WORK LESS Institute of Technology

LEISURE AS A FACTOR OF PRODUCTION IV

Life is excessively complicated

by the Sandwichman

The stories textbook authors tell are simple stories – the simpler the better.... Supply and demand are fictional characters added for pedagogical or ideological purposes. - David Colander

In his article, “On the Genesis of the Canonical Labor Supply Model,” Laurent Derobert (2001) mentioned Chapman’s theory in connection with Hicks’s description of it as “the classical statement of the theory of ‘hours’ in a free market...” Although Derobert rightly dismissed the Chapman analysis as a source for the canonical model, he overlooked the significance of Hicks introducing his so-called simplifying assumption. Whether intended by Hicks or not, it helped clear the way for the opportunity cost analysis and the butt-on-a-slide model because it papered over otherwise irreconcilable discrepancies and thus side-stepped a confrontation that could have revealed the inadequacy of the then as-yet-to-become-canonical approach.

Again, Derobert was technically correct when he argued, “there would be no use reproducing it [Chapman’s analysis] here,” but only to the extent that Derobert’s article was narrowly about the canonical butt-on-a-slide model and not broadly about the comparative merits of various theories of labor supply.

In the course of his very brief discussion of Chapman’s theory, though, Derobert blithely cast several unsubstantiated aspersions: that it more closely resembled an amalgam than a synthesis, that it became “excessively complicated,” and, most egregiously, that “it lies somewhere between Jevons’s analysis and the canonical model. (emphasis added)” The latter point is accurate only chronologically. Analytically, Chapman’s theory is not a step away from Jevons and toward the eclipse of labor in the canonical model. It is, rather, a further step in the qualitative direction and toward a more realistic treatment of labor. The canonical model is a big step back from Chapman.

Unquestionably, Chapman’s diagram -- which Derobert chided for its “over-abundant use of construction devices” -- suffered from trying to do too much all at once. But on the other hand, it was not intended as a snapshot of some sublime moment of equilibrium where the fictional characters, supply and demand, are viewed from a controlling perspective, consummating their mutual lust. It is more like a cubist composition, breaking up the subject matter and then re-assembling it in a way that shows multiple perspectives and positions in space simultaneously.

I'm not saying the diagram was self-consciously cubist. But it was produced in the same era as the pioneering works of Braque, Picasso and Duchamp. It seems to me (others may disagree) that the diagram "comes to life" when viewed juxtaposed to Duchamp's Nude Descending a Staircase.

(Compare Chapman's dynamic cubism to the static neo-classicism of the butt-on-a-slide model):

Notwithstanding my eccentric suggestion of a cubist rhythm in Chapman's diagram, its complicated appearance is not beyond remedy. It can be decomposed into four diagrams, each displaying a different but complementary facet of Chapman’s analysis. To the extent that Chapman’s analysis is “complicated’ or an “amalgam,” that might more constructively be laid to the nature of the problem Chapman was analyzing rather than any fault of his own.

But even that may be conceding too much. What is "simple" and what is "complicated" may have far more to do with expectations and conventions than the inherent clarity or difficulty of the subject matter. Who today would find the Duchamp painting particularly difficult, let alone scandalous? As Andre Breton cited Francis Picabia on a famous dada sandwichboard: «Pour que vous aimiez quelque chose il faut que vous l'ayez vu et entendu depuis longtemps tas d'idiots.»

Next: Queensland's summary of Sandwichman's synopsis of Chapman's theory.

LEISURE AS A FACTOR OF PRODUCTION III

The marginal utility of evasion

by the Sandwichman

Hick's simplifying assumption - that the given length of the working day be assumed to be optimal - is utter nonsense. It is not actually a simplification at all but an abnegation of the theory. Here's why: one immediate implication of Chapman's theory is that leisure time, no less than work time, is itself a factor of production. The worker who shows up for work tomorrow morning is, after all, as much a product of today's labor and leisure combined (and added to those of yesterday and all the days before) as are the commodities produced today a product of that worker’s labor power combined with raw materials, machines and overhead.

That, of course, is an old insight of Marx's, and it is a view reiterated in the 1920s by John Maurice Clark’s analysis in Studies in the Economics of Overhead Costs. To illustrate, Clark posed the extreme example of what would happen if a worker was compelled to continue working around the clock. Eventually, that worker’s hourly productivity would decline to zero as he or she collapsed from exhaustion. Clark doesn’t mention it, but of course the same thing would happen if, after a moderate day of work, the worker was to spend the rest of the day and night at the tavern repeatedly “bending an elbow.” To attribute a recuperative power to leisure assumes the leisure is being used in one way and not another, that it is put to the task of recuperation.

But never mind Marx or Clark, Chapman's theory confirmed the same thing from the standpoint of marginalist analysis. One doesn't resolve the difficulty of calculating the returns to various factors of production by simply asserting, without grounds, that one of them is something else. To do so is to replace analysis with fairytales.

By treating leisure as a "normal good" rather than as a factor of production, the canonical labor supply model opposes a theory not with another theory but with an evasion of what the earlier theory says. Is it a way of ploughing ahead with a failed and inherently ideological paradigm by simply putting on a particularly clumsy and obtrusive set of blinders? Or is there some inherent difficulty in communicating "genuine theory" that can readily be avoided by teaching pseudo-theory?

Next: Life is excessively complicated

LEISURE AS A FACTOR OF PRODUCTION II

Nemo contra deum nisi deus ipse
(Second of a seven part, plus epilogue, series)

by the Sandwichman

All those gripes I talked about in part one are very well and good. But as I said, they miss the mark -- namely, the standard of nemo contra deum nisi deus ipse: “no one can stand against a god unless he is a god himself.” You can’t replace something with nothing. You can’t substitute a nullity for a theory.

Except, it seems, in economics.

S.J. Chapman's theory of the hours of labour (1909) was described in 1932 by Hicks as "[t]he classical statement of the theory of 'hours' in a free market..." Expounded in Chapman's article, "Hours of Labour" and restated in Pigou's Economics of Welfare, "[t]here is very little that needs to be added to the conclusions of these authorities."

Twenty-four years later, in "Hours of Work and Hours of Leisure," H. Gregg Lewis (1956) described "the theory of the demand for leisure viewed as a consumption good" (that is to say the butt-on-a-slide model) as the "orthodox approach."

So, how did this allegedly "orthodox approach" supplant the prior "classical statement"? You tell me.

According to Nyland (1989), Hicks introduced a "simplifying assumption" that the given working day be assumed to be optimal for production. The purpose of the simplification was to get around the difficulty of calculating the returns to the various factors of production posed by the fact that worktime has both a temporal and an intensive dimension. Anyone familiar with the Cambridge capital measurement controversy (see Cohen and Harcourt 2003) might recall that the ambiguous temporal/intensity character of labour is not the only obstacle to calculating the returns to capital, nor was it the one highlighted in the controversy.

To be sure, Hicks himself cautioned about the "very considerable dangers" of his simplifying assumption and the need to "think back our arguments into a more cumbrous but more realistic form as frequently as possible." But it would seem that when you introduce an assumption that is diametrically opposed to the more realistic one, it's not so easy to do the thinking back and "as frequently as possible" decomposes into never.

Next: The marginal utility of evasion.

LEISURE AS A FACTOR OF PRODUCTION I

The butt-on-a-slide model
(first of a seven part, plus epilogue, series)

by the Sandwichman

Anyone who has taken an introductory economics course has heard of something euphemistically called the "labor supply model." I won't expound on it here in detail because it’s common as dirt. If you don’t already know it, you can find out without difficulty.

All I really need you to know is that, paradoxically, there is no labor in the labor supply model. There's only income, leisure and the preference of the individual for more of one or the other. Leisure is assumed to be a "normal good". Labor is just what's left over after the individual "consumes" all the leisure he or she wants. Working resolves itself into the vacuous double negative of not-not-working.

One more thing: the canonical labor supply model is typically graphed to look like a minimalist line sketch of a butt on a slide. So I’ll call it the butt-on-a-slide model to afford it the disrespect it deserves.

Thankfully, this butt-on-a-slide model has come in for criticism lately. But it seems to me that this criticism has so far missed the mark.

Spencer (2003, 2004a, 2004b) objected that the model effaces the qualitative dimension of both work and leisure, a dimension that was specifically addressed in the approaches of Jevons and Marshall, not to mention Marx. Philp, Slater and Harvie (2005) disputed the epistemological coherence of the model's microfoundations, concluding that "the indifference curves which underpin labor-leisure preferences are themselves founded on axioms which have been shown to be problematic elsewhere in neoclassical economics. Derobert (2001) questioned the pedigree of the model, noting the paradoxical disappearance of labor, documenting the bibliographic anomalies in the model's transmission and discovering that the model's final codification (by Scitovsky 1952) was accompanied by a heralding of its defects: "the above argument also shows the kind of pitfalls one might fall into by regarding leisure as a commodity."

Next: Nemo contra deum nisi deus ipse