An NBER conference on The Rise of the Megafirm: Causes and Consequences for Labor and Product Markets took place October 19 in Cambridge. Research Associates Kathryn L. Shaw of Stanford University and John Van Reenen of MIT organized the meeting, sponsored by the Smith Richardson Foundation. These researchers' papers were presented and discussed:
Chinhui Juhn, University of Houston and NBER; Kristin McCue, U.S. Census Bureau; Brooks Pierce, Bureau of Labor Statistics; and Kathryn L. Shaw
The Use of Performance-Based Pay versus Wage Insurance within the Megafirm: Implications for the Within-Person Volatility of Income
Nicholas Bloom, Stanford University and NBER; Fatih Guvenen, University of Minnesota and NBER; Benjamin S. Smith, University of California, Los Angeles; Jae Song, Social Security Administration; and Till M. von Wachter, University of California, Los Angeles and NBER
The Disappearing Large Firm Premium
Large firms have paid significantly higher wages for over a century. Based on administrative data Bloom, Guvenen, Smith, Song, and von Wachter document that the large-firm wage premium (LFWP) has declined steadily over the last 30 years. Decomposing pay into worker and firm fixed effects, they document that the LFWP can be largely explained by a rise in firm effects with firm size. The dramatic decline is due a reduction in these firm effects at large firms. These changes have been concentrated at very large employers. In contrast, worker composition has changed little. The researchers also find the majority of the change occurs within industries.
Robert E. Hall, Stanford University and NBER
New Evidence on Market Power, Profit, Concentration, and the Role of Mega-Firms in the US Economy (NBER Working Paper No. 24574)
Hall studies how markup of price over marginal cost reveals market power. The distinction between marginal and average cost is key. Average cost is easy to measure, but the price/average cost ratio understates the price/marginal cost ratio when fixed costs are present. In particular, in free-entry equilibrium, where revenue equals cost, the price/average cost ratio is always one, while the price/marginal cost ratio may be above one. The idea here is to calculate marginal cost as the ratio of the adjusted expenditure on inputs to the adjusted change in output. The first adjustment is to remove the change in expenditure that arises from the changes in input costs. The second adjustment is to remove the change in output attributed to productivity growth. Application to KLEMS productivity data finds a typical markup ratio of 1.3. Markup ratios grew between 1988 and 2015. For mega-firms, the paper uses employment at firms with 10,000+ workers. Substantial heterogeneity occurs across sectors and in growth rates. There is no evidence that mega-firm intensive sectors have higher price/marginal cost markups, but some evidence that markups grew in sectors with rising mega-firm intensity.
Sharat Ganapati, Georgetown University
The Modern Wholesaler: Global Sourcing, Domestic Distribution, and Scale Economies
Nearly half of all transactions in the $6 trillion market for manufactured goods in the United States were intermediated by wholesalers in 2012, up from 32 percent in 1992. Ganapati discusses how seventy percent of this increase is due to the growth of "superstar" firms - the largest one percent of wholesalers. Structural estimates based on detailed administrative data show that the rise of the largest wholesalers was driven by an intuitive linkage between their sourcing of goods from abroad and an expansion of their domestic distribution network to reach more buyers. Both elements require scale economies and lead to increased wholesaler market shares and markups. Counterfactual analysis shows that despite increases in wholesaler market power, intermediated international trade has two benefits for buyers: directly through buyers' valuation of globally sourced products, and indirectly through the passed-through benefits of wholesaler economies of scale and increased quality.
Xavier Gabaix, Harvard University and NBER, and Ralph Koijen, University of Chicago and NBER
In many settings, there is a dearth of instruments, which hampers economists' ability to investigate causal relations. Gabaix and Koijen propose a general way to construct instruments: "granular instrumental variables" (GIVs). In the economies studied, a few large firms or countries account for a large share of economic activity. As they are large, their idiosyncratic shocks (e.g., productivity shocks) affect the aggregate. This makes those idiosyncratic shocks valid instruments for the aggregate shocks. The researchers provide a methodology to extract idiosyncratic shocks from the data, this way creating GIVs. Those GIVs allow then to estimate parameters of interest, including causal relations. The researchers first illustrate the idea in a basic static setup. They then show how, even in the classic supply and demand framework, they can achieve a novel identification of supply and demand elasticities. In contexts of social or economic influence (with the "reflection problem"), the researchers show how to achieve identification where it previously seemed hopeless. Then, Gabaix and Koijen show how the procedure can be adapted to handle many enrichments, such as arbitrary feedback loops, heterogeneity, and multiple factors. They illustrate the procedure in detail in some applications. First, they study the impact of mega-firms on aggregates, such as aggregate competition (as idiosyncratic shocks to large firms change concentration ratios). This allows the researchers to find a causal relationship between the rise of mega firms and lower investment. Second, they measure how shocks to domestic banks causally affect sovereign yields. Gabaix and Koijen document how negative shocks to Italian banks adversely affect Italian government bond yields, and vice-versa. This gives the first causal measure of the "doom loop" between banks and sovereign yields. The researchers delineate how GIVs can offer ways to identify a host of causal relations, e.g. growth spillovers from one country to the rest in the Eurozone, supply and demand elasticities for oil, going from local to general-equilibrium impact in cross-state regressions, the effects of shocks to financial intermediary on asset pricing. Gabaix and Koijen conclude that the GIVs are a good candidate to supplement the empirical economists' toolbox.
José A. Azar, IESE Business School; Ioana Marinescu, University of Pennsylvania and NBER; and Marshall I. Steinbaum, Roosevelt Institute
Labor Market Concentration (NBER Working Paper No. 24147)
A product market is concentrated when a few firms dominate the market. Similarly, a labor market is concentrated when a few firms dominate hiring in the market. Using data from the leading employment website CareerBuilder.com, Azar, Marinescu, and Steinbaum calculate labor market concentration for over 8,000 geographic-occupational labor markets in the US. Based on the DOJ-FTC horizontal merger guidelines, the average market is highly concentrated. Using a panel IV regression, the researchers show that going from the 25th percentile to the 75th percentile in concentration is associated with a 15-25% decline in posted wages, suggesting that concentration increases labor market power.
Ryan Decker, Federal Reserve Board; John C. Haltiwanger, University of Maryland and NBER; and Ron S. Jarmin and Javier Miranda, U.S. Census Bureau
Changing Business Dynamism and Productivity: Shocks vs. Responsiveness (NBER Working Paper No. 24236)
The pace of job reallocation has declined in all U.S. sectors since 2000. In standard models, aggregate job reallocation depends on (a) the dispersion of idiosyncratic productivity shocks faced by businesses and (b) the marginal responsiveness of businesses to those shocks. Using several novel empirical facts from business microdata, Decker, Haltiwanger, Jarmin, and Miranda infer that the pervasive post-2000 decline in reallocation reflects weaker responsiveness in a manner consistent with rising adjustment frictions and not lower dispersion of shocks. The within-industry dispersion of TFP and output per worker has risen, while the marginal responsiveness of employment growth to business-level productivity has weakened. The responsiveness in the post-2000 period for young firms in the high-tech sector is only about half (in manufacturing) to two thirds (economy wide) of the peak in the 1990s. Counterfactuals show that weakening productivity responsiveness since 2000 accounts for a significant drag on aggregate productivity.
John Van Reenen
Increasing Differences between Firms: Market Power and the Macro-Economy
A rich understanding of macro-economic outcomes requires taking into account the large (and increasing) differences between firms. These differences stem in large part from heterogeneous productivity rooted in managerial and technological capabilities that do not transfer easily between firms. In recent decades the differences between firms in terms of their relative sales, productivity and wages appear to have increased in the US and many other industrialized countries. Higher sales concentration and apparent increases in aggregate markups have led to the concern that product market power has risen substantially which is a potential explanation for the falling labor share of GDP, sluggish productivity growth and other indicators of declining business dynamism. Van Reenen suggests that this conclusion is premature. Many of the patterns are consistent with a more nuanced view where many industries have become “winner take most/all” due to globalization and new technologies rather than a generalized weakening of competition due to relaxed anti-trust rules or rising regulation.