angle-up angle-right angle-down angle-left close user menu open menu closed search globe bars phone store

Skills and Earnings in the Full-Time Labor Market
Neeta Fogg, Paul Harrington, and Ishwar Khatiwada

Educational Attainment and Literacy and Numeracy Proficiencies and Earnings

Formal educational attainment represents one measure of human capital and the productive capacity of workers. Literacy and numeracy proficiencies represent a major component of the actual skills that workers supply and employers purchase in the labor market. Our examination of earnings by literacy and numeracy proficiencies presented in a previous section found that workers with higher level proficiencies have sharply higher earnings than those with lower level proficiencies. Our examination of the link between earnings and educational attainment above found that earnings of workers increased with the level of educational attainment. However, we also found sizable variations in the level of literacy and numeracy proficiencies among workers with the same level of educational attainment, particularly the surprisingly large share of college-degreed workers with skills below level 3 on the PIAAC scale.

Given the strong link between skills and earnings, these findings raise a question: Do the earnings of workers with the same level of educational attainment vary by the level of their literacy and numeracy proficiencies? We explore this question by examining variations in the monthly earnings of 25- to 54-year-old full-time employed workers within each educational attainment level by their literacy and numeracy proficiencies.33 The data suggest that the monthly earnings of prime-age, full-time employed workers rise with literacy scores as well as numeracy scores within most educational groups (figures 9 and 10).

These findings provide important insights into the limitations of estimating returns to human capital based exclusively on educational attainment, along with work experience, as the measure of the stock of an individual's human capital. The combination of educational attainment, work experience, and cognitive skills provides a more complete measure of human capital.

However, because of the widespread availability of educational attainment data and limited availability of measures of cognitive skills in most national household databases, education is frequently used as the only measure of human capital. Large sample surveys such as the decennial U.S. Census and Current Population Surveys (CPS) contain data on background socioeconomic traits, earnings, and educational attainment and are frequently used by researchers to estimate human capital earnings regressions. One of the shortcomings of human capital earnings regression models estimated from these databases is the lack of cognitive skill measures. The result is an exclusive focus on educational attainment with too little attention paid to the acquisition of abilities, knowledge, and skills—traits that are rewarded in the job market. In some instances, this reliance on educational attainment as the measure of human capital may have led to overemphasizing policies designed to increase the level of educational attainment of the population with little regard to developing basic skills in that education process.

Figure 9: Mean Monthly Earnings by Literacy Proficiency Levels of 25- to 54-Year-Old Full-Time Employed Workers in Each Major Educational Group, 2012-2014

Figure 10: Mean Monthly Earnings by Numeracy Proficiency Levels of 25- to 54-Year-Old Full-Time Employed Workers in Each Major Educational Group, 2012-2014

The social and economic consequences of misinterpreting the educational attainment measure and overrepresenting its influence on outcomes can be substantial. Resources that might be better spent on alternative ways to bolster skills such as early education or career and technical education could get short shrift. If the only dimension of human capital that is measured is the quantity of education, then it is not surprising to conclude that the quantity of schooling is central to policies to improve employment and earnings outcomes. Because of the wider availability of the education measure of human capital and less of the skill measure in most large-scale databases, there are considerably fewer studies that utilize measures of skills to estimate the impact of human capital on earnings and other labor market outcomes of workers.

Only few large-scale nationally representative surveys, such as the National Longitudinal Surveys, contain measures of cognitive ability (one measure is the Armed Services Vocational Aptitude Battery score). Yet the evidence is quite clear that cognitive skills are important determinants of labor market outcomes of workers.34 Both cognitive and noncognitive skills are found to influence a variety of labor market and behavioral outcomes.35 Research on labor market returns to schooling are frequently based on educational attainment and background traits without cognitive and noncognitive skills due to a lack of reliable and large-scale data.36

Availability of skills in national datasets for the adult population began in the early 1990s when the NCES conducted a national household survey of adult literacy—the National Adult Literacy Survey (NALS). Findings from this survey revealed that the literacy proficiencies of workers were positively and strongly associated with their weekly and annual earnings.37 In 2003, the NCES launched the nationally representative 2003 National Assessment of Adult Literacy (NAAL) survey, which assessed proficiencies of adults aged 16 and older. Labor market findings from this survey were similar to those from NALS; in comparison to adults with higher levels of literacy, adults with low levels of literacy were likely to have considerably lower wages and higher rates of reliance on public assistance.38 Internationally, in the early 1990s the Organisation for Economic Co-operation and Development (OECD) launched the International Adult Literacy Survey (IALS) in 15 countries. In 2003 and 2006, the OECD launched the Adult Literacy and Lifeskills (ALL) survey with a goal to assess the literacy and numeracy skills of adults that are required in the workplace.

Building upon the framework of these previous surveys, the OECD in 2010 launched the PIAAC survey in 22 countries. PIAAC tested adults in literacy, numeracy, and problem solving in technology-rich environment skills that were comparable across those countries. Unlike the IALS and NAAL surveys, PIAAC had substantial sample sizes in each participating country.39 PIAAC found that in each participating country, workers with higher literacy and numeracy skills outearned peers with lower literacy and numeracy skills.40 Hanushek, Schwerdt, Wiederhold, and Woessmann found that among full-time workers (between ages 35 and 54) in the 22 nations studied, an increase of one standard deviation unit in numeracy skills increased the hourly wage by an average of 18 percent. The study found a wide variation across these countries in the returns to numeracy skills, with the highest returns to numeracy skills observed among workers in the United States (28 percent).41

The rich contents of the PIAAC database, especially the availability of data on three key measures of human capital—educational attainment, skills, and years of work experience—allow for a comprehensive measure of human capital. Using the PIAAC database for the United States, this study examines the connection between human capital and the earnings of American prime-age, full-time employed workers. The study relies on earnings regressions that are designed to estimate the independent effect of human capital variables on earnings after statistically controlling for other variables that are known to affect earnings of workers and are included in the regression as explanatory variables. These earnings regressions are often referred to as human capital earnings regressions,42 where human capital corresponds to education and skills that determine the productivity of workers.43

In the remainder of this report, we present findings from multivariate regression analysis of monthly earnings (human capital earnings functions) of 25- to 54-year-old full-time employed workers in the United States. The human capital earnings functions estimated in this report are based on Jacob Mincer's framework, with the dependent variable in these regression models consisting of the natural log of earnings (the natural log of the monthly earnings of 25- to 54-year-old full-time employed workers) and measures of human capital included as explanatory variables.44

An important measure of human capital included in these regressions is the skills of workers. In this study, the skills of workers are measured by their literacy and numeracy proficiencies. Although not a direct measure of ability, proficiency scores of workers are included in the regressions to measure the independent effect of basic skills and education on earnings. Gary Becker contends that the regression-estimated earnings premium to schooling tends to be biased upward and attributes this bias to ability. He contends that a college education accounts for only part of the earnings premium of college-educated workers over high school graduate counterparts. He attributes about 12 percent of the earnings premium of college graduates relative to high school graduates to the fact that college graduates possess a greater ability over high school graduates that would result in higher earnings among them even if they did not graduate from college.45 A similar upward bias in the rate of return to schooling was estimated by Griliches and Mason.46

Higher ability is rewarded in the labor market because it improves worker performance. Furthermore, people with higher abilities are more trainable because higher abilities are associated with a higher aptitude for learning. Although the literacy and numeracy proficiencies of workers measured in PIAAC data might not exactly measure the "ability" of workers to perform in a specific work setting, they do provide an important measure of cognitive skills that are required to perform effectively across the wide array of jobs in the labor market.

The other two measures of human capital, educational attainment and work experience, are also key determinants of the level of earnings of workers. Educational attainment represents a measure of formal investment in human capital, while work experience represents additions to individual productive capabilities that are acquired through post-school on-the-job learning. These gains in the productive abilities of individuals are most often acquired informally through everyday work activities but sometimes are acquired in more formal programs of on-the-job training including apprenticeship and cooperative education. Wages are expected to increase with additional years of work experience because workers acquire additional skills and move on to higher paying positions as they gain occupational and workplace knowledge and develop key workplace social skills.

Notes

33 See Appendix tables C-1 and C-2 for mean earnings and standard errors of mean earnings estimates for each education-proficiency subgroup of workers presented in figures 9 and 10.

34Richard J. Murnane, John B. Willett, Yves Duhaldeborde, and John H. Tyler, "How Important are the Cognitive Skills of Teenagers in Predicting Subsequent Earnings," Journal of Policy Analysis and Management 19, no. 4 (2000): 547-568; Gonzalo Castex and Evgenia Kogan Dechter, "The Changing Roles of Education and Ability in Wage Determination," Journal of Labor Economics 32, no. 4 (2014): 685-710.

35 James J. Heckman, Jora Stixrud, and Sergio Urzua, "The Effects of Cognitive and Noncognitive Abilities on Labor Market Outcomes and Social Behavior," Journal of Labor Economics 24, no. 3 (2006): 411-482.

36 Claudio E. Montenegro and Harry Anthony Patrinos, Comparable Estimates of Returns to Schooling around the World, Policy Research Working Paper No. WPS7020 (Washington, DC: World Bank Group, 2014), http://documents.worldbank.org/curated/en/830831468147839247/Comparable-estimates-of-returns-to-schooling-around-the-world.

37 For evidence, see Andrew Sum, Literacy in the Labor Force: Results from the National Adult Literacy Survey (NALS), A Report Prepared for the National Center for Education Statistics (NCES), Center for Labor Market Studies, (Boston: Northeastern University, September 1999).

38 See William C. Wood, Literacy and the Entry-Level Workforce: The Role of Literacy and Policy in Labor Market Success (Washington, DC: Employment Policy Institute, June 2010).

39 See Technical Report of the Survey of Adult Skills (PIAAC), (Paris: OECD Publishing, 2013), https://www.oecd.org/site/piaac/_Technical%20Report_17OCT13.pdf.

40 See Marguerita Lane and Gavan Conlon, "The Impact of Literacy, Numeracy and Computer Skills on Earnings and Employment Outcomes," OECD Education Working Papers No. 129 (Paris: OECD Publishing, 2016).

41 See Eric Hanushek, Guido Schwerdt, Simon Wiederhold, and Ludger Woessmann, "Returns to Skills around the World: Evidence from PIAAC," European Economic Review 73(C) (2015): 103-130.

42For a review of the key theoretical underpinnings of human capital earnings functions, see Jacob Mincer, Schooling, Experience, and Earnings (New York: National Bureau of Economic Research, 1974); and Solomon W. Polachek and W. Stanley Siebert, The Economics of Earnings (Cambridge, UK: Cambridge University Press, 1993).

43Jacob Mincer, "Investment in Human Capital and Personal Income Distribution," Journal of Political Economy 66, no. 4 (1958): 281-302.

44Mincer, Schooling, Experience, and Earnings.

45Becker, Human Capital.

46Zvi Griliches and William Mason, "Education, Income, and Ability," Journal of Political Economy 80, no. 3, Part 2 (May-June, 1972): S74-S103, https://doi.org/10.1086/259988.