Thursday, November 27, 2014

Marx and the Mechanical Turk

BERKELEY – The economist Suresh Naidu once remarked to me that there were three big problems with Karl Marx’s economics. First, Marx thought that increased investment and capital accumulation diminished labor’s value to employers and thus diminished workers’ bargaining power. Second, he could not fully grasp that rising real material living standards for the working class might well go hand in hand with a rising rate of exploitation – that is, a smaller income share for labor. And, third, Marx was fixated on the labor-theory of value.

The second and third problems remain huge analytical mistakes. But, while Marx’s belief that capital and labor were substitutes, not complements, was a mistake in his own age, and for more than a century to follow, it may not be a mistake today.

Think of it this way. Humans have five core competencies as far as the world of work is concerned:

· Moving things with large muscles.

· Finely manipulating things with small muscles.

· Using our hands, mouths, brains, eyes, and ears to ensure that ongoing processes and procedures happen the way that they are supposed to.

· Engaging in social reciprocity and negotiation to keep us all pulling in the same direction.

· Thinking up new things – activities that produce outcomes that are necessary, convenient, or luxurious – for us to do.

The first two options comprise jobs that we typically think of as “blue collar.” Much of the second three options embody jobs that we typically think of as “white collar.”

The coming of the Industrial Revolution – the steam engine to generate power and metalworking to build machinery – greatly reduced the need for human muscles and fingers. But it enormously increased the need for human eye-ear-brain-hand-mouth loops in both blue-collar and white-collar occupations.

Over time, the real prices of machines continued to fall. But the real prices of the cybernetic control loops needed to keep the machines running properly did not, because every control loop required a human brain, and every human brain required a fifteen-year process of growth, education, and development.

But there is no iron law of wages that requires technologies of power and matter manipulation to advance more rapidly than technologies of governance and control. The direction of technological progress today is toward moving very large parts of both the blue-collar and white-collar components of overseeing ongoing processes and procedures from humans to machines.

How many of us can be employed in personal services, and how can such jobs be highly paid (in absolute terms)? The optimistic view is that those, like me, who find ourselves fearing the relative wage distribution of the future as a source of mammoth inequality and power imbalance simply suffer from a failure of imagination.

Marx did not see how the replacement of textile workers by automatic looms could possibly do anything other than lower workers’ wages. After all, the volume of production could not possibly expand enough to reemploy everyone who lost their job as a handloom weaver as a machine-minder or a carpet-seller, could it?

It could, but Marx’s mistake was not a new one. A century earlier, the French physiocrats Quesnay, Turgot, and Condorcet did not see how the share of the French labor force employed in agriculture could possibly fall below 50% without producing social ruin. After all, in a world of solid farmers, useful craftsmen, dissolute aristocrats, and flunkies, demand for manufactured items and flunkies was limited by how much of each aristocrats could use. Thus, a decline in the number of farmers could produce no outcome other than poverty and widespread beggary.

Neither Marx nor the physiocrats could imagine the great many well-paid things that we could find to do once we no longer needed to employ 60% of the labor force in agriculture and another 20% in hand spinning, handloom weaving, and land transport via horse and cart. And today, the optimistic view is that those with excess wealth will continue to think of lots of things for everyone else to do to make their lives more convenient and luxurious, and that the ingenuity of the rich will outstrip the supply of labor by the poor and turn the poor into the middle class.

But, given the rapid development of technologies of governance and control, the pessimistic view deserves attention. In this scenario, pieces of option three remain stubbornly impervious to artificial intelligence and continue to be mind-numbingly boring, while option four – engaging in social reciprocity and negotiation – remains limited. Welcome to the virtual sweatshop economy, in which most of us are chained to desks and screens – so many powerless cogs for Amazon Mechanical Turk, forever.

Editor's note: A shorter version of this commentary was previously published by the New York Times.

  • Contact us to secure rights


  • Hide Comments Hide Comments Read Comments (10)

    Please login or register to post a comment

    1. CommentedGerald Silverberg

      Brad is barking up the wrong tree here. Machines automating human supervision go right back to Watt's governor and Roberts' self-acting mule, nothing new here with digital computers. See my blog "Brad DeLong: The Mechanical Turk Was a Fraud, But the Watt Governor and the Roberts Self-Acting Mule Weren’t!" at

    2. CommentedG. A. Pakela

      You can only get rich if you can produce something that other people want or are willing to pay for your services. My guess is that the dissolute rich who have inherited their wealth is very small in the U.S., although perhaps not insignificant in 19th century Europe. If you got rich by being an entrepreneur, you likely created more jobs and generated more business revenue that in total, dwarfs the income received by that entrepreneur.

    3. CommentedJohn McDonald

      Of course Marx was right. Just look at Maoist China, the USSR, North Korea, Communist Eastern Europe, Cuba, pre-Thatcher UK, pre-Reagan US, and...
      Well you get the idea.

        CommentedHarland Brown

        Marx was a historian, not a politician. That is like confusing Adam Smith with Ronald Reagan. The academic may serve as a justification for political action, but they are operating in fundamentally different areas.

    4. CommentedRE Mant

      Of course Marx was right, the shame was that few others saw the mechanism of the industrial revolution.

    5. CommentedTom Walker

      So, Brad DeLong's understanding of Marx's "mistakes" is based on something Suresh Naidu once remarked to him. Too bad Brad didn't bother to read enough Marx to realize that this second-hand account was dead wrong about what Marx supposedly believed. thought, could not grasp or was fixated on.

      But I wonder also if DeLong's account of what Suresh Naidu "once remarked" to him are accurate.

    6. CommentedMarc Laventurier

      It's indicative of the scope and quality of the conventional economist's mind that it misses entirely the philosophical and revolutionary tenor of Marx's work. While scholarship demands a close and critical reading of the lot, Marx really begins just where Brad leaves off...what are you going to do about the social order that supports, at best, distinct inequality and, at worst, exploitation and deprivation? Scholarship indicates that Engels and Lenin, but not Marx, wrote that existing democratic institutions provided the easiest path towards a robust socialism (and I'd add, a system not built on a crumbling foundation of self-reinforcing financialization and artificial, game-theoretic incentives requiring such absurdities as infinite Demand and eternal Growth.) The transition from such late capitalism to a 'dictatorship of the proletariat' as described in the Communist Manifesto was predicted to be violent, and be followed by a classless society, a workers' paradise. Other scholarship has suggested that what Marx really had in mind was something like the Nordic welfare state model, probably the most viable known because it doesn't hide structural incentives for short term business profits in public and private deficits, keeping society in thrall to the money game that is capitalism.

      (For a treat, check out this brief tune by Randy Newman: )

    7. CommentedPeter Blok

      But what if future labor will consist of knowledge, owned by the rising class of knowledge workers, as the most vital part of this knowledge is tacit? Couldn't this lead to a whole new conceptualisation of Marx' principles?

    8. CommentedVal Samonis

      RE: engaging in social reciprocity and negotiation – remains limited.

      Rather UNLIMITED, as the Internet tends to reduce transaction and collaboration costs to zero and therefore produces unlimited opportunities for the explosion of global entrepreneurship.

      Val Samonis
      Vilnius U and Royal Roads U

    9. Commentedjames durante

      Marx was not wrong on any of these points. What he WAS wrong about was his idea that the state was completely captured by the bourgeoisie and incapable of delivering any relief to labor. DeLong and all the ortodoxos on this site zealously omit history from their analyses. Incomes and wealth became more evenly distributed roughly from 1920-1970 because of four primary factors: violent clashes between unions and agents of the ownership class (pinkertons, state forces, etc.); the development of labor parties that enacted reforms; a tremendous public commitment to reconstruction after ww ii; and a deep distrust of the employer class following on the devastation of the great depression along with a broad political consensus about social welfare. Now that those are disintegrating or gone inequality is returning with a vengeance and will only be exacerbated by the technical changes coming online.