There is only one problem. The conventional wisdom upon which this mandatory exclusion is based is rotten to its core. Although difficult to estimate and impossible to measure, raising the retirement age will have a negative effect on youth employment. It is uncertain whether countervailing economic trends will mitigate the effect, swamp it or intensify it. That uncertainty has been taken as a license by mainstream economist to argue that raising the retirement age will have the opposite effect and actually increase youth employment. Harvard economist, Edward Glaeser, presented a glib pop-culture version of this "counter-intuitive" conventional wisdom in a New York Times opinion piece published last fall:
It’s counterintuitive, but the forever work life of older Americans may turn out to be a good thing for young workers. The “lump of labor fallacy” envisions an economic order in which there is a fixed amount of work to be done. But we can make more or less, buy more or less, and most important, we can create new lines of enterprise. Over time, growth and innovation can create plentiful new work opportunities.There are more scholarly excursions into the same territory. See, for example, "Social Security in Theory and Practice (II): Efficiency Theories, Narrative Theories, and Implications for Reform," by Casey B. Mulligan and Xavier Sala-i-Martin (1999). The anthology, Social Security Programs and Retirement around the World: The Relationship to Youth Employment, edited by Jonathan Gruber and David Wise (2010), contains several references to the alleged lump-of-labor fallacy, including an entire section on "Lump of Labor Fallacy and Youth Employment in Belgium" and a discussion by Banks, Blundell, Bozio and Emmerson of debates in the U.K. around the lump of labour fallacy in which they cite this author's research (Walker 2007).
Banks, et al., it should be noted, found "no evidence of long-term crowding out of younger individuals from the labor market by older workers. The evidence, according to a variety of methods, points always in the direction of an absence of such a relationship." It would be prudent, however, to read their "points always in the direction" as a somewhat reluctant admission that their results were inconclusive – an admission, moreover, stated more forthrightly in the conclusion to section 11.5 of their chapter, upon which "our results have to rest":
…we do not find conclusive evidence of at least some substitution between young and older workers. This does not, again, prove that these schemes have had no effect on the youth unemployment, but despite our best effort, we have not been able to identify, even in the short run, their presumed positive (negative) impact on youth employment (unemployment).It is indeed difficult to find conclusive evidence from time-series data of the remote effects of a policy intervention, especially when there are a myriad of intervening variables. That is why basing a conclusion of "evidence of absence" on the absence of conclusive evidence is dicey and may even shade into a classic argumentum ad ignorantiam.
But what does the finding of "absence of evidence" imply for my above statement that "raising the retirement age will have a negative effect on youth employment"? In the first place, I am not talking about a one-to-one correspondence between jobs vacated by older workers and jobs opened up to younger ones. The notion of college grads stepping into the employment shoes vacated by their grandparents sounds rather bizarre to me. Nor am I assuming, more loosely, that a larger aggregate labour supply will necessarily entail a higher unemployment rate, the consequences of which will fall disproportionately on the young. The latter circumstance is a possibility, though, that should not be ruled out.
What I am basing my prediction on is a policy orientation indicated by the rhetoric of the retirement age debate that is biased against wage-earners and full employment. It is not so much that older workers will crowd out younger workers as that regressive economic policies, based on dishonest rhetoric, will crowd out better policies. The evidence for this corrupt policy orientation can be found not in the time-series data but in the frequent references in the literature to the supposed lump-of-labor fallacy.
As Cecil Pigou stated in 1913 and Maurice Dobb explained in 1928, the fallacy claim is itself fallacious. But the incessant repetition of the debunked fallacy claim, without any acknowledgement of the rebuttal, nudges the allegation into the category of fraud, either witting or unwitting.
The history of the fallacy claim is a long and tortuous one that commences in an anonymous pamphlet published in 1780 in response to rioting the previous year in Lancashire, England, titled, "Thoughts on the Use of Machines in the Cotton Manufacture." A Bolton magistrate, Dorning Rasbotham is believed to be the author. The pamphlet and Rasbotham were highly praised by classical political economist J. R. M'Culloch in an 1827 Edinburgh Review article in a footnote that concluded with the admonition that "There is, in fact, no Idea so groundless and absurd as that which supposes that an increased facility of production can under any circumstances be injurious to the labourers."
M'Culloch overlooked one detail. At the top of page 18 of his pamphlet, Rasbotham committed a version of the same (alleged) fallacy that he accused others of at the bottom of the same page. This peculiar feature of "fighting fallacy with fallacy" was noted by economist Charles Beardsley in his critique of a similar fallacy claim made by John Rae in 1892. I already mentioned the Pigou and Dobb criticisms.
What these repeated episodes of the pot calling the kettle black indicate is the impossibility of making economic generalizations without some kind of arbitrary fixed (or ceteris paribus) assumptions. What they also show, though, is a sense of entitlement among many conventional economists who imagine (quite wrongly) that their arbitrary, unrealistic and unacknowledged assumptions don't stink. Furthermore, the most insistent fallacy claims have emerged historically where evidence and analysis about "means to given ends" do not suffice. That is to say, the thrust of these ritual fallacy allegations has been to limit and foreclose politically contentious debate whose outcome is otherwise uncertain.
Ostensibly, the aim of raising the retirement age is reducing government expenditures or, viewed from the perspective of the recipients, lowering the social wage. This might be rationalized by proponents as necessary to prevent excessive transfers from young workers to older ones. Again, however, the history of the deployment of the lump of labour fallacy claim is instructive. In the 19th and early 20th century, the fallacy claim was essentially an argument about the illegitimacy of trade union efforts to improve the wages and working conditions of workers. It surfaced in the violently anti-trade union tract by E.C. Tufnell (1834), Character, Object and Effects of Trades Unions and at the turn of the century became a mainstay of the anti-union rhetoric of the U.S. National Association of Manufacturers. The fallacy claim has lately disappeared from college textbooks but remains a staple of "free market" think-tanks, such as the Fraser Institute in Canada; the Cato Institute, the American Enterprise Institute and the Ludwig von Mises Institute in the U.S. and the Institute for Economic Affairs in the U.K.
This is not to say that all purveyors of the fallacy claim are right-wing ideologues, only to highlight the particular appeal the argument has to that audience. What, then, is the "excluded variable" mentioned in the title of this essay. Superficially, it is youth employment and the possibility that delayed retirement could "crowd out" younger people from the work force. But more fundamentally the excluded variable is the ideological slant of the mainstream literature, which through its uncritical embrace of a bogus fallacy claim has abdicated any claim to impartiality.