Monday, February 18, 2013

How to Think About Aggregate Labor Markets

The remarkable Tyler Cowen has me scratching my head again.  Here he is at the beginning of a critique of minimum wage laws and sticky-price Keynesianism.  (The latter is an oxymoron, as anyone who knows the history of Keynes’ dispute with the “Treasury view” knows, but we’ll let it pass.)
Let’s say your labor is worth $10 an hour but you won’t go back to work for less than $12, thereby leading to the unemployment of you. 
In essence you are self-imposing a minimum wage on that market, but the employer is responding by leaving you jobless.
You can guess where this is headed.

The interesting thing is that Cowen apparently has no inkling that most people would find his opening sentence insulting.  It implies that some significant portion of the unemployed are simply worth less than they think they are.  Imagine going up to someone who’s been without a job for a while and saying, “I’m sorry, but have you considered the possibility that your abilities are really not very valuable, and your job search has failed because of this delusion?”  If you insist on saying this, I’d advise doing it from a distance.

Now, of course some people have an inflated sense of self-worth, and others are too bashful.  It might be an interesting research project to see whether the distribution of these types is correlated with employment status.  I don’t have any priors about which would predominate where—do you?

What makes this interesting to an economist is that the popular perception of unemployment actually fits how we model the aggregate labor market pretty well.  Let me explain.  The view of most unemployed people, according to the interviews I’ve seen, goes something like this: “I’m looking for a job, and I’m willing to take something that’s worse than what I used to have, but I haven’t found anything yet.”  The unemployed person hopes that the job is out there but that the connection hasn’t been made.

This formalizes to the now-standard model of search and matching, for which Peter Diamond and especially Dale Mortensen and Christopher Pissarides split a Nobel.  Equilibrium in such models does not occur where the Beveridge curve crosses the 45-degree line, which it would if the criterion were supply equals demand, but depends on a larger array of factors.  The model is used to explain why the ratio of unemployed workers to vacant jobs is typically greater than one, even in “full employment”.

This is how knowledgeable economists study aggregate labor markets today.  Supply and demand, as deployed by Cowen, is a special and highly unlikely case that assumes away the complications that make the theory empirically relevant.  If you see someone drawing supply and demand curves for labor and trying to explain unemployment as a result of too-high wages, you know they are employing outmoded methods.

What’s striking is that the more high-powered model actually conforms better to popular intuition.  You have to have a rather uncharitable view of human behavior to believe that excess unemployment is due to people overestimating their true worth.


johnhaskell said...

This better articulates my problems with the current S/D arguments vis-a-vis the labor market. Some argue the S/D analysis as though it is fungible across markets, which assumes equal bargaining power, ability to trim costs, and so forth. For instance, if a business wants to hire someone at $9, but that person wants to be hired at $10, who is in the better bargaining position. To answer that question, we need to look at the respective situations each are in. Presumably, a business can let that position sit idle, and either increase productivity through the rest of the workforce, or by reducing total output. An job seeker, meanwhile, short of being wealthy independent of that position, will have little choice but to accept that position--but will do so reluctantly. The worker has little to no bargaining power, and any attempts to unify that little bargaining power (unions) is being stripped away.

What really bothers me about the arguments put forth by the Cowen's regarding S/D and the labor market is there is an assumption that businesses are paying employees based upon productivity. In other words, if a position sits idle, it will just cut output, or increase productivity through the rest of the workforce and increase their pay in some equal measure to what was offered to fill that idle position. But we have seen an overall stagnation of wages in the face rapidly rising productivity. If pay is redistributed within the firm either through a new position or to other workers because that is that the business can afford based on productivity, then we shouldn't be witnessing such a disparity between the two.

Greg Hill said...

I think it's worth distinguishing between labor market search models in which bargaining over net benefits usually leads to poor outcomes and search models, like Diamond's in which there are externalities and multiple equilibria.

A bit more here: