American Workers: Smarter but Poorer

Despite the neoliberal boomlet (the so-called "New Economy") of the 1990s, real hourly earnings are still lower than the peak of the mid-1970s. Whose fault is that? If you ask technocrats, the culprits are the working class themselves -- especially students and teachers -- who have allowed educational standards to decline and failed to match employers' demands for higher skills: "[t]he fall in real wages of people with low skills and widening earning differentials since the 1980s are also evidence of upskilling in Canada, the European Union and the United States (OECD 1996b)" (Literacy in the Information Age: Final Report of the International Adult Literacy Survey, 2000, p. 8). Especially in the United States, alarming soundbites are used to create a sense of national crisis of education: "Nearly half of all adult Americans cannot read, write, and calculate well enough to function fully in today’s society, and people in their early 20’s have poorer literacy skills than did those in a 1985 survey, according to a federal study" (Debra Viadero, "Half of Adults Lack Skills, Literacy Study Finds," Education Week, September 15, 1993). Is that really true?

The federal study in question, the 1993 National Adult Literacy Survey, however, said nothing of the sort, as Regie Stites clarifies in a working paper prepared for the National Center for Education Statistics: "First, the statement that half of all adult American lack adequate writing skills is an apparent misunderstanding of the nature of the NALS document literacy measures. Second, the association of performance at NALS Levels 1 and 2 with functional illiteracy is never made in the NALS report" ("'How Much Literacy is Enough?' Issues in Defining and Reporting Performance Standards for the National Assessment of Adult Literacy," National Center for Education Statistics Working Paper No. 2000-07, March 2000, p. 23).

There is no evidence to support the claim that Americans -- especially young Americans -- today have lower skills than those in the past. If anything, American workers are more educated and literate than ever, according to Michael J. Handel, who meticulously examined the data provided by not only the NALS but a battery of other tests.

Taking education as the measure suggests that the workforce is considerably more skilled than in the past. In 1964, prior to the perceived deterioration of public education, the share of all Americans who were high school dropouts was 47% and the proportion of young people age 24-29 with less than high school was 31%, compared to roughly 13% for both groups in 1997 (author's calculations, March Current Population Survey). . . .

. . . . . . . . . . . . . . . . . . . .

IQ tests show large gains for Americans throughout this century, including every postwar decade for samples as recent as 1995, the most current, and there is no obvious recent change in the slope of growth (Flynn 1998: 27,35ff.). . . .

. . . . . . . . . . . . . . . . . . . .

One of the most frequently cited are college entrance exams, such as the SAT, whose decline beginning in the mid-1960s initiated the recent concern over the state of public education in the U.S. However, less widely reported is that math SAT scores started rising around 1980 and exceeded 1971 levels by the mid-1990s, despite the growing share of high school students taking the exam, though verbal scores did not recover, while the pattern for the rival American College Test (ACT) shows English scores exceeding earlier levels while the rebound in math did not fully offset the earlier decline (Economic Report of the President 2000:148; Boesel & Fredland 1999:72). Contrary to popular impression, the SAT and ACT test score declines are highly cohort-specific and ceased or reversed long ago.

. . . . . . . . . . . . . . . . . . . .

The best time series of inter-cohort data is the U.S. Department of Education's National Assessment of Educational Progress (NAEP), sometimes called the Nation's Report Card, which has a continuous series of reading and math scores for representative samples of 17-year olds since the early 1970s; the test instrument has remained roughly the same over this time. Reading scores did not change significantly between 1971-1999, in contrast to SAT scores. Math scores dipped about 2% between 1973-1982 and then rose until reaching a level in 1999 about 1% higher than 1973 (Campbell et al 2000). Overall test score inequality for both math and reading also declined since the 1970s, due mostly to gains at the lower percentiles, contradicting the impression that the lower part of the distribution is losing ground (Campbell et al 2000:9ff.). Black math and reading scores rose and closed roughly one-half of the black white gap during the 1980s, also contradicting popular impressions, before losing some ground in the 1990s, a development that is still poorly understood (Campbell et al 2000:36ff.)

. . . . . . . . . . . . . . . . . . . .

Contrary to popular impression, age is negatively associated with test scores even for those under 65, lending no support to the idea that younger Americans have poorer literacy skills than older ones. Those age 55-64, who would have graduated high school in the reputed golden years of American education (1946-1955), clearly have lower scores than more recent cohorts, who supposedly bear the effects of less rigorous schooling. Many fewer in the older group actually finished high school but a significant gap remains even after controlling for educational attainment (Smith 1995:214; Freeman & Schettkat 2001; cf. OECD & Stat. Can. 2000:147f.). The NALS data provide no evidence that more recent cohorts have lower cognitive skills than older cohorts. (Michael J. Handel, "Skills Mismatch in the Labor Market," December 2002 [published in Annual Review of Sociology 29, August 2003], pp. 11-14, 17)

Moreover, the supply of college graduates is larger than the number of jobs that require college education (the requirement that is due more to credentialism than to actual skill demands of the jobs or actual skills supplied by college education):

Hecker (1992:4) found that the percentage of college grads in occupations not requiring a college degree or unemployed rose from about 12% (1967) to 18.6% (1980) during the "over-education" years, and continued to rise modestly to 19.9% during the years of ostensible shortage (1990).1 Further, the supply of college graduates grew 62 percent between 1979-90, while employment in managerial, professional, technical, and high skilled detailed occupations in other broad occupational categories grew only 57 percent (Hecker 1992:7). (Handel, p. 37)

In the 1970s, when workers were on the offensive on the class struggle as well as social movement fronts, the power elite in fact fretted that workers were overeducated and overskilled relative to available jobs, breeding discontent due to relative deprivation: "A prominent government report considered the dilemma of how to make work more satisfying when job complexity at all levels seemed to fall short of workers' rising education levels and aspirations for meaningful work (US Department of Health, Education, and Welfare 1973)" (Handel, p. 3).

Given that the rises in workers' educational attainment (even producing an oversupply of college graduates) and test scores took place at the same time as the proportion of well-paying manufacturing jobs contracted and that of poorly-paying service jobs expanded in the US economy, one would think that the question of "how to make work more satisfying" is even more urgent now than in the 1970s.

Instead, technocrats continue to cling to the "skills mismatch" theory, asserting that workers' skill levels fall short of the demand of the post-Fordist job market. Evidence for statistically significant upskilling in the last couple of decades -- the period of real wage stagnation and widening income inequality -- is as scant as evidence for a national literacy crisis, however: "Sum (1999) and Barton (2000:15,19) replace DOT [Dictionary of Occupational Titles] measures of occupational skills with mean NALS scores and find job literacy requirements almost completely unchanged over the periods 1990-2005 and 1986-2006, respectively, based on BLS occupational data and projections for the terminal years" (Handel, p. 27).

Besides, when it comes to hiring, employers themselves appear to value docility and servility more than anything else: "Cross-sectional studies often suggest employers are less concerned about cognitive skills deficits than what they consider poor work habits, motivation, demeanor, and attitudes" (Handel, p. 25). Lucy Kellaway put it more bluntly:

Think what characterises the really intelligent person. They can think for themselves. They love abstract ideas. They can look dispassionately at the facts. Humbug is their enemy. Dissent comes easily to them, as does complexity. These are traits that are not only unnecessary for most business jobs, they are actually a handicap when it comes to rising through the ranks of large companies." ("Companies Don't Need Brainy People," Financial Times, November 22, 2004)

The "skills mismatch" story is a red herring that seeks to deflect our attention from the real problems: union-busting, outsourcing, and subcontracting that have led to the declining rate of unionization, which, in combination with social programs starved by budget austerity, has made workers poorer and put them into a weaker bargaining position vis-a-vis employers. It's time to bury the myth.