DOWNLOAD ACROBAT PDF VERSION
Wage growth slows despite employment gains
The strong economy continued to benefit low-wage and minority workers in the second quarter, but wage growth appears to have slowed this year compared to 1997 and 1998.
While overall unemployment fell a slight 0.1 percentage point over the past year (Table 1), jobless rates among young workers with at most a high school degree fell by 0.5 percentage points, driven by a 0.8 percentage-point decline for females. The underemployed made even larger gains. Overall, underemployment among this group dropped by 1.0 percentage points since the same quarter in 1998; for women, it dropped 2.1 points. Even with these recent gains, however, about one-fifth of young persons with high school degrees or less are still underemployed. (For a discussion of the concept of underemployment, see the data appendix.)
Despite this tightening in the labor market, which has been generating inflation-adjusted wage gains for all workers since about 1996, wage growth was generally flat over the last quarter (Table 2 – note that wage changes of less than 1% are statistically insignificant in these data). Other data sources have found a similar trend. For example, the BLS production worker wage series shows an unexpected deceleration of wage growth despite the continued low unemployment rate.
Measured over the last four quarters, from 1998:2 to 1999:2, real wages have grown significantly in almost every wage category (last column of Table 2). Low- and middle-wage males (those at the 10th and 50th percentiles) each gained slightly more than 3%, while high-wage (90th percentile) males gained 4.1%. This pattern of wage inequality-high-wage earners pulling away from the rest of the pack-has been the norm for males for the past decade. Wage growth among women over the past year has been fairly uneven throughout the distribution. Real wages among women grew more than 2% at the bottom and top of the wage scale but were flat at the median.
Average wage trends by education level and race (Table 3) reveal no evidence of wage growth over the past quarter, but they show real gains for most groups over the past year. The wages of male, entry-level (i.e., age 20-30), high-school-educated workers grew by 2.1%; females gained 2.3%. Among the high school educated, only Hispanic wages have failed to respond to the tightening job market.
Unemployment and underemployment rates for young blacks and Hispanics with a high school degree or less have mostly fallen since a year ago (Table 4). While most of the changes in unemployment over the past quarter are too small to be statistically significant, the pattern of falling rates for most groups is indicative of continued job gains. Over the past year, young, minority females have made significant gains; unemployment rates dropped 3.9 percentage points for blacks and 2.0 points for Hispanics. On the other hand, Hispanic male unemployment is at about the same level of one year ago.
Yet despite these gains for most low-wage workers, their levels of unemployment remain much higher than the national rate of 4.3% (see the first line of Table 1). For young minorities with a high school education or less, unemployment rates in 1999:2 were between two and one-half and five times the overall rate.
Underemployment rates fell precipitously for young minority females but rose for young Hispanic males. Overall, over one-fifth of the young, minority population with at most a high school degree remained underemployed in 1999:2 (these rates are driven mostly by high school dropouts).
Turning back to wages, Table 5 illustrates the trend over the current business cycle (from 1989 to the first half of 1999) for certain hourly wage deciles (the same as those in Table 2). Real wages for each decile are indexed to their 1989 level, so the table shows percent changes relative to 1989.
The top section of the table (“All”) shows that each decile has surpassed its 1989 level, though only by 1.9% at the median. For males, however, the median remains almost 2% below the 1989 level, despite fast real growth since 1997.
Hourly wages were slow to respond to the recovery, but, as Table 5 shows, they finally began to rise around 1996, when unemployment was heading down below 5%. In numerous deciles, this rise decelerated over the first half of 1999. For example, for all workers, real wage growth decelerated at the 10th, 20th, and 90th percentiles.
Thus, at the close of the first half of 1999, the labor market remains tight for low-wage workers, and jobless rates continue to edge downward. There is, however, some preliminary evidence that real wage growth may be slowing. In this regard, it is worth recalling that over the past few decades real hourly wage gains have been the exception, not the rule. Of course, a few quarters do not constitute a lasting trend, and it is too soon to tell whether wage growth will continue to slow or instead respond positively to the tight labor market.
For more information, see the data appendix.