Paul Mattick and validating Marx’s law – a critique of The Long Depression

Last weekend, at a session of the Left Forum in New York, I presented the basic theses in my book, The Long Depression.  My arguments were then subjected to critical analysis by my invited discussants, Paul Mattick Jnr and Jose Tapia.  The whole session was video recorded and will appear on You Tube – Left Forum in a few weeks.

Now, all three of us are agreed that Marx’s law of the tendency of the rate of profit to fall is the underlying foundation and main cause of crises under capitalism.  This view remains a minority one among Marxist economists, let alone other heterodox economics.  But even though we agree on this, there is still much to debate about how to explain and validate Marx’s law.  In this post, I shall respond to the profound critique of my book that Paul Mattick presented at the weekend.  In the next post, I shall look at Jose Tapia’s critique.

Paul Mattick is emeritus professor of philosophy at Adelphi University, New York and the son of that eminent contributor to Marxian economic analysis, Paul Mattick snr, who explained so well Marx’s theory of crisis in the post-war period and exposed where mainstream economics, particularly Keynesian theory and policy, fell short. Paul Mattick jnr has continued his father’s work just as successfully.  His book, Business as Usual is, in my view, the best analysis of the global financial crash and the Great Recession that is easily accessible to non-economists – essential reading.

Paul’s criticism of my book boils down to how to validate Marx’s law of profitability as the theory of crises.  As he says in his commentary paper (Roberts Panel) , he agrees that the world economy is in what could be described as a long depression and I am right that Keynesian and monetary policies of the mainstream have failed to get capitalism out of this depressed state.

However, he considers my attempt to validate Marx’s theory of crises by trying to measure the rate of profit in a Marxian way as impossible and unnecessary. “This is not possible, fundamentally because of the fact that value is represented only by prices, which move independently of values”. The Marxian rate of profit can only be measured in value terms (average labour time) and yet all official statistics are in prices; and worse, in the market prices of one currency usually.  Such price measures can and will vary well away from Marx’s modified values (prices of production), let alone value.  As such, all my (and other people’s) ‘Marxian’ measures of profitability are a waste of time.

Moreover, most of the measures of profitability made by me and other scholars are national rates of profit and usually just for the US.  Such measures, says Paul, tell us nothing about the movement of profitability in global capitalism.  And Marx’s law is one based on a world economy. But a world rate of profit with a proper calculation of total surplus value globally is impossible to measure.

Indeed, Paul argues we don’t need to ‘test’ Marx’s law and his theory of crises with such statistical measures.  We can validate Marx’s theory by the very fact that capitalist economies go into recurring slumps; that there are periods of prosperity and growth that give way to periods of depression, as now.  Capitalist accumulation cannot deliver harmonious and sustained expansion of production and, most particularly, accumulation of capital; and it cannot deliver full employment etc.  Paul says: “These are all features recognizable without a significant statistical apparatus; in Marx’s own work, historical data for the most part function to illustrate, not to test, theoretical ideas.”  There is no need to get into all the details of measuring rates of profit; that is an exercise in futility.

Well, I don’t agree.  Using general events or trends to ‘illustrate’ the validity of a law can help.  But that is not enough.  Slumps in capitalism could be explained by other theories like Keynesian ‘lack of demand,’ or from ‘underconsumption’ through low wages and rising inequality etc (post-Keynesian); or by the failure of consumer sectors to grow in line with capital goods sectors (disproportion theory); or by excessive debt (Minsky) or ‘too much profit’ that cannot be absorbed (Monthly Review).  The ‘features’ of capitalist crises can be used just as well to ‘illustrate’ these alternative theories.  Indeed, they are the more  dominant explanations in the labour movement and the same illustrative events are used to validate them.

To justify Marx’s law of profitability, I reckon we need to go further scientifically.  That means measuring profitability and connecting it causally with business investment and growth and slumps. Then we can even make predictions or forecasts of future crises.  And only then can other theories be dismissed by using a body of empirical evidence that backs Marx’s law.

Such an approach may be ambitious, but it is not impossible.  First, Marx’s law may be based on labour values, but it is expressed in prices.  Capitalists start with money capital and finish with more money capital in the capitalist mode of production for profit.  Money is the nexus between value and price.  Yes, market prices can and will necessarily vary from prices of production and from commodity values.  But they are still tied like an anchor or a yo-yo to value over time, even as value incessantly changes due to changes in the productivity of labour.  Total value still equals total prices.

Several scholars have shown empirically the close connection between market prices and value in production.  And money prices can be checked empirically against value in labour time.  For example, Cockshott and Cottrell broke down the economy into a large number of sectors to show that the monetary value of the gross output of these sectors correlates closely with the labour concurrently expended to produce that gross output: Anwar Shaikh did something similar.  He compared market prices, labour values and standard prices of production calculated from US input-output tables and found that, on average, labour values deviate from market prices by only 9.2 per cent and that prices of production (calculated at observed rates of profit) deviate from market prices by only 8.2 per cent.

And G Carchedi in a recent paper showed that the validity of Marx’s law of value can be tested using official US data, which are deflated money prices of use values.  He found that money and value rates of profit moved in the same direction (tendentially downward) and tracked each other very closely.  Carchedi and I used this in a joint paper to show how the rate of profit in the sectors that create value and surplus value in the US economy is not so far out of line with the overall rate of profit in the ‘whole economy’.

That brings me to another criticism of my approach by Paul.  He says that the Marxian rate of profit is the total surplus produced by productive labour in an economy and unproductive labour should be measured as part of total surplus value.  But, says Paul, many measures of the rate of profit by scholars fail to take account of interest, rent and financial profit which are also parts of total surplus value. In Marxian terms, the rate of profit should be a measure of total surplus value against capital advanced, not profit as defined in capitalist statistics.

Paul is right.  Indeed that is why I use what I call a ‘whole economy’ measure.  This defines surplus value (for a national economy) as annual gross national income (less the annual depreciation of the means of production) less the cost of wages and benefits.  Then, to get the rate of profit, this surplus value is divided by the total capital advanced for means of production (fixed assets) and circulating capital (raw materials or inventories) in the productive sectors and variable capital (labour) in the productive sectors.  This then encompasses Paul’s critique of some measures that exclude rent, interest and financial profits.  In my view, all these moving parts can be measured to deliver a meaningful rate of profit using official statistics.  We can measure ‘productive’ capital and we can incorporate all forms of surplus value.  And several scholars have done so for different countries.  If you read this blog regularly, you know who they are.

What gives support to these attempts to do the impossible (in Paul’s view) are the results.  However the rate of profit is measured, the general trend is the same. Take the post-war period, using official statistics for the US, you can measure the rate of profit for the ‘whole economy’, for the corporate sector alone, for the non-financial corporate sector or even for a more accurately defined ‘productive’sector’ and the general trend is the same.  There was a high rate of profit immediately after 1945, which holds up to the mid-1960s.  Then there is a profitability crisis that lasts until the early 1980s.  Then there is a ‘neo-liberal’ recovery in profitability that comes to an end about the late 1990s or by 2001.  After that, the rate of profit does not return to the level of the 1990s and certainly not to that in the 1960s. I think these measures are robust (because they are similar) and thus provide powerful validations of Marx’s law of profitability.  So they are not to be dismissed, as they help to refute alternative theories of crises.

Yes, these measures are just national and do not show the ‘world rate of profit’, which would be necessary to support Marx’s law fully.  As Paul says: “An increase in the profitability of American capital tells us nothing determinate about the Marxian rate of profit”.  But attempts there are being made to measure such a ‘world rate’ by various scholars, including me.  They are by averaging national rates, not an ideal solution theoretically, but nevertheless, again they deliver similar results to the US measures as described above.

Paul says that you cannot get a direct connection between profitability and investment in a capitalist economy, because returns on investment in the stock market are really reflections of fictitious capital, not the rate of profit on productive capital.  As Paul quotes a securities analyst: “the security’s price does not have a direct relationship to the surplus value currently being exploited from the productive workforce”.

Clearly stock market returns can and do vary sharply from the return on productive capital.  In The Long Depression, and in my previous book, The Great Recession, I make this point as well.  Dividends and capital gains from stock market purchases are not the Marxian measure of profit as they are distorted by the fictitious nature of financial capital – much of profits accumulated in the financial sector are fictitious, particularly gains from the purchases of government bonds.  Government borrowing and the printing of money deliver a continual stream of fictitious profits.

But this can be accounted for and several scholars have done so.  We can delve into the data and begin to show the clear causal connection between the movement in the profitability of capital in the productive sectors, investment in productive capital and economic growth – indeed, the movement in the mass of profit in an economy is a very good guide to the likelihood of a change in business investment and a slump in capitalist production (The profit investment nexus Michael Roberts HMNY April 2017).  The other discussant at Left Forum, Jose Tapia, has shown just that for the US economy  (does_investment_call_the_tune_may_2012__forthcoming_rpe_), as has a recent and comprehensive paper by G Carchedi on Marx’s law and crises.

So I think we can go further than just use anecdotal evidence to ‘illustrate’ Marx’s law of crises. We can provide hard evidence based on robust empirical data to support Marx’s law of profitability and its relation to recurrent crises under capitalism.  Yes, the task of defining our categories and massaging the data so that we measure things accurately is formidable.  But nobody is arguing that science is easy (and we often get it wrong), but I am arguing that it is not impossible.

Moreover, it is necessary.  Otherwise, alternative theories to Marx’s theory will continue to claim validity and hold sway.  And that is bad news because these alternative theories deliver policies that look to ‘manage’ or ‘correct’ capitalism rather than replace it. And these policies don’t work and will not work in the interests of the majority (the working class) and will instead perpetuate the iniquities and horrors of capitalism.

Paul is also sceptical of my proposition that there are discernible cycles or regularities in capitalist accumulation, although he reckons that whether there are or not does “not matter very much from a Marxian point of view.”  In the next post, I shall deal with the critique of my view on cycles in capitalism as presented by Jose Tapia, the other discussant at the Left Forum panel on my book.


9 Responses to “Paul Mattick and validating Marx’s law – a critique of The Long Depression”

  1. fosforos17 Says:

    ” Total value still equals total prices.” This is total nonsense. Equality between quantities denominated in different units of measure is conceivable only after conversion of one unit into the other. Minutes of socially necessary productive labor time are not dollars of NNP. To convert those dollars into those minutes you need a value for the *labor quantity of the price unit* prevailing on average during the relevant accounting period. I presented such a calculation for the US economy in my 1963 dissertation. Nobody, as far as I know, has even attempted such an exercise since then,

  2. sartesian Says:

    total value = total prices =total nonsense due to a lack of a conversion factor?

    Not exactly. A conversion factor as fosforos presents it, and desires it, would require that individual values = individual prices; that the factor exist prior to the conversion, and that is precisely not how the socially necessary value, surplus value, is distributed among capitals.

    It’s possible to produce a conversion factor for the annual aggregate of production time units into aggregate prices; and we can produce an average of the ratio of production time to price, but you cannot establish an immediate conversion as the labor-time is purchased and consumed privately, not socially, and hence the product has to be marketed, exchanged, to establish the socially necessary labor time.

    That’s kind of fundamental to Marx’s critique.

  3. Wal Buchenberg Says:

    Paul said: “ in Marx’s own work, historical data for the most part function to illustrate, not to test, theoretical ideas.”

    “To test theoretical ideas” is not the fundamental question here.
    Statistical data must not test “theoretical ideas”. Statistical data show changes in the capitalist economy. Karl Marx used very often the sparse statistical data of his time, especially in his journalistic work to explain economic and social developments.
    The data together with the underlying laws make developments of capitalism understandable. You need both: the data and the theoretical ideas to explain and to understand capitalism. To use only one of them will not help.

    Wal Buchenberg

    • Wal Buchenberg Says:

      in German:
      Der Vorwurf von Paul Mattick jun. an Michael Roberts war, dass Marx keine statistischen Daten verwendete, um die Gesetze zu beweisen, die die kapitalistische Ökonomie bestimmen (Wertproduktion durch lebendige Arbeit, Durchschnittsprofitrate, Fall der Profitrate etc.).
      Das ist sehr richtig.
      Marx schrieb im Vorwort zum “Kapital”: “Bei der Analyse der ökonomischen Formen kann außerdem weder das Mikroskop dienen noch chemische Reagentien. Die Abstraktionskraft muss beide ersetzen.” (MEW 23, 12)  Es ist also eine Sache der Einsicht, nicht der Empirie, diese ökonomischen Formen zu entdecken und zu erklären.
      Sind diese ökonomischen Formen einmal gefunden, geht es nun nicht mehr darum, sie ständig neu – mit oder ohne Empirie und statistischen Daten – zu beweisen, es geht vielmehr darum, die wirkliche Entwicklung des Kapitalismus anhand der statistischen Daten und mittels der von Marx entdeckten ökonomischen Gesetze zu analysieren und zu verstehen. Dafür braucht es  empirische Daten. Und Marx selbst hat in seinen journalistischen Arbeiten ständig und ausgiebig statistische Daten benutzt. Darüber hinaus war es ihm wichtig, innerhalb der IAA die dort organisierten Arbeiter zur Erhebung statistischer Daten über ihre Arbeitsbedingungen anzuhalten. Vergl. das Programm für den Kongress der Internationale in Mainz, Punkt VI. “Über die Pflicht der Arbeiterklasse, zur Verfertigung einer allgemeinen Arbeitsstatistik zusammenzuwirken gemäß dem Beschluss des Genfer Kongresses von 1866.” (MEW 16, 435).
      Die Kritik von Paul Mattick geht daher an den Thesen von Michael Roberts’ “Long Depression” völlig vorbei.

      Wal Buchenberg

    • Wal Buchenberg Says:

      Marx began the third part of the capital with the words: “The various forms of capital, as evolved in this book, thus approach step by step the form which they assume on the surface of society, in the action of different capitals upon one another, in competition, and in the ordinary consciousness of the agents of production themselves.” MEW 25, Part I. Chapter 1.
      This surface and its changes are shown in statistic data. And it is our task to analyze and understand the surface and the data by means of the hidden capitalist laws. It is no longer our task to discover and “prove” or “test” these laws. This task has been done by Marx.
      Wal Buchenberg

  4. CecilTheLion Says:


    Is your thesis uploaded into an academic database? I’d be curious to look at it.

  5. Virgens Kamikazes Says:

    “This is not possible, fundamentally because of the fact that value is represented only by prices, which move independently of values”

    Well then the very first thing mr. Roberts should’ve asked to mr. Mattick is: do you know how to read?

    Then, if the answer is “yes”, the immediate next thing mr. Roberts should ask from mr. Mattick is: then, please, sir, read Capital book III again before stating blatant lies.

  6. Publicstat (@Publicstat) Says:

    “Several scholars have shown empirically the close connection between market prices and value in production. ”
    It is not strange at all: The “problem” is not in the transformation itself, but in the equalization of the rate of profit, (when departing from the existing sectoral rates in labour values in a simple sraffian system expressed in labour values)
    Hence, in this case, with no equalization, individual production prices=labour values and profits=surplus value.
    Equalized (of the marxian) rate of profit makes prices differ from values and the two equalities to be different.

Leave a Reply

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

You are commenting using your 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 )

Google+ photo

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

Connecting to %s

%d bloggers like this: