A Critique of ‘Human Capital’ theory

One of the staples of contemporary neoclassical theory is the use of the concept of ‘human capital’, by which it broadly means all investment into skills and education as applied to individuals or an entire population. In particular in popular neoclassical growth theories, such as those developed by Robert Solow and refined by the likes of Elhanan Helpman, human capital plays a key role. It often appears as an essential component in those theories because they tend to see growth as reliant primarily on increases in productivity, which in turn are based on the interaction between the quantities of the ‘factors of production’ (capital, land and labor, though usually just capital and labor) and the state of technology. The virtue of human capital as a concept within these theories is that they allow the technological level to be determined endogenously to a greater or lesser degree, that is, that they enable the theory to take non-given and non-constant levels of technological increase into account, and model variations in technological improvement between countries.

However, the concept of human capital also seems to play an important role in the ideological legitimations of neoclassical economics, since it allows the marginal factors of productivity approach to take on a slightly more ‘human face’. After all, the theory relies entirely at the microeconomic level on understanding the interaction of given individuals and their labor with given quantities of capital and given quantities of technology; the income earned by each factor equal to its marginal product. This implies, practically speaking, that the income of labor (wages) is equal to the marginal productivity of the unit of labor – the infamous theory which essentially in a Panglossian exercise of benevolence declares that everyone by definition earns the wages they deserve, since wages are held to equal the extra productivity the laborer adds to the firm. Not earning enough? Should have been more productive! Now this theory does not appeal very much to anyone except the best-earning wage earners, who can pride themselves on their innate superiority, all the more since the technology being given there is not much a worker can do to improve their productivity in the workplace, except perhaps work more intensely or longer hours. But our friend, the neoclassical economist, can now produce a soothing balm for our wounded pride: in fact there is also such a thing as human capital, residing in labor, and this is also a form of capital (i.e. a means of production). Moreover, since it resides in labor itself, it can be improved by the person involved, or even by the government on his or her behalf! This sounds much nicer, indeed. The problem is stated, the solution is offered. What more can one demand of the economist?

Unfortunately, although the concept sounds appealing and has even been tempting to some of the social-democratic trend of thought because of its implications for state investment and its correlation with growth – simply put, spend more on educating your citizens and you get not just increased output, but because higher skill and education levels themselves improve technology endogenously, one can even get positive externalities from investment in education (1). Also, in fields such as education studies and in sociology the concept has become popular here and there, often being used as a way to fill in a gap in existing theories or by representing that which is considered desirable in terms of personal skills or development in terms of human capital. Notwithstanding this ‘left’ application of neoclassical growth theory, there are good reasons to object to the use of ‘human capital’ as a concept altogether, and it is quite more nefarious than it may seem at first instance.

The most important thing to realize about the concept of human capital is the direct meaning of the term itself: human capital. Conceiving of skills and education as an aspect of labor in fact reinforces the understanding of labor itself as a commodity, which happens to reside in human beings individually. This has two important effects. First, the individual human’s labor has under neoclassical theory already been assumed to be a commodity, a mere instrument for the production of capital. Now, even the aspects of the carrier of labor, the human indidivual, which were hitherto considered to be given and outside the sphere of neoclassical capital theory are now subsumed within it. All skills, all education, in fact any number of intangibles relating to character, outlook and so forth can now be subsumed into the theory as means of production. While this is only an abstraction in theory, it reflects the very real subsumption of such attributes to capitalist production in actual life. Secondly, although it is considered separately as a ‘factor of production’, the use of labor as a factor in reality only differentiates itself in that it is considered to apply to employees as opposed to employers, with the income on marginal labor therefore being wage and the income on marginal capital being profit. But it is not otherwise differentiated – each is equally seen in terms of being a means of production. This means that it is in principle not even necessary for such ‘human capital’ to reside in individual labor at all – it may as well be an attribute of society altogether, or of culture, or any number of vague and general determinants. This is in fact often the way it is viewed in growth theory, where human capital mainly shows itself in the shape of general ‘stocks of knowledge’. The consequences are very clear in terms of theory: on the one hand, human capital production is as much as any other means of production the result of private, individual investment, even if such is done by the state. If you are poorly educated relative to the market, you have made bad investment choices; equally, if your skills are out of date, or even your character or culture is not conducive to profit, your human capital levels are by definition low, since they constitute small increases in productivity as an attribute of your labor. Even when measured in terms of years of schooling, human capital measures still weight these by productivity under the normal conditions of declining returns on investment. In other words, it’s yet another form of individualizing the social processes of human life and fostering a competitive, blame the victim mentality. This is also reflected eventually in the schooling and skill-building processes themselves, which under the influence of human capital theory (which first became truly influential in the 1960s) have aimed to foster the buildup of ‘stocks’ of such human capital. This means in practice that they orient themselves to teaching skills which will make these units of labor, humans, more productive in the workplace.

This is just the minor objection, but it brings us at once to the greater objection. The nefariousness of this manner of thinking resides specifically in the general failure of neoclassical economic theory to distinguish between labor and labor-power as attributes of workers.(2) Since it does not distinguish different types of input, especially in aggregate, and reduces them all to undefined ‘units’ of means of production, it completely ignores the real-worldly attributes of the ‘factors’ it deals with. A worker can only ever sell their labor power by contract to a capitalist for a given amount of time; his labor, that is the specific, qualitative aspects of work (which exist under all circumstances and in all civilizations) is another thing. Failing to distinguish between these means in the context of human capital that those attributes seen as improvements in the qualitative nature of the factor labor are in reality ‘enhancements’ of labor-power. When your schooling and your skills increase your human capital, what they do in reality is enhance your abilities to produce capital, in the real and precise sense of the word, during the period in which you apply your labor-power. Any ‘improved’ human capital means improved from the point of view of capital, not from the point of view of the human involved. In fact, for capital it does not matter at all whether a human is involved or not as productivity is concerned, which is why its ideology, neoclassical economics, cannot distinguish human from nonhuman inputs. This raises a second issue, namely that because labor itself is done by humans, labor-power also resides in humans. The real struggles in the workplace between labor and capital take place precisely because the human and his labor are not commodities, and cannot be commodities except under conditions of actual slavery. In the qualitative process of labor, those humans will therefore attempt to resist their labor-power being used solely to produce capital and as much as possible in as little time as possible. Far from being mere inputs into a process, to be bought and discarded at will by ‘entrepreneurs’ and to be ‘rewarded’ for their productivity in helping make a profit by a wage equal to their ability to offer such ‘help’, human beings resist being commodified and resist working for someone else’s benefit. The concept of human capital therefore inherently passes by the class struggle and the real forms it takes in the workplace. In fact, workers can have high skills and great knowledge of production processes they are involved in at their work, but low human capital, because they are not inclined to maximally use them to help their employer make more profits!

Finally, human capital as a form of capital is subject, like all neoclassical capital theory, to the Cambridge capital critique. Although it would be too technical to go into the details here, the gist of the matter is that although growth theory and neoclassical macroeconomics generally rely on aggregates of the factors of production it uses, in fact these factors of production are mutually interdependent and qualitatively differentiated. In order to aggregate them it becomes necessary to understand all of them as one single, undefined ‘capital’. For any modelling of such aggregates, the component parts can only aggregate to a whole when each of the parts has the exact same attributes in its exogenous aspect (usually technology, but basically any qualitative aspects of production). Since those qualitative aspects differ strongly per branch of industry and even per firm, for example in terms of capital intensity (ratio of capital goods per unit of labor) the aggregation is not possible except under wildly unrealistic assumptions. Now why is this relevant? Because human capital as a notion of a quality of the input labor, to be aggregated, requires exactly the same assumptions. This is quite obvious in practice – even measurements in something as easily measured as years of schooling will fail altogether if one aggregates years studied at Harvard with years studied at Phoenix University online, as everyone will readily recognize. Even more so when human capital is interpreted broadly to include any traits and skills that may increase one’s ability to produce output – how does one aggregate the measure of people’s enjoyment of overtime work, say? Clearly any growth models based on such assumptions must fail altogether.

The idea of investing in human capital, either individually or as state policy, in order to produce readily aggregable units of human capital is offensive to all notions of human development and flourishing. Since one can wonder what the point of any economic policy whatever is if it is not conducive precisely to these traits, any growth model or any policy aiming to increase human capital must be rejected as falsely stating the question. Human capital is a notion deeply retrograde from the point of view of seeing ‘the full development of each as the prerequisite for the full development of all’, it attempts to further subsume all aspects of the human being and his or her personality to the accumulation of capital, and it reduces all human attributes to qualities of a single commodity. The actual one thing that all capital has in common is that it is dead labor dominating live labor, as Marx pointed out and the Cambridge controversy eventually also concluded.(3) As we resist the rule of dead labor over our own, so we must resist any outlook which further enables it to do so by further commodifying all aspects of life. Resist human capital.

1) See: Robert Lucas, “On the Mechanics of Economic Development”, in: Journal of Monetary Economics 22:3-42 (1988).
2) See also: Samuel Bowles & Herbert Gintis, “The Problem with Human Capital Theory–A Marxian Critique”. American Economic Review (May 1, 1975), p. 74.
3) Piero Sraffa’s work, the main basis of the Cambridge capital critique, in fact determined the fact of the rate of profit on capital to reside not in aggregates of prices of capital, but in a rent based on the social power that holders of capital have over those who do not hold capital – in other words, in the monopoly of the means of production, exactly as Marx had said more than half a century earlier.