Long and Variable Lags

I wrote this in a comment to a post on Economist’s View, but it’s sufficiently critical to my world-view to deserve a post of its own in my blog. The launching point is a Vox EU piece by Tito Boeri and Luigi Guiso, in which they denominate three causes underlying the current financial crisis:

• The low financial literacy of US households;
• The financial innovation that has resulted in the massive securitisation of illiquid assets; and
• The low interest rate policy followed by Alan Greenspan’s Fed from 2001 to 2004,

of which the third is “by far the most important.”

So far I don’t particularly disagree, but Boeri and Guiso are then quick to paint Alan Greenspan, along with “the Keynesian sirens” whose song he chose to follow, as the villains in this matter. In my own view,

If there was a monetary policy mistake in 2002-2003, the mistake was a failure to anticipate the lag between policy action and results. When a monetary stimulus is transmitted through the housing market, the lags are apparently longer than the typical lags associated with monetary policy historically. Greenspan kept easing because he wasn’t seeing results from his earlier easing, and in retrospect, it appears that he just wasn’t waiting long enough for those results.

If that was the mistake, then Bernanke may have made the same mistake in the other direction (and if so, he is apparently still making it). The housing bust that is taking place today is more or less a deliberate result of policy: the Fed wanted to slow down the economy, and the only way they could do it was to put the brakes on the housing market. Bernanke kept tightening in the first half of 2006 because he wasn’t seeing results from Greenspan’s earlier tightening (because the housing market was still booming, or it had just begun to slow down, and the macroeconomic effects weren’t yet apparent). Starting in the second half of 2006, Bernanke got what he was looking for: the monetary tightening from 2004 and 2005 finally was affecting economic growth.

If we look at the time lag from the beginning of the tightening (June 2004) to the point where the effect was first apparent (July 2006), it suggests that about half the impact of the tightening has yet to be felt. (Apply the same time lag to the end of the tightening, June 2006, and it suggests a peak effect starting in July 2008.) The hope, I suppose, is that we haven’t yet, even now, seen the full effect of Greenspan's original loosening outside the housing market, and that the residual effect of the loosening will counterbalance the effect of the tightening via the housing market. (For example, the dollar seems to have weakened very slowly, and we are beginning to see the effects on the trade balance.) When I think about it, that scenario seems pretty optimistic. It seems more likely that Bernanke has indeed repeated Greenspan’s mistake in reverse.

From this passage, it may begin to become clear what I meant when I said, “The Fed should encourage the continuation of a housing boom (or something of that nature).” I realize that, at the peak of the boom, the stimulus was stronger than what the US economy needed, but the policy today seems to be one of keeping interest rates high enough to erase, ultimately, a large part of the boom, perhaps bringing the US economy back to where it was in mid-2004. (Even that is only if we assume no overshoot on the downside. And the current financial crisis is already laying the groundwork for just such an overshoot.).

I guess some economists were perfectly happy with 2004, but I certainly wasn’t: the unemployment rate was 5.6% in June 2004; by a typical recent estimate, that’s almost a full percentage point above the natural rate. (The Philadelphia Fed’s most recent Survey of Professional Forecasters puts the median natural rate estimate at 4.65%. Anyone who has been following this blog since the beginning will know that I put the natural rate even lower.) So, while I don’t think the Fed should try to bring back the boom conditions of early 2006, I also don’t think it should deliberately “puncture” the “bubble.” Scrape off the froth, but don’t pour half the beverage down the drain.