I. The Problem – A spectre is haunting the industrial and post-industrial world — the spectre of job insecurity. While globalization explains some job insecurity (jobs have been shifted from the more- to the less-developed parts of the world), the greater threat emerges from artificial intelligence (AI), computerization, and robotization, which threaten jobs everywhere. At many McDonalds, one can order a Coke and pay with a Visa card without interacting with human beings via computerized interfaces. At some restaurants it is literally a tablet that has displaced your friendly waiter. Robots weld together cars more efficiently and more reliably than did human welders. Even in dentistry, 3-D printers have driven traditional creators of dental appliances from the market. Whose job is next for elimination?
A. Theoretically – The central thesis to be elaborated here is that the internal logic of our ever-modernizing system of production is carrying us towards a future where virtually everything we think of as “work” will vanish. The central insights of the most prominent political economists will be employed to understand better this dilemma. The dilemma is:a) either we stop modernizing to save jobs or b) we keep modernizing and eliminate vast swaths of jobs. Either way, there is a danger of systemic collapse with unpredictable social consequences for which we must prepare. The US election of 2016, Brexit, the rise of right-wing parties in Europe, and recent French protests illustrate the beginnings of this dynamic as legions threatened by globalization and modernization angrily expressed their angst.
At the beginning of the capitalist era, Adam Smith proposed that the wealth of nations is dependent upon (actually, is a function of) the productivity of its workers. The productivity of workers, in turn, was a function of:a) the quality of labor (skills, literacy, work ethic, etc.) and b) investment in capital goods (machinery which enabled workers to produce finer goods more quickly). This insight became the foundation of the “Labor Theory of Value”. The conundrum will arrive when, at a tipping point, capital will prefer to purchase “smart” machines, replacing labor on a massive scale.
We are rapidly approaching the day when, alongside cars “smart” enough to drive themselves, the production lines could also become autonomous. Robots already exist that could, in principle, create and build other robots, and diagnose and repair one another. Artificial intelligence (AI) only compounds the difficulty. We could all wake up one morning and find that we need not report to (manufacturing and related) jobs; labor could become irrelevant. This creates three related problems. First, if workers will not be needed for production, would they have any claims for wages? A new type of theory of social justice would need to be developed. Second, unless the irrelevant workers were to have some source of income, they would not be able to generate the effective demand that drives production in an economy in which 66-68% of economic activity is domestic consumption. The economic system would become so efficient that it would collapse under its own weight; without effective demand, productive machinery would rest idle. Third, and most seriously, although our initial reaction might be celebratory (Yay, every day is a weekend day!), the greatest concern is that surplus workers would lose self-esteem because they could not produce anything of value. But, as Marx saw correctly, humans need to produce something of value to feel that their lives are worthwhile. This is the central point of the principal type of Marxian alienation, namely alienation from oneself as a productive being. The principal sin of capitalism is that we have to take what is a sacred part of our “species-being” — our productive capacity — and sell it on a labor market. This dehumanization is akin, Marx continues, to selling one’s reproductive capacities on a street corner. The crucial difference between feudal exploitation and capitalist alienation is that now we must be our own pimps. But if modernization succeeds in eliminating our jobs, we may not be able to pimp ourselves any more; every employer will have robo-lovers
It would appear to be a peculiar kind of folly, during a period of high employment rates, to be concerned about this loss of jobs. Nevertheless, jobs of all sorts are disappearing. For example, between Thanksgiving and Christmas of 2018, General Motors announced the elimination of 14,000 positions. Of these, 6,000 were blue-collar and 8,000 white-collar, so we can no longer assume that only low-skill workers are endangered by the effects of globalization, mechanization, robotization, and artificial intelligence (AI). Some were engineers who had devoted their lives to figuring out how to make internal combustion engines cleaner and more efficient, skills no longer needed with the advent of electric and autonomous cars. While unemployment is low by recent standards, none of us is immune to these threats of modernity. Even jobs that appear immune may not be; engineers could be replaced by AI, educators could be replaced by “distance” education, physicians and surgeons may succumb to robotization and data-based diagnostics. Bank tellers and meal servers are beginning to be replaced by ATMs and tablet computers. In the middle of a shortage of truck drivers, autonomous trucks are already delivering freight. Computerized robots, via networking, “talk” to one another in production lines that have become cyber systems. And, if we want superior batteries to power cars and trucks, it may be that AI can get there before a team of engineers.
According to the Demographics Research Group, Census data reveal certain overall patterns of occupational loss. On the basis of census data, between 1910 and 2000, the percentage of farm laborers dropped from about 18% to about 2% of the population. In the same period, the percentage of farmers dropped from about 16% to about 2%. Private household servants dropped from about 7% to about 1%; laborers dropped from a bit over 10% to about 1%. Others, like boarding-house keepers, switch-board operators, manufacturers of horse drawn carriages, and elevator operators have disappeared. Similarly, manufacturing and other blue-collar occupations dropped dramatically from 1920-2010; only sales, clerical, and professional occupations showed relative growth. More recent developments, like the elimination of white-collar jobs in the automotive industry suggest that even the professional and managerial occupations are endangered. While innovators, professional and managerial positions will become an ever larger proportion of the work force, they, like the workforce in general, will be declining in absolute size. Can an economy employing only relatively few innovators and certain service professionals be sustained?
II. Tools to Understand the Problem More Deeply
To address this question in a systematic way, let us consider what the great political economists have taught us and use their insights to investigate the possibility that, as capital investment in automation increases, many occupational paths will become irrelevant. It might seem utopian to believe that most of us need not appear for work on any given day, but two intransigent problems make the picture more dystopian. First, there is the question of distributive justice:if most people have no jobs, how will they support themselves? What impact would there be on an economy that depends on domestic consumption for 66-68% of its economic activity? Could we pay people not to work just to keep the engine running? Or would we take a Social Darwinist perspective and simply put them out of sight? Second, and more importantly, what would be the damage to the self-esteem of superfluous, unemployable masses? How could they retain their dignity, their sense of self-worth without creating something of social value?
As mentioned, Adam Smith argued that the wealth of nations depends directly upon the productivity of their workers which, in turn, depends upon the collective skills of workers and a willingness of entrepreneurs to invest in the machinery that maximizes outputs while minimizing inputs (the definition of economic efficiency). We may be, as the Moody Blues put it, “on the threshold of a dream,” a dream in which no one needs to appear for work any more. Smith’s all-consuming optimism about the viability of a capitalist system provides us with little guidance in a new capitalism in which there will be declining need for workers to create things of value. In Smith’s world, given growth, a lack of increasing demand for labor is inconceivable. When labor is not required for the production of goods, the Labor Theory of Value will implode. Smith cannot provide much help in confronting this new quandary.
Malthus, by contrast, was darkly pessimistic about Smith’s vision of capitalism. In his famous essay on population, he posited that it would grow geometrically while the means to support population growth would only increase arithmetically; he also foresaw episodes in which markets would be faced with “gluts” which would result in negative growth (recessions). In his vision of the future, populations would suffer terribly, by means of famine, wars, and pestilence. In our post-industrial world, the problems merged:we can produce more than we (can afford to) consume while the “gluts” that he feared as occasional dangers of a capitalist economy will become endemic. Malthus leaves us with a gloomy feeling that there is no solution to our post-modern quandary.
Ricardo, who shared Smith’s optimism but sought only to “correct” Smith, effectively revised away Smith’s paradigm. Smith had assumed, perhaps incorrectly, that rent on land and profits for entrepreneurs were relatively constant, so that the amount of skilled labor needed to create a good was the prime determinant of its value. Yet, although value defined this way would be resistant to change, Ricardo observed that prices vary; how can that be? Ricardo “improved” upon Smith by distinguishing between several notions of value as they impacted the Labor Theory of Value. Without intending to, he laid siege to its foundations. After all of these distinctions, one can rightly wonder how much of the Labor Theory of Value remains intact. Most importantly, Ricardo recognized that the “market price” of goods may at times differ from “their primary and natural price.” Paradoxically, if markets drive prices, the Labor Theory of Value is of even less utility and validity. If, then, the demand (or need) for labor approaches zero, then, in a post-job capitalist system, the price of labor approaches zero, and the marginalized workers who may be lucky enough to find jobs would find that their salaries are exceedingly low, perhaps not even enough for subsistence. All of the proceeds of sales would flow to capital. Ricardo, then, despite sharing Smith’s enthusiasm, offers us little help in dealing with our current dilemma.
Marx interrupted this line of thought by working out some of the inner contradictions of capitalist theory. Using the assumptions of Smith, Ricardo, and Malthus, Marx 1) developed theories of several types of worker alienation and exploitation and then 2) demonstrated that capitalist markets are inherently inclined toward periods of boom and bust. (Following Marx, we now call this series of alternations “the business cycle.”) One implication of the business cycle is that, during periods of “bust,” marginal firms will fail and their capital will be acquired by ever-growing behemoths. At some point, this widening gap between the hyper-wealthy and the oppressed workers would motivate a social revolution that would remake the mode of production. So far, however, workers of the world have not united. Nevertheless, if a socialist workers’ paradise were one without productive work for all who can and want to work, it would fail to meet our human needs, as Marx described them. (Signs of mental health are correspondingly negative for involuntary joblessness).
Partly in response to Marx, “neo-classical” economists sought to restore Smith’s optimism and to resolve his “paradox of water and diamonds.” In Smith’s “paradox,” the demand for water — a necessity of life — should be quite high but is not. (Let us abstract for the moment from concerns about supply because, for instance, there are many very scarce but fine gemstones — like tourmaline and peridot — whose price is low compared to diamonds.) Meanwhile, the demand for diamonds (seen as mere trinkets or ingredients for industrial bits) should be very small but is not. Concentrating on the demand for a good at the edges or margins of the market, where an individual is trying to decide whether or not to buy one more, explains this anomaly. What drives decisions at the margins is the comparative utility of marginal units, or marginal utility. Normally, the marginal utility of water (the utility of the next gallon) is quite low and we will be disinclined to spend hundreds of dollars for small amounts. But for those in love but lacking a ring, the marginal utility of a diamond is quite high and we are inclined to spend hundreds of dollars for a small amount. Interestingly, then, aggregate markets are frequently defined at their edges by individual decisions. Less exotically, the prices of sugar and salt are not set simply by global supply of and demand for sugar, commodities for which prices are relatively stable. They are set at the point at which consumers derive less utility from marginal units of sugar than from marginal units of salt. Alfred Marshall developed this powerful notion of “marginality,” the belief that pricing is set at the margins of the market, as one is deciding to buy or sell an additional-marginal-unit (or units) of a good.
Marshall’s famous Law of Diminishing Marginal Utility — the closest thing in the social sciences to the Law of Gravity — illustrates but does not resolve our problem. The law mandates that there is always a point at which additional units of any good do not provide as much utility (or joy) as previous units had; it is frequently graphed as a rising curve that peaks and then declines, implying that too much of the good might actually have negative marginal utility. (There is such a thing as too much ice cream.) Applied to our problem, it seems apparent that the investments of the capital class in units of capital have — in certain locales — reached a point at which capital would prefer to purchase a device over hiring another worker. (Devices do not need wages, insurance, vacations, sick days or pensions; investing in them is a comparatively good investment even if the first year’s expense is significantly higher than the annual cost of a marginal unit of labor, especially given that its cost can be amortized over decades).
Addressing concerns about the Marxian business cycle, both Keynes and Friedman sought ways to minimize its extreme swings. In other words, both assumed that the business cycle is an integral propensity of capitalist economies; they merely disagreed on the method of controlling it, so that we might approach long-term, stable growth. Keynes preferred to control the economy by prudent adjustments to governmental expenditures; when the economy is “cold,” the government should spend more (via deficit spending) to augment effective demand; when it is “hot,” governments should reduce spending by turning surpluses to dampen excessive demand. Friedman, by contrast, believed that when an economy is cold, government should increase effective money supply; when it is too hot, government should reduce money supply. However, both operate at such a macro level, concerned with systemic growth, that they lose sight of workers and their job security; neither offers a solution to job insecurity caused by mechanization.
We have, in short, a dilemma: if capital does not invest in new technology, worker productivity will stagnate and other economies will out-compete us; if capital does invest in new technology, workers will become irrelevant and earn no salaries, again leading the economy to implode for lack of effective demand.
Practically, it is hard to predict when this crisis will manifest itself. Currently, there are parts of the (developing) world where human labor is still cheaper than the machines to replace them; the point of diminishing marginal utility of labor lies still ahead of us. Already, though, Chinese labor is becoming more expensive; even Vietnamese labor is rising in cost. At some point, machines will be less expensive than manual labor; additional units of labor will have negative utility while additional units of machinery will be cost-effective. Are we prepared for all of the social consequences?
Robert Kocis is a professor emeritus from the Department of Politics at the University of Scranton. Email: firstname.lastname@example.org.
- The central irony of the early capitalism — mercantilism — was to explain that while Spain brought back galleons of gold and silver from the new world, this money quickly wound up in the coffers of England — because Spanish consumers demanded English goods presumed to be cheaper and of higher quality. And English goods were superior and cheaper at that time because conditions in England made workers far more efficient than their Spanish counterparts.
- Some occupations are more resistant to automation than are others. The innovators, for instance, may remain necessary; the only threat to them is AI. Certain service providers (social workers, psychologists and psychiatrists) and professionals (workers in the system of justice) and supervisors (to oversee the robots and direct them to make goods that are in demand) are instances of those more immune to job insecurity. (A somewhat more advanced AI could eventually threaten even these occupations). For a somewhat different, and less worrisome, perspective,
- Federal Reserve Bank of St. Louis: https://fred.stlouisfed.org/series/DPCERE1Q156NBEA.
- Marx, Karl. Robert Tucker, ed. The Marx-Engels Reader 1978 New York:W. W. Norton & Co. Ò1844 Manuscripts;Ó pp. 74-78.
- Ibid., 93-94.
- Morning Call, 8 December 2018, p. 1.
- If, as disciples of Smith and Ricardo assume and Locke argues (Second Treatise, Chapter 5), one’s compensation is some function of the wealth one creates, how is it that the wealth creators (workers) haven’t seen real wage growth since the 1960s, despite unprecedented increases in their productivity? http://www.pewresearch.org/fact-tank/2018/08/07/for-most-us-workers-real-wages-have-barely-budged-for-decades/ and https://www.epi.org/productivity-pay-gap/ .
- Among the types of labor-based theories of value that Ricardo identified are:value in use; value in exchange; amount of labor “in” the good; amount of labor the good can command on the market; and the amount of labor a worker can command for it on the market.
- The Principles of Political Economy and Taxation, (1817), London:John Murray; Chapter 4.1, p. 32
- Kaster & Moser, “Unemployment Impairs Mental Health:Meta-analyses,” Journal of Vocational Behavior. Vol. 74, #3, June 2009, Pages 264-282
- In Friedman’s optimal situation, the money supply would increase slowly and continuously, in a range from 2 to 5%, thereby eliminating the business cycle.