Monday, May 31, 2010

OECD Report

I'm taking my lead from Krugman's New York Times column this morning (previewed in his blog), and going straight to the OECD Economic Outlook document. This is a rather bland piece of work. Most of it reads like mainstream intermediate macro textbooks (Mankiw or Abel and Bernanke), and is more or less innocuous for the same reasons. You can quarrel with the basic modeling framework this is based on, but it's some mix of standard Old Keynesian and New Keynesian economics, and is therefore at least internally consistent.

Where the thing actually takes a stand, and gets a reaction from Krugman, is here. There is this:
In this unsettled financial environment, we think governments need to get out ahead of markets, because otherwise they will be hostage to them.
Then there is this:
Moving ahead of markets requires a long term view, that allows all the policy actions to unfold and have an impact. This is expecially true now when markets seem to be returning to a short term view.

On inflation, the issue is not whether it is a risk today—it is not—but whether it will be a risk in two years’ time. Monetary policy needs to be forward looking and, on our projections, things will be rather tighter by then. This means easing up on monetary stimulus. To be clear, we are not arguing for contractionary policy, but for progressively less stimulus. In fact, stimulus should not be withdrawn completely until the economy returns to full employment. But the process should be started fairly soon, to take into account the well known long and variable monetary policy lags. Raising interest rates above zero would also signal a commitment to price stability that would help to prevent inflation expectations from drifting up. If expectations do start to rise, policy will eventually have to tighten more to bring them down again, as we found out in the 1980s.
I think Krugman and I agree here, but for different reasons. It is clear that, when we formulate monetary policy, we have to start in the future, and work backward, in an attempt to weigh the risks of alternative current policies. However, what I think the OECD is recommending here is wrong, for the Fed or the ECB. The large quantity of reserves sitting on the Fed's balance sheet do not represent an inflation threat. Under the current regime, the Fed has all the power it needs to control inflation by setting the interest rate on reserves, which is currently the key policy instrument. There is no reason to raise that rate currently, as inflation is very low, and anticipated inflation is low, as measured by the difference in yields on Treasury securities and Treasury inflation-protected securities (TIPS). There is plenty of time to tighten when we actually start to see more inflation. For me, the key risk is the maturity mismatch on the Fed's balance sheet. The Fed should be selling mortgage-backed securities now to reduce that risk. That would not represent a tightening of monetary policy - it's an asset swap that is essentially irrelevant. For an earlier take on this view of things, applied to the ECB, see this.


  1. TIPS are linked to the CPI that intentionally excludes investment items, and may fail to capture an increase in money and credit relative to GDP that only shows up in a specific asset class (as the CPI, through its design, did not highlight unsustainable increases in housing prices relative to incomes).

  2. Yes, if the CPI is a poor measure of the price level, the anticipated inflation measure I back out from the TIPS yields will not be any good either.