Explaining the downturn

Profitability and the dynamics of capitalism

The faster the rise in productivity, concludes Nick Rogers, the more likely it is that the rate of profit will fall

In the first article last week1 I discussed the debate in Marxist economics about the role of Marx’s law of the tendential fall in the rate of profit (LTFRP) in cyclical capitalist crises. I outlined Moshé Machover’s critique of the LTFRP, particularly his thesis that the future trend in the rate of profit is indeterminate. Machover argues that capitalists strive to make savings on all the inputs into the production process - equipment and raw materials as much as workers - and that we cannot predict the impact of increases of productivity on the ratio of living labour to overall investment.

I developed a hypothesis that the savings capitalists make on the fixed-capital element of constant capital - particularly investments in machinery and equipment - add to the instability of capitalism. The profits of investors at the beginning of the business cycle are progressively undermined by competitors, who invest later in the cycle in cheaper and more efficient versions of the same technology (as a result of continual productivity increases). I proposed that this “moral depreciation” (as Marx termed it) may be critical in generating the short-term, five-to-10-year business cycle acknowledged by bourgeois economists and mainstream politicians alike. The suggested mechanism is a widening of the spread of the profit rates of individual capitalists from the average during the upturn phase of the cycle, as moral depreciation takes hold, with the downturn serving to narrow the spread, as those capitalists with lower profit rates are forced out of business and the technological advances of the previous period are standardised (the impact of moral depreciation is thus temporarily blunted).

My hypothesis builds on a finding of Machover and Emmanuel Farjoun in their Laws of chaos that the “standard deviation” from the average profit rate during the crash of 1981 was four times higher than usual.2 Further analysis of fluctuations in the distribution of individual profit rates, testing whether they correlate with the business cycle, and attempting to isolate the effect of moral depreciation in the fluctuations, would be needed to lend weight to the hypothesis.

A merit of the hypothesis is that it is consistent with the role Marx attributed to the rate of profit and “revolutions in value” in determining the behaviour of the capitalist economy: capitalism’s incessant search to increase the productivity of labour and reduce constant capital inputs periodically checks growth and the accumulation of capital.

Long waves

In this article I intend to demonstrate that the impact of productivity changes on composition of capital (best defined as the ratio of living labour, or new value created, to capital invested rather than the ratio of constant to variable capital that Marx deployed as a measurable proxy) remains relevant to the dynamics of capitalism. For not all downturns in the business cycle are equally destructive and not all upturns display the same exuberance. In the 1950s and 1960s, for instance, the business cycle was heavily dampened: recessions (officially at least two successive quarters of economic decline) were short and shallow, while robust economic growth was quickly resumed. Downturns in Britain and elsewhere sometimes registered as slower growth rather than an actual fall in output.

Since the 1970s-early 1980s decade of stagflation, which witnessed two sharp contractions (1974-75 and 1981-82), the contours of the economic landscape have shifted to become much more unstable, as the business cycle returned with a vengeance. Most recently, the very deep slump of 2008-09 has been followed by such a weak recovery that in Britain output did not regain its 2008 levels until 2014. Across the United States, Europe and Japan unprecedentedly low interest rates close to zero percent and the printing of money by central banks (previously a taboo option that, according to the economic textbooks, is supposed to lead to hyperinflation) coexist with prices that threaten to actually fall. The next downturn in the business cycle will strike economies that are already on their knees.

So, as far as the post-war history of the global economy is concerned, the phenomenon of decades-long economic phases, straddling multiple business cycles, which share common features of economic behaviour, is real enough. Many economic historians discern similar periods stretching back to something like the 1780s, with the first long boom of the early industrial revolution. It is not my intention here to discuss the relative merits of Kondratiev’s approach (and his eponymous cycles) and that of Trotsky in the 1920s, or Ernest Mandel’s discussion of ‘long waves’ in the 1970s - let alone Paul Mason’s interpretation in his recent book.3 Whether the transition from a period of relative stagnationary growth to one of relative expansionary growth is due to features internal (‘endogenous’) to capitalism or whether external (‘exogenous’) shocks are responsible, and then whether there is anything automatic about the supposed 50- or 60-year length of such ‘cycles’ or ‘waves’ (in fact, whether there are cycles or waves at all lasting decades that mimic the business cycle) is not my concern in these articles. I simply intend to agree with Mandel (and also Mason, as it happens) that the rate of profit is the most important determinant over any given timescale as to whether economic growth is going to be relatively fast or slow. And, that being so, to assert that Marx’s LTFRP can help us explain successive turning points in economic history.

The reason why the rate of profit is the critical variable in explaining historical economic trends is quite simple. There is a direct relationship between profit and the resources capitalists have available for investing in the growth of their businesses. The rate of profit - the return on total capital - is the same ratio as the maximum potential rate of accumulation - new investment as a proportion of existing capital. And the rate of accumulation directly determines the rate of growth of the economy.

It is important to remember that the maximum potential rate of accumulation serves as a ceiling - it would be reached only if all profits were productively reinvested. The actual rate of accumulation will never match it and, therefore, the economy will never achieve its maximum potential growth. Nevertheless, the rate of profit is crucial to analysing aggregate economic behaviour.

Indeed it is Andrew Kliman’s contention4 that it was the lower rate of profit in the 1980s and 1990s compared with the post-war boom (rather than any tendency for the rate of profit to fall yet lower over this period) that meant there was a smaller pot of resources to invest. This explains the fall in growth rates compared with the post-war boom. Lower profitability also encouraged capitalists to resort to speculative investments in an attempt to improve their individual rates of return and made debts more difficult to pay off. Greater economic instability was the consequence.

Kliman does not maintain that the LTFRP proves that the rate of profit must fall over the long-term - the thesis that Machover sets out to disprove. Kliman’s interest is in investigating the role of the LTFRP in capitalist crisis.

Alan Freeman, another temporal single system interpretation (TSSI) theorist - in a paper that succinctly analyses the role of the rate of profit in post-war history and leans heavily on moral depreciation - is scathing about the argument that “the rate of profit must fall for ever”. He regards this as a “catastrophist view” that has taken on a “millenarian form”, finding “theoretical currency in those places where the practical prospects for political change were weakest: namely, the Anglo-Saxon countries.”5

Marx himself does not discuss the falling rate of profit in these terms. For him a fall in the profit rate “promotes overproduction, speculations and crises”6 and he discusses the impact of the LTFRP on the capitalist economy exclusively in terms of cyclical crises. As I explained in my first article, the actual behaviour of industrial capitalism over more than two hundred years - a regular, short-term business cycle mapped onto longer phases of variously vigorous or more feeble economic performance - would not lead us to expect a continual decline in the rate of profit over those two centuries.

That is why I question Esteban Maito’s findings that the global rate of profit during most of the second half of the 19th century was higher than at any point in the 20th (even than the post-1945 period, which saw the highest growth rates in the history of capitalism).7 Maito’s 19th century growth rates do not correlate with actual economic performance. His measurements for the UK, for instance, give no indication that profit rates contributed to the post-1873 long depression. Tellingly, his US profit rates, for which the best records probably exist, do substantially follow broader economic trends.

Kliman’s and Freeman’s thesis, with which I agree, is that the level of the rate of profit is the decisive factor in determining the rate of economic growth and therefore the relative strength of the economy. When the economy is recording consistently high growth, the impact of the business cycle (driven in my hypothesis by the losses attributable to the technological obsolescence suffered by investors early in the business cycle), is fairly limited.

On the other hand, when average profitability is already comparatively low, the downturn phase of the business cycle will have much more damaging effects, with a greater number of companies going bust, more workers being laid off and the contraction in economic output deepening. Kliman notes that Farjoun’s and Machover’s exposition about the spread of profit rates represented by the average (rather than the uniform profit rate that is usually assumed) accentuates the impact of low or falling profitability. Those firms with profit rates below the average are more likely to be tipped into absolute losses.8

In conditions of generally low profitability even the upturn phase of the cycle will see only relatively weak economic growth. And, the more profits are directed to speculative investments in an attempt to compensate for lower returns on productive investment, the more economic instability will ratchet up. Assets and other forms of ‘fictitious capital’ will form bubbles waiting for any sign of panic (perhaps occasioned by a widening distribution of profit rates and therefore an increase in the number of capitalists experiencing low or no profits as a result of moral depreciation) to burst and precipitate a new crash.

Counteracting factors

If profitability and its interaction with all the other phenomena of capitalism is the key to understanding both the regular cyclical return of crisis and successive phases in the economic history of industrial capitalism, it is obviously important to understand what determines the rate of profit. This is where Marx’s LTFRP comes into its own. For not only does Marx uncover the way in which rises in labour productivity can depress the rate of profit, but he examines the “counteracting factors” that can increase it. This is exactly the kind of dynamic (we could say dialectical) interaction that we would expect to underlie the complex economic history of capitalism.

In chapter 14 of the third volume of Capital, Marx identifies a total of six counteracting factors. This fascinating, but all-too-brief discussion should be read alongside chapters earlier in the same volume and other discussions throughout his writings on political economy. A fully rounded discussion of the LTFRP gives us the tools to model real economic behaviour.

I suggest that, broadly speaking, there are five ways in which the rate of profit can be raised in the face of capitalism’s drive to raise productivity by squeezing living labour from production.

For example, privatisations that shift labour-intensive processes, such as education, health and social care, from the non-commodity-producing state sector to private hands lower the economy-wide composition of capital. The emergence of the Chinese economy on the world stage in the 1980s and 90s, based on low wages and generally more labour-intensive production techniques, would have had a similar impact.

However, the faster turnover of capital impacts primarily on circulating capital. This plays a less important role in determining the rate of profit than fixed capital - and, as turnover speeds up, will form a yet smaller component of total invested capital.

Equally, a reduction in the capital allocated to unproductive sectors will increase the proportion of total capital that is generating surplus value. Marx thought that the specialisation of some capitalists as merchants (we would think in terms of wholesalers, retailers and advertising companies) and financiers would allow unproductive but essential functions to be carried out more efficiently and with a smaller proportion of aggregate capital. If, on the contrary, the resources allocated to unproductive (non-surplus value-generating) activities has increased since 1945, as many believe, this counter-development will have served to lower the average rate of profit.

The relationship between the counteracting factors and the rate of profit is highly dynamic and contradictory. Some operate over the very long term - equivalent to the life-cycle of capitalism. For instance, Marx points out how the concentration and centralisation of capital can lead to savings in the use of constant capital.10 Others have a fluctuating or cyclical impact on profitability.

None of the counteracting factors invariably work in the interests of capital. Nor are capitalists aware of concepts such as the composition of capital, or of the way labour-saving changes can impact on overall profitability - the fetishistic veil behind which commodity production obscures the source of profit makes sure of that.

Take attempts to raise the rate of surplus value. Workers’ hours can be increased without a compensatory rise in wages (absolute surplus value) or wages can be forcibly reduced (below their value, as Marx says). Increasing wages by a smaller percentage than the rate of inflation has this effect. A rise in consumer debt, with workers paying more of their wages in interest payments, effectively transfers a portion of the new value previously accruing to workers to the interest-earning faction of capital and so increases economy-wide exploitation.

But attempts to increase the extraction of surplus value from workers by raising their productivity (the method most intrinsic to the nature of capitalism) are the source of the tendency of the rate of profit to fall - a more productive worker will tend to process more constant capital and thereby raise the composition of capital. Early adopters of new technologies can capture a surplus profit from their laggard competitors, but, once the technology is taken up by most capitalists, everyone’s profits will tend to sink towards the average. The rate of profit of capital as a whole benefits only if the productivity rise contributes to the production of workers’ consumption goods. If the price of these commodities falls (because less labour is involved in their production) and workers do not secure wage increases that reflect their increased production of use-values, then the economy-wide rate of surplus value can rise.

However, Marx demonstrated that an increase in the rate of surplus value can only “check the fall in the profit rate”: “two workers working for 12 hours a day could not supply the same surplus value as 24 workers each working two hours, even if they lived on air and hence scarcely need to work at all for themselves”.11 But, in combination with other factors, changes in the rate of surplus value can contribute to the swing from one phase of capitalist development to another. Mandel argues that a sustained rise in the rate of surplus value (due to the smashing of labour movements under fascism) was the principal factor behind the launch of the post-war golden age. The profit-squeeze theorists argue that a rise in working class militancy in the 1960s, and its success in raising wages, improving conditions and placing tight constraints on the freedom of manoeuvre of capitalists, cut the rate of profit and brought the long boom to an end.

Neither tells the whole story. Mandel does not explain increased profitability in the United States (the core economy in the post-war world), where workers were newly emboldened - nor the ‘spirit of 45’ in Britain. And Alan Freeman demonstrates that a fall in the rate of surplus value explains only part of the fall in the rate of profit during the 1960s. Shifts in the rate of surplus value, nevertheless, did serve to amplify other developments.

Composition of capital

It is the impact of those counteracting factors constraining or reversing rises in the composition of capital that are most critical to the direction of the rate of profit. To really get to grips with the dynamic at play, we need to return to Machover’s summary of capitalism’s capital-saving tendencies12 that I discussed in the first article.

Here it is helpful to be more specific about Marx’s definition of the composition of capital. Marx called the ratio of physical quantities of means of production to workers the technical composition. The ratio of constant to variable capital (two value categories) was the value composition. Marx defined the organic composition in these terms: “I call the value composition of capital, in so far as it is determined by its technical composition and mirrors the changes in the latter, the organic composition of capital.”13

Machover refers throughout his article solely to the organic composition of capital. I think he means the same thing as the value composition. Kliman means by the organic composition a special “constant-price” version of the value composition that has a direct and equivalent relationship with the technical composition of capital. To avoid confusion, I will dispense with the use of the term, ‘organic composition’, and refer only to the technical and value compositions. Kliman’s constant-price value composition can usefully serve as a measure of changes in the technical composition: ie, the use-values involved in production. I take the value composition as fully measuring changes in the values of the composition of capital.

Over time the value composition will tend to diverge markedly from the technical composition. Marx discusses this divergence:

The value of the capital employed today in spinning is 7/8 constant and 1/8 variable, while at the beginning of the 18th century it was 1/2 constant and 1/2 variable. Yet in contrast to this, the mass of raw material, instruments of labour, etc that a certain quantity of spinning labour consumes productively today is many hundred times greater than at the beginning of the 18th century. The reason is simple: with the increasing productivity of labour, the mass of means of production consumed by labour increases, but their value in comparison with their mass diminishes. Their value therefore rises absolutely, but not in proportion to the increase in their mass.14

Remember that Machover specifies two ways in which the value of constant capital (and therefore the composition of capital) can be reduced. First, new productivity-boosting technologies might be cheaper than the equipment and machinery they replace - indeed, capitalists do their best to save on all constant capital inputs. Marx was very aware of these tendencies - he discusses at length the various ways in which capitalists try to reduce their expenditure on constant capital.15 Reductions in the physical quantities of constant capital reduce both the technical and value compositions of capital - there is a direct correlation between the two.

I question whether entirely capital-saving changes increase the productivity of labour, as Machover asserts. They will definitely reduce the quantity of ‘dead labour’ entering the production process and therefore reduce the value and price of the commodities that emerge, but they will not in themselves increase the number of commodities (in physical use-value terms) that each worker produces, which is surely the only meaningful definition of labour productivity. However, by reducing the quantity of inputs that are required to produce each commodity, they may reduce the effort and quite possibly the time involved in their production and facilitate an increase in labour productivity.

Machover makes a wider point about the nature of technology. He cites the example of the rapid reduction in the price (and size) of printing technology, as the hot-metal typesetting printing press gave way to the electronic photo-printer. Actually, all technologies pass through life-cycles in which their price tends to fall and the extent to which they are adopted rises. As Marx explains, “When machinery is first introduced into a particular branch of production, new methods of reproducing it more cheaply follow blow upon blow, and so do improvements which relate not only to individual parts and details of the machine, but also to its whole construction.”16

In simple terms this life-cycle could be said to involve initially the emergence of relatively expensive (and often physically bulky) prototypes that are not widely adopted, followed by the development of cheaper, more efficiently constructed machinery and equipment that spread rapidly through an industry (or create new ones), and finally a phase of rapid price reductions, including reductions in the quantity of physical constant capital and the ubiquitous distribution of the technology.

Machover’s description of the printing industry follows something of this pattern - all information technologies certainly do. It is likely that the transition from the first to the second of these phases will involve the shedding of labour and an economy-wide increase in the technical and value compositions. The transition from the second to the third phases may well involve a fall in the technical and value compositions, as more savings are made in constant capital inputs than workers. This pattern overlaps with technologies in different sectors to be found at varying stages of evolution and is constantly recreated.

Perhaps this schema can point us towards discovering the holy grail of economic history: an explanation of the foundations of the post-war boom. I think we can improve on Mandel’s thesis about a rise in the rate of surplus value - or the main alternative: a fall in price (devaluation) of the fixed capital elements of production resulting from the slump of the 1930s. The latter has no effect on technical composition and, therefore, can only temporarily lower the value composition of capital, which, all things being equal, should rapidly return to its pre-slump levels. Although falls in the price of constant capital inputs (and in workers’ wages, for that matter) are part of the process by which capitalism overcomes downturns in the business cycle, such an explanation is a thin theoretical basis for understanding a boom that lasted for close to three decades.

What needs to be explained is Kliman’s finding:

US corporations’ physical stock of fixed capital fell during the depression and didn’t rebound thereafter, even during the war. It was smaller at the end of 1945 than at the end of 1929. But, right after the war ended, corporations were producing roughly twice as much output in physical (inflation-adjusted) terms, as they did right before the depression. This means that physical capital investment per unit of output fell by about one-half. This helped raise the rate of profit tremendously.17

Such a dramatic shift in the technical composition of capital must have been built on a suite of new (cheaper) technologies being released into civilian production after their turbo-charged development during World War II. War production was tightly controlled by the state and directed to military (department two) output - so outside the normal circuits of the reproduction of capital. Technology that had been in phase one of its life-cycle, or was just a twinkle in the eye of a laboratory scientist, was tested and matured and already falling in price by the time capitalists were freed to invest on a large scale after 1945.

Machover’s second example of “capital-saving” technological changes is related to the way that productivity increases in any industry producing means of production (Marx’s department 1), lowering the price of inputs into the rest of the economy (other department 1 industries and the consumer goods industries of department 2). The composition of capital may rise in the industry introducing the new productivity-boosting technology, but it will fall in the industries that purchase its cheaper output. To be precise, the technical composition of capital (the actual physical quantities involved) across the whole economy rises, but the value composition (the values involved in production) falls - in the case of this type of capital-saving change there is an inverse relationship between the technical and value compositions.

However, the impact of the effects Machover discusses are not instantaneous. It takes time for the new, cheaper commodities to be produced and sold. This is where a temporal analysis is useful in understanding what happens. In Reclaiming Marx’s ‘Capital’ Kliman uses the corn model beloved of classical economists to model the impact of increases in labour productivity at the level of the whole economy. His purpose is to demonstrate that physicalist and simultaneist measures of the rate of profit both measure the increase in use-values as a result of increases in productivity.18 Their rate of profit must always rise. Once values based on labour time are calculated, the rate of profit (in value terms) can fall as a result of increases in labour productivity and the consequent fall in the prices of outputs (measured in values).

The point in the context of the present discussion is that Kliman is measuring what happens to the rate of profit as a result of cheaper outputs feeding into the next production cycle. In an extreme simplification, all corn outputs are recycled as seed corn inputs into the succeeding cycle. So the whole economy is devoted entirely to producing means of production (only department 1 exists), workers receive no wages (the capitalists have secured the maximum rate of surplus value) and all surplus value is reinvested (the rate of accumulation equals the rate of profit). Highly unrealistic, but it is an example that allows Kliman to isolate very specific effects. And it allows us to test what happens when productivity increases in department 1 feed through into the rest of the economy - note also that this test maximises the impact of two important counteracting factors, creating the most favourable conditions for a rising rate of profit.

I reach a conclusion that is the opposite of Machover’s supposition that “even if all technological change were labour-saving, overall organic composition can still go down, provided the productivity of labour in department one increases sufficiently fast”.19 Guglielmo Carchedi, who examines the same process in Behind the crisis, reaches a conclusion similar to Machover’s that higher productivity increases are more likely to lower the value of means of production in subsequent production cycles and increase the rate of profit than lower productivity increases. Carchedi goes on to argue that ultimately the cost of means of production will tend to zero and we can then expect all labour to vanish from production and there will be “no profits and no profit rate”.20 However, this is a long-term prediction (and a debatable one) about the fate of capitalism and has no bearing on the role of the rate of profit in capitalism’s cyclical crises.

In contrast, using Kliman’s example, I find there is a continued direct relationship between productivity increases and the value composition of capital (and therefore an inverse relation to the rate of profit), so long as the increase in the productivity of workers is maintained at a high enough pitch in each successive production cycle. Only if the increase in productivity slackens (or temporarily halts) do profit rates rise - this is the behaviour that Carchedi models. So, the faster the rate of continuous increase in labour productivity across the economy (even if it is concentrated in department 1), the more likely it is that the tendency for the rate of profit to fall will outstrip those counteracting factors that tend to raise profitability.

Once fixed capital is added back into the mix (Kliman’s example excludes it), moral depreciation comes into play, not only providing a possible explanation for the short-term business cycle, but also serving over longer periods, if there is an acceleration in the rate of innovation, such as the information technology revolution, to undermine existing investments and their profitability.21

The fact that rapid increases in the productivity of labour can impact on profitability, but that there are also a range of counteracting factors that can serve to raise it, gives us greater confidence that Marx’s analysis of the dynamics of capitalism (especially in the disputed volume three of Capital) provides a solid foundation for explaining capitalism’s real behaviour. Given that the contradiction between what is possible as a result of the achievements of human creativity and what capitalism delivers is becoming increasingly stark, a Marxist economics enriched with a cogent and powerful explanation of capitalism’s crisis tendencies and the turning points in its life-cycle would provide the working class with a powerful theoretical weapon in the struggle for human liberation.


1. ‘“Revolutions in value” and capitalist crisis’ November 19 2015.

2. E Farjoun and M Machover Laws of chaos: a probabilistic approach to political economy London 1983, p174.

3. N Kondratiev The long wave cycle New York 1984; L Trotsky, ‘The curve of capitalist development’ (1923): www.marxists.org/archive/trotsky/1923/04/capdevel.htm; E Mandel Late capitalism London 1975 and Long waves of capitalist development: a Marxist interpretation London 1995; and P Mason Postcapitalism: a guide to our future London 2015.

4. A Kliman The failure of capitalist production London 2012, chapter 5, pp74-101.

5. A Freeman What makes the US profit rate fall (2009), p15: https://mpra.ub.uni-muenchen.de/14147.

6. K Marx Capital Vol 3, London 1981, p350.

7. EE Maito, ‘And yet it moves (down)’ Weekly Worker August 14 2014 and The historical transience of capitalism: the downward trend in the rate of profit since the 19th century (2014): www.academia.edu/6849268/Maito_Esteban_Ezequiel_-_The_historical_transience_of_capital_The_downward_trend_in_the_rate_of_profit_since_XIX_century_final_draft_.

8. A Kliman op cit pp17-19.

9. David Harvey discusses in The limits to capital (Oxford 1982, p 199) the work of the French theorist of state monopoly capitalism, Paul Boccara, who speculated that state-owned industries which produce commodities and generate surplus value but operate at close to zero profit rates could play a similar role.

10. K Marx Capital Vol 3, p172.

11. Ibid p356.

12. M Machover, ‘Saving labour or capital?’ Weekly Worker October 6 2011.

13. K Marx Capital Vol 1, London 1990, p762.

14. K Marx Capital Vol 3, chapter 25.

15. Ibid chapter 5, pp170-99.

16. K Marx Capital Vol 1, p528.

17. A Kliman, ‘Crisis, theory and politics’ Weekly Worker September 27 2012.

18. A Kliman Reclaiming Marx’s ‘Capital’: a refutation of the myth of inconsistency Lanham 2006 - see tables 7.1 to 7.5, pp120-32.

19. M Machover op cit.

20. G Carchedi Behind the crisis: Marx’s dialectics of value and knowledge Chicago 2012, pp95-97.

21. A Kliman The failure of capitalist production: underlying causes of the great recession London 2012, pp138-48.