The New Dynamics of Labor Market Institutions, Skills, and Productivity

Paper Session

Saturday, Jan. 7, 2017 10:15 AM – 12:15 PM

Hyatt Regency Chicago, McCormick
Hosted By: Econometric Society
  • Chair: David Autor, Massachusetts Institute of Technology

Training Programs, Skills, and Human Capital: A Life-Cycle Approach

Gueorgui Kambourov
,
University of Toronto
Colin Caines
,
University of British Columbia
Florian Hoffmmann
,
University of British Columbia

Abstract

Economic studies of the effectiveness of government-sponsored worker training programs in fostering career progression is traditionally based on models with one-dimensional skills and human capital. Since training is an upfront human-capital investment, it is predicted to depress the rate at which workers reallocate across jobs. In this paper we analyze if this view is consistent with observed life-cycle labor market dynamics of workers with and without a training degree. To this end we focus on Germany's apprenticeship program, which offers occupation-specific training to high-school graduates together with government-sponsored general education and which is currently the largest training program of its kind in the world. We rely on a rich administrative worker-level panel data set that follows employees from labor market entry on until 25 years into their career. We document a number of striking facts: First, the large majority of apprentices are observed in just about a dozen of occupations even though training programs are offered in more than 500 occupations. In contrast, the employment distribution across occupations is much more even for high-school students who do not enter an apprenticeship program. Second, when using data on occupation-specific task usage, we find that apprentices are concentrated in occupations that predominantly require non-routine rather than routine tasks, while non-apprentices are more likely to work in routine occupations. Third, workers with an apprenticeship degree are quite mobile. However, in contrast to workers without a formal degree, their mobility patterns are "directional" in the sense that they clearly reflect either upgrades or downgrades in the occupational skill space.
We argue that standard models with one-dimensional skills and human capital cannot explain these distinct patterns. Instead we develop a model in which human capital is occupation-specific, but in which non-routine occupations require upfront occupation-specific human capital built-up.

The Performance Pay Premium, Human Capital, and Inequality: Evidence From Over Forty Years of Microdata

Christos Makridis
,
Stanford University

Abstract

This paper studies the rise of performance pay contracts and their aggregate effects on the labor market. First, using the Panel Study of Income Dynamics (PSID) and National Longitudinal Survey of Youth (NLSY), I document several stylized facts: (i) the share of performance pay workers grew from 15% in 1970 to nearly 50% by 2000, (ii) performance pay workers experience higher earnings levels and growth rates, work longer hours, and invest more in human capital, and (iii) performance pay workers face lower (higher) permanent (transitory) income shocks, relative to their fixed wage worker counterparts. Second, using the National Compensation Survey (NCS), I show that increases in performance pay are associated with increases in inequality at the micro-level and accelerate the rate of skill-biased technical change. Third, I structurally model the rise of performance pay contracts by solving a dynamic model with unobserved person-specific heterogeneity, discrete sector-occupation job choices, time-varying sector-occupation probabilities of performance pay, and human capital accumulation. The model is estimated using simulated method of moments. Fourth, I use the model to characterize the contribution of performance pay to aggregate inequality and examine the counterfactual effects of making the U.S. marginal tax code as progressive as the one in France.

Revisiting the Relationship Between Unemployment and Wages

Giovanni Gallipoli
,
University of British Columbia
Yaniv Yedid-Levi
,
University of British Columbia

Abstract

We investigate the empirical relationship between wages and labor market conditions. Following work histories in the NLSY79 we document that the relationship between wages and unemployment rate differs across occupations. The results hold after controlling for unobserved match quality. This suggests that evidence about history dependence of wages obtained from pooled samples conceals significant differences and provides an imprecise description of earning dynamics. We examine these discrepancies and offer new evidence suggesting that the sensitivity of wages to current unemployment is linked to the prevalence of performance pay.

The Fall of the Labor Share and the Rise of Superstar Firms

David Autor
,
Massachusetts Institute of Technology
David Dorn
,
University of Zurich
Lawrence Katz
,
Harvard University
Christina Patterson
,
Massachusetts Institute of Technology
John Van Reenen
,
Massachusetts Institute of Technology

Abstract

The fall of the labor share of GDP in the US and many other countries has been well documented, but its causes remain controversial in part because empirical assessments have relied on industry or macro data. In this paper we use firm and establishment panel data, focusing on the US Economic Census since 1982 (but also drawing on international micro data) to describe some stylized facts that seem consistent with a new explanation of the fall in the labor share based on “superstar firms." If globalization or technology mean that product markets are increasingly “winner takes most," industries will become increasingly dominated by superstar firms with high profit margins. As the importance of such firms increases, the aggregate labor share will fall. The model has a number of predictions. First, industry sales will be increasingly concentrated in a small number of firms. Second, those industries where concentration rises the most should have the largest falls in the labor share. Third, the fall in the labor share should have a large between firm component (rather than being primarily a within-firm phenomenon, where the labor share falls similarly across most firms). Fourth, this between firm fall in the labor share should again be greatest in the sectors that are concentrating the most. Fifth, these patterns should be observed not only in US firms, but also internationally. We find support for all these predictions of the superstar firm model, particularly in sectors where technical change is more rapid, as measured by the growth of TFP or patent intensity.
JEL Classifications
  • E0 - General