Sunday, June 17, 2012

Economic Fallacy #112: The "Lump of Labor"

Much of the talk about job creation assumes there is only so much work to be done, so it needs to be "spread out" between workers to avoid unemployment.

This "lump of labor" mental model has led The Guardian call for a reduction in the number of hours worked per week in the UK down to thirty, in order to supposedly address their record number of unemployed youths. 

The same mode has also been used to justify union "job creation", or the hiring of more people to do unnecessary or unproductive jobs. (I must admit, one activity unions are highly productive at is Search Engine Optimization - pro-union pages dominate the top ranks of Google)

While this fallacy emerges frequently in the popular press at the behest of union leaders, it has been debunked many times already. From Henry Hazlitt's Economics in One Lesson (1952):
It is not only in reducing scheduled working hours that union policy has worked against productivity... Most of these policies have been followed under the assumption that there is just a fixed amount of work to be done, a definite "job fund" which has to be spread over as many people and hours as possible so as not to use it up too soon. This assumption is utterly false. There is actually no limit to the amount of work to be done. What A produces constitutes the demand for what B produces. (Chapter 19)
In other words, the potential stock of work to be done is constantly rising as worker income rises, because those workers now demand products they previously could not afford. Human wants are infinite: someone who supposedly has everything still wants new services in order to save time and have more leisure. Human lifespans are finite so time is always a limited resource, from the individual's perspective.

There are legitimate reasons to worry about technology and stratified worker productivity levels leading to unemployment and social divisions. However, those concerns should be framed through the lens of marginal worker productivity, not the over-worked lump of labor fallacy. 


  1. Nick, the "lump-of-labor fallacy" is a myth. The only people who make an assumption of a fixed amount of work are the ones who allege the fallacy and falsely attribute the assumption to others. Unless you or Henry Hazlitt or The Economist (or Dorning Rasbotham for that matter) can document that unions or the Guardian or whoever actually assumes "there is only so much work to be done" you are arguing with a straw man. See

    1. Hi Sandwichman. Thanks for stopping by.

      I think the Guardian article I link to above, which quotes some people in the UK asking for a 30 hour work week to reduce unemployment (among other things), demonstrates the lump of labor fallacy in action. I'll agree that nobody calls it the "lump of labor" argument when they do so, because it sounds silly... but that is often the implicit assumption behind their claims.

      If one didn't think there's "only so much work to be done", how would it be possible to believe that a shorter working day could reduce unemployment?

    2. Quoting that Guardian article:

      "Without the advances of a shorter workweek, vacation time, earlier retirement and later labour force entrance, the economies of the OECD would never have attained the "golden age" of high employment that prevailed after the 1930s depression. Between 1870 and 1970, hours of work fell roughly in half. These countries have re-balanced the labour market by re-distributing work to make its allocation fairer."

      I'm thinking of that last sentence in particular.

      In reality, what led to lower working hours and higher living standards was greatly increased productivity between 1870 to 1970. Thinking about worker product in terms of marginal revenue is a much better way to understand the issue.

    3. Nick,

      How would it be possible to believe that a shorter work day could reduce unemployment? My impression is that most people who advocate a shorter work week believe that the rate of increase in the amount of work to be done doesn't match the rate of increase in the labour supply. That's not the same thing as believing that the amount of work to be done is fixed. If the amount of work to be done increases by 5% while the number of people in the labor force increases by 10%, there will be more unemployment. That addresses the fallacy claim.

      I think the situation is somewhat more complicated than the typical advocate of shorter hours assumes. My view is based, in part, on Sydney Chapman's theory of the hours of labour in which he demonstrated that hours of work determined in a competitive market were likely to be too long from the standpoint of productivity and in part on Robert Prasch's model of an "inverted 'S' shaped" labour supply function, in which workers tended to supply more hours with a reduction in wage in order to maintain a subsistence standard.

      Without going into all the details, my view is that reducing the hours of work performs a hygienic function of increasing productivity, worker well-being and income share all of which are positive for macroeconomic performance. The problem with my view is that you would actually have to be familiar with the history of economic thought to comprehend it. Some of it is "counter-intuitive." But -- returning to the fallacy claim and what typical advocates of shorter hours actually believe -- whether employment keeps pace with labour supply is an empirical question and the evidence is clear that at times job creation lags behind labour force growth.

      If you think that lower hours are ONLY a result of higher productivity, you really ought to have a look at Chapman's theory, in which the relationship is much more complex. Shorter hours are not only the result of higher productivity but also the cause of further increases in productivity.

    4. A few responses.

      1. I can tentatively agree with the claim that labor force is increasing faster than amount of work to be done - loosely speaking, this is the argument in book "Race Against the Machine" which I thought was pretty good. But, I object on ethical/longterm reasons to spreading around the work as a solution to this. High productivity should be rewarded if we want to get more of it, not hobbled.

      2. I'm not that familiar with Chapman, and just read a brief summary. But here's my initial reaction.

      a) if lower hours increase productivity, we'd expect to see businesses support that voluntarily. It seems Chapman's argument is that businesses will bid away workers to make them work more hours, but my game-theoretic intuition is that some sort of "collusion" could be maintained between employers to avoid this, if they all benefit from higher productivity and not as high wages by doing so.

      b) the U.S. has been moving toward shorter working hours, which employers use to avoid giving benefits, but at the same time productivity has been stagnating. This cuts against the core claim of Chapman's argument.

      3. Regarding the S-shaped labor supply function: this may be true for each individual worker, but not for the labor market as a whole, because higher wages cause more people to enter the workforce who otherwise would not. Also, in developed countries, almost no workers are as low as the subsistence level, so the impact is likely to be small.

      I'll have to look deeper into Chapman's theory to speak with more confidence on this. But I'm not sure we're really in disagreement on the core point, which is that talking about a "fairer allocation" of work, as the Guardian article does, is a bad model. We can disagree over the specifics of marginal productivity theory and labor markets, but the fact that we're both talking about productivity at all means we're somewhat in agreement

  2. Nick,

    I wouldn't go so far as to say the usual, Guardian-type argument for shorter work time is a "bad model." I would just say it is an incomplete and rather conventional model -- maybe a weak model. The Chapman theory is a more sophisticated but also very uncertain model -- by which I mean it factors in uncertainty.

    I would also argue that the standard "lump of labor fallacy" response to the usual argument for shorter work time is worse than a bad model. In fact, the fallacy claim is the fallacy -- technically a "false dilemma" fallacy but in addition one that relies on attributing to one's opponent a belief he or she doesn't necessarily have to make the opponent's argument seem even worse than it is.

    What you need to take into account with Chapman's theory is that it is based on assumptions of a "competitive labour market" and compensation at marginal product value (which in turn assumes full employment). It is thus a heuristic rather than a description of how the world actually is. I should mention, though, that Pigou, Marshall, Lionel Robbins and J.R. Hicks, all accepted Chapman's theory as canonical.

    One can take the analysis one step further following Thorstein Veblen's critique of neoclassical economics. That would be to view the firm's objective as being to limit production in order to attain a "target" price. Under such a Veblenian regime firms might even set unproductively long hours specifically to "sabotage" output -- especially if they could pass on the cost of doing so in the form of lower wages (which the inverted S supply curve suggests they might).

    I would also suggest a different interpretation of "subsistence" that would encompass a much larger component of the workforce in developed countries. I would argue that high levels of indebtedness might induce a subsistence-like response when wages fall to a level that threatens the ability to both service the debt and maintain a standard of living.

    1. P.S. -- at the other end of the spectrum from Chapman's analysis, I think you ought to have a look at the ORIGINAL "lump-of-labour" fallacy claim by Dorning Rasbotham in 1780. It wasn't called the lump of labour in those days, but the argument, which appears on page 18 of the pamphlet, "Thoughts on the Use of Machines in the Cotton Manufacture" is unmistakable:

      "There is, say they, a certain quantity of labour to be performed. This used to be performed by hands, without machines, or with very little help from them. But if now machines perform a larger share than before, suppose one fourth part, so many hands as are necessary to work that fourth part, will be thrown out of work, or suffer in their wages. The principle itself is false. There is not a precise limited quantity of labour, beyond which there is no demand. Trade is not hemmed in by great walls, beyond which it cannot go. By bringing our goods cheaper and better to market, we open new markets, we get new customers, we encrease the quantity of labour necessary to supply these, and thus we are encouraged to push on, in hope of still new advantages. A cheap market will always be full of customers."