Further questions? If not, let me get started. We are seeing what appears to be the first indication in this Committee of an increased variance of opinion which, if you go back over the previous meetings, you will find was virtually nonexistent. The reason that our opinions are only now beginning to spread is that the policy stance we’ve taken over the last year or two has been very unusual. During this period, we have had a very substantial and surprising growth in productivity, which drove unit labor costs down and created a view that inflation had very little probability of emerging. Therefore we could be fairly well convinced that a low funds rate, a highly accommodative funds rate, could be maintained for quite a significant period of time. As a result, our general policy and the variety of terminology that we began to use, including “patient” and “considerable period,” became the central mantra of this Committee.
I think that period is coming to an end. It’s not that we suddenly found some extraordinary capability to communicate to the markets our views, which would then craft their general expectations. It’s that the last couple of years were a very special and unusual period and not likely to be continued into the future. We’re going to move back, I suspect, far closer to our more general position, which at one point was that we never projected the future, we merely tried to give a sense of the balance of risks. That assessment often didn’t convey very much because we didn’t behave as though we really viewed it as a projection; we responded largely to the data as they evolved. My suspicion is that we’re moving back in that direction, and we’re beginning to see the reasons for that in the context of structural productivity gains.
If, indeed, structural productivity growth is 2¾ percent and if the general response of the wage market to the unemployment rate continues where it is, we will end up with, as the projection in fact implies, a 4 percent increase in compensation per hour and around 1¼ percent for the core inflation rate. Were we able confidently to project that indefinitely into the future, we could be fairly well convinced that inflation will be contained.
The only problem is that our ability to make judgments about separating cyclical from structural output-per-hour changes is questionable or, I would say, modest at best. This issue is coming forward, for example, in the context of the slowdown in high-tech investment, which a number of people have mentioned. Now, if there is, indeed, a slowdown in high-tech investment or, as the Vice Chair of the Board of Governors indicated publicly, if the slower rate of decline in high-tech prices is suggesting a reduction in the extent to which the high-tech area is contributing to productivity, then we have to ask where we should place the long-term productivity number. And what is our confidence level for that number? It’s quite likely at this stage that 2¾ percent may well be the right number. I say that in part because, despite the 337,000 increase in employment in October, if you examine the hires less separations, which theoretically should be coming from the same data source, that series is showing a smoother trend of gradual decline in the rate of job gains. The latest data we have on that obviously are for September. That series shows less volatility than the payroll numbers but, on average, is not terribly different from the payroll series since the beginning of the year; both indicate a gradual slowing. If that were, indeed, the trend, then the productivity outlook is better. If it is not and we are looking at an acceleration of employment, then chances are we are getting closer to a level of structural productivity growth that is lower than 2¾ percent.
Now, the reason I raise this issue is that, although we may choose 2¾ as our number, the argument that 1¾ may actually be what we’re looking at is not readily dismissible. Making the distinction between what is happening short-term in output per hour and what is happening to underlying productivity is very difficult. Indeed, the data for the numerator of output per hour come from a source wholly different from that of the denominator. I think we felt a sense of comfort that a goodly part of the productivity growth had to be structural when we had this very high rate of growth. That was almost certainly the case because the numbers were so high. As a result, what we were observing was definite downward pressure on inflation, to the point in fact where we were at the edge of being really quite concerned about the onset of deflation. That is no longer the case. And I think we’re beginning to confront problems that really bring us back to the way we used to do forecasting, which had much wider areas of variance. Clearly, the variance of the views of this group is beginning to reflect that.
Earlier we discussed the issue of the dollar. We’re not sure exactly how to translate the exchange rate into unit import costs. Indeed, for those of us who were in Basel the last couple of days, the startling new piece of information was that the fascinating decline in the rate of pass- through of exchange rates into unit import prices, a decline that is progressing in the United States at a fairly remarkable pace, is exactly what they are finding in the United Kingdom, according to Mervyn King at the Bank of England. When he raised that point and I mentioned our findings on that, everybody looked around, and there was general agreement that we were all seeing the same phenomenon. It all started, as you may recall, in the failure of the devaluations in Brazil and Argentina to create increased inflation, which they did not. Something very unusual is going on with respect to the exchange rate translation issue. And this is obviously a very important question for how we view currency depreciation and inflation and, therefore, what the implications are for monetary policy.
We have very significant problems with private saving. The household saving rate has come down very dramatically and now is close to zero. We have to resist the notion that there’s some floor at zero, which the saving rate can’t go below, and that therefore it must flatten and turn up. The idea of having a negative personal saving rate is not out of line with the way the world works. Remember, the average household does not look at the book value of its equity holdings, which is what the saving rate is. The average household looks at the market value. And we can readily have, as our data have shown in the past, a negative saving rate with a significant part of the population believing that they are saving at a fairly pronounced rate, which has been the experience of recent years. We don’t know how that is going to come out.
We do know that Australia, as I understand it, has had a negative saving rate for a while. Our flow-of-funds accounts here do not show a negative saving rate, but they indicate that almost all of the change in the saving rate in the last decade has been in the upper quintile of the income distribution, and there it is driven very substantially by asset values. People in the upper quintile look at the market value of what they own. Adjusting their saving behavior in response to the book value of their net worth would never enter anybody’s mind. Hence, this is a synthetic number, and let’s be careful not to presume very much about what it is telling us.
The fiscal issue is very critical because it is the only issue in which policy per se has a clear positive contribution to make. It is going to be very interesting to see how this plays out. I must say that I’ve turned a little optimistic in the last week or so listening to some of the Administration people in this respect. I’m not sure whether it will translate into something of significance or whether in fact, Ned, as you put it, “we all know how likely” that is.
When it comes to policy, with the increased uncertainties that we have about how the economic situation is evolving, we have to acknowledge to ourselves that our forecast is going to be wrong. It always is. We expect it to be wrong. The question is, Where do we want to take the risks? If we project continued low core inflation because of a 2¾ percent projection for structural productivity growth and that growth is in fact at 2 percent or less, inflation expectations will rise, bond yields will rise, and asset prices will fall, as, of course, will the dollar in that context. We could face problems of balance sheet losses if the change is rapid. Our ability to separate cyclical from structural changes in output per hour, as I said before, is modest at best. Should inflation expectations rise, the higher the funds rate, the better position monetary policy will be in to address unexpected inflation increases.
If we signal today a pause for December, the ten-year note yield will decline, engendering an even larger correction should underlying inflation unexpectedly rise in the months ahead. Capital gains do not impair balance sheets. Capital losses do, however, especially if they are rapid. If rates move up and the economy slows, the ten-year yield will decline and possibly equity prices as well, but balance sheets will not be impaired. The structural damage obviously would be quite limited in that regard. Should the economy weaken, of course, we always have the option of moving rates lower.
It strikes me, therefore, that even if we believe the risks are symmetric with respect to how the economy is going to emerge, a cost–benefit analysis of the consequences of our taking various actions leads me to conclude that we would be far better positioned today not only to move 25 basis points but also to signal nothing about December. The best thing that can happen is for the market to perceive that we are implying another rate increase in December because the further beyond the 1 percent range and into a positive real rate that we can get, the better off I think we will be.
I want to repeat that we don’t know very much about the interaction of very large productivity gains, wage rates, unit labor costs, and the NAIRU. We’ve been experiencing these gains only in recent years. I personally would feel a lot more comfortable if the federal funds rate were back in a range that is consistent with much of our historical data rather than on the edges of a number of our econometric functions of how the economy works. The edges are significant in the overall picture, but I wonder how much emphasis we can put on the outer edge. Their significance depends on whether or not the functions are bending in one way or another, whether they are still linear, or what their nature is at those points. Therefore, overall, I think we are far better positioned to get the funds rate up as fast as we can this year and then recognize that we have time to discuss pausing as we get into next year. Now, in the December meeting we may want to suggest a pause for February 2, which I guess would be the particular day of the meeting. That will very much depend on what is evolving. If, for example, the weakness in the high-tech area continues into next year, I’m not sure how to read that. If it says that we’re getting much less in the way of technological advance, it suggests that the productivity growth rate is lower. If it is suggesting that capital investment is going to slow, we get a potential decline in the outlook for real GDP.
How we should play this is going to depend very much on how these numbers evolve, and at this particular stage, I think that we have to start preparing for a communication transition as we get into the early weeks and months of next year. We do have our semiannual report on monetary policy in which to make a very broad statement. But we may have to decide something in December about how we want to position ourselves for February, especially if, as now seems likely, we move another 25 points in December.
What I’d like to propose is that we move 25 basis points today and have a statement that is as little changed as possible, so we seem neutral with respect to the move we might or might not make in December. Then we can proceed to observe how the data evolve over the next number of weeks. It is very unlikely to turn out that an upward move in December will be inappropriate and yet the markets will not have made that judgment already. If we deem a move in December inappropriate, it will be because of the types of evidence that will move the federal funds rate and indeed the whole financial structure. So the concerns that we would have about moving in a manner that would shock the market would not be relevant. I don’t think that is going to be a problem, but I do think we have to start thinking of the word “measured” as having a fairly short life expectancy. We have to begin to reflect on how to approach our communications in a way that will be more consistent with an increased variance of opinion in this Committee in coming months, as I think is going to be the case. President Hoenig.