In most U.S. industries, the biggest firms have a higher market share than they did three decades ago. For example, one study found that 75% of U.S. industries have become more concentrated since the 1990s and that the average size of the largest players in the economy has tripled. One potential concern with this rise in industry concentration is that it reduces workers’ employment options, and thus gives employers the ability to lower wages. However, research also suggests that when firms make outsized profits — as they might when they have a large share of the market — they share some of it with workers in the form of higher wages. Which of these effects is more significant? Research suggests it depends on the industry and the type of work. Specifically, a new study finds that workers in finance may see higher wages when the industry becomes more concentrated, whereas workers in other industries see lower wages.
In most U.S. industries, the biggest firms have a higher market share than they did three decades ago. One study found that 75% of U.S. industries have become more concentrated since the 1990s and that the average size of the largest players in the economy has tripled. A potential concern with this rise in industry concentration is that it reduces workers’ employment options, and thus gives employers the ability to lower wages. However, research also suggests that when firms make outsize profits — as they might when they have a large share of the market — they share some of it with workers in the form of higher wages.
Which of these effects is more significant? My research suggests it depends on the industry and the type of work.
The chart below plots trends in industry concentration measured by the Herfindahl-Hirschman Index (HHI) from 1990 to 2008. HHI is calculated by squaring the market share of each firm competing in the industry; it rises when fewer firms have more of the market. The data is broken out between financial and nonfinancial industries. On average, the employer concentration has kept increasing since 1990 in both financial and nonfinancial industries. Specifically, the average HHI concentration measure has increased by approximately 40% in finance and about 40% in nonfinance.
If industry concentration reduces workers’ outside options and gives firms bargaining power to lower wages, we should expect flat wage growth in both financial and nonfinancial sectors. But interestingly, the next chart shows that the average real wage in financial sectors has increased by 23.38% from 1990 to 2008, which is almost three times more than in nonfinancial sectors.
To identify the marginal difference in the correlation of industry concentration measured by HHI and wages in financial and nonfinancial industries, my research conducts a cross-sectional analysis using data from the U.S. Census. I show that higher industry concentration is associated with lower wages in nonfinancial sectors. However, in financial sectors, a one-standard-deviation increase in industry concentration is associated with 13.78% higher wages. The difference between financial and nonfinancial sectors is still economically significant when I control for macroeconomic trends. Also, the positive correlation between concentration and wages in finance is not driven by the fact that financial firms, on average, hire relatively better workers in terms of education and working experience.
Why is industry concentration in finance associated with a smaller wage-dampening effect? My research argues that industry concentration has two competing effects on wages: a labor-market monopsony effect and a rent-sharing effect. (Monopsony refers to a market where one large buyer has substantial power.) First of all, industry concentration increases firms’ market power in the labor market. As workers have fewer possible employers in a more concentrated market, firms possess more bargaining power and so are able to pay less. This is called the labor-market monopsony effect. Meanwhile, industry concentration also increases firms’ market power in product markets. By increasing product prices and/or lowering nonlabor input costs, firms are able to extract rents from their product markets. When firms share rents with workers, wages increase. This is called the rent-sharing effect. The labor market monopsony and rent-sharing effects coexist in all industries, but it appears that the rent-sharing effect dominates the labor-market monopsony effect in financial sectors. My research provides evidence on two non-mutually exclusive mechanisms for this.
First, I show that higher market power is associated with a relatively higher return on assets (ROA) in financial industries, as compared with nonfinancial industries. Compared with other profitability measures, ROA has been shown to be a better measure for capturing firms’ abnormal profitability. In other words, financial firms aren’t just capturing more of the market — they’re securing more market power, which means they have relatively more rents to share with workers. (My findings are also consistent with the idea that when financial institutions capture a large share of the market they become “too big to fail” and are then able to borrow at lower cost, as investors assume the government would bail them out if they faced bankruptcy.)
Second, compared with nonfinancial firms, the way financial firms work may give certain employees considerable bargaining power. Compared with workers in most nonfinancial industries who generally work in teams, financial workers such as traders are closer to final products and their performance is directly linked to firm performance. This feature increases the external visibility of these financial workers and thus increases the likelihood of workers’ getting poached by rivals, giving them greater bargaining power. The availability of online rankings of the best financial workers can act as a catalyst of this process.
In addition, Claire Celerier and Boris Vallee show that high-skill workers in financial industries are matched with larger-scale tasks, as compared with high-skill workers in nonfinancial industries. For example, based on the Hedge Fund Survey, each hedge fund manager is, on average, in charge of $70 million in 1990, and the amount increased to $190 million in 2006. The high scalability gives high-skill finance workers bargaining power to ask for a larger fraction of rents. In support of this mechanism, I find higher market power is associated with relatively much higher wages for high-skill workers in finance as compared with high-skill workers in nonfinance. Specifically, a one-standard-deviation increase in a firm’s market power in finance is associated with higher wages of $1,468 per quarter, whereas the increase is only $300 per quarter in nonfinancial sectors. Because rents are disproportionally distributed to high-skill workers within financial firms, I also find that higher market power is associated with relatively higher within-firm inequality in finance.
In my analysis, high-skill workers are individuals whose earnings are above the 90th percentile of the wage distribution within a firm in a given quarter. Based on this definition, high-skill workers are likely to include most executives and managers. People may think that these workers get larger shares of total rents because of bad corporate governance in finance. If this is the case, we should observe that higher market power is associated with worse firm performance in finance. But what I find is the opposite. This supports the story that finance firms share more rents with workers to create incentives for them to work harder.