Share
Wage inequality has increased dramatically in the US since the 1970s, largely driven by within-firm earnings inequality. This column uses combined data from three large micro datasets to study the drivers and implications of within-firm inequality. It shows that employees across the entire wage distribution gain from higher firm productivity, but these gains are unequally distributed. The top 1% of employees by pay have a fourfold higher response to productivity than the bottom 1%. Aggregating these effects suggests that the increase in labour productivity over 1980-2013 can explain around 40% of the rise in the gap between CEO and median worker pay.
There has been a dramatic rise in wage inequality in the US since the 1970s, spurring an enormous body of research attempting to understand the causes of this trend. Song et al. (2019) find that about two-thirds of the overall point-in-time earnings inequality in the US since the 1970s can be explained by within-firm earnings inequality. Thomas Piketty (2013) has argued that “the primary reason for increased income inequality in recent decades is the rise of the super-manager” (p.315). As corporations have become more tolerant of generous pay packages, executive pay has surged, driving up overall inequality. This insight does not address several key facets of wage inequality: Is the extraordinary rise in executive pay linked to firm performance? How has pay at other levels of the firm wage distribution evolved alongside executive pay?
New evidence
To investigate the nature of within-firm inequality more deeply, in Wallskog et al. (2024) we combine three massive micro datasets at the US Census Bureau. We use detailed quarterly labour earnings data from 2003 to 2015 for millions of US employees, matched to their employers, from the Longitudinal Employer-Household Dynamics (LEHD) programme. We match these data to employment and revenue information for firms across the US from the Longitudinal Business Database (LBD). Finally, we incorporate information about firms’ use of structured management practices relating to performance monitoring, targeting, and incentive setting from the Management and Organizational Practices Survey (MOPS), a supplement to the Annual Survey of Manufactures for 2010 and 2015.
Analysing the combined data from these three programmes yield three major findings.
First, employees at more productive firms have substantially higher pay across every percentile of the earnings distribution. Not only are executive earnings higher at successful firms, but so are employees at every earnings level. However, this pay-performance link is not evenly distributed throughout the wage distribution.
Figure 1 shows how firm productivity gains translate into worker wage increases for different employees at a firm. In Figure 1a, we plot the coefficient of a regression of individual pay on firm productivity for each percentile of the employee pay distribution. Notably, the top 1% of earners, which often includes senior executives, sees the largest increase in earnings, with a coefficient of 0.13. This means that for every 10% increase in productivity, the top 1% (99th percentile) see about 1.3% more pay. This compares to a more modest 0.7% increase for median earners (coefficient of 0.07). Interestingly, even the lowest-paid workers, typically entry-level or frontline employees, experience some benefit. Those in the 10th percentile of earnings see their earnings grow by about 0.5% for every 10% productivity gain (coefficient of 0.05).
Figure 1 Firm productivity and worker wages
When we zoom in on top executives, these differences are particularly stark. Figure 1b plots the increase in pay from an increase in productivity by pay rank within companies. The far-right point in Figure 1b is for the top-ranked employee – typically the CEO – which shows that for every 10% increase in productivity, the CEO sees about 1.5% more pay (coefficient of 0.15). If we look at the 10th ranked employee – typically a very senior manager – they see about an 1.1% pay increase for 10% higher productivity (coefficient of 0.11). The 50th ranked employee – typically more of a middle manager – would receive about 0.9% higher pay (coefficient of 0.09).
The fact that these figures are upwards sloping highlights that while every employee gains from higher productivity, some gain more than others. For example, the top 1% of employees by pay have a fourfold higher response to productivity than the bottom 1% (coefficient of 0.013 vs. 0.03). In other words, inequality grows as productivity rises.
Figure 2 plots these results by the level of pay. These results control for additional employee characteristics, but the main message is the same as in Figure 1: higher paid employees see greater gains from rising productivity. In particular, employees earning $500,000 a year see a pay increase from higher productivity that is twice that of the increase seen by those earning $20,000 a year.
Figure 2 Impact of productivity on pay growth by pay level
Second, we find that this pay-performance link is particularly strong in public firms. Figure 3 plots the response of pay to productivity for public firms (blue circles) and private firms (green triangles). We see that among the highest-paid executives (top 50), those leading public companies receive nearly double the pay increase compared to executives at private companies for a comparable increase in firm productivity. As an additional check, we use the public pay information in Execucomp, which harvests the data in executive pay statements for S&P 1500 firms. We see a very similar result here to our Census data on public firms.
Figure 3 Pay-performance link for public versus private firms
Why public firms provide stronger pay-performance rewards to their executives than privately held firms is an important research question. Perhaps the pressures of being publicly listed or the ability to offer stock options and grants generates this tighter pay-performance link.
Our third result explores the mechanism by which higher firm productivity translates into increased pay, starting by examining bonus behaviour. We can’t directly see bonus amounts, but by looking at how employee pay fluctuates throughout the year (pay volatility), we can estimate bonus patterns. Interestingly, we find a strong link between how much top earners’ pay responds to performance and the volatility of their earnings over the course of a year. This suggests that more productive companies might also have more aggressive management styles. They might monitor employee performance more closely and offer incentive pay schemes with bigger bonuses. While this can lead to higher overall pay for top earners, it also creates more uncertainty in their earnings. Indeed, in Figure 4 we show that the greater adoption of structured management practices for manufacturing businesses measured in the Management and Organizational Practices Survey (MOPS) is associated with higher pay volatility, particularly for senior executives.
Figure 4 Within-year pay volatility and management score
Aggregate implications
Quantifying these effects, we estimate that the increases in labour productivity from 1980 to 2013 can explain about 40% of the rise in the gap between CEO and median worker pay. This CEO-to-median-worker pay ratio is particularly salient because the Dodd-Frank Act requires publicly traded firms to publish it annually from 2018 onwards.
Conclusions
Using detailed microdata, we find evidence that higher firm productivity does appear to lead to higher pay across every percentile of companies’ pay distribution. This response is four times higher for the top 1% than the bottom 1%, meaning that rising productivity generates rising within-firm inequality. It seems that all workers’ compensation is dependent on firm productivity, but the degree to which pay is linked to performance varies depending on workers’ relative positions in their firms’ pay distribution, with top earners’ pay much more dependent on firm performance compared to those at the lower end. This dependency is reflected in greater wage inequality for more productive firms.
Now, if we’re interested in understanding what drives changes in firm productivity, Faggio et al. (2007, 2010) offer an explanation. They suggests that technological differences between firms may be the underlying cause of these productivity variations, and consequently, a contributing factor to rising inequality.
source : voxeu
