Scott Sumner  

The lump of labor fallacy

Michael Novak RIP... Taxes and Deadweight Loss...

Andy Puzder was one of the few Trump appointees that I sort of liked (I say "sort of", because even he had ethical issues.) He was pro-immigration and anti-minimum wage. But in the end even many conservatives opposed him so he withdrew his name from consideration for Labor Secretary.

Reihan Salam was one of the conservative opponents of Andy Puzder:

Puzder has also been an influential critic of minimum-wage hikes and overtime regulations, warning that such measures would force employers to replace low-wage workers with machines. He seems animated by the Luddite conviction that productivity-boosting automation is necessarily a bad thing, despite the fact that rising productivity levels are essential to wage growth.
This has things exactly backwards. People who subscribe to the lump of labor fallacy (there are a fixed number of jobs) are exactly the people who favor these four bad public policies:

1. Restrictions on automation
2. Higher minimum wage rates
3. Protectionism
4. Lower levels of immigration

People suffering from this fallacy think there are a fixed number of jobs, which allows the government to arbitrarily raise the minimum wage without hurting employment. This view also suggests that there is only so much to be produced, and if more is produced overseas, or by immigrants, or by robots, then less will be produced by American born workers.

In the past, commenters have objected when I claim that deporting illegals would devastate the California fruit and vegetable industry. They insist that someone will do the work, that high quality farmland won't lie fallow. That's missing the point:

Some farm jobs, like tomato picking, could be automated fairly easily in the 1960s. And ending the bracero scheme seems to have accelerated mechanisation in the tomato fields of California. Much the same happened with cotton and sugar beet. Other crops, like lettuces and asparagus, still required human pickers. Production of some such crops simply declined.

. . . In California, America's most important farming state, politicians have ensured that workers will receive at least $15 an hour by 2023. And Manuel Cunha, a citrus grower who is president of the Nisei Farmers' League, complains about other costly reforms, such as mandatory overtime pay for people who work more than eight hours a day.

In response, he says, farmers are moving from crops that require careful handling, like apricots--"just look at an apricot and it will turn brown"--to crops that can be harvested by machine. Almond trees are spreading across California. In spring the fields are white with their blossom. In September great machines shake the nuts to the ground and sweep them up.

There's simply no way that California fruit and vegetable producers could pay enough money to attract America workers. They'd go out of business, and their output would be replaced by imports. Instead they'd switch to crops that do not require significant farm labor. Thus deporting illegals will not create new jobs for American workers.

Screen Shot 2017-02-17 at 9.41.35 AM.png
A dramatically higher minimum wage will make America more like southern Europe. Today, Hispanic Americans are employed at a fairly high rate. Do we want to make our labor market more like France, where large numbers of Arab immigrants are unemployed, and often resentful of the country they live in?

The same article also reported:

Michael Clemens and Hannah Postel of the Centre for Global Development, and Ethan Lewis of Dartmouth College, have used archived records of American agricultural jobs and wages to test whether Kennedy was right. Did ending the bracero scheme in 1964 in fact lead to higher wages and more work for Americans in the fields?

The answer is a firm no. In states where farmers had relied heavily on foreign labour--a group that includes California and Texas--American natives found a few more farm jobs in the mid 1960s. But the rise was small and temporary; within a few years the long decline in agricultural jobs had resumed. And the trend was almost identical in states where there had been no braceros. Similarly, farm wages rose in states where there had been lots of migrant workers, states where there had been few migrant workers and states where there had been almost none (see chart). Ending the bracero scheme seems to have affected American workers not a bit.

COMMENTS (18 to date)
Kevin Erdmann writes:

It's such a strange argument, because it seems to rest on an unspoken assumption that real capital can just be conjured out of thin air - as if the capital that would replace those workers isn't being pulled from some other use. In fact, I suspect this is partly why it's hard to detect some of the employment damage, because, to the extent that capital is taken from other sectors in order to replace low wage workers, it reduces productivity and employment in other sectors, so that, in the end, the new minimum wage creates a minor negative supply shock that is basically diffused through the labor market in a way not that unsimilar to other shocks.

Maybe we should cut everyone's legs off at the knee, then we'll all get those cool prosthetics that the para-athletes use. It will be a boon!

MikeP writes:

Ah, yes. The broken legs fallacy.

With all that demand, prosthetic leg technologies will improve to practically bionic levels. And it will be pretty much the first thing anyone spends money on. It's all win!

Scott Sumner writes:

Everyone, People often don't think deeply enough about the unintended side effects of policies like the minimum wage.

john hare writes:

Scott Sumner writes:
Everyone, People often don't think deeply enough about the unintended side effects of policies like the minimum wage.

You could have stopped at four words. People often don't think, about anything outside their bubble. Or even inside for that matter. Watch how most people do their jobs or run their lives. I hire people on the low end of the economic and educational scale and find ~10% worth retaining. I am convinced that untrainable is a choice.

Not all. There are plenty that think. They appear to swim in an ocean of those that try to avoid it as far as I can tell. It is those individuals that think that move everything forward. Not an original thought of mine obviously, just something I thought needed to be repeated in the context of this post.

The last quoted bit states that "Ending the bracero scheme seems to have affected American workers not a bit."

But, it did affect American workers: it reduced their vegetable consumption.

TMC writes:

Part of the issue is that the farmers do not have to pay for the total cost of the immigrants. We import people, and their families, and descendants, not just workers. I'd much prefer an Australian immigrant policy.

Michael writes:

John Hare: LOL 😀

to continue in this vein: your observation "untrainable is a choice" is not just about the lowest decile (economically and education-wise). On the contrary, I'm afraid, the people selling us minimum wages may well be in the highest decile (just showing that wage = adde value is a gross idealization)

James Alexander writes:

Didn't it used to be organized labour who promoted the minimum wage to keep out cheaper competition?
Now it's just people who mean well?

Sandwichman writes:

You are right, Scott, that the fallacy claim has been used against all four of those "interventions." But Reihan Salam is also right in pointing out that Puzder's response to minimum wage increases partakes of the Luddite (AKA lump of labor) fallacy.

How could you (and Puzder) and Salam both be right? Because the fallacy claim is incoherent. Essentially, there is a contradiction in the fallacy claim that those who invoke it cannot wrap their heads around. A.C. Pigou pointed out that discrepancy more than a century ago. Maurice Dobb extended that critique in the late 1920s.

The fallacy claim is a self-contradictory critique of the old wages-fund doctrine that assumes on the one hand the very relationship is it criticizing as fallacious on the other. The "fixed amount of work" corresponds to a supposedly fixed amount of wages. Sometimes emphasis is placed ("ceteris paribus") on one of these spurious assumptions and sometimes on the other. But the argument evaporates if you recognize the two components as interdependent.

To put it another way, the choice is not between a fallacious fixed amount of work and a logically-imperative one-to-one match between labor supply and labor demand. The alternative to a one-to-one match could be any number.

If labor supply increases by 100 and labor demand increases by 90 -- has the amount of work remained fixed? No. Has unemployment increased? Yes, by 10.

If the minimum wage increases by 10% and minimum wage employment falls by 5% are minimum wage workers better or worse off? Obviously 95% of them are better off -- but what about the rest? Aren't they worse off? Not necessarily. But if we assume that "everything else remains the same," we would have to conclude that they are worse off (at least in terms of employment).

If we fully embrace the fixed wage-fund assumption, we may even conclude that an increase of 10% in the minimum wage will result in a 10% decrease in minimum wage employment. A very unlikely scenario.

See what happens here? The economist cannot avoid ceteris paribus to reach conclusions yet disparages an alleged ceteris paribus assumption by opponents as a fallacy.

Don't let the Latin fool you. What is good for the goose is good for the gander. The lump of labor is indeed a fallacy; the lump of labor fallacy is also a fallacy.

mm writes:

Doesn't your fear of a France like scenario only occur if we allow high levels of immigration (legal or illegal)-ie open borders.

JMCSF writes:

Also, California is an expensive state. Land costs, taxes, everything in addition to labor costs.

So in addition to easily automated crops, I would expect a move to high margin crops - organic, wine, etc.

CC writes:

[Comment removed for supplying false email address. Email the to request restoring this comment and your comment privileges. A valid email address is required to post comments on EconLog and EconTalk.--Econlib Ed.]

Luis Pedro Coelho writes:

@mm "Doesn't your fear of a France like scenario only occur if we allow high levels of immigration (legal or illegal)-ie open borders. "

Yes, but only in the specific point that unemployment affects mostly immigrants.

Without immigration, it's the natives who'd be unemployed. France's rust belt has many regions with almost no immigration and high unemployment rates.

Don Boudreaux writes:

Perhaps I missed it in the comments and in Scott's post, but isn't there an even more straightforward criticism of Salam's quoted objection to Puzder? I would respond to Salam by saying:

Yes, real wages rise as worker productivity rises. And yes, mechanization generally raises worker productivity. And yes, the sorts of machines that employers would put in place in response to minimum-wage hikes would raise the productivity of the workers who remain employed. But to admit all of this is to miss Puzder's objection to the minimum wage - namely, it puts some workers out of jobs. That is, Puzder points to the option to mechanize in order to help people understand that employers have options other than simply keeping the same number of workers and paying them more when minimum wages rise.

Put differently, Puzder points to the mechanization option to help explain why in reality raising minimum wages does indeed destroy jobs for some low-skilled workers. For Salam to infer from this explanation a Luddite fallacy is, well, fallacious.

Bill Woolsey writes:

If there is an increase in the demand for labor, the immediate effect is a shortage. The simple story is that each firm raises wages to attract workers to them instead of other firms. As the wages increase, the quantity of labor demanded decreases. This decrease in the quantity of labor demanded will at least partly be due to the introduction of labor saving devices that were relatively too expensive when labor was cheaper. The higher wage also increases the quantity of labor supplied. The new equilibrium includes a higher wage and more employment and more labor saving equipment.

A decrease in the supply of labor would involve a similar account. An initial shortage, higher wages, a decrease in the quantity of labor demanded partly due to the introduction of what was excessively expensive labor saving equipment and an increase in the quantity of labor supplied. The end result is more labor saving equipment, less employment and higher wages.

Raising wages when there had been no prior increase in demand or decrease in supply disrupts coordination and while the introduction in labor saving technology does reduce the disruption to those buying the products--it is a kind of harm reduction, it exacerbates the harm of unemployment.

More saving, investment, and capital almost certainly raises labor demand and so leads to the first process--higher wages, employment and labor saving technology.

Innovation--new or improved labor saving technology being developed is a different when ambiguous effects on labor demand and so wages and employment. Labor will tend to get a smaller share of a bigger pie which is an ambiguous result.

Jon Murphy writes:


Doesn't your fear of a France like scenario only occur if we allow high levels of immigration (legal or illegal)-ie open borders.

No. France doesn't have open borders or a high level of immigration, and yet they still have problems. The issue is not with immigration but the inability of people to integrate into a workforce because of tight labor laws.

Interesting historical tidbit: in the US, minimum wage was passed partly to discourage immigration to the US. By raising the wage and reducing opportunities for immigrants, the promoters of the Wage were hoping less immigrants would come to America.

Scott Sumner writes:

Sandwichman, I'm afraid I didn't follow that. What exactly is wrong with my post?

Also, you ended up with:

"Don't let the Latin fool you. What is good for the goose is good for the gander. The lump of labor is indeed a fallacy; the lump of labor fallacy is also a fallacy."

Is there a typo there?

mm, Not at all. You could get the same result with whites in Appalachia, with blacks, with American born Hispanics, with native Americans, etc.

Don, I agree.

Sandwichman writes:


I understand why it would be difficult to follow my critique. Those who make the fallacy claim have a fairly sophisticated model of the economy in mind and compare their model with what they suppose is an extremely crude model ("a fixed amount of work") assumed by advocates of, for example, a higher minimum wage. To follow my critique would require you to suspend belief in your own model, which is difficult for anyone to do.

Your model of economic reality rests on a bundle of assumptions about rationality, information, prices, equilibrium, etc. The logical deduction from these assumptions is -- to put it crudely -- that "supply of labor creates its own demand." Those assumptions and the conclusions deduced from them may even be approximately correct, in that an increase in labor supply initiates a series of responses that ultimately result in an increase in demand.

Nevertheless, your model is not reality. "All models are wrong," as the saying goes. One of the ways in which all models are wrong is that they inevitably misconceive alternative perspectives. The fixed-amount-of-work (lump of labor) assumption is something attributed to other perspectives by your model. It is not actually a feature of those other perspectives. The other perspectives may even be wrong -- but not for the reasons you think they are!

To a degree, the lump of labor fallacy claim makes the valid point that "things are not so simple." But, almost invariably, those who make the claim step into their own trap by their insistence that their model faithfully and completely represents reality.

If you would like to pursue this further, please see my article, "Why Economists Dislike a Lump of Labor" in the Sept. 2007 issue of Review of Social Economy and/or my earlier chapter, "The 'lump of labor' case against work-sharing: populist fallacy or marginalist throwback?" in Working Time: International trends, theory and policy perspectives (2000).

Comments for this entry have been closed
Return to top