Thursday, July 21, 2011

More on wages, labor surplus, and that wage distribution graph Krugman shared

So the other day Krugman shared a great graph from Barrattieri, Basu, and Gottschalk that appears to show some wage rigidity. If you define wage rigidity as "wages have a tough time falling" it not only appears to show wage rigidity - it actually does show it:

The idea is that given a typical supply and demand graph, if demand falls wages and employment should fall. Since only employment is falling, people conclude (1.) wages are rigid, and (2.) that creates a labor surplus which reduces employment by even more than it would otherwise be reduced. The solution is "let wages fall". One of the easiest ways to do that if you think workers are being intransigent is to raise the price level.

As I said, if we just define "wage rigidity" as this sort of censored wage growth distribution, then obviously we have "wage rigidity". What follows much less clearly, though, is that this creates a "labor surplus" and that that surplus is somehow coterminous with "unemployment".

But consider the possibility that spending, long-term investment, and employment decisions are made simultaneously. Take my wife and I, for example. We're five years out of college. Immediately after leaving school we lived relatively cheaply because Kate went immediately to get her master's degree. She took out student loans and I just had an entry-level salary at a non-profit research group, so we didn't buy a house, we decided not to start a family yet, and we lived fairly frugally. Two years later I had been promoted and got a big wage bump, Kate was out of school and working herself, and our consumption changed. We lived in a nicer neighborhood, ate out more, our wine rack was always full and we would go to wineries in Loudon and Charlottesville regularly. Now I'm going back to school and while I have a stipend it's a big wage cut. So we moved to cheaper neighborhood and are living more cheaply than we did just two years ago.

Right now our labor supply function is fairly elastic because our spending commitments aren't very stringent commitments. But in five or ten years this won't be the case. I expect we'll have a house, a few kids etc. Clearly there are legal (to say nothing of moral) obligations associated with the maintenance of both of those things. Now, at that time (i.e. - prime working age for workers) when I think about making labor market decisions my reservation wage is going to be much less flexible. Most labor isn't bought on a spot market, after all, and it would benefit me to wait a little to get a wage that can actually maintain my mortgage and feed my kids than immediately lowering my reservation wage to get a job. Thus far, of course, this is just standard search theory type stuff.

Now, the point is these sorts of consumption commitments (kids, houses, etc.) are also committed to with specific employment expectations in mind. So if you go through several years of decent labor markets, the labor supply curve to the left of the equilibrium point is going to get increasingly elastic. People get into certain consumption habits - particularly debt-financed consumption habits - that are hard to extract themselves from, and therefore which are going to increase people's reservation wages and increase the value of searching (should they ever lose their job). In a robust labor market, the shape of that labor supply curve to the left of equilibrium doesn't bother us all that much. But if labor demand drops we're going to see the brunt of the labor market adjustment carried in employment.

So why this censored distribution that Krugman shows? Well, different sectors are behaving differently. While they're probably growing slower than they would have otherwise, health and education are actually not doing so bad through the crisis. If you have elastic labor supply to the left of equilibrium because people tie themselves up in consumption commitments when times are good, and if you assume different sectors are experiencing different levels of labor demand, then you can get Krugman's censored wage growth distribution without any (1.) labor surplus or (2.) puzzling questions about why workers won't take lower wages:

So to recap, here's one story on that wage distribution:

1. During tight labor markets, people feel like they can make long term consumption and life plans, so they do.

2. A lot of these life plans are necessarily debt financed - mortgages, car loans, sending kids to school, etc. Those plans that aren't debt financed are still tough to get out of (it starts to get awkward to leave a kid in a basket with a note on it at a church doorstep when they're seven or eight years old).

3. These consumption commitments and life plans make maintaining a certain income level for extended periods very important - so important that it makes sense to forgo employment a little longer and maintain a higher reservation wage than take a job immediately that will provide an income that can't support your commitments. This results in a relatively elastic labor supply curve for those currently employed.

4. During a period of weak demand, when all sectors are weaker than they would be otherwise but there is still some dispersion in growth in different sectors, that will result in a censored distribution of wage growth, a la Barrettieri, Basu, and Gottschalk.

5. No labor surplus.

6. We're at an equilibrium, so no great puzzle around what workers are thinking not taking lower wage rates.

7. Inflation to lower real wages makes less sense here. Inflation to relieve debt burdens does make sense because presumably that would make labor supply less elastic, but

8. It's still not really an answer because while the more elastic labor demand is, the more of an employment boost you'll get from #7, there's still no guarantee you're solving the underlying effective demand problem. A clearing labor market is not the same thing as a labor market at full employment (hell, if it were we could have stopped all this at #6).


  1. "A clearing labor market is not the same thing as a labor market at full employment."

    1 worker for 1 job Full employment

    2 workers for 1 job Unemployment

    1 worker for 2 jobs What is this called? Is this the situation in which the labor market is at full employment, but not cleared? Others?


  2. Daniel,

    So according to your example, younger people should be more flexible with their demands with respect to quality, location and wages from a job. But unemployment rate amongst teens are very high compared to folks in their 30s. Why?

  3. Ed -
    Well for your third example, why wouldn't it be cleared? To say a market clears is just to say that quantity demanded and supplied at a certain price is equal, right? So 2 workers 1 job could be a clearing market or it could not be. 1 worker 2 jobs could be a clearing market or it could not be. As to what I would call your third example - presumably that's a "tight labor market", and if the one worker in the economy has a job it's at full employment, right? To determine whether the market clears or not you'd have to know something about the supply and demand schedules.

    Subhi -
    Demand for younger workers, for a variety of reasons, is lower than demand for older workers.


All anonymous comments will be deleted. Consistent pseudonyms are fine.