"Will Stimulating Nominal Aggregate Demand Solve our Problems?"
There has been a bit of a pushback, both implicit and explicit, to calls to implement policies to accelerate hiring. For example, Jim Hamilton recently noted an old theory of his where some types of unemployment cannot be overcome through standard stimulative policies (this was in response to a question about whether Arnold Kling's recalculation model can explain asymmetric adjustment, but I am focusing on the technological and physical constraints present in both Hamilton and Kling's model, not whether the asymmetries can be explained):
Will stimulating nominal aggregate demand solve our problems?, by Jim Hamilton: ...[I]n 1988 ... I presented a model in which unemployment arises from a drop in the demand for the output of a particular sector. The unemployed workers could consider trying to retrain or relocate, or might instead decide to wait it out in hopes that the demand for their specialized skills will come back. ...[T]he key kind of unemployment that I think this sort of model describes-- waiting for an opening in the particular area in which you've specialized-- is caused by drops in demand...
Insofar as the frictions in that model are of a physical, technological nature, increasing the money supply would simply cause inflation and not do anything to get people back to work. I should emphasize that I built that monetary neutrality into the model not because I think it is the best description of reality, but in order to illustrate more clearly that there is a type of cyclical unemployment that stimulating nominal aggregate nominal demand is useless for preventing.
My personal view is that real-world unemployment arises from the interaction of sectoral imbalances with frictions in the wage and price structure of the sort documented by Truman Bewley and Alan Blinder. The key empirical test, in my opinion, is at what point inflationary pressures begin to pick up. If Krugman is correct, we could have much bigger monetary and fiscal stimulus without seeing any increase in inflation. If the sectoral imbalances story is correct, it would be possible for inflation to accelerate even while unemployment remains quite high. ...Thus, according to this view, some part of the sectoral imblances in of a "physical, technological nature," and standard demand side policy does not help. Policy may be able to induce people to stop sticking around for jobs that will never materialize and move on, but those typically aren't the kinds of policies typically associated with stimulating employment, e.g. tax credits to encourage hiring.
A new colleague of mine, Nick Sly, emails that it is not always optimal, from a long-run economic growth point of view, to provide incentives for firms to hire workers, how those incentives are structured is crucial:
There is a paper on my website called Intraindustry Trade and the Composition of Labor Market Turnover. (It is a heavily revised version with more of a trade focus.) The highlights of the paper are:
1. Because of constant turnover in labor markets, hiring costs are persistent for all firms.
2. Turnover and Hiring occur both because firms update their workforce (job creation costs) and to replace workers who leave for reasons unrelated to the firm (worker hiring costs). These phenomena are distinct 3. (KEY) I show (theoretically and confirm empirically) that each source of turnover has the opposite effect on the incentives of firms to adopt state-of-the-art production techniques. As a consequence industries with different compositions of labor mobility have varying degrees of engagement of foreign markets.
The act of hiring workers could be the result of demand side (firms creating new jobs) or supply side (workers need to be replaced) incentives. We may not want to jump to quickly to put people back to work if it means employing less productive production methods. The short term gains can be lost as poor matching of workers and adoption of weak production methods alter the recovery path.
I believe that the timing of the hiring tax credits, and the sort of hiring it promotes (i.e. creating new vacancies versus filling previously existing positions), will determine the long-run consequences of such a policy.
Let me try to express the main point a different way. When firms hire workers, as they are constantly doing, they have a choice between using old or new technology, and the way in which hiring incentives are structured can affect this choice. As we think about putting programs to induce firms to hire workers in place, we need to be sure that we are not giving firms the incentive to use old rather than new technology so that economic growth is maximized, and we also need to be sure that we don't distort the choice firms make toward labor intensive rather than growth maximizing change.
Our economy faces lots of adjustments as it recovers from the recession, far more than in some past recessions when we could return, pretty much, to what we were doing before the shock hit. But not this time. We have adjustments in the auto, finance, and housing sectors just for openers, and there are other underlying adjustments that are in progress as well (e.g. in the manufacturing sector). As these adjustments occur, it's important that we don't impede the necessary change, or induce firms to make suboptimal choices as we attempt to induce them to hire more workers.
But if we give firms the time they need to make the changes that are needed, there will be excess labor during these adjustment periods, both from sectoral reallocations and from technological change. The question is what we are going to do to help people who lose their jobs or are otherwise negatively affected by these transitions.
One choice is to induce firms to house the excess labor during this time period through tax or other inducements, but the danger is that in doing so you distort the choices of firms away from the optimal trajectory. Another choice is for the public sector to absorb much of the burden by providing jobs to the unemployed and providing the aid needed to carry workers through the adjustment period (and we can also provide incentives for workers to relocate in areas where they have a better chance of finding employment).
Even better, though, is to structure the incentives so that the technological change is encouraged by the hiring of new workers. For example, Nick Sly suggests that the hiring credit be only for "new" jobs offered by firms, somehow defined, because this gives firms an incentive to both hire new workers and to employ the latest technology. Thus, the best choice of all is to provide incentives to employ workers that have, as a byproduct, and inducement to maximize technology and economic growth, and then use public employment (e.g. infrastructure) or aid to help those who remain unemployed.
No matter what we do, however, there will be those who cannot find employment during these time periods, and we need to do a better job than we do in helping those who, through no fault of their own, are caught up in the tumultuous change that sometimes occurs in modern economies.
Posted by Mark Thoma on Monday, October 12, 2009 at 02:59 PM in Economics, Productivity, Taxes, Technology, Unemployment |
You can follow this conversation by subscribing to the comment feed for this post.