Data from the National Longitudinal Survey of Youth indicate that the employer often pays the explicit costs of not only on‐site training but also off‐site general training. Although few of these costs appear to be passed on to workers in the form of a lower wage while in training, completed spells of general training paid for by previous employers have a larger wage effect than completed spells of general training paid for by the current employer. A model where contract enforcement considerations cause employers to share the costs and returns to purely general training can explain these findings.

Hires occur for two reasons - to grow a business and to replace those who have left (churn). Churn is an important part of employment dynamics, allowing workers to move to their most productive use. We present evidence on churn from the Job Openings and Labor Turnover Survey (JOLTS). Churn is procyclical. During the 2007-09 recession, four-fifths of hiring reductions are associated with reduced churn, not with reductions in job creation. We estimate that the cost of reduced churn is about two-fifths of a percentage point of GDP annually throughout the three-and-one-half year period since the beginning of the recession.

Analyses of survey data highlight observable person characteristics, such as education and occupation, as critical factors for rising earnings inequality, while industry has an offsetting effect. In contrast, analysis of administrative data highlights that rising between-firm earnings dispersion and, in turn, between-industry earnings dispersion dominates the rise in earnings inequality. We construct a novel integrated dataset based on CPS micro data linked with LEHD administrative records. We find that most of the rise in earnings inequality is accounted for by rising between-industry inequality. This finding reflects a substantial contribution of increased sorting and segregation of observable person characteristics between industries.

The typical Mincerian wage equation examines wages in relation to the education, potential experience, and other personal characteristics of job incumbents. These characteristics serve as proxies for the job holder’s skill level, but do not indicate what specific skills are being rewarded. A number of recent papers make use of data on occupational skill requirements, which have proven useful for understanding shifts in labor demand (see, for example, David Autor, Frank Levy, and Richard Murnane 2003; Maarten Goos and Alan Manning 2007) as well as for understanding the relationship between wages and specific job skills (see, for example, Beth Ingram and George Neumann 2006; Autor and Michael Handel 2008). To the extent that the labor market does a good job of matching individuals to jobs for which they are well suited, such analyses can shed new light on how workers’ job skills are valued in the labor market. Studies that focus on job skills generally begin with data from the Current Population Survey (CPS) or another household survey that contains information on the detailed occupation in which people work. Information on required job skills is attached to the survey records according to reported occupation. There is considerable evidence, however, of significant errors in the coding of occupation in household survey data. Wesley Mellow and Hal Sider (1983) find disagreements between the occupation recorded in CPS data compared to that based on information supplied by individuals’ employers for 19 percent of jobs at the major occupation level and 42 percent at the detailed occupation level. Nancy Mathiowetz (1992) reports similar findings for the employees of a large manufacturing firm. Comparisons of aggregate data on the number of jobs in each of 19 broad occupations from the New Evidence on the Returns to Job Skills

Most of the rise in overall earnings inequality from 1996 to 2018 is accounted for by rising between-industry dispersion. The contribution of industries is right-skewed with the top 10 percent of four-digit NAICS industries dominating. The top 10 percent are clustered in high-paying high-tech and low-paying retail sectors. In the top industries, high-wage workers are increasingly sorted to high-wage industries with rising industry premia. In the bottom industries, low-wage workers are increasingly sorted into lowwage industries, with rising employment and falling industry wage premia.

The average wage is a key aspect of economic well-being in a society. Several alternative wage measures based upon official government statistics are available to analysts. Although these series always have differed in their relative levels, since the mid-1970's their trends have diverged markedly. This paper, which summarizes and updates our earlier work (Abraham et al., 1998), has several goals: to describe several commonly used hourly wage series; to characterize differences in their trends; and to evaluate alternative explanations for these differences.

Using a large data set that links individual Current Population Survey (CPS) records to employer-reported administrative data, we document substantial discrepancies in basic measures of employment status that persist even after controlling for known definitional differences between the two data sources. We hypothesize that reporting discrepancies should be most prevalent for marginal workers and for marginal or nonstandard jobs, and we find systematic associations between the incidence of reporting discrepancies and observable person and job characteristics that are consistent with this hypothesis. The paper discusses the implications of the reported findings for both micro and macro labor market analysis.

Good information on self-employment is needed to inform the ongoing discussion of the rise of the gig economy and its implications for workers. Tax data show significant growth in self-employment not captured in the Annual Social and Economic Supplement to the Current Population Survey (CPS-ASEC). The growing gap reflects both self-employment in tax data missing from the CPS-ASEC and self-employment misreported as wage and salary work. We document consistent patterns in the discrepancies between the tax and survey data but are able to explain only a modest share of the growing disagreement between them.

Productivity Differences across Employers: The Roles of Employer Size, Age, and Human Capital by John C. Haltiwanger, Julia I. Lane and James Spletzer. Published in volume 89, issue 2, pages 94-98 of American Economic Review, May 1999