No. 15 — April 14, 2000


Weekly Review

--- by Douglas Ostrom

Is Japan's economy expanding or contracting? The Management and Coordination Agency has thrown another monkey wrench into the ongoing debate among economists on both sides of the Pacific over that seemingly simple question. At the time of its March 31 announcement that seasonally adjusted unemployment had risen to a record 4.9 percent in February, MCA supplied details that render suspect the conventional, and recently repeated, interpretation of the higher jobless number. An alternative explanation has implications for the argument about where the economy currently stands.

Unemployment typically is a lagging indicator. As such, it can increase even if, as has been the case since mid-1999, industrial production and other broad aggregates swing upward. One reason is that potential workers, hearing that employment prospects are improving, may start looking for jobs. They then are added to the labor force, increasing the denominator of the unemployment rate. If their job searches initially are unsuccessful, however, they will be counted as jobless and lift the unemployment rate. Should economic conditions continue to improve, the pool of potential workers eventually will shrink, reducing the increase in the ranks of the measured jobless. At the same time, some of those seeking employment will get hired. Accordingly, the unemployment rate will fall a few months after economic activity picks up.

Last year, this theory seemed to be playing out in Japan. Although unemployment rose despite a spring pickup in industrial activity, it was declining by late summer. After peaking at a seasonally adjusted 4.8 percent in June and July, joblessness dipped to 4.6 percent in November. However, in December, it rose to 4.7 percent and remained at that level through January. It was up again in February to 4.9 percent.

On the face of it, the traditional explanation that lower unemployment lags rising output does not appear to hold, especially in view of jumps in factory production in January and February on a seasonally adjusted, month-to-month basis. As such, the numbers seem to support analysts who focus on gross domestic product rather than supply-side figures to assess economic conditions. In their view, the simultaneous drop in GDP in 1999's fourth quarter and the rise in unemployment toward the end of the year are indicative of a recession (see JEI Report No. 11B, March 17, 2000).

A closer look at the numbers suggests, however, that the economy might have experienced a sort of minidip last fall and winter, to which the currently higher unemployment rate could be a delayed response. If that is the case, it follows that the number of people seeking work would be increasing and that, in turn, the rise in unemployment is only temporary. Indeed, some analysts have pointed to this possibility. Yet MCA statistics show that the seasonally adjusted labor force, which includes those working or seeking jobs, shrank on a month-to-month basis in December, January and February, experiencing a cumulative decline of 0.4 percent.

Therefore, according to conventional wisdom, the behavior of the labor market is a puzzle. However, at least two other explanations are available. The more popular one holds that in terms of restructuring, corporate Japan finally has gotten religion. In this view, firms accept the almost-universal judgment of outsiders that they have redundant workers on their payrolls and are moving to trim those numbers — increased demand for their output and a tradition of permanent employment notwithstanding. This theory runs into a serious problem, though. It must explain why businesses did not act this way as recently as last summer, a time when they were adding workers.

The other explanation interprets the lag in unemployment as a response to the three stages in which companies raise output. First, they very tentatively respond to an increase in business even while their personnel numbers continue to drop. This is possible in part because past efforts to retain redundant workers to whom the firm feels a long-term commitment have left companies with enough people to step up output. As such, the phenomenon may be unique to Japan and its idiosyncratic labor practices. At this point, the ranks of discouraged workers and the unemployed keep on growing while the labor force shrinks. In the second stage, steady increases in demand persuade companies to resume hiring, but the surge in job-seekers initially prevents unemployment rates from dropping. In the third stage, unemployment rates fall as hiring and output continue to increase.

This proposition explains the mid-1999 pattern as well as more current data. The labor force shrank in the early part of last year, bottoming out one month before the unemployment rate hit its earlier peak in July. The recent decline in the size of the labor force probably is a response to weak economic conditions late in 1999. The shrinkage in February is a sign that unemployment could head higher even if production remains strong. This explanation suggests that if output continues to grow, the contraction in the labor force will end within a few months. Unemployment could resume its downward course as quickly as a month after that.

While this argument leaves open the possibility that firms are restructuring, the recent labor market statistics can be explained without resort to it. Moreover, the three-stage expansion theory seems to fit the data better. If correct, the short-term prospects for the Japanese economy are better than under the restructuring explanation since individual incomes — and consumer spending — will rebound more quickly. The long-term outlook is worse, however, because limited restructuring leaves corporate Japan with low rates of return and little ability to grow if, as is likely, it increasingly has to pay market rates of interest or earn competitive rates of return on investors' equity.

Admittedly, the three-stage explanation of the labor market is not perfect. According to MCA data, all of the employment cutback over the winter occurred at small firms. Like the restructuring argument, though, the multiphase expansion scenario applies primarily to big companies, not small ones, which have only the weakest tradition of permanent employment in the first place.

Whatever their drawbacks, both theories are consistent with the belief that the Japanese economy is growing despite GDP figures for the third and fourth quarters of 1999 that put it in a recession. However, if the economy were expanding rapidly, GDP, industrial output and employment data all would be pointing in the same direction: up. That clearly is not the case. The employment statistics suggest that this unsettled state of affairs is likely to continue for at least several more months.

The views expressed in this report are those of the author
and do not necessarily represent those of the Japan Economic Institute

Issue Index aaaa 2000 Archive Index aaaa Subscriber Area aaaa Home