Monday, August 18, 2014

Immigration analysis on job market "refuted" with red herring

Dear Bob Birrell and Carla Wilshire,

In her attempt to refute Bob Birrell's analysis, Carla Wilshire makes the following claim:
Birrell chooses to ignore the dynamic effects of the labour market. The assumption that there are only so many jobs to go around has been roundly rejected. Labour economists have long known the number of jobs is not fixed. According to Nobel Laureate Paul Krugman, this lump of labour fallacy “encourages fatalism” and “feeds protectionism”. The trouble with promoting such notions is that policy-makers stop thinking about ways to create jobs.
There are so many errors in that brief paragraph, it is hard to know where to begin. But let's start with Paul Krugman. Krugman has indeed criticized the lump-of-labor assumption but he has also criticized, as "equally fallacious," the counter assumption (often referred to as Say's Law), that demand for labour increases with its supply. In fact, Paul Krugman has spent far more time recently refuting Say's Law than he has the lump of labor.

The problem with the supposed refutation of the supposed lump-of-labour fallacy goes deeper than that, though. Ultimately the refutation is based on the same hidden assumption as the alleged "fallacy" -- what was known in classical political economy as the wages-fund doctrine. In short, the fallacy and its supposed refutation are simply two sides of a paradox that arises from the static nature of analysis, similar to Zeno's paradox of the arrow.

Wilshire uses the word "dynamic" incorrectly. Wilshire is making the mistake of assuming that the colloquial connotation of "dynamic" is the same as the technical one. This is not acceptable in the context of claims about economic theory (see John Maurice Clark's discussion of "The Relation between Statics and Dynamics" or Alfred Marshall's discussion in "Distribution and Exchange").

What she presumably means by "dynamic effects" are the 'long term' effects of economic growth. But "what labour economists have long known" about the number of jobs in a growing economy is itself based on another static analysis, not on a dynamic analysis. The assumptions of the latter static analysis may be more realistic to the extent that they are based on empirical observation but that doesn't make the analysis dynamic. Often, though, the refutation is not based on empirical observation, in which case it is pure, empty assertion. There is nothing "inevitable" about economic growth.

Ms. Wilshire may have an alibi for her misuse of dynamic in that economists often make the same error of referring to static analyses as "dynamic" just because they incorporate a few rudimentary moving parts ( for example, DSGE: "dynamic" stochastic general equilibrium models). This is like referring to an animated GIF file as a "full-length moving picture."

Of course, Bob Birrell will be quite aware that he didn't assume a "fixed amount of work" (which is not the same thing as "only so many jobs to go around"). If at time "t" there are 10 jobs and 11 workers and at time "t+1" there are 12 jobs and 14 workers, then the amount of work is clearly not fixed but the number of unemployed has doubled and the unemployment rate has increased from 9% to 14%. The allegation of a belief in a "fixed amount" of work here is clearly inapplicable.

Accusing one's opponent of committing a lump-of-labour fallacy is such a hackneyed, inadequate argument, that it immediately raises suspicions that the person making the claim has no substantive argument to make.

Yours sincerely,

Tom Walker, author of
"Why economists dislike a lump of labor" and
"The "lump-of-labor" case against work-sharing: populist fallacy or marginalist throwback?"

No comments: