FOR DELIVERY: 9:30 A.M., E.D.T. FRIDAY, OCTOBER 4, 1996 Advance copies of this statement are made available to the press under lock-up conditions with the explicit understanding that the data are embargoed until 8:30 a.m. Eastern Daylight Time. Statement of Katharine G. Abraham Commissioner Bureau of Labor Statistics Friday, October 4, 1996 Good morning. I am pleased to have this opportunity to comment on the employment and unemployment data released this morning. Nonfarm payroll employment was virtually unchanged in September after seven months of steady growth. The unemployment rate, at 5.2 percent, also was little changed over the month. Manufacturing employment fell by 57,000 in September, with losses spread through most of the component industries. Employment in transportation equipment fell by 12,000, mostly because of losses in automobile manufacturing; industrial machinery also lost 11,000 jobs. In nondurable goods manufacturing, food processing lost 7,000 jobs. Losses in this industry have been accelerating in recent months. Employment in printing and publishing also continued to decline, falling by 5,000. Elsewhere in the goods-producing sector, growth in construction was weak, owing in part to losses in heavy construction in the wake of hurricane Fran. Mining employment resumed its long-term downward trend. In the service-producing sector, government employment fell by 81,000 over the month, due largely to a decline of 67,000 in local education. This decline partly offset unusually large gains recorded over the summer. Local education is an industry in which there are large seasonal swings in employment over the course of the year. Due to changes in school schedules, there were larger returns to payrolls prior to September this year than were expected by our seasonal factors. This accounts for the large increases, on a seasonally adjusted basis, in local education employment over the summer and for this month’s sharp decline. Taking a longer view, local education employment was up by 157,000 compared to its year-ago level. The long-term decline in federal government employment continued in September. Employment growth in services, which continued to slow in September, was only 54,000. This is less than half of this industry’s average monthly increase during the first half of 1996. Smaller-than-usual increases in the help supply component of business services and a decline in private educational services contributed to the overall weakness. On the other hand, health services, which grew by 30,000 over the month, improved after 3 months of slow growth. Half of the increase in health services was in hospitals. Motion pictures and engineering and management services both were quite strong. Retail trade was sluggish for the second straight month. Employment in eating and drinking places rose by 15,000, and there were smaller increases in building materials stores, auto dealers and service stations, and furniture and home furnishings. Employment declined, however, in general merchandise, apparel, and miscellaneous retail stores. Average weekly hours of production and nonsupervisory workers on nonfarm payrolls increased by 0.2 hour to 34.7 hours in September. The factory workweek edged up by a tenth of an hour to 41.8 hours. Average hourly earnings rose by 6 cents and were up by 3.5 percent over the year. Before moving on to the data from the household survey, I would like to call your attention to a technical correction that affects the payroll employment numbers presented in this month’s news release. During the preparation of the data for September, we identified a coding error in the X-12 ARIMA computer program that has been used since June to seasonally adjust the data from our establishment survey. The seasonally adjusted employment data for August and September released today have been produced using corrected factors. We are in the process of recalculating all of the employment series from March through July and the hours and earnings data from March through September, and plan to release the revised data series on October 18. Although we are making these corrections in the interest of ensuring that our published series provide as clear a picture of recent trends as possible, I would note that the error we have discovered had only a very small effect on our historical estimates of employment and appears to have had a negligible effect on our estimates of hours and earnings. For the period from February through July of this year, correcting this programming error will result in small changes in individual months’ employment and a net cumulative change in estimated employment growth of less than 20,000. The current estimate of job growth for that period is about 1.1 million. Turning to data from the household survey, the unemployment rate was 5.2 percent in September, little different from its August level. The jobless rate for adult men, which fell sharply last month, rose to 4.5 percent. Teenage unemployment dropped to 15.6 percent. In summary, the labor market was somewhat weaker in September. Payroll employment was flat, as all major industries showed either slower growth or declines. The unemployment rate, at 5.2 percent, showed very little movement. My colleagues and I now would be glad to answer your questions. 5 4