FRBSF Economic Letter
1996-16 | May 17, 1996
This Economic Letter is adapted from recent speeches given by Robert T. Parry, President and Chief Executive Officer of the Federal Reserve Bank of San Francisco.
A controversy about the national economy has been in the headlines recently. For example, in the Christian Science Monitor, one headline was “What’s the Economic Speed Limit?” As if in answer, the New York Times featured an article entitled “It’s a Slow-Growth Economy.” At the heart of the controversy is something called the “potential growth rate”–that is, the rate the national economy can sustain in the long run without generating inflationary pressures. Potential growth is important because it is a measure (albeit an imperfect one) of the pace at which the average standard of living grows in the U.S. Even a small change in the potential growth rate will accumulate into a large difference in living standards when sustained over decades.
The potential growth rate is controversial for several reasons. First, it appears that potential growth in the 1990s is running in the 2 percent range–a number that many people think is very low. Also, this estimate represents a slowdown from earlier periods. In fact, some estimates going back to the 1960s show that the rate has dropped from 4 percent to just over 3 percent from 1973-1979, to just under 3 percent from 1979-1990. This decline runs counter to people’s expectation that potential growth should increase over time, as computer technology improves and spreads and as corporations streamline their organizations. A final reason for the controversy over potential GDP is that some people blame the Fed for slow economic growth. They assert that the reason that growth hasn’t been higher is that the Fed hasn’t let the economy grow any faster.
I can’t give you a complete explanation for the pattern of growth in the U.S., but I can provide some insight into some of the controversies underlying this issue, as well as a perspective on how it relates to monetary policy.
As I said, potential growth is the rate that the economy can sustain in the long run without generating inflationary pressures. It’s measured as the long-run trend of economic growth–that is, abstracting from the ups and downs of the business cycle. It depends on two things: the trend in the number of people available to work–that is, the size of the labor force, and the trend in how much output those people can produce an hour–the productivity of the labor force.
The official Commerce Department data on GDP suggest that for the 1990s, there’s no change in the productivity trend, but there’s a noticeable slowdown in the growth rate of the labor force. This slowdown is due in part to the fact that the working age population is growing somewhat more slowl–in the 1980s, the annual growth rate was around 1-1/4 percent, and in the 1990s it dropped to about 1 percent.
But a more significant development has to do with the role of women in the workforce. While the pattern of men’s labor force participation has been pretty much the same for the past 35 years, the pattern for women has changed more dramatically. From 1960 to 1990, women’s participation rate–the proportion of working age women who are either employed or looking for work–rose from under 40 percent to just under 60 percent. But since then, it’s stayed right about there. Whether this more or less flat growth in women’s labor force participation rate is likely to continue depends on a number of things, many of which will be discussed in next week’s Economic Letter (Motley 1996).
To summarize those points briefly, it appears that since 1990, women’s participation rate has leveled off because of a leveling off in both the number of women who have entered the workforce at all and the number of weeks they’ve worked per year. There still is a lot of upside potential for the number of women who choose to enter the labor force at all during a given year. But there’s not much upside potential in expanding the number of weeks worked per year–for either women or men. For both, the average is around 48 weeks per year. Overall, then, we can’t say for sure what will happen to women’s participation in the future, but we can say that the proportion of women working and the number of weeks they’re working recently has reached a plateau.
This story about a change in women’s participation in the workforce would seem to explain why the potential growth rate has slowed in the 1990s. But a lot of people don’t accept that explanation. It’s not because they aren’t convinced about a change in women’s participation rate. It’s because they’re not at all convinced that the growth rate of productivity hasn’t changed in 20 years, as the data suggest.
According to the official Commerce Department data, the trend growth rate is still about 1 percent a year, with no sign of a pickup in the 1990s. A lot of analysts would argue that productivity growth is underestimated. They’d also say actual productivity growth has been rising faster in the 1990s. And they’d cite a couple of reasons. One has to do with computers. People argue that current measures aren’t accounting for the improvements in productivity that computers have made. They focus on the fact that much of the growth in computer use has been in the service sector, where it’s hard to measure exactly what the output is–for example, how do you measure the output of a lawyer, or an economist? And if it’s hard to measure output itself, it’s even harder to measure the improvement in quantity or quality made by a faster, more powerful computer.
Unfortunately, there’s no consensus on whether computers have led to a major productivity surge. For example, some economists have done elaborate studies looking for evidence of a pickup in productivity growth due to computers. But they haven’t found it. Their reasoning is that computers make up only about 2 percent of the U.S. capital stock, and therefore can’t make a large, direct contribution to overall productivity even if each computer is highly productive. Specifically, these studies show that even if we assume extremely high rates of return to computer hardware and software, communications equipment, photocopy equipment, scientific instruments, office and accounting equipment, we obtain an increase in potential GDP growth of only 0.5 percent per year.
At the same time, other economists don’t find these results to be definitive. For example, even though the direct contribution of the stock of computers to productivity may be relatively small, computers also may contribute to productivity through technological advances–that is, through the development of entirely new ways of doing things. These effects are extremely difficult to capture.
Another thing people point to as contributing to increased productivity in the 1990s is the apparent trend toward downsizing. But here, too, there’s controversy. Despite all the news we hear about big companies laying off thousands, the evidence shows that in the 1990s the proportions of workers who have been laid off hasn’t been strikingly higher than in previous periods. On top of that, a study of individual firms showed that downsizing has not been consistently associated with increased productivity. What does seem to be new is that downsizing is concentrated more in white collar jobs now, while before it was focused mainly on blue collar jobs. Finally, even if downsizing improved the productivity of the corporation doing it, that would not be sufficient to establish an increase in productivity for the economy as a whole. The latter question also depends on how productive the terminated worker is once he finds a new job. Given that job skills often are tied to one’s current job, it would not be surprising to find a fall-off in productivity in a person who involuntarily changes jobs.
To sum up, we are left with a lot of uncertainty about productivity–and that means there’s also uncertainty about potential GDP.
Does this uncertainty about potential GDP have a big effect on monetary policy? That is, if potential growth is underestimated, does that mean monetary policy’s likely to be too restrictive? The answer’s “no.”
The Fed uses potential GDP as a kind of benchmark. Although we don’t have a target for real GDP growth, we know that we’re going to end up with higher inflation if actual GDP grows above the potential rate for too long. So what we look at is the gap between actual and potential GDP. And since potential GDP is simply estimated from the GDP data itself, it’s probably not crucial if those data underestimate productivity. Both figures are likely to underestimate productivity to about the same degree, so the gap isn’t likely to be seriously affected by measurement problems.
Moreover, although potential GDP is an important benchmark for the Fed, it’s certainly not the only thing we look at. So if it turned out we weren’t reading potential GDP correctly, we’d get signals about that from other indicators. For example, if actual GDP had been below potential GDP in recent years, we’d have seen increases in unemployment rates and unused industrial capacity, as the demand for workers and capital fell short of potential gains. At some point, this would have translated into downward pressure on inflation. But, in fact, we haven’t seen this happen.
Let me conclude by saying that the Fed certainly would be just as happy as anybody to see faster growth in potential GDP. But it’s important to remember that the Fed can’t create it for long by stimulative monetary policies. If we were to try–by consistently pushing GDP growth beyond its potential–the inevitable result would be higher inflation. What we can create is an environment of low, stable inflation. It’s that environment that allows our market economy to function as efficiently as possible and that leads to investment in capital and labor in the long run. Ultimately, that’s the kind of environment we need to promote productivity and improve the living standards in the U.S.
Robert T. Parry
President and Chief Executive Officer
Oliner, Stephen D., and William Wascher. 1995. “Is a Productivity Revolution under Way in the United States?” Challenge (November-December).
Opinions expressed in FRBSF Economic Letter do not necessarily reflect the views of the management of the Federal Reserve Bank of San Francisco or of the Board of Governors of the Federal Reserve System. This publication is edited by Sam Zuckerman and Anita Todd. Permission to reprint must be obtained in writing.
Please send editorial comments and requests for reprint permission to
Attn: Research publications, MS 1140
Federal Reserve Bank of San Francisco
P.O. Box 7702
San Francisco, CA 94120