Tuesday, 21 November 2017

Simon Wren-Lewis and the job guarantee.

Job Guarantee (JG) is a name for an idea which has been around for a very long time. It’s the idea that there are an infinite number of useful jobs to be done, thus instead of paying the unemployed to do nothing, government could pay them to do something useful. Far as I know, Pericles in Ancient Athens was the first to implement that idea (about 2,600 years ago). The WPA in the US in the 1930s was another example of JG.

Wren-Lewis published an article (1) on JG a few days ago and made the point that if pay for JG work is too generous, that will attract people way from regular jobs, i.e. the effect will be to cut job search efforts, which equals cutting aggregate labour supply to the regular jobs market, which is inflationary. Hence aggregate demand has to be cut, which destroys regular jobs: hardly the object of the exercise.

That point by Wren-Lewis is far from original: indeed it’s a fairly obvious common sense point (which I and others were making at least twenty five years ago (2)). Maybe Pericles was aware of the point as well.

However, two of the more enthusiastic and I think naïve advocates of JG are having none of it. Neil Wilson (3) and Brian Romanchuk (4) argue that a generous JG wage will not have the above adverse effect: rather, a generous JG wage will induce employers to offer higher pay for the low paid, i.e. JG in that capacity would work much like minimum wage laws. And as evidence to support that, Neil Wilson cites a study which show that raising minimum wage rates has no effect on employment. (There are of course other studies which claim that raising the min wages DOES increase unemployment, but that’s not central to the argument here.)

The flaws in the “Wilson / Romanchuck” argument is thus.

First, the broad claim that raising the minimum wage never stokes inflation or raises unemployment must become nonsensical at some point, as the minimum wage is raised. E.g. if the minimum wage was $200/hr the effect would be catastrophic. Thus there is what might be called a “maximum feasible minimum wage”.

Now if we’re going to have any sort of minimum wage, clearly the best way to effect that is the conventional way: that is, min wage laws. I.e. it’s frankly a bit odd to set up a form of employment (JG) which may make work easily available to absolutely everyone, but probably won’t, and which pays the desired min wage, and then hope that anyone working for less gets themselves a JG job. Put another way, if we think that working for less than $X/hr is unacceptable, then the obvious and simplest way to enforce that is a law which says “no one shall be paid less than $X/hr”.

Moreover, that $X level of pay ought to be at the above mentioned “maximum feasible level”.

But that puts JG into check mate in the following sense. JG cannot pay more than $X/hr because that puts the pay above the “maximum feasible level”: i.e. the effect of pay in excess of $X/hr would be to induce employers to shed a significant number of low paid “regular” workers.

In short, there are no logical circumstances in which the “Wilson / Romanchuk high JG pay doesn’t destroy regular jobs” theory holds.

And finally my latest and needless to say “seminal” paper on this subject is here.


1. Article title: Some Thoughts on the Job Guarantee.
2. Article title: Workfare: A Marginal Employment Subsidy…”;.
3. Article  title: Thoughts about the Job Guarantee: A Reply.
4. Article: On Using NAIRU to Analyse a Job Guarantee.

No comments:

Post a Comment

Post a comment.