There is a growing contingent at the Fed advocating interest rate increases sooner rather than later. I continue to think that is a mistake.
The reasoning from those who think it's time to begin reducing monetary stimulus is that the natural rate of output -- the full employment level of output -- has fallen so much that even though the recovery to date has been slow, nevertheless we are nearing potential output. Thus, any further push to increase output further could be highly inflationary.
Why do I think this is incorrect? I believe there are several types of shocks that can hit the economy. There are both permanent and temporary shocks to aggregate demand, and there are both permanent and temporary shocks to aggregate supply. As I explained here, analysts who conclude we are almost back to potential output may very well be confusing permanent and temporary shocks to aggregate supply.
As Charlie Plosser explained to me recently, it is difficult to sort aggregate demand and aggregate supply shocks. Aggregate demand shocks can produce supply shocks, and supply shocks can have an effect on demand. The explanation I was given by Plosser was, I think, intended to convince me that what look like aggregate demand shocks are actually the result of supply shocks. However, I think the explanation works better in the other direction. For example, repeating a previous argument:
When there as a large AD shock in the form of a change in preferences, say that people no longer like good A as it has gone out of fashion and have now decided B is the must have good, then there will be high unemployment in industry A and excess demand for labor and other resources in industry B. As workers and resources leave industry A, our productive capacity falls and it stays lower until the workers and other resources eventually find their way into industry B. When this process is complete, productive capacity returns to where it was before, or perhaps goes even higher. Thus, there is a short-run cycle in productive capacity that mirrors the business cycle.
Even a standard business cycle type AD shock will temporarily depress capacity and produce similar effects. Suppose that interest rates go up, taxes go up, government spending goes down, investment falls --pick your story -- causing aggregate demand to fall. When, as a result, businesses lay people off, close factories, etc., productive capacity will fall. It can be cranked up again, and will be when the economy recovers, but rehiring labor and taking equipment out of mothballs takes time. In the interim the natural rate of output falls and, just as with a change in the preference for good A versus good B, a negative aggregate demand shock can cause "frictions" on the supply side that temporarily increase the natural rate of unemployment. And there are many other ways this can happen as well.
The point is that there can be short-run cyclical AS effects, and failing to account for these can lead to policy errors.
So it may be true that productive capacity has fallen, but I beleive the fall is largely temporary, not permanent. (To be clear, I think there is a permanent component, but it is nowhere near as large as the inflation hawks are assuming -- i.e. the full employment target, once temporary effects have been cleared out of the way, is higher than the estimates that are behind the hawkery. Essentially, what I am arguing is that the temporary supply shocks are, in part, a function of AD shocks, but the effect of the AD shocks on AS wanes over time.)
If this is correct, policymakers should not be concluding that the shocks are permanent, throwing up their hands, and saying there is nothing more we can do. Instead, if, as I believe, much of the fall in productive capacity is temporary, then the job of policymakers is to make sure that employment recovers as fast as the temporary supply shocks wane. That won't be easy, employment so far has been very slow to recover and if that continues it's entirely possible that productive capacity will recover faster than employment. If policymakers try to freeze employment at a level that is too high out of misguided worries about inflation, then they will hold back the recovery and make this problem worse. That's the opposite of what they should be doing.
I could be wrong, which is what I'd like the hawks to consider. That is, what are the costs of being wrong versus the costs of being correct? My view is that the costs of doing too much -- the inflation cost -- is much lower than the costs of doing too little, i.e. the costs of higher than necessary unemployment (though see David Altig). I'm aware that we differ on this point, those in favor of relatively immediate interest rate increases see the costs of inflation as very high and it's this point that I hope will generate further discussion. In reality, how high are the costs of a temporary bout of inflation -- I have faith that the Fed won't allow an increase in inflation to become a permanent problem -- and are they so high that they justify erring on the side of doing too little rather than too much? I don't think they are, but am willing to listen to other views.