ByAdam M. Zaretsky
The U.S. economy could not decide at what pace it wanted to produce output last year. In the first quarter of 1998, real GDP posted an unexpectedly high 5.5 percent annual growth rate. This was followed by a slower 1.8 percent annual growth rate in the second quarter. Many analysts, who at this point had already begun to acknowledge the effects that both the Asian crisis and jittery financial markets were having on the national economy, were then surprised by the early number for third-quarter real growth—a startling jump to 3.9 percent. Fast, slow, fast—just what has been happening in the U.S. economy? And has the economy of the Eighth Federal Reserve District been following suit or marching to the beat of a different drummer?
To answer the first question, the national economy has slowed since 1997, despite the appearance of a rebound in the third quarter. As always, the announced third-quarter GDP growth rate is only a preliminary number, subject to revision. In the first and second quarters of 1998, for example, the preliminary numbers showed real growth of 4.8 percent and 1.6 percent, respectively. After revision, the final numbers revealed that real growth was actually 5.5 percent in the first quarter and 1.8 percent in the second.
On top of this, the initial data show that much of the increase in the third-quarter GDP growth rate came from an unexpected rise in inventory accumulation, due to a rebound at the end of the General Motors shutdown and a stronger U.S. dollar. Although the end of the GM shutdown in late July led to a rebuilding of dealer inventories, the actual change in total automotive inventories was still negative. The rate of decline of these inventories, however, slowed substantially. Add to this the many businesses—particularly retailers—that sought to take advantage of the relatively low prices of Asian goods resulting from weak Asian economies and an appreciating dollar, and the jump in inventory investment is accounted for.
The Bureau of Economic Analysis—the government agency responsible for GDP data—stated in its third-quarter GDP report that inventory investment accounted for almost a full percentage point of the 3.9 percent growth rate. Actually, this percentage point represents almost all of investment's contribution to real output. In the second quarter of 1998, in contrast, inventory (dis)investment reduced real GDP growth by almost 2.7 percentage points, making total investment's contribution to real output negative.
At the same time inventories were increasing, consumer spending was moderating. In particular, spending on durable goods—like cars and household appliances—and spending on nondurable goods—like food and clothing—increased slightly, but substantially less than in the second quarter. Meanwhile, spending on services—like entertainment, health and financial—maintained its brisk second-quarter pace. Thus, third-quarter growth in spending on services essentially drove growth in consumption. Earlier, growth in spending on durables had played a much more important role.
Along with the recently erratic GDP numbers has been a slowing in payroll employment growth. Between June and October of 1998, employment growth has averaged about 189,000 jobs a month. During the first five months of the year, however, employment growth averaged 255,000 jobs a month.1 Actually, as the accompanying table shows, U.S. payroll employment grew at a 2 percent annual rate in the third quarter of 1998, almost a percentage point less than the first quarter. In the second quarter, jobs grew at a 2.3 percent annual rate. Thus, job growth has been steadily slowing all year. These employment data do not necessarily contradict the output data, though, because third-quarter GDP growth would have been roughly 2.9 percent (about the long-run average growth rate) without inventory investment. The continued slowing in employment growth, coupled with the recent inventory accumulation, could therefore signal a further slackening of output growth in the coming quarter—most likely, a return to a more average growth rate.
The picture isn't all glum, though. Tight labor markets across the nation have made it tough for employers to fill vacancies. Without new workers, there can be no job growth. Nonetheless, the nation's nonmanufacturing sector, which employs about 85 percent of all workers, continues to show strong employment growth. The third quarter's 2.9 percent annual growth rate in nonmanufacturing employment is comparable to the 2.8 percent rate posted in the second quarter and the 3 percent rate in the first.
The U.S. manufacturing sector, however, is bearing the brunt of the employment decline. In the third quarter, employment in this sector fell at a 3 percent annual rate, much sharper than the 0.5 percent decline posted in the second quarter. Employment declines at manufacturing firms, however, are not that unusual since these firms tend to be affected by the business cycle more. But manufacturing employment declines are not necessarily the norm, either. Manufacturing employment grew at a 1.5 percent annual rate in the first quarter of 1998, marking the end of a two-year positive growth trend. Taken together, then, the currently available data seem to suggest that the slowing in national output growth that began in the second quarter of 1998 is probably still with us and will likely stick around into next quarter.
Whether the District economy has been following suit or marching to the beat of a different drummer is a more difficult question to answer. The principal reason for the difficulty is that state-level output data are not as timely as GDP. In fact, the latest gross state product data (the state-level equivalent of GDP) are from 1996—almost three years old! Thus, much more reliance must be placed on employment data.
Payroll employment growth in the District has actually been following a different beat since 1997.2 For example, payroll employment grew at a 0.3 percent annual rate in the third quarter of 1998, after rising at a 1.9 percent rate a quarter earlier. And that rate was up from the 1.4 percent gain in the first quarter. Thus, in the third quarter of this year, District employment growth was more than 1.5 percentage points less than the nation as a whole. Like their national counterparts, however, District labor markets are extremely tight.
Unlike the rest of the nation, though, employment growth in the District's nonmanufacturing sector, which employs about 85 percent of all District workers, has been slowing. In the third quarter of 1998, new jobs in this sector were created at only a 0.9 percent annual rate, after rising at a 2.4 percent rate in the second quarter. That's a 1.5 percentage point swing in one quarter—the biggest monthly drop in this growth rate since the second quarter of 1995. That said, it is only one quarter, and the decline came on the heels of a 1 percentage point gain between the first and second quarters of 1998.
At the same time, the District's manufacturing sector has followed a pattern of steady decline, like its national counterpart. District manufacturing employment fell at a 2.5 percent annual rate in the third quarter of this year, after a drop of 0.5 percent in the second quarter. As with the national numbers, though, it's not unusual to see manufacturing employment decline from time to time. Still, coupled with the severe turnaround in nonmanufacturing employment, the trend becomes more troubling. With both major employment sectors in decline, therefore, the outlook for District output growth is probably a bit weaker than it is for the rest of the country.
What is the answer to the question, then? The District economy has been marching to the beat of a different drummer, but it's not completely out of step. While District manufacturing employment seems to be mimicking the nation rather well, District nonmanufacturing employment appears to be doing its own thing. Furthermore, if the swings in District employment growth rates are a reliable indicator of expected movements in District output growth rates, output growth is probably going to remain slower than it was six months or a year ago. Unfortunately, the output data to confirm this won't be available for two more years.