TCS Daily

Is Fed Acting To Halt A Recession, Or A Vacation?

By Kevin Hassett - January 29, 2001 12:00 AM

Three big pieces of news last week all point to a half- point interest rate reduction at this week's Federal Reserve Open Market Committee meeting to head off recession.

First, orders and shipments for December durable goods were pretty weak once again. The top line increase of 2.2 percent -- featured in most news accounts -- was much less positive than you might think. The gains were mostly attributable to a big swing in transportation equipment. In short, one extra airplane could explain the whole thing. The Federal Reserve usually excludes transportation equipment when it analyses this important series, so the Fed looked at the release as confirming the view that we are close to recession.

In addition, unemployment claims stayed very high. And finally, the Employment Cost Index, Fed Chairman Alan Greenspan's favorite measure of inflation, showed inflation pressures were declining.

So the Federal Reserve's Open Market Committee members will enter the grand boardroom on 20th and C Street with mostly negative economic news to review, and no fear of inflation. Another rate cut is about as certain as these things get.

That does not mean a typical recession is obviously on the way. Indeed, the data remain remarkably strong considering. Activity has slowed from the torrid pace of earlier in the year, but it is still well above a level that would traditionally signal a recession. What's going on? Are we experiencing a new kind of recession?

One way to think about this is to relate the current episode to some brilliant work on recessions done by University of California at San Diego economist James Hamilton. Professor Hamilton observed about 10 years ago that economic growth has weird, herky-jerky patterns that are inconsistent with the kind of empirical models that most economists were using to forecast the economy. He went on to develop a new model that redefined what we mean by recession in an important way, one that sheds an interesting light on our current experience.

Here's some of the thinking behind that.

The economy is very random; it sometimes goes up, sometimes it goes down. So, Hamilton set out to understand better that randomness. One simple conceptual model for gross domestic product growth that many people used 10 years ago was the "urn model." Economists would think of GDP growth as being like a random drawing taken from an urn that has colored balls in it with various growth rates written on them. The average of all the numbers in the urn was about 2.5 percent, but the actual number we experience could be larger or smaller than that depending on luck. According to this model, a recession happens when a really unlikely ball with a big negative growth rate is drawn.

Hamilton showed that this simple way of thinking about things missed something very, very interesting. When we are in a "boom," the good times from quarter to quarter look very similar. When we are in a recession, the bad times look very, very similar. Using extremely complex computations, Hamilton was able to demonstrate that the "urn" model had to be discarded, and that the correct model of a recession had to have, at the very least, two urns!

So, here's how Hamilton's model works: A blindfolded person is spun around in a room that has two urns. After being spun, he walks over to the nearest urn, and draws a ball. If he draws a ball from the "good times" urn, then we get GDP growth in that quarter of about 2.5 percent. If he draws a ball from the recession urn, then we get GDP growth of about minus-1 percent. Our person also is more likely to draw from the urn that he visited in the previous draw as well, so good times and recessions are rather persistent.

If Hamilton's model were revised to fit the information technology economy boom years, then it would change somewhat. Lately, good years have not been centered on 2.5 percent, but, rather, around 4 percent or so. We have yet to see a new economy recession. Last week's data, which were bad but not too bad, suggested, though, what it might look like. The 1.5 percentage points that we added to boom years will be added to recession years as well. So, a recession will be a time when GDP growth is about .5 percent, which will feel far less disruptive than recessions used to -- more like a vacation than a recession.

There is one other change we will have to consider. In the old Hamilton model, the probability of drawing a good ball is much higher if we did so last time. The probability of drawing a bad ball is much higher if we did so last time as well. Did the new economy change the probability of switching urns? It might have, since firms are so much more flexible now that bad strategies can be discarded quickly. As in so many things, though, only time will tell.

TCS Daily Archives