FOR DELIVERY: 9:30 A.M., E.S.T. FRIDAY, FEBRUARY 7, 2003 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 Standard Time. Statement of Kathleen P. Utgoff Commissioner Bureau of Labor Statistics Friday, February 7, 2003 Nonfarm payroll employment rose by 143,000 in January, after seasonal adjustment. This follows a loss of 156,000 jobs in December (as revised). The unemployment rate was 5.7 percent in January, down from December but not much different from the average jobless rate for last year. On a net basis, more than two-thirds of the January employment increase occurred in retail trade, where a job gain of 101,000 (after seasonal adjustment) offset a decline of similar magnitude in December. Much of the over-the- month employment swing was in eating and drinking establishments, an industry frequently subject to large employment fluctuations. Several other retail components, notably department stores and miscellaneous retailers (e.g., toy stores), also showed employment gains over the month after seasonal adjustment. In these industries, holiday season hiring had been below normal. Hence, fewer workers than normal were let go in January, when most of the post- holiday cutbacks usually occur. This resulted in a seasonally adjusted employment rise. Elsewhere in the service-producing sector, employment in services was up by 35,000 in January. Health services employment rose by 18,000, close to average monthly growth for the industry in 2002. By way of contrast, there was no job growth in help supply services. Taking a longer perspective, employment in help supply has trended slowly downward since last June, after showing some signs of recovery between March and June. Employment in transportation and public utilities overall was little changed in January, as a large job gain in air transportation was offset by a decrease in communications. The increase in air transportation follows a decline of similar size in December; air couriers accounted for most of the change in both months. Over the longer term, air transportation employment has been trending downward since the beginning of 2001. The January decline in communications employment marks nearly 2 years of continuous job losses; the industry has shed 179,000 jobs since its peak in April 2001, a decline of about 10 percent. In finance, hiring continued in mortgage banking. Employment in real estate was little changed over the month, and the job totals in wholesale trade and in government also held fairly steady in January. In the goods-producing sector of the economy, construction employment rose by 21,000 in January, with gains split between heavy construction and special trade contractors. Manufacturing employment edged down by 16,000, following a large job loss in December. Within manufacturing, employment continued to decline in industrial machinery and electrical equipment, and there were small losses among several other component industries. There was a sizable job gain of 11,000 in motor vehicles in January, offsetting a loss from the prior month. Fewer auto plants were idled than is typical in the January survey reference period. Both the manufacturing workweek and factory overtime edged down by 0.1 hour in January, to 40.8 hours and 4.1 hours, respectively. Average hourly earnings of private production or nonsupervisory workers were unchanged in January at $14.98; over the past year, average hourly earnings were up by 2.7 percent. Turning to the data from our survey of households, the unemployment rate was 5.7 percent in January, down over the month but in line with the jobless rate that prevailed for much of 2002. As previously announced, we implemented several changes in our household survey this month. These include the introduction of new questions about race and Hispanic ethnicity in accordance with OMB guidelines, benchmarking to new population controls that reflect Census 2000 and updated information on net migration, the introduction of new industry and occupational classification systems, and improvements to our seasonal adjustment procedures. The result of a multi-year effort by the staff at BLS and our partners at the Census Bureau, the changes help to ensure that our household survey provides reliable data that reflect the evolving nature of the U.S. population and labor market. While necessary to keep our data relevant and reliable, changes such as these can complicate comparisons with earlier periods. This is particularly true for this month and next. Due to the introduction of the new questions on race and ethnicity, for example, we had to make a slight modification to our normal estimation process for January (discussed in more detail below). This change alone makes most measures from our household survey for January not strictly comparable to those for December and earlier months. However, for the most closely watched measure from the household survey, the unemployment rate, the effect of this change in estimation is very small. Taking a closer look at the impact of some of the changes, our research indicates that the new questions on race and ethnicity introduced this month also had a negligible effect on the overall unemployment rate. The January jobless rate for Hispanics or Latinos, however, may have been somewhat higher than it otherwise would have been. This is due to a revamped question on Hispanic ethnicity that is more direct and identifies some people as Hispanic who previously were not identified as such. This conclusion is based on data from a special survey conducted in May 2002; we will conduct additional research to confirm it. As for the effects of some of the other changes to the survey introduced this month, estimates of the total number of people employed and unemployed are impacted, mainly by the introduction of the new population controls based on Census 2000 results. The Census 2000-based controls increase the size of the civilian noninstitutional population age 16 and over by more than 3 million and thereby raise the estimated number of people employed and unemployed. As previously announced, data for the January 2000-December 2002 period were revised to reflect the new Census 2000-based population controls. In addition to the new population controls based on Census 2000, the Census Bureau recently introduced an additional upward revision to the population that reflects more recent information on net migration. The increase (+941,000) was added to the January 2003 civilian noninstitutional population. The increase in population raised the estimated levels for the labor force by about 615,000, employment by about 575,000, and unemployment by about 40,000. However, the overall unemployment rate, employment-population ratio, and other percentages generally were not noticeably affected by this population adjustment. Levels (and most percentages) from the household survey in January also are affected by the inability to use our regular composite estimation procedures this month (the temporary change in methodology stems from the introduction of the new race and ethnicity questions). This complicates comparisons of levels between December 2002 and January 2003; indeed, this also will limit our ability to look at changes in levels between January and February next month, since we’ll be returning to our monthly composite estimation procedures for the February data. Finally, I would note that in January we updated the methodology we use for seasonal adjustment of household survey time series data and issued the annual revision of 5 years of seasonally adjusted data. The upshot of all of this is that discerning changes in employment, unemployment, and other levels from the household survey over the December 2002 to February 2003 period is more complicated than usual. Fortunately, as I noted above, comparisons of the overall unemployment rate are not problematic, but it is clear that we will have to rely even more heavily on the payroll survey for our analysis of employment change over this period. Additional information about all of these changes and their impact can be found in our Employment Situation news release and in articles that will appear in our Employment and Earnings publication and on our Web site. In summary, total nonfarm payroll employment rose in January, after seasonal adjustment. Much of the increase was in retail trade, where seasonal layoffs were smaller than usual. Employment edged up in construction and services, and job reductions continued in manufacturing. The unemployment rate was 5.7 percent, down over the month but in line with figures recorded during much of 2002.