Monday, June 28, 2010

Replies to Comments on "Kocherlakota Sells Out"

In a previous post, I gave Narayana Kocherlakota a hard time about his stand on New Keynesian economics. A couple of commenters characterized this as unnecessarily vitriolic and personal. Of course it was personal. I have an enormous amount of respect for Narayana, but he is a public figure now, with some influence on public policy in this country, and I expect a lot from him. I have said nice things about him before (see this piece), but if I disagree with what he says, I think I should say so. Economists need thick skins to survive, and I can assure you that Narayana has one as well.

Now, some of you seem to think, like Narayana, that New Keynesian economics is coherent. A while back I took the time to study what New Keynesians were doing. This research program was getting a lot of attention, and I wanted to see if there was something to it. I came to the conclusion that New Keynesian Economics had nothing to offer in its current form, and since then I have taken the opportunity to criticize this stuff whenever the opportunity arises. Here are some critical pieces I have written (two of these with Randy Wright, which also provide a constructive alternative):

1. Richmond Fed Economic Quarterly piece.
2. "New Monetarist Economics: Methods"
3. "New Monetarist Economics: Models"

What is incoherent about New Keynesian Economics? As represented in Woodford's "Interest and Prices," the paradigm has the following problems, among others:

1. Firms whose prices are "stuck" in the current period behave suboptimally, and in ways that drive all of the results.
2. These models are stated to be immune from the Lucas Critique, but for a number of reasons (e.g. fixed decision rules for price-setting) they are not.
3. The framework is supposed to be about analyzing monetary policy, but does not contain the most basic elements of what monetary policy is about, i.e. the quantities on the central bank's balance sheet. Recent work by Curdia and Woodford tries to correct this problem, but in a somewhat pathetic way.

Some people wondered what I meant by the current predictions of New Keynesian models being laughable. I'm not sure why this was not obvious. The example I gave, which is typical, was of a New Keynesian model that was currently predicting that the nominal interest rate should be less than -5%. What would you think if I offered up a model that predicted that the unemployment rate should be -5%, and tried to interpret this in terms of what policymakers should actually do? Should you take me seriously? Of course not. We all know that arbitrage dictates that "the" nominal interest rate cannot fall below zero. The zero lower bound should be in the model, as should "quantitative easing," if we think that is important.

In my comment in reply to a commenter to "Kocherlakota Sells Out," I mentioned that I wasn't currently concerned about the unemployment rate, which of course drew the reaction that I was insensitive to the plight of the poor. As I have pointed out in a number of previous posts, I think that the US economy is recovering nicely. However, the behavior of the labor market and productivity is unusual. My conjecture, based on readily available aggregate observations and nothing more, is that this unusual behavior (including the high unemployment rate) is due to sectoral reallocation in the United States, away from manufacturing and housing, and toward services, and within the manufacturing sector. As such, there is little the government can do about it, except to wait for the reallocation to work itself out, and ease the transition with relatively generous unemployment insurance benefits.

10 comments:

  1. Stephen,

    You write:

    In my comment in reply to a commenter [me] to "Kocherlakota Sells Out," I mentioned that I wasn't currently concerned about the unemployment rate, which of course drew the reaction that I was insensitive to the plight of the poor. As I have pointed out in a number of previous posts, I think that the US economy is recovering nicely. However, the behavior of the labor market and productivity is unusual. My conjecture, based on readily available aggregate observations and nothing more, is that this unusual behavior (including the high unemployment rate) is due to sectoral reallocation in the United States, away from manufacturing and housing, and toward services, and within the manufacturing sector. As such, there is little the government can do about it, except to wait for the reallocation to work itself out, and ease the transition with relatively generous unemployment insurance benefits.

    End Quote

    But here's the thing.

    The Great Depression lasted about 10 years and who knows how long it would have gone if not for World War II.

    Japan has had a lost decade, and now two decades of at least very poor performance.

    There's this Austrian explanation that the economy is taking time to restructure, that it's finding more efficient ways to allocate recourses, but 10 years? 20 years?

    Unlike most economists I have a lot of experience and training in business, MBA from one of the top schools and successful entrepreneur (in addition to having most of the econ PhD coursework from completing all but dissertation for a finance PhD, and a great deal of independent study). I often think of business as micro-microeconomics, because you study individual businesses, managers, workers, and consumers in depth rather than whole industries and markets, or in macro the aggregation of industries and markets. This business knowledge can be valuable sometimes in thinking about economics. Here, I can tell you that it's hard to think of businesspeople taking 10 or 20 years to decide how to reallocate their recourses in response to a change in the economy like the bursting of a housing bubble, or the crash of the stock market in 1929. Time is money, and it just does not take that long to re-evaluate the economic landscape for a manager, months maybe, if you really really push it a year or two, but 10 years? 20 years? I've had the same top MBA training as is common of CEOs and other top managers, and from that I can tell you it just does not take anywhere near that long to decide where to re-hire and re-allocate idle people, to find something for them that's more productive than doing nothing at all.

    I know you've criticized the sticky prices assumption as non-optimizing behavior, but I find the existence of sticky prices compelling because as a business person I know consumers hate price increases and jumpy prices. You don't want to cut prices if you're going to have to raise them again quickly and anger customers (and my experience is that business people tend to be overly resistant to cutting prices). It does work this way for commodities like oranges and tomatoes but there's no brand to tarnish there, for consumers to become angry at, plus that culture of jumpy prices has been established for things like produce, but not for other products. In addition, the empirical research is clear that workers will resist a nominal wage cut far more than a wage freeze that equates to a real cut. They won't adjust their required wage down very quickly or easily at all.

    Models that assume people only care about real consumption, with no context or emotion attached, won't include this, but the real world certainly does.

    ReplyDelete
  2. And I think it's important to clearify/verify this: Krugman wrote on Austrian explanations of recession/high unemployment:

    My view is that the fatal flaw in Austrian economics is that it can’t explain unemployment — or, worse, that it thinks that it can explain unemployment, but is deluding itself. The Austrian view is that unemployment in a slump results from the difficulty of “adaptation of the structure of production” — workers are unemployed as resources are painfully transferred out of an overblown investment-goods sector back into production of consumption goods.

    But this immediately raises the question, why isn’t there similar unemployment during the boom, as workers are transferred into investment goods production?...

    ...And the key to all this, I believe, is that the Austrian abhorrence of explicit models, even for the purposes of clarifying thought, leaves them unaware of the holes in their account.

    at: http://krugman.blogs.nytimes.com/2010/04/07/martin-and-the-austrians/

    So, are there coherent models that give an Austrian reallocation mechanism for recession/high unemployment? If so can you please cite them and/or other good explanatory articles.

    ReplyDelete
  3. I know Krugman wants us to worry that we are in the middle of the Great Depression, but I don't see the evidence for it. And I don't think anyone should think that sticky prices and wages have much to do with the phenomena at hand.

    ReplyDelete
  4. I'm not sure I understand what you are talking about

    ''The example I gave, which is typical, was of a New Keynesian model that was currently predicting that the nominal interest rate should be less than -5%.''

    The research agenda in the last few years has exactly been about the problem of the zero bound. See e.g. here

    http://www.newyorkfed.org/research/economists/eggertsson/BrookingsPaper.pdf

    ReplyDelete
  5. Ok but here's the information that I really think is needed:

    You wrote that the high unemployment:

    "is due to sectoral reallocation in the United States, away from manufacturing and housing, and toward services, and within the manufacturing sector. As such, there is little the government can do about it, except to wait for the reallocation to work itself out, and ease the transition with relatively generous unemployment insurance benefits."

    What models, papers, and/or other writings do you base this Austrian view on? And that it can take years or even decades for businesspeople to make the decision on how to reallocate their resources in a way that's better than just leaving large amounts of human capital producing nothing.

    ReplyDelete
  6. See this story from today's New York Times. It's just anectodal, but consistent with my story:

    http://www.nytimes.com/2010/07/02/business/economy/02manufacturing.html?_r=1&src=me&ref=homepage

    ReplyDelete
  7. Ok, that was an interesting article, although anectdotal.But as I was asking, can you cite any models behind the Austrian explanation you give for today's unemployment level?

    ReplyDelete
  8. See Robert Shimer's web page:

    http://sites.google.com/site/robertshimer/

    and his article on mismatch

    http://www.jstor.org/pss/30034085

    The model is fairly stylized, but gets at the idea I have in mind.

    ReplyDelete
  9. Thanks, here's a big reason I asked:

    Krugman wrote:

    And the key to all this, I believe, is that the Austrian abhorrence of explicit models, even for the purposes of clarifying thought, leaves them unaware of the holes in their account.

    at: http://krugman.blogs.nytimes.com/2010/04/07/martin-and-the-austrians/

    ReplyDelete
  10. The Shimmer model at JSTOR at first glance basically looks like an explanation that sometimes search and find and utilize time is longer, sometimes the matching and utilizing of the available human capital process is slower and more difficult. I considered that before. I think it definitely could happen to some extent, but:

    1) How large? How much of a force, or explanation is it, because some of the empirics I've seen show it to be not much of a force at all right now. For example, here's a graph from an ongoing survey by The National Federation of Independent Businesses, a small-business trade association:

    http://voices.washingtonpost.com/ezra-klein/2010/07/why_businesses_arent_hiring_in.html

    Only a few percent of businesses list "labor quality" as their single most important problem, and that percent has plummeted since the start of the recession, and it's the lowest it's been since the graph begins in 1986 – even though unemployment is still today very high by historical standards. Meanwhile, the problem "sales" (a Keynesian looking explanation) is approximately ten times larger, has skyrocketed since the start of the recession, and during the recession is by far the highest since the graph begins in 1986.

    2) This is a reason for the government to spend far more on education and retraining. This decreases unemployment two ways. The unemployed become "employed" as students, and after they finish they can now fit the jobs needed better.

    3) This increased matching time doesn't seem like it can be nearly big enough and lasting enough to alone explain depressions and slumps that have lasted ten, twenty years.

    ReplyDelete