Tuesday, October 6, 2009

Assorted Musings

Here are some more assorted musings:

(1) Caroline Baum asks a probing question: if the Fed is not able to identity an asset bubble and prick it in a timely fashion, how then is it able to know what are appropriate spreads in the credit market as it expands it balance sheet to shore up the financial system? In the former case it claims ignorance and refuses to intervene while in the later case it claims prescience and readily intervenes. Baum notes this asymmetry is typical of Fed policy in recent times.

(2) Has the Fed's independence already been compromised? Nouriel Roubini and Ian Bremmer argue yes and its not because of congressional probbing. Rather, it is because of its bailout of large financial institutions last year. Roubini and Bremmer also explain that if the Fed is not careful it could set the stage once again for the next bubble, a point recently made by Peter Boon and Simon Johnson.

(3) Roberto M. Billi has a new article that examines whether monetary policy was optimal in past deflation scares. He looks at Japan in the period 1990 to 1995 and the United States from 2000 to 2005. Using a Taylor Rule he concludes that monetary policy was too accommodative in the case of the United States. While I concur with his conclusion and have said so before, I also would like to note several things. First, the article assumes that deflation is always economically harmful. Deflation, however, can arise for reasons other than a collapse in aggregate demand. As I have noted before, positive aggregate supply shocks can also generate benign deflationary pressures and this form has far different policy implications than deflation arising from a collapse in nominal spending. Second, when constructing the federal funds rate prescribed the Taylor Rule one needs a measure of the output gap. There are, however, different measures of the output gap and, as result, different implications for the Taylor Rule. A popular version for the United States is the CBO's output gap measure. John Williamson of the San Francisco Fed , however, argues that the CBO measure is flawed since it doesn't allow for short-run fluctuations in the growth rate of potential output. Here is his preferred measure (LW) graphed along with the CBO measure:

The CBO measures show a negative output gap during the housing boom while the LW measure shows a positive output gap. The LW makes more sense for this period. Now plug the LW measure into a Taylor Rule and there is even a stronger case that monetary policy was too loose during the housing boom period. Finally, there are other ways to learn the stance of monetary policy. Here is one measure I like.


  1. On Federal Reserve "independence". Would it not be better to outsource monetary policy to the BIS ? In the 1970s, the UK effectively outsourced policy to the IMF during the sterling crisis under the Labour govt. Italy and Greece adopted the euro to "tie the hands" of their dysfunctional governments. America might wisely copy this strategy given the fairly dreadful quality of governance at all levels in your country.

  2. I like your idea. I will run it by my congressman and the Dallas Fed!

  3. Greetings David,

    I will ask here rather than on the post you refer to in case you are no longer watching its comments. What is the reason why you use nominal GDP growth as a benchmark for short-term interest rates (I have seen this used many times in market economists' research, but the justification is rarely given)? Is it some Ramsey golden rule type idea? Thanks.

  4. Hey RebelEconomist,

    I first came across this idea in the Economist magazine back in 2004. They explained the way to “interpret this [metric] is to see America’s nominal GDP growth as a proxy for the average return on American Inc. If the return is higher than the cost of borrowing, investment and growth will expand [and vice versa].

    I would look at it this way. The fundamentals behind the natural rate of rate of interest and the natural rate of output rate are similar. In the former case productivity growth, population growth, and intertemporal preferences determine the natural interest rate. In the latter case, productivity and population growth also matter. When either of these accelerates both the natural interest rate and the natural rate of output increases. Throw in inflation to get the nominal versions of these series(i.e.NGDP,ffr) and one should see similar movements in them given (1) the Fed is not causing the market interest rate to deviate from the natural interest rate and (2) intertemporal preferences don't experience big swings. If, however, the Fed does cause the market interest rate to deviate from the natural interest rate it is likely to also cause the economy's growth rate to deviate from its natural rate level. In this case, moneary policy has not been neutral.

    I am actually planning to do a paper with a colleague where we formally model and estimate this relationship and hopefully better account for the influence of changes in intertemporal preferences.

  5. Thanks David. With all due respect to you, it seems to be, as I thought, one of these economic ideas that emerges from a stylised analysis that gets picked up and used uncritically as a rule of thumb by many who have not followed the theory. I would imagine, for example, that GDP growth as a country absorbs more people (eg by immigration) has different implications for interest rates than GDP growth arising from increases in total factor productivity.