Friday, December 30, 2011

The Weak Recovery is a Policy Failure

In my last post I argued that nominal GDP targeting does not depend on bank lending to work.  Instead, its success depends on the Fed using the nominal GDP target to manage expectations such that the portfolios of the non-bank sector rebalance in a manner that shores up aggregate demand.  Bank lending may respond to this process, but is not essential to it. I mention this again because I just came across an interesting paper by Edward Nelson and David Lopez-Salido that lends support to this view.  The authors show that, contrary to the claims of Reinhart and Rogoff (2009), recoveries following financial crises are not inherently weaker. Rather, they depend on policy.  From their abstract [emphasis mine]:
We find that the regularity that recoveries are systematically slower in the aftermath of financial crises does not hold for the postwar United States. The pace of the expansion after recessions seems to reflect deliberate aggregate demand policy. A weak lending outlook does not appear to pose an insurmountable obstacle to the functioning of stimulative aggregate demand policies.
The implication of this paper and my previous post is that the weak economic recovery is a failure of policy to fully restore aggregate demand, nothing more.  All the talk about how recoveries following financial crises are typically weaker and that we are now in a balance sheet recession distracts from this fact. There was nothing inevitable about the Great Recession and subsequent lack of robust recovery.  A nominal GDP level target is the best way to fix this policy failure.

P.S. Edward Nelson also has a recent paper on Milton Friedman that provides a nice discussion of the portfolio channel. 


  1. David,

    Did you have a chance to read this paper?

    What are your thoughts?

  2. To borrow a phrase from Thomas Sowell, MMTers don't think past stage one.

    When a central bank purchases assets, banks receive additional reserves. If this was all, then monetary stimulus would depend on bank credit, but this is a very simplistic understanding of the situation. It's time to think past stage one.

    Actually, it's probably more appropriate to return to stage zero, because monetary stimulus does not begin with asset purchases, but rather the announcement that such purchases are planned. (Maybe it even begins with the expectation of that announcement, or even the expectation that other people expect that announcement, and so on.)

    Anyway, there are people holding assets that would rather have money from the central bank. How do we know they would rather have money? Because the central bank will offer them however much money it takes for them to accept the deal. That is, there is some x where $x would be sufficient inducement for each exchange to be made.

    These asset holders don't have to sell directly to the central bank, and, indeed, many of them can't. However, privileged financial institutions can, and they may begin purchasing particular assets in anticipation of central bank purchases. The rebalancing of portfolios, the shift of assets and money, begin even before the central bank has made its first purchase.

    Money is a unique asset. We accept money in exchange even though we don't actually want money, because we know it can be readily exchanged for something we do want.

    For monetary stimulus to be wholly ineffective, it would require every asset seller involved in the process to hold onto all the money like it was a bond. So long as some of the people who receive additional money from the cascade of exchanges instigated by monetary stimulus wish to exchange that money for something else, then monetary stimulus can boost aggregate demand.

    The trick for the central bank is to avoid buying assets that are very close substitutes for money, e.g. short-term government bonds with a near zero interest rate.

    Basically, MMTers implicitly assume that the liquidity trap holds everywhere and always. That is, they assume that if people accept money in an exchange, it's because they want to hold that money as a savings vehicle. Even if the central bank bought every asset in the economy, it would "just exchange assets" and spending would remain unchanged.

    Once you thinks beyond (and before) stage one, it's clear the monetary stimulus doesn't just create more bank reserves. Many different people, including households, may find themselves with extra money and looking for somewhere to spend it. The recovery of bank lending is likely to be a consequence, rather than the cause, of monetary simulus.

  3. In her recent NYT article it seems Christy Romer had the Nelson Salido paper in mind:

  4. Excellent post.

    There is a recondite word, "teleology." In modern use (I think) it means we find what we are looking for. If we determine that economies recover slowly after financial collapses, that is what we find when we comb through the record. Events are explained by our theory or idea.

    To which my battle cry is, as always (in recessionary, low inflationary times), "just print more money."

    Show me the money.