Antitrust enforcement in the United States has become a populist rallying cry and a rare example of bipartisan agreement. The Trump administration unsuccessfully sued to block the $85 billion merger between AT&T and Time Warner Cable, after Trump said while campaigning, “Deals like this destroy democracy.” Meanwhile, Democrats including Senator Elizabeth Warren (of Massachusetts) have rallied against corporate concentration.
In economic circles, an argument has gained traction that corporate market power is surging, resulting in skyrocketing markups, a falling labor share, and other negative consequences for consumers and workers. But some researchers are pushing back, emphasizing weaknesses in the argument and urging policy makers to be cautious before taking any actions.
Proponents of the market-power argument often rely on one of two methodologies, one that calculates and compares total revenues and costs at the economy-wide level and another that uses company-level data. University of Minnesota’s Loukas Karabarbounis and Chicago Booth’s Brent Neiman focus on the first of these two, in which the economy is considered a pie that is made of up three slices: the labor share (which goes to workers), the capital share (costs incurred to use factories, equipment, software, etc.), and economic profits. Economic profits are calculated by finding the difference between revenues and costs, including the cost of capital faced by companies to fund their assets used in production.
The labor piece has been shrinking, economists agree. Karabarbounis and Neiman demonstrated this in 2014, arguing that roughly half the decline is due to automation and robots. But a counterargument observes that since 1980, revenues have increased faster than estimated costs and assumes this rising difference is all economic profits. This interpretation holds that rising profits and market power are the primary culprits for workers’ woes—and bolsters the idea of promoting more competition to reverse the trend.
Karabarbounis and Neiman dispute these claims, arguing that it is difficult to properly measure the capital share and therefore wrong to simply assume the estimated gap between revenues and costs is economic profits. Rather, they refer to this gap as “factorless income” and consider several alternatives for what it comprises. Of the alternatives, they are least convinced that the gap entirely represents economic profits. In fact, when they extend this interpretation back to the 1960s, they find that the results imply that market power was typically higher in the 1960s and ’70s than it is today.
The findings suggest both academics and policy makers have a long way to go before reaching definitive conclusions.
The most promising explanation for factorless income, they find, is that the standard way economists measure a company’s cost of capital is missing some important elements. Namely, there’s a changing gap between the economy’s cost of capital and the Treasury yields often used when measuring the capital share. Their work suggests the possibility that this gap reflects risk premia, which have fluctuated significantly and risen since the early 1980s. If this were the explanation for changes in factorless income, it would dampen any argument for tougher antitrust laws or enforcement, carry implications for monetary and fiscal policy, and reassert that automation is a primary driver of the labor-share decline.
In closely related work, Chicago Booth PhD candidate James Traina challenges the second methodology used by market-power proponents—the one that pulls operating data from publicly traded companies. Traina notes that analyses of readily available financial statements almost always exclude private companies, which are often smaller and may have less market power. But public companies are an important part of the economy, and Traina finds that even they haven’t experienced skyrocketing markups.
One key driver of earlier works’ conclusion that market power has increased is the decline in the cost of goods sold (a cost category that typically includes materials and low-skilled labor) relative to sales. But Traina argues that this decline in the costs of goods sold doesn’t reflect an increase in markups and market power because at these same companies, the decline has been largely offset by increases in marketing and management expenses. This fact is consistent with companies changing their production to keep up with changing technology. Karabarbounis and Neiman apply Traina’s critique to company-level data for countries other than the US and draw the same conclusion.
The possibility of rising market power remains nevertheless persists among researchers, and since it’s a critical topic for policy, Neiman argues, the findings suggest both academics and policy makers have a long way to go before reaching definitive conclusions.