Facts or Fallacies? Part II: In which Paul Krugman takes his lumps and eats them too while Jamie Galbraith runs afoul of the notorious lump-of-labor f

by Tom Walker
(Sandwichman at Ecological Headstand)

Facts or Fallacies? Part II: In which Paul Krugman takes his lumps and eats them too while Jamie Galbraith runs afoul of the notorious lump-of-labor fallacy-fallacy

In comments in response to Part I at Angry Bear, it was suggested that Paul Krugman has also made the lump of labor fallacy claim and that perhaps I should be talking about his arguments (or those of Paul Samuelson) instead of those of conservative think-tankers. Both of those liberal Keynesian economists have indeed advanced the fallacy claim. I would love to discuss the matter with Professor Krugman and sent him an invitation.

Meanwhile, indulge me while I rehearse my debating points on some archival material. I’d also like to bring in a few big names on my side: James K. Galbraith, Dean Baker and… Paul Krugman! Just to even things out, the pre-recession Krugman gets Bruce Bartlett and Larry Summers on his tag team.

(Update: Crooked timber picks up the lump theme.

Here’s what Krugman wrote in his 2003 column, titled Lumps of Labor .

Economists call it the lump of labor fallacy. It’s the idea that there is a fixed amount of work to be done in the world, so any increase in the amount each worker can produce reduces the number of available jobs. (A famous example: those dire warnings in the 1950’s that automation would lead to mass unemployment.) As the derisive name suggests, it’s an idea economists view with contempt, yet the fallacy makes a comeback whenever the economy is sluggish.

Sure enough, the lump-of-labor fallacy has resurfaced in the United States — but with a twist. Traditionally, it is a fallacy of the economically naïve left — for example, four years ago France’s Socialist government tried to create more jobs by reducing the length of the workweek. But in America today you’re more likely to hear lump-of-labor arguments from the right, as an excuse for the Bush administration’s policy failures.

Do I think “any increase in the amount each worker can produce reduces the number of available jobs”? Certainly not. But clearly large gains in productivity require some kind of adjustment. The adjustment process can be slow or fast, smooth or rough, complete or incomplete. Reducing the length of the workweek can be an important part of the adjustment process. So what motivates the “derision” and “contempt” of economists?

Personally, I prefer a piece Krugman wrote six years earlier, The Accidental Theorist in which he chided William Greider for being a “dull boy” and employed a playful hot dogs and buns economy to show why Greider’s alarmism in One World, Ready or Not: The Manic Logic of Global Capitalism was unwarranted. In that column, Krugman didn’t call it a “lump of labor fallacy” but the intent was clearly the same. Thirteen years later, the following comment appeared in reply to Krugman’s column:

Wow. Reading this after recently rereading the chapter entitled “The Alchemists” on the follies of big finance, written by Greider 13 years ago confirmed what I was recently becoming aware of: It turns out that Greider was articulating with shocking prescience what would happen 10 years later while economists like Krugman were mocking him for not consulting them. Well, Greider was right!

To be fair, when the facts change, Krugman changes his mind, as he did in November, 2009 when he endorsed Dean Baker’s proposal for a work-sharing subsidy:

Just to be clear, I believe that a large enough conventional stimulus would do the trick. But since that doesn’t seem to be in the cards, we need to talk about cheaper alternatives that address the job problem directly. Should we introduce an employment tax credit, like the one proposed by the Economic Policy Institute? Should we introduce the German-style job-sharing subsidy proposed by the Center for Economic Policy Research? Both are worthy of consideration.

So much for the non-accidental hot dogs and buns theory… But what about the specific question of older workers and the lump of labor? Couldn’t it be that some lumps are lumpier than others?

In contrast to the delayed retirement age prescriptions of Jason Kuzicki, Francois Melese or Andrew Biggs (see Part I, Jamie Galbraith has been advocating a temporary suspension of the early retirement penalty as a way to open up more jobs for the young. Let Old Folks Retire Early and Make Way for the Young was Galbraith’s contribution to Dan Froomkin’s Huffington Post series on job creation ideas. One comment in response to Jamie’s suggestion declared Galbraith is a victim of the “lump of labor fallacy.

Unfortunately, the commenter, “bgladish”, did not elaborate on his declaration. However, in a Forbes column published last February, Bruce Bartlett explained why “early retirement, work sharing and tax credits won’t boost employment.” To his credit, Bartlett cited not only the ubiquitous lump-of-labor fallacy as the reason why not but also a pair of studies: one from the Social Security Administration and the other by economists at the International Monetary Fund.

The problem is, the IMF study (which I had critiqued two years earlier on EconoSpeak misrepresented the standard rationale in Europe for early retirement — which was not so much to open up jobs for the young as to divert workforce reduction away from the low-seniority, younger workers. The IMF study did helpfully point out, however, with regard to the alleged lump-of-labor fallacy, that “Those who make the fallacy claim fail to offer specific evidence of the supposed belief in a fixed amount of work.”

The main conclusion of the SSA study was, to put it bluntly, utterly irrelevant to the case Bartlett was trying to make. Larry DeWitt, the SSA historian, rejected the hypothesis that the retirement earnings test of social security was designed to open up jobs for the young on the grounds that “the aged had already been forced out of the workforce” so they were “not a major factor in blocking opportunities for the young…” Other points raised by DeWitt were the “ineffectiveness of the supposed incentives,” the gradualness of implementation of the program and, finally, the limited coverage of the original Social Security program. None of these points spoke to the issue of whether or not in today’s circumstances early retirement might be effective in boosting employment among the young.

In his column, Bartlett described work-sharing (which Krugman had endorsed a few months earlier) as “another bad idea making the rounds” and cited the lump of labor fallacy as his rationale for why work-sharing wouldn’t boost employment.

The problem is that the amount of income being produced would still be the same. While some unemployed workers would gain jobs and income, current full-time workers would become underemployed and see a reduction in their incomes.
True to form for lump-of-labor claims, Bartlett failed to cite specific evidence
of the supposed belief in a fixed amount of work.

So I made a point of asking Dean Baker whether or not his work-sharing proposal assumed a fixed amount of work.

Here is Dean Baker whether or not his work-sharing proposal assumed a fixed amount of work. Here is his reply:

Actually, I don’t assume a fixed amount of labor, but in the context of an economic downturn, we definitely are in a situation where there is deficiency of labor demand. In this context any reasonable person would ask whether it is better to have more workers employed at fewer hours per workers or fewer workers (more unemployed people) employed 40 hours a week.

In a more general context, I think we should definitely be trying to have the U.S. follow the rest of the world in promoting shorter workweeks, longer vacations, paid time off for family leave, sick days etc. There is nothing natural about the current workweek. In fact, one of the main reasons that we have not followed the rest of the world in moving toward shorter workweeks is because of the high overhead costs (most importantly health care) associated with hiring workers. Firms would often rather pay a worker time and a half for overtime hours, or even double-time, rather than incur these overhead costs by hiring another worker.

Since this is a rehearsal for a debate with Paul Krugman, I’ll leave the last word to the good Professor:

Should America be trying anything along these lines? In a recent interview in The Washington Post, Lawrence Summers, the Obama administration’s highest-ranking economist, was dismissive: “It may be desirable to have a given amount of work shared among more people. But that’s not as desirable as expanding the total amount of work.” True. But we are not, in fact, expanding the total amount of work — and Congress doesn’t seem willing to spend enough on stimulus to change that unfortunate fact. So shouldn’t we be considering other measures, if only as a stopgap?