Transcripts of the monetary policymaking body of the Federal Reserve from 2002–2008.

Thank you, Mr. Chairman. I’ve talked recently to most of my usual contacts. My Wal-Mart contact said that he views the situation as a bit confusing. I’ve mentioned before the midmonth paycheck cycle that has been evident in the pattern of sales at Wal-Mart. Wal-Mart had sales on January 16 that were 12 percent above sales on January 14 as a consequence of the arrival of paychecks in the middle of the month. My contact said that the ratio is about as large as they ever see; a normal number is more like 4 percent or even zero when the economy is really strong. However, overall sales in January so far are coming in at the top of the expected range—about 5 percent above year-ago numbers for same-store sales. Wal-Mart is concerned that, if job creation does not occur soon, their sales growth may taper off because in their view the consumer is liquidity constrained.

My UPS contact said that UPS had, in his words, “an explosive December.” Business was well over projections. International shipments, especially electronics from Asia to the United States—as well as computers and TVs—were extremely strong and particularly so in the four or five days before Christmas. He said that fuel prices are going to play a major role going forward, although UPS is significantly hedged against the price increases. To give you an idea of the magnitude of the miss here, UPS had planned on jet fuel at about 84 cents a gallon, and it is coming in at 99 cents per gallon, a significant upward revision. UPS expects to be passing along that higher cost in fuel surcharges, although there is a lot of customer resistance. Nevertheless, he said that it looks as if it is going to be a great year. They are expecting a super year in terms of growth, but they are managing costs very, very tightly, deferring projects until the point where they just have to do them. He also reported that they are seeing headhunters prospecting for UPS employees to move to other firms; I thought that was rather interesting.

My contact in the trucking industry at J.B. Hunt said that business is much better now than it was last January. Volume is running about 8 percent above January a year ago, and the increased activity is across the board by region and by industry. But he said it is a bit hard to sort out how much of this is a consequence of the improving economy and how much is due to the reduced supply of carriers. A lot of the trucking companies are no longer in business. At the end of last year, I talked about the fact that there were new constraints on the number of hours that drivers can drive. I asked my contact if there had been any impact from that and he said “no” because the Department of Transportation has permitted a two-month delay in the enforcement of the new rule. J.B. Hunt, by the way, also has imposed fuel surcharges.

Our contact at Bank of America said that the bank has been experiencing a moderate pickup in small business lending and that middle-market lending has flattened after sustained declines. Larger commercial lending has yet to turn up.

I would like to comment briefly on two other issues relevant to the national outlook. We had some discussion earlier about the employment numbers. Bob McTeer raised a question about the household employment series, which had to do with statistical population controls, and an issue on the payroll number, which had to do with births and deaths of business firms. An enterprising staff member in St. Louis, Kevin Kliesen, contacted four states—New York, California, Georgia, and Texas—from which he was able to get some data on new incorporations. One of the problems is that the payroll numbers in the birth–death models, which apparently resemble fancy ARIMA models, are not tied to current information on actual firm births, which would be relevant at the present time. The Texas numbers applied to the first eight months of last year and were converted to an annual rate. In all four states, new business incorporations grew at a slower rate in 2003 than in 2002. I think that sample at least tells us that there is no smoking gun in terms of the payroll survey failing to pick up new firm births. If these numbers had gone the other way, that would raise more questions in my mind.

I want to comment briefly also on the inflation outlook. Bob Rasche and Jeremy Piger did a very nice memo for me on this subject. They used a standard Phillips curve equation—a Bob Gordon type of equation, which I don’t think is too far from what is in the Board staff’s model—which makes the rate of inflation dependent on inflation expectations. They did their analysis in several different ways, on the CPI and on other measures. The Gordon models and the Rasche and Piger model have a distributed lag on the actual rate of inflation, a gap term, a shock term, and a random term. That is a pretty standard formulation. Now, suppose you leave the gap term out of the equation and then compare that equation with one that has a gap term in it. If you compare the standard error squared—you can sort of convert that to an r-squared kind of measure—it turns out that the gap adds, depending on exactly how you measure it, only perhaps 8 to 15 percent to the total predictive value. A lot of the variance over the sample period comes from the high inflation years in the 1970s, when we had some big swings, and that is when we saw a major impact from the shock terms as well—the food and energy prices.

What I am driving at here is that a great deal of the inflation picture—most of what we can talk about systematically—is tied to the gap term. But historically some very important changes in inflation have come about that can’t be explained that way. We shouldn’t forget about that fact if we start to get some rise in inflation. The significance of that will depend on how one wants to interpret the expectations term. If you think of it as just a pure distributed lag, then you get a lot of warning about rising inflation because the expectations feed through slowly. On the other hand, if you think about that as econometrically proxying for expectations that could move pretty quickly, we could find ourselves in a situation in which we would have a surprise. We could have more inflation on our hands than we would predict right now. I just want to emphasize this point: Most of our discussions have focused on the gap term, but historically that’s not where we find most of the predictive value in an inflation equation.

Keyboard shortcuts

j previous speech k next speech