The labor share, the fraction of value added that goes to workers as wages, is one of the most widely used indicators in economics. It is often interpreted as a measure of how the gains from economic activity are distributed between capital and labor. However, while the labor share has declined in many advanced economies over recent decades, this trend hides substantial underlying variation. In particular, its meaning at the firms' level is not straightforward. A low labor share may reflect that a firm uses relatively little labor in production, for example because it is capital-intensive or highly productive. But it may also mean that the firm pays wages that are low relative to workers' contribution to output, possibly due to limited competition in the labor market. These are two very different mechanisms leading to similar outcomes. To interpret the firm-level labor share and use it to understand inequality, productivity, or market power, it is essential to disentangle these mechanisms.
This project takes an empirical approach, using firm-level data to understand the variation in labor shares across firms. In the first phase, I will jointly analyze productivity and wage-setting behavior, giving a more granular account of why firms differ in their labor shares. While prior work has typically examined either productivity or wage-setting power in isolation, this project relates the two systematically, across firms and over time. In a later phase, I will turn to life-cycle patterns, investigating how wage-setting power evolves as firms grow. This provides new evidence on how labor market power develops and under what conditions it contributes to wage inequality and inefficiency. Together, these steps help clarify the mechanisms behind firm-level differences in the labor share, and contribute to ongoing debates on firm heterogeneity, market power, and wage dynamics.
Źródło finansowania | Financing: Narodowe Centrum Nauki, PRELUDIUM 24
Projekt realizowany | Timeline: 02/2025 -- 02/2029
Kierownik | Principal Investigator: Sebastian Zalas
Budżet łączny | Total budget: 167 092 zł
- stypendia dla młodych badaczy | scholarships for young scholars: 54 000 zł
- komputery i oprogramowanie | hardware and software: 9 800 zł
- konferencje i inne wyjazdy | conference travels: 49 900 zł
- pozostałe koszty | other costs: 23 293 zł
- opłaty publikacyjne | publication fees: 2 700 zł
- koszty pośrednie dla FAME | overheads for FAME: 27 399 zł
This project addresses the following research question: to what extent do technological efficiency and wage-setting distortions shape firm-level labor shares, and how does the relative importance of these forces evolve as firms age and grow?
To address this question, the project advances two core hypotheses.
- Hypothesis 1: Firm-level labor share reflects two distinct components: total factor productivity and wage markdowns. Variation in labor shares across firms is driven by two forces: technological efficiency, proxied by total factor productivity (TFP), and monopsony power, proxied by wage markdowns. Although the labor share is often read as a single indicator, it can mask whether a low value results from high productivity or from wages set below workers' marginal product. To separate these channels, I will estimate firm-level production functions to approximate the marginal revenue product of labor (MRPL) and then compute wage markdowns as the ratio of wages to MRPL, with values below one indicating monopsony power. Combining markdowns and productivity estimates allow to characterize firms by whether low labor shares reflect efficiency, market power, or both.
- Hypothesis 2: Wage markdowns increase with firm age and size, conditional on productivity, accumulating over the firm life cycle. This proposes that monopsony power is not a static firm characteristic but evolves as firms grow. Younger firms may need to offer competitive wages to attract workers, while mature firms may be able to pay below marginal productivity due to reduced local competition, higher worker switching costs, or accumulated firm-specific human capital. Markdowns may therefore widen systematically over the life cycle, contributing to lower labor shares among older and larger firms. A life-cycle perspective is essential here. By tracking the same firms over time, I can test whether markdowns rise with age and under what conditions, rather than relying on cross-sectional comparisons. This helps determine whether declining labor shares in mature firms stem from rising technological intensity or growing monopsony power. It also offers a structural explanation for labor share heterogeneity and a novel mechanism for rising wage inequality: if mature, dominant firms accumulate wage-setting power, they may contribute disproportionately to both suppressed wage growth and falling aggregate labor shares, even in the absence of any change in underlying technology. Empirically, I will estimate the MRPL for each firm-year, compare it to observed wages to obtain annual markdowns, and track these across age cohorts to test whether productivity gains and wage-setting behavior evolve differently as firms mature.
Taken together, these hypotheses support two central innovations. First, the project uses harmonized European firm-level panel data to analyze productivity and wage-setting behavior jointly. Thus, it provides a more granular account of why firms differ in their labor shares. By contrast, prior work has typically examined either TFP or monopsony power in isolation, whereas this project relates them systematically across firms and over time. Second, it introduces a life-cycle framework to investigate how wage-setting power develops as firms grow. This approach provides new evidence on how labor market power emerges and when it contributes to wage inequality and allocative inefficiency. It also clarifies the mechanisms behind firm-level labor share differences and adds empirical evidence to debates on firm heterogeneity, market power, and wage dynamics.