G.A. Cohen on the output bias

(Originally drafted for a conference at Frankfurt in 2018 to mark the 40^th anniversary of Karl Marx’s Theory of History: A Defence. I’ve done a bit of editing of my conference script and added a few footnotes etc, but it isn’t necessarily produced to the scholarly standards one might require of a journal article.)

In Karl Marx’s Theory of History, G.A. Cohen attributed many of the ills of capitalism to the market mechanism. Later in his career he came to see the market as practically ineliminable. Insofar as he was right about the market in his earlier work, it may turn out that the alternatives to capitalism he championed at the end of his life will also generate the pathology he deplored: the systematic bias in favour of output over leisure and free time. The following explores some of these tensions.

Introduction

In the second half of his career, G.A. Cohen concentrated his discussion of capitalism on its wrongs and injustices. According to his diagnosis, the primary injustice in capitalism arose from the combination of private property and self-ownership, which enables capitalists – who own the means of production – to contract with workers – who own only themselves and their labour power, on terms massively to the capitalists’ advantage. The workers, who produce nearly all of the commodities that possess value in a capitalist society, see the things that they have produced appropriated and turned against them as tools of exploitation and domination by the capitalists. But the wrongness and injustice of capitalism, the theft of what rightfully belongs to workers, is only one part of what is to be deplored about capitalism. In chapter 11 of Karl Marx’s Theory of History, a chapter where he went beyond the expository and reconstructive work he undertook earlier in the book, Cohen articulated a different critique, this time focused not on injustice but on the ills to which capitalism gives rise. In that chapter he attacks capitalism for stunting human potential through a bias towards the maximization of output, a bias which condemns human beings to lives dominated by drudgery and toil. Relatedly, he attacks capitalism both for stimulating demand for consumption that adds little of real value to people’s lives and because for damaging of the natural environment through pollution. In developing this critique, Cohen also notes that the bias towards output he identifies is celebrated by Max Weber as exemplifying rationality itself, a celebration which Cohen thought ideological and mistaken.¹

Though both the wrongness and the badness of capitalism arise from the conjunction of private property and the market, it seems natural to emphasize the role of private property more in the production of injustice and to stress market relations more in the genesis of its badness. It is the fact of what the capitalists own that gives them decisive leverage over workers in the labour market, making exploitation within the workplace consequently possible; it is the market that compels everyone, capitalists and workers both on pain of extinction, to act in ways that end up being so destructive for human and planetary well-being.

This difference in effect between private property in the means of production on the one hand and the market on the other is potentially consequential in the following way: the elimination of private property in the means of production, as envisaged by market socialism, or its wide dispersal among individuals (as envisaged by some versions of a Rawlsian property-owning democracy) could put an end to capitalist injustice and exploitation. But insofar as market relations cannot be eliminated, the ills that arise purely or largely from those relations may also be ineliminable.

In chapter 11, Cohen attributes the output bias to market relations. Because at that stage of his career he had thought less about the arguments around planning and markets than he did later – as reflected, for example in his final work Why Not Socialism? – the chapter is far too sanguine about the prospects for escaping drudgery and excessive production. This optimism seems even more marked when we notice that Cohen is tacitly and unjustifiably presupposing in this chapter a framework for understanding social relations and political action that has as its principal players capitalists and workers on the territory of a single state and therefore neglects the contribution of interstate rivalry and other dimensions of the international order to the output bias.

Cohen’s argument comes in two parts. First, he diagnoses the bias towards greater output and consumption over leisure as having its roots in market competition. Second, he asks why this bias, which is a mere systemic tendency for certain effects to be realized, is in fact realized rather than counteracted by other pressures such as political and trade union action.

The output bias as a systemic tendency

In chapter 11, Cohen writes that he presents “contemporary ‘affluent’ capitalism as the climax of the history of use-value and exchange-value, and as the prelude to the suppression of exchange-value as regulator of the social organism.” (297). After an opening section in which he gives an idealized account of a sequence that leads from the very first human production (which aims at use-value) to modern capitalism that aims at the “appropriation of value, of money, of abstract wealth.” (302, citing Theories of Surplus Value ii, 503). In such a system, he argues,

Whatever goals a corporation has, it must, on pain of bankruptcy, obey the imperative: seek to expand the exchange value at your disposal. This holds whether firms aim at profit or at growth, for they are different modes of increasing exchange-value, and no firm can aim at neither. It also holds whether or not managerialist theses are true: whoever wields ultimate corporate power is constrained to favour decisions which enlarge the difference between M and M¹. (302)

Capitalism, Cohen argues (following Marx), develops technological power “on an unprecedented scale, progressing at an unprecedented rate” “because the competitive position of its industrial decision-makers compels them to increase the productivity of the production processes” (303). He notes that such increases in productivity could in principle be directed in two ways. On the one hand, they could be used to reduce toil and to give people more leisure for the same level of output. On the other, those gains in productivity could be used to generate more output, with workers labouring just as much as they did before. (And, of course, some combination of the two is possible, with workers working a bit less than they had previously to produce more than they did before, though not as much as they could if their hours had remained undiminished.)

Cohen then makes the following claim, which he italicizes throughout to emphasize its importance to his argument:
Now capitalism inherently tends to promote just one of the options, output expansion, since the other, toil reduction, threatens a sacrifice of the profit associated with increased output and sales, and hence a loss of competitive strength. (304)

Supposing this claim to be true so that capitalism does indeed generate the output bias with its attendant pathologies, this does not establish that the abolition or supercession of capitalism would eliminate it. As Cohen puts it (313) capitalism might be a sufficient condition for the bias, but it might not be a necessary one. He observes that

Capitalism is a society whose production is governed by a capitalist principle and which exhibits a division between a capitalist and a labouring class. But capitalism’s preference for output depends largely on the first feature. We may envisage class undivided societies whose production follows the capitalist principle. (314)

He invites us to consider three such societies: first, “a market economy of self-employed producers who hire no labour”; second, “a set of firms each of which is wholly owned by its employees, who share fairly equally the firm’s income”; and a third, which Cohen calls “people’s capitalism” and which to my mind resembles Rawls’s “property-owning democracy”.² This is a society which “differs from the second in that share ownership is not restricted to the employee’s firm. In addition, in this “Rawlsian” society, in order to keep things egalitarian, there is a duty of all to work and an income ceiling aimed at preventing the formation of distinct classes.” (314–5)

Cohen dismisses the first form of society as “incompatible with really advanced technology” and focuses his attention on forms two and three which he considers together.

In KMTH he argues both that there are no historical examples of such societies and also asserts that they are “not reasonable options now” (315). The argument to this effect is brief, but it is essentially that some firms would be more successful than others and that the unsuccessful ones would either be driven out of business or would end up as subordinated subcontractors of the successful ones. Although there could be institutional mechanisms, such as insurance schemes, that might mitigate these effects, they would either be unsuccessful in doing so or they would have the effect of undermining the fundamentally market character of the society.³

Contrasted with both capitalism and these hypothetical non-capitalist market societies is a non-market society with comprehensive economic planning in which “decision-makers are free, as far as systematic economic constraints are concerned, to choose between expanding output and reducing labour, when there is progress in productivity.” (315) Cohen concedes, in passing that actual planned economies have been rather disappointing in their preference for output over leisure. (And in fact we now know, perhaps better than we appreciated in 1978, that in their attempts to compete with the West, Soviet planners both increased burdensome toil and artificially restricted consumption whilst generating enormous environmental problems.) But he claims that “Political change in the Soviet Union could lead to altered priorities with no dramatic reshaping of its economic system.” (316)

Could social democracy mitigate the output bias?

I shall assume that Cohen is right about the tendency of market societies, including, and especially, capitalist societies to promote output over leisure and hence to produce longer working hours than would obtain if productivity gains were channeled into freeing people from burdensome toil rather than maintaining the toil and producing more stuff. Still, as mentioned earlier, this is a mere tendency which might not be realized in practice if countervailing forces were to prevent this from happening. Such countervailing forces might have their origins in the desire of workers to have more leisure and forego consumption to get it, and such desires might become effective through social-democratic legislation, through trade union action or even through the market power of some, probably skilled, workers who could insist on a shorter working week as one of their contractual terms. Cohen suggests that if such restrictions on working time could be generally enforced, then they might take place with little or no detriment to capitalist firms because the restrictions would apply to all such firms and no firm would be put at a competitive disadvantage by them. ⁴
That capitalism has the bias that Cohen attributes to it in actuality and not just in theory seemed evident to him from an examination of recent history. Though workers were not working such long hours in 1978 as they did in 19^th century factories, working hours had not fallen all that much since the 1920s and these are reductions from an extreme lengthening of the working day that capitalism itself had caused. Allowing for things like religious festivals in the Middle Ages, Cohen claims that modern workers now (that is, in 1978) work as long as medieval peasants did.

Nor have things let up much since the publication of Karl Marx’s Theory of History.⁵ Enormous gains in productivity consequent on revolutions in computing and information technology have typically been directed not at the relief of burdensome toil, but, among other things at increasing the intensity and sometimes duration of labour through greater control and surveillance of the workforce, as with pickers in Amazon warehouses. Sarah O’Connor reports in the Financial Times :

In the UK, the government funds a vast survey of UK workers every five years to keep track of how the quality of work is changing. The latest results show Britons are working harder than at any time in the past 25 years, to tighter deadlines and with less autonomy.⁶

Even as the nature of the workforce has changed, with fewer people in advanced capitalist economies engaged in industrial production than before and more in “white-collar” jobs, the character of jobs in the expanded sectors has become more routinized and proletarianized in character and experience. There were far fewer university professors, for example, in 1950 than there are today, but those that there were worked shorter hours and were subject to less controlling management than the expanded army of university teachers are today (and we are, of course, a relatively privileged sector).

Capitalism, Cohen argues, once had a progressive role. It was an “indispensable engine for producing material wealth from a position of scarcity.” But because of the output bias, capitalism cannot “realize the possibilities of liberation it creates”. “It brings society to the threshold of abundance and locks the door.” (306-7).

One way of explaining the failure of workers to push back on working time and to favour leisure over consumption is that they are merely duped by ideology to want more and more things rather than increased free time, even though more free time would be better for them than more things. Sensibly, Cohen doesn’t rely on such a claim though he thinks it has more going for it than people generally suppose.

What then is the explanation? Why do workers typically push for more consumption rather than for shorter hours? Here I think Cohen’s argument is untypically unclear. He tells us that workers do indeed want more goods (other things being equal) but it is also plain that they dislike the toil then need to engage in to get those goods and evidences that claim by making the point that if workers were granted money such that they didn’t need to work at their jobs, very few of them would choose to do so gratis, and there would be a dramatic decline in the amount of work done. His suggestion seems to be that since there is a permanent propaganda campaign in support of people getting more stuff and no corresponding campaign in favour of free time of the kind one can enjoy without extra spending, extra consumption always has a salience for people that additional leisure lacks.

I think Cohen’s uncharacteristically feeble discussion of this point can be explained in part by a blindspot in his thinking. First of all, he neglects the way in which states support and encourage consumer spending and simultaneously encourage a culture of work and effort through moralizing campaigns, stigmatization of “shirkers” and the like. Take the first of these elements: states help establish and support structures of lending to help people to do things like buy their own homes and then furnish their homes. They do this in a variety of ways, by providing support to banks and other financial institutions but also by giving borrowers all kinds of tax breaks. In order to buy the things they want, consumers have little choice but to borrow, but once they have borrowed they have to repay and repaying requires income, which requires work. Now perhaps workers could simply refuse the blandishments of capital and refuse to borrow in order to consume and then the option to work less would be more available to them. But I think this neglects the competitive and social aspects of consumption that Cohen rather dismissively refers to as “keeping up with the Joneses”. One need not be of a particularly competitive or comparative disposition to be caught up in social standards of self-presentation that are inherently comparative in nature, as Adam Smith knew in that famous passage about labourers being ashamed to appear in public without a linen shirt or, in some countries, leather shoes.⁷

Second, there is the question of why states support the bias towards output over leisure in the way that they do. Cohen would certainly have one part of the answer readily to hand given that the states with which he is concerned are capitalist states and capitalist states have as their primary function, in the historical materialist story he tells in KMTH, the defence and promotion of the interests of capital. Since capitalist firms prefer greater output to less, want their workers to put in more time and effort, to be “committed” and have the right attitudes, it is unsurprising that capitalist states echo and reinforce this preference. But were capitalist states isolated from one another, we could expect this tendency to be moderated. After all, the capitalist state also has the function of saving capitalists from large-scale collective action problems in which the pursuit of profit by each leaves them all worse off than they otherwise might be, and to this end states impose common standards and regulations on capitalists. Given that the output bias is systemically driven by competition among capitalist firms, it might be moderated by state-imposed restrictions on working hours that hold for all capitalists. (And indeed, historically, the state has imposed such limits on the employment of labour in order to prevent inter-capitalist competition from degrading, through overwork, the human capital on which capitalism depends.)

Unfortunately, Cohen writes, in this chapter at least, as if capitalist states and societies were isolated entities. The dramatis personae in his story are capitalist firms and workers co-existing within a single society. But societies are not isolated from one another and this matters in various different ways. First, inter-capitalist competition takes place across national boundaries, so the ability of states to mitigate that by enforcing common standards is limited. Second, the fact that capital is mobile across international boundaries strengthens the bargaining position of capitalists against workers, such that demand by workers to earn more or work less can often be credibly countered by a threat to move production to a different jurisdiction. Third, and I think this is of the greatest importance, states are and conceive themselves to be, in competition with one another. This means they care about their relative economic strength and wealth because this tends to translate to relative power in the international arena. This will lead them to favour increased output over increased leisure for their populations, at least up to the point where the physical capacities of potential soldiers are not too badly damaged. Insofar as increased consumption fosters economic development more generally, it will also lead them to favour consumption over leisure. But if it is necessary to suppress consumption also, perhaps to divert effort into infrastructure development or defence spending, states will do that too. Though Cohen’s attention was focused on pressures arising from competition among capitalist firms, it is plain that these competitive pressures among states also apply to non-market societies, hence the choice of the Soviet Union (as mentioned above) to promote hard work and labour discipline.

In the final substantive section of the chapter, Cohen takes issue with Max Weber for some remarks Weber makes which appear to identify the bias in favour of output over leisure with rationality as such. Weber comments on how, only a generation previously it was

futile to double the wages of an agricultural labourer in Silesia … in the hope of inducing him to increase his exertions. He would simply have reduced by half the work expended because with this half he would have been able [to earn as much as before]. (321)

Cohen was surely right about this, and indeed we see this Silesian behavior reproduced in more recent periods of capitalist society by individuals such as doctors in general practice in the UK’s National Health Service who, after securing a large increase paid for their services, were able to hire locums to do much of their work and to use the time saved to play some extra rounds on the golf course.

Cohen embraces the market
Central to Cohen’s case that capitalism exhibits a bias towards output and against leisure was the market mechanism. And we have seen that in Karl Marx’s Theory of History he dismisses as infeasible market-based alternatives to capitalism whilst suggesting that they too would exhibit the same bias as productive units were disciplined by market forces. There, it is a system of comprehensive planning that stands as feasible alternative to capitalism and as a society in which the choice to favour leisure over output is genuinely socially available. But at the end of his life Cohen had basically conceded that market mechanisms are indispensable in a complex society. In Why not Socialism? he writes:

… once we pass out of the sphere of need, or, more generally, of goods that everyone can be expected to want, to the wide sphere of optional commodities, and we pass increasingly to that as economies progress and as life therefore becomes easier and more elegant, it also becomes difficult to know what to produce, and how to produce it, without the device of market signals: very few socialist economists would now dissent from that proposition. (WnS, 60).

He had also cooled on central planning, writing:

Central planning, at least as practiced in the past, is, we now know, a poor recipe for economic success, at any rate once a society has provided itself with the essentials of a modern productive system. (WnS, 67).

In the light of the failures of central planning, Cohen’s revision of his opposition to the market is understandable. But as far as I know, he never revisited his claims about the output bias in the light of his subsequent embrace of the market, and that is a silence that his admirers have reason to be disappointed by.

Two conclusions

There seem to be two conclusions open to us on the basis of this discussion: call these the optimistic conclusion and the pessimistic conclusion. According to the optimistic conclusion, the variation in labour-time reduction in some countries and the remarkable gains made in some of those countries during the twentieth century, often at the behest of social-democratic governments and trade unions and made in the context of what have been essentially market-based economies, tells us that Cohen was wrong in some of his exposition. Even if market economies do have a bias towards output and consumption over leisure, there is more that we can do to tame them than Cohen thought and workers can and will use political action and trade union bargaining to reduce hours and to improve working conditions so that work is less onerous. And presumably we can expect that the same political levers that are effective in reducing burdensome toil can also be brought to bear to mitigate the environmental damage inflicted by consumption.

We will reach the pessimistic conclusion, on the other hand, if we agree with Cohen about the strong consumption bias of market economies, think that markets are something we just cannot do without, and think that political action to suppress the bias will always be ineffective, at least in the long-term. And if Bowles and Park are right in the paper cited above, if we fail to suppress capitalist inequality, that will shape worker’s desires in consumption (and therefore output) -favouring ways that will make pressure for shorter hours less likely. If we also believe that inter-state competition will lead state policy to favour greater output and that states that refuse to play this game will find themselves outcompeted and undermined by those that do, rather in the manner that English capitalism as depicted in the Manifesto used its cheaper and better goods to force countries stuck in antiquated social forms to transform the way they were organized, then will perhaps think that Max Weber (whose views on rationality and work Cohen rightly castigates) was a better predictor of the future than Marx when he wrote in The Protestant Ethic and the Spirit of Capitalism:

This [capitalist] order is now bound to the technical and economic conditions of machine production which today determine the lives of all the individuals who are born into this mechanism, not only those directly concerned with economic acquisition, with irresistible force. Perhaps it will so determine them until the last ton of fossilized coal is burnt. In Baxter’s view the care for external goods should only lie on the shoulders of the “saint like a light cloak, which can be thrown aside at any moment.” But fate decreed that the cloak should become an iron cage.⁸