mainstream economics in a confusion about what just does drive inflation in the prices of goods and services.  In this post, I want to argue that mainstream theories of inflation falter because they are not based on the law of value that operates in the capitalist mode of production.  Both the monetarist and Keynesian theories fail because of this.

Marx opposed both these mainstream theories. The quantity theory of money was opposed by Marx for two reasons: 1) money is endogenous, created by banks etc, not by state fiat; 2) overall, money represents value in commodity production and is not independent of it.

So returning to the quantity theory of money equation, MV=PT (see the previous post); for Marx, the basic causal direction is from PT to MV, not the other way (ie from prices to money, not money to prices).  Money is endogenous to capitalist production and prices of production are formed from value creation not from money creation. Money supply generally will follow price changes, so deliberate attempts to alter the money supply will fail to determine price inflation.

Cost-push theories were also rejected by Marx because wage rises do not cause price rises. As Marx put it in Value, Price and Profit, when he debated with trade unionist Weston who argued that wage rises would cause inflation: “a struggle for a rise of wages follows only in the track of previous changes, and is the necessary offspring of previous changes in the amount of production, the productive powers of labour, the value of labour, the value of money, the extent or the intensity of labour extracted, the fluctuations of market prices, dependent upon the fluctuations of demand and supply, and consistent with the different phases of the industrial cycle; in one word, as reactions of labour against the previous action of capital (my emphasis).”

“By treating the struggle for a rise of wages independently of all these circumstances, by looking only upon the change of wages, and overlooking all other changes from which they emanate, you proceed from a false premise in order to arrive at false conclusions.”  Broadly speaking, argued Marx, “A general rise in the rate of wages would result in a fall of the general rate of profit, but not affect the prices of commodities.”

Marx never developed a comprehensive theory of inflation, but can we develop one based on Marx’s value theory?  Italian Marxist economist, Guglielmo Carchedi has come up with one. His work will be fully published later this year.  But let summarise his main arguments.

Capitalist production continually strives to increase the productivity of labour ie produce more units per worker.  But this means that the labour time per unit will fall.  As only labour creates value, while there is a general tendency for the supply of units of goods and services to rise, there is also a general tendency for the value of commodities to fall, over the long term. This is because capitalist accumulation is a labour-saving process, so the value of commodities will fall alongside a rise in the productivity of labour.  Use values are produced at greater amounts than the value contained in them.  So, if prices of production depend on value, there is an inherent tendency for prices of commodities to fall not rise. as total value will fall relatively to total production over time.

The demand for commodities depends on the new value created in production.  New value commands the demand or purchasing power over the supply of commodities.  New value is divided by the class struggle into wages and profits.  Wages buy consumer goods and profits buy capital or investment goods.

But new value will tend to decline: first, because total value declines relatively to the supply of commodities…

Source: author’s calculations from NIPA GDP data

… and second because of the rising organic composition of capital (c/v).  Capitalist accumulation is labour-saving, so the value of machinery, plant, and raw materials etc (c) will tend to rise relative to the value of labour power (v).  As the price of production in value terms is made up of constant capital (c) and new value (v+s), a rising c/v will tend to reduce the share of new value in the price of production.

Source: author’s calculations

Total value will decline relatively to use value production and new value will decline relatively to total value.  So there is an underlying deflationary or disinflationary pressure on the prices of commodities over the long term.

But there are counteracting factors that can exert an upward pressure on prices over the long term; in particular, the intervention of the monetary authorities with their attempts to control the supply of money.

Carchedi’s theory of inflation is that there is a value rate of inflation (VRI), which combines the impact of changes in the purchasing power of wages and profits (new value) and the money supply, measured as cash deposits in banks (M2).  The former factor is the determining one and will tend to drive price inflation down, while the latter is the counteracting factor that will tend to push inflation up, but with no permanent success.

The value rate of inflation (VRI) = % change in wages and profits (CPP) + % change in money supply (M2).  Using data from the US since 1960, we find that the VRI falls over the long term.  This is because the combined purchasing power (CPP) of wages and profits grows more slowly and any changes in money supply (M2) have been insufficient to stop the VRI slowing.

Source: author’s calculations

But is there a close correlation between the VRI and consumer price inflation?  Yes.  Between 1960 and 1979, the VRI rose and so did US CPI inflation; between 1980 and 2019, the VRI slowed and so did CPI inflation.

Indeed, if we model the VRI forecasts for each year against actual CPI inflation, there is a close correlation over the long term.  In the graph below, the VRI inflation model forecast for US consumer price inflation (orange lines) is a pretty good fit for actual CPI inflation (blue lines).  It offers a much better result than monetarist forecasts or the Phillips curve, especially since the early 1990s, the period of so-called disinflation that has puzzled monetarists and Keynesians.

The value theory of inflation thus explains the slowdown in annual consumer price inflation since the 1980s, unlike mainstream theories which are nonplussed.  Even though central banks pumped more money into the economy and M2 money supply growth accelerated, especially from the 1990s and after the Great Recession, because new value growth kept slowing, the slowing combined purchasing power of wage and profit growth continued to drive down inflation.

Can we forecast where inflation is going in the COVID and after?  If Carchedi’s theory is right, then whether inflation returns after the COVID depends your forecasts for new value and M2 money growth and thus on the forecast for the value rate of inflation.  When he reads this, Carchedi will complain that the value theory of inflation is long term and cannot be used to forecast inflation over a few years or less.  But nevertheless, let’s have a go.

This year, 2020, has seen a huge rise in M2 money supply, up 25% yoy so far. But we can expect a fall in profits of about 25% and in wages of about 20% – so a big drop in the combined purchasing power of new value.  The VRI model translates into US consumer price inflation this year of about 0.5-1.0%, an annual rate not seen since the depth of the Great Recession.  Currently US CPI annual inflation is at 1.0% in July after falling to 0.7% in June.

If we assume that in each of the two next years, 2021 and 2022,the nominal wage bill rises by 5% and profits rise by 10% and 15% respectively, while M2 money growth slows to 10% a year, then the VRI model forecasts 3.0-3.5% US CPI annual inflation over the next two years, not deflation as some expect.

Of course, that result depends on the assumptions.  More important, what the value theory of inflation shows is that mainstream theories of inflation fail because of their ignorance of value theory.  Once changes in value, not money or employment, are analysed, we can understand the trajectory of inflation under capitalist production.

This post in no way covers all the points and arguments in the Value Theory of Inflation.  They will be developed in detail in an upcoming academic paper and as part of the jointly authored book, Through the Prism of Value that Carchedi and I will publish next year.

44 thoughts on “A Marxist theory of inflation

  1. Marx’s rather simple theory of inflation was, that prices in gold are determined by the labour theory of value. This would also apply for paper money as long as paper money just replaces gold money. If more paper money is issued than the amount of gold money necessary for the circulation of the produced commodities, then prices expressed in paper money start to rise. This is then similar to the quantitiy theory of money.

  2. Great job. I think it will be more interesting to integrate Chris Harman’s theory of inflation in his article “Do wages causes inflation?” 1979 which relies on difference between transition from value to prices in more competitive non-monopoly and monopoly capitalism. And Also to look at the bigger picture in which less-developed countries experience higher rates of inflation in crisis as way of non-wage exploitation for monopolies while developed countries in crisis resulted similarly falling tendency of rate profit, suffer from deflation or lower rates of profit. I think it has some thing to do with super-profits in global economy.

    1. 1) Regardless of the concentration and centralization of capital, “monopoly” conditions simply have no relevance to the long and short-term movements of prices, and profits. If anything, competition is greater now, intra and inter sector; and internationally.

      2) re “Do wages cause inflation?” Didn’t Marx answer that over 150 years ago in his response to the esteemed Citizen Weston?

      3) re inflation in “less developed countries”– actually plays a role in separating the direct, small producers from the means of production.

    2. Farid Saberi is repeatedly calling for attention to differences in inflation dynamics between developed and less developed economies. I’d also like to emphasize it as well. All studies that I have read neglect the inflation dynamics in the so call third world. It appears as if VIR in less developed countries is mostly dependent on M2 and the new value component impact seems negligible. When central banks in those countries print money (increase M2) prices rise almost immediately. Why? I don’t know but allow me to bring to your attention the two departments 1 and 2 in capitalism. In advanced economies department 1 goods are both produced and consumed in those countries whereas in less developed countries it is mostly the consumption not so much production of capital goods. Department 1 goods are mostly imported from advanced economies. OCC of less developed tend to always be below those of the developed, a mechanism of transferring surplus value from less developed to developed which I believe Michael has alluded to in his previous posts. The other observation is the status of reserve currency. Countries with reserve currencies seem to have a different inflation dynamic than the rest.
      I don’t subscribe to super-profit, Harman’s inflation theory, non-wage exploitation, etc. But I do agree with Farid that there are differences in inflation dynamics.

  3. There seem to be some similarities but probably also differences to Anwar Shaikh’s “Capitalism”, p. 697ff.

    1. No Thomas, this differs substantially from Shaikh’s account.

      Carchedi and Roberts suggest that because prices reflect values, the level of prices will move with the level of value embodied in commodities. Given the trend in OCC, there will be a tendency for the price level to fall, barring counteracting influences.

      Shaikh says the price level is measured relative to the money commodity (gold). Since the value embodied in the gold commodity is moving along with all other commodities, there is no tendency for the price level to fall. At least not for the commodity money – fiat money is discussed separately.

      Shaikh presents data to argue the level of prices measured in gold has not moved secularly up or down over the past couple centuries (pp188-9, 199, 696). Instead it has fluctuated around a kind of centre of gravitation. This contradicts what you would expect if Carchedi and Roberts are right to suggest that CPI will track VRI – because we expect that VRI has a secular downward trend given OCC.

      For Shaikh, the era of secular inflation (as opposed to ups and downs in price level) begins when the international money system moves from gold standard to fiat money.

  4. An interesting approach, but I do not see any pressing need to anchor a theory of inflation in just one process. There seems to be a long-run effect (or habit) such as that described, but I am more skeptical about the order of all the “intermediate” factors given in the Marx quotation cited. And why should the first relationship (value production and the rate of exploitation) be taken as necessarily the 1st-order cause? Mr. Saberi’s comment makes me think of the exchange rate pass-through, which need not, for instance, be anchored in any particular value creation process — in the short- to medium-term anyway. So explain to me why one should believe that Carchedi’s approach should be preferred to viewing inflation rates as just a continuation/compilation over time of a diversity of possibly heterogeneous short-term effects? Clearly, there are occasions when inflationary expectations get embedded in workers and capitalists views as part of the class struggle, but at least on the workers’ side it is not clear that this is always the case.

    1. Very interesting post! But I have many questions coming to my mind. Leaving away expectations and exchange rate pass-through (regarding the latter I find Anwar Shaikh’s theory very convincing), I have several questions because I do not understand:
      1) How is newly created value calculated? Is it not that higher value creation involves more money because production has to be financed. Hence value creation should be a driver of M2. However, value creation should itself not be inflationary because more money is met by more output. Why then is inflation not M2 minus new value? I know this would be the old story again…
      2) How does Garchedi’s theory deal with the fact that M2 still does not tell anything about causality due to money endogeneity? There could still be demand pulls or cost pushes driving the amount of money. They then still would be the relevant factors.
      3) M2 still does not tell us how much of it is spent on financial markets vs. goods markets. How to include this?

      1. Good questions, Basil. Not sure that I can answer them all satisfactorily. 1) New value is created by labour power and then divided between wages and profits by class struggle. New value must be represented by money and so money is ‘created’ to match. But in a fiat currency economy and with the intervention of banks and central banks, money supply growth will not match new value growth. Carchedi’s theory is that new value growth tends to slow for reasons explained in my post, so there is a disinflationary or even deflationary pressure on prices. However, a rising money supply growth can counteract that. So we have a combination to give us VRI. 2) Money supply is still endogenous to accumulation and prices of production in value terms. 3) Good point, that shall be dealt with in the final paper.

  5. I have a paper that presents a simple theory of inflation: all else constant, if the rate of profit falls, the inflation rate. Obviously, the “all else constant” clause is crucial (as is the way that the profit rate is measured). Alas, I haven’t updated this research.
    “The Rise and Fall of Stagflation: Preliminary Results,” Review of Radical Political Economics, 32(3), 2000: 398-407.

  6. There is a very neat model that shows how inflation and full employment functioned together in Sweden on https://www.internetional.se/pgbhow.pdf
    As i have shown the result measured by workingtime/worker and workingtime/person was outstanding.
    I think the reason why inflation in itself is belived to create jobs is that the effects of taxation is forgotten.

  7. I don’t think Marxism needs a theory of inflation, but ok – it’s always nice to counter-attack and conquer terrain from vulgar economics.

    On a side note: we can demonstrate value has deflationary tendencies by analyzing the capitalist world when the gold standard existed alongside the industrial revolutions (mainly, 18th and 19th centuries’ Europe). Gold almost always tended to valorize, so the pressure was on reducing production (as gold cannot be printed), thus rising the value of the other commodities so prices could fall properly.

  8. My understanding is that the value of a good or service is determined by the socially necessary labour time embodied in it and that the price of the commodity fluctuates with supply and demand of and for the good or service.

    Marx makes the argument that money itself is a commodity so, its price will also fluctuate with supply and demand of and for its use.

    Am I wrong?

    With regard to the rise in the market price of stock, you have made the argument that the recent rise in the U.S. stock market has been due largely to the injection of money into corporate coffers via tax cuts and stimulus payments and it use by corporate boards to buy back their own stock thus, decreasing the supply stock and increasing its price.

    Is that a fair summary?

    1. As far as I understand it your’e right. But money has at least two different use values. To buy commodities with and as capital, a tool to extract surplus. Those values can be changed by taxes and by subventions of goods and services. Maybe one should also have a look at money as a tool for measuring the amount of work put into commodities, where gold seems to be a good standard. And money is usful to measure the effects of supply and demand on a market. Fiat money has the extra use value to be easily manipulated?

      1. My understanding is that capital is a commodity. Indeed, the social relation of Capital is based on the ownership of commodities, on the one side the owners of capital and on the other, the ownership of labour power.

        As for gold, it has a price determined by both its value and the supply and demand for it in the marketplace of commodities.

      2. I would rather claim that money has at least two distinct use values. The first is general, all money can be used for exhange. The other use depends on if the owner is capitalist and can use the money as capital. To be able to use money as capital you must have quite a lot of money. I would say something like a million in US dollars (excluding value of home, car and such). Then you can start using money as capital, as a means to get more money. Below that level money can only be used for exchange – or hoarding.
        The dichotomy was still physical in ancient Mesopotamia, where they used two currencies. Barley and silver. Barley coud of cause be used as a commodity – food – and as capital. For a time. It woud rot. While silver couldn’t feed neider people nor cattle nor be brewed to beer. But silver was established as money that could be used as capital. And by and by silver could be exchanged even for food (or beer). But there still has to be surplus of food – compared to death by starvation – for the exchange money (paper or silver) for food to function.

  9. Michael, do you agree that the rise in the organic composition of capital is a better indicator of crisis than the falling rate of profit? If not, why?

    1. If you read my book Marx 200, I explain three laws of capitalist accumulation and a rising c/v is the second and very important. It is the main driver of the tendency for the rate of profit to fall. In that sense, a rising OCC is proof of Marx’s law. But there are counteracting tendencies that decided the actual rate of profit and the movement in the rate of profit is the main arbiter of crises

  10. If Marxists can explain why there is no inflation, then this is also an explanation why rising government debt and printing money is no problem for bourgeois politics (apart from ideological prejudices). So Keynesianism must be critisised along other roads.

  11. The way I see it, the decline in prices over the last 40-50 years is due to globalization. Globalization has meant production has increasingly come from low wage countries, lowering costs of production. Migration to the West has also meant that labour rates are depressed by workers unaccustomed to western conditions – they work for less. So wage rates have been depressed in the West. This has also lowered the income of workers, a factor depressing prices in the West.

    Does this fit at all into inflation theories from a Marxist perspective?

  12. Value rate of inflation (VRI) = % change in wages and profits + % change of money supply.
    How to derive this equation? The monetary value of production = the real value of production x average price.
    In terms of the rate it is: variation in the monetary value of production = variation in the real value of production + variation in prices.
    If the price variation is equal to the M2 variation, there is the VRI. A tautology.

  13. The problem your theory has to cope with, is that there is not one rate of inflation but many, and, your theory has to explain all of them. Henry Rech has a point about globalisation. If we look at goods production versus service provision a gulf emerges (the former tends to be globalised the latter localised).

    First on the issue of globalisation as reflected in the Import Price Index. All data from the St Louis FED. The base year is 2000 when the IPI stood at 100 compared to July 2020.
    Capital goods have fallen to 88.5
    Consumer goods have risen to 106.8
    Durable consumer goods have fallen to 90.6

    This is reflected in the CPI for goods in urban areas which has fallen to 83.4 Thus in the realm of goods production there has generally been deflation.

    On the other hand there has been substantial inflation in the service sector which has increased to 172 in urban areas driven mainly by the rise in medical costs 218 and housing costs to 159

    I would suggest that your VRI model relates to only one sector of the economy, what Marx called Department 2B or the luxury goods sector which has expanded vastly with inequality. I prefer to call it the bespoke sector, because it is necessarily labour intensive and not price dependent. I have previously shown the disparity between cheap mass-produced watches and hand-crafted watches such as Rolexes. Further the issue of demand is absent as long as capital gains are in play and when they are, cashing them is a necessary driver of M2.

    I will confine my remarks to the above but there is a lot more to inflation than this.

  14. Michael, excuse my arrogance in what I’m going to say, but I don’t think anyone will be able to explain what will happen in the economy in six months or a year, simply because the factors that are driving the economy are EXOGENOUS to her. You see, China that managed to quell the epidemic may even grow this year or at least be tied, simply because it managed to suppress the exogenous factor.

    To say this I am basing myself on something that I know and that I studied for some time as a hydraulic engineer, hydraulic transients.

    The situation of the economy is similar to what occurs in any pipeline that is subject to exceptional forcing and therefore reacts differently than it would react if this “energy” introduced in the flow had a longer period of variation. What causes this strong introduction of an external energy in the system, which would have to be analyzed from a dynamic and POLITICAL point of view and not by crisis models that come from ENDOGENOUS phenomena.

    In any type of physical system, or natural or even more SOCIAL, if the preconditions are of instability, a minimum disturbance can produce a breakdown of the balance, however in systems in which these conditions of instability are present, but there is still a capacity to dampen the same, the same does not happen, because in weak conditions of instability the balance can be stabilized by countercyclical movements. But therein lies the problem, countercyclical movements in situations that created the so-called global imbalances, will never have the same reaction because the economic and political system itself will look for a NEW equilibrium point, in theory of stability of dynamic systems, produced instability the system may migrate to one of the other points of stability, which are called strange attractors.

    In short, economists or any simulator of current reality has no means of predicting what will happen in the short, medium and long term, everything is merely speculation.

  15. Thank you for this post. I find VRI intriguing, but I’m a little confused at how you reach the no-deflation conclusion. If 2020 wages are -20%, profits -25%, and money supply is +25%, can you explain how that translates into .05-1% inflation? I’m missing the math. Thank you!

  16. Hi Michael,
    I think it is a daring undertaking to relate the phenomena “inflation” and “deflation” to changes in the value of goods. This is daring because it involves price developments that can change completely independently of the value of the goods.
    Karl Marx took care of two price relationships: on the one hand, the price development of industrial goods compared to handicrafts and manufactured goods (1), and, on the other hand, the price relationships between the real wages that the wage workers receive and the prices of the goods that are normally in the reproduction of wage laborers (2).
    A global inflation figure that includes all or most goods in an economy interested neither Marx nor bourgeois economists. Their inflation figure also measures only a part of the total commodity production and circulation.
    Karl Marx:
    ad (1) „Eine vergleichende Analyse der Preise handwerks- oder manufakturmäßig produzierter Waren und der Preise derselben als Maschinenprodukt ergibt im Allgemeinen das Resultat, dass beim Maschinen-produkt der dem Arbeitsmittel geschuldete Wertbestandteil relativ wächst, aber absolut abnimmt. Das heißt, seine absolute Größe nimmt ab, aber seine Größe im Verhältnis zum Gesamtwert des (Einzel-)Produkts, z. B. eines Pfundes Garns, nimmt zu.“ K. Marx, Kapital I, MEW 23, 411.
    ad (2) „Der Wert der Arbeitskraft löst sich auf in den Wert einer bestimmten Summe von Lebensmitteln.“ K. Marx, Kapital I, MEW 23, 186.
    Der „wirkliche Lohn … sind die dem Arbeiter zur Verfügung gestellten Lebensmittel“. K. Marx, Kapital I, MEW 23, 584.
    „Der Arbeitslohn ist vor allem noch bestimmt durch sein Verhältnis zum Gewinn, zum Profit des Kapitalisten – verhältnismäßiger, relativer Arbeitslohn.
    „Der Reallohn drückt den Preis der Arbeitskraft im Verhältnis zum Preise der übrigen Waren aus, der relative Arbeitslohn dagegen … den verhältnismäßigen Wert von Lohnarbeit und Kapital, den wechselseitigen Wert der Kapitalisten und Arbeiter.
    Der Reallohn mag derselbe bleiben, er mag selbst steigen, und der relative Arbeitslohn kann nichtsdestoweniger fallen.“ K. Marx, Lohnarbeit und Kapital, MEW 6, 413.
    „Die Stellung der Klassen zueinander bedingt mehr durch das relative Gewicht der Löhne als durch die absolute Höhe der Löhne.“ K. Marx, Theorien über den Mehrwert II, MEW 26.2, 420f.

      1. The graph shows the difference between the average input of working time for the workers (employed), and the output of commodities to each inhabitant. The later also measured in hours worked. I first used it as a way to measure the value-transport in the restauration of the Swedish private capitalistic fortunes after 1990.
        When the function line moves toward the lower left corner capital is used for investments in faster production (less working hours/worker) and at the same time better products, less working hours needed per inhabitant. http://www.fredtorssander.se/fredpress/2020/02/29/a-non-monetary-graphical-description-of-the-productivity-growth-in-sweden-1870-2017/

  17. The rise in prices is made dependent on the growth of the money supply. This is Hume or Ricardo, but not Marx

  18. Mr Roberts thank you for the dense and quite simple representation of this interesting inflation theory. I find myself confused though and I hope you could clarify things on the following issue.
    It is certain that as labour productivity rises, commodity value is falling (due to the falling socially necessary time to produce) in relation to the use value. Though when referring to the total new value produced (v+s) we refer to aggregate notions of total labour spent on commodity production and not the relation of exchange value objectified into a single commodity and thus a ratio of exchange value/use value. Bearing that in mind I am thinking that saying that the “Total value will decline relatively to use value production” is actually assuming that the use values produced are constant. The latter though is contradicting my general perception of capitalism as an innovative system that is constantly expanding and thus is always introducing new use values into the market sphere.

    1. Yorgos, yes more use values means more value in total too, And there is more value being created by labour, but the increase in value does not match the increase in commodities over time. In effect, there is overproduction of supply relative to value,

  19. Hi Michael! I was wondering whether you think it might be possible to derive at least some sort of proxy indicator for the organic composition of capital from the latest Penn World Table? I tried to relate capital stock and output-side GDP, the result of which returned a positive correlation of increasing capital stock in relation to GDP to sinking rates of profit. But then again, GDP is probably a less than optimal proxy for value generated … I would be thankful for any insight!

      1. Hi Michael, just to make sure: how exactly did you calculate the OCC based on the Penn Tables for your latest post?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: