An innovative new study finds substantial, increasing deadweight losses resulting from oligopolistic behavior and points to the important role that startup acquisitions—particularly by large tech firms—played in driving this trend.

It is no secret at this point that the American economy has a concentration problem. Nearly every American industry has experienced an increase in concentration in the last two decades, to the point where (as Barry Lynn wrote in 2011) sectors dominated by two or three firms are not the exception, but the rule.

In recent years, a growing number of studies have linked this increase in concentration to a decline in competition, arguing that higher prices, increasing inequality, and sluggish productivity growth, along with labor’s falling share of income and rising health care costs, can all be at least partly explained by an increase in market power. Ordinary Americans, who face a limited set of alternatives whenever they go to the bank, use a credit card, book an airline ticket, or check into a hospital, have responded positively to this emerging narrative and as a consequence, antitrust enforcement has become one of the key issues in the 2020 presidential election, and one of the few that has received bipartisan support.

Among economists, however, there is still some skepticism regarding the so-called “market power narrative,” particularly among Industrial Organization (IO) economists. Despite the studies listed above, some of the leading IO economists argue there is still no “smoking gun” that allows us to determine conclusively that market power is indeed responsible for trends such as record corporate profits, declining dynamism and firm entry, and lower investment rates.

A new study, however, seeks to help resolve this debate by studying the impact of oligopolies on resource misallocation and consumer welfare. The study, written by Bruno Pellegrino, a PhD candidate at UCLA, finds that US industries have indeed become more oligopolized, and that rising oligopoly power is associated with significant welfare costs. The study also points to the important role that startup acquisitions—particularly by large tech firms—played in driving this trend.

Oligopolies Cause Significant Inefficiencies – to the Detriment of Consumers

Part of the reason some economists are hesitant to accept the market power explanation is the scarcity of data that allows them to gauge the intensity of competition between firms. Pellegrino strongly agrees with this criticism, telling ProMarket that more data and a nuanced analysis of product market rivalries are needed in order to measure “who competes with whom.”

In the past, macroeconomists have created models that sorted companies into buckets: one set of firms produce shoes; others make cars or pharmaceuticals. But this type of structure doesn’t tell us the whole story: Apple and Microsoft, for instance, are part of the same industry, but they don’t produce exactly the same things.

The extent to which a firm behaves like a monopolist depends among other things on how easily consumers can substitute its products for others. This means that in order to really understand whether oligopoly power is on the rise, we need to take into account how the mix of available products in the economy has changed over time. What could be going on, some economists argue, is that within or across industries, firms’ product portfolios are becoming more similar: every firm competes more aggressively with everyone else, and this is driving the weakest competitors out of business. The reason concentration has increased, according to this interpretation, is actually that there’s more competition, not less.

In his paper, Pellegrino offers a solution to this conundrum, in the form of a new model. Its crucial innovation is that it dispenses with the notion of industries and sectors, allowing Pellegrino to view the product market not as a rigid assortment of sectors, but as a network that changes over time as firms become more or less competitive with each other, thus drifting farther or closer apart. To implement this framework, Pellegrino took advantage of an extensive dataset developed by Gerard Hoberg and Gordon Phillips, which measures similarity between firms using a computational linguistic analysis of regulatory forms filed by US public companies between 1997 and 2017.

Figure 1: the Hoberg-Phillips data, as visualized by Pellegrino. Every dot in the diagram is a publicly-traded US firm. The distance between nodes reflects product similarities between firms.

The idea behind the model, Pellegrino explains, is that firms continuously change the products they offer consumers, becoming more or less “similar” in the process. “Take, for example, Apple, Dell, and Motorola. As Apple enters the mobile business, over time it becomes less similar to Dell. When Motorola spins off its mobile business, the similarity between Motorola and Apple goes to zero.”

Figure 2: Apple’s similarity to Dell and Motorola over time

Using this model, Pellegrino was able to provide new confirming evidence that the US economy has indeed become more oligopolized, and that increasing concentration is causing significant inefficiencies—to the detriment of consumers. Moreover, Pellegrino was able to compute the loss of economic efficiency (otherwise known as “deadweight loss”) resulting from oligopoly power, finding that oligopolistic behavior causes a significant deadweight loss of 13.3 percent of total potential surplus. Meaning, if companies were to act competitively and not as oligopolists, the total economic surplus would go up by more than 13 percent. The deadweight loss, he finds, has increased over time as the economy has become more concentrated.

Another consequence is that the share of the surplus that goes to consumers has decreased, while the share that goes to oligopoly profits has increased. “The consumer is losing twice,” says Pellegrino. “Less surplus is being produced (as a percentage of the surplus that could be produced), and a smaller share of that goes to the consumer.”

Figure 3: Breakdown of total potential surplusUS public corporations

Pellegrino’s model also suggests that higher levels of concentration contributed to higher markups. The issue of markups and market power has been the subject of contentious debate among economists in the past two years, but to Pellegrino, it’s not so much the increase in markups that matters, but their distribution. “When you compute markups in my model, you do get a slight increase in markups over time, but what increases noticeably is the dispersion of markups. A big part of the increase in concentration and market power, as well as many of the inefficiencies, has to do with the increase in this dispersion.”

From IPOs to Acquisitions

The model also allows Pellegrino to simulate different policy and market scenarios: namely, merge firms or break them up. That way, he is able to shed light on the possible factors that contributed to rising concentration in the US economy.

While many factors likely contributed to this trend, one factor Pellegrino identifies is the growing rate of venture capital-backed startups being acquired, particularly by large tech firms. Before the early 1990s, he explains, successful startups would opt for IPOs over buyouts. Over time, however, startups began to ditch IPOs in favor of acquisitions, until the trend has completely reversed: nowadays, most startups choose to be acquired. When simulating what would have happened if the rate of IPOs had frozen at the rate it was in 1996, Pellegrino found that many negative trends associated with oligopoly simply disappear. While studies have previously pointed to mergers as a major contributing factor in the oligopolization of the US economy, Pellegrino’s findings show that the acquisition of startups contributed to rising concentration as well.

Figure 4: US venture capital exits by year and type

Which brings us to the five big tech firms—Alphabet (Google), Amazon, Facebook, Apple, and Microsoft. If startup acquisitions have indeed contributed to the increase in concentration, then the Big Five played a special role in this process. The Big Five tech firms have acquired more than 600 startups in the last two decades. Google alone has acquired one company per month, on average, for the past 17 years. This amounts to a disproportionate share of all startups acquisitions in the US.

While Pellegrino’s study does not contain any explicit policy recommendations, his findings highlight the challenges faced by antitrust enforcers in the age of digital platforms. In the past decade, the large tech platforms have made hundreds of acquisitions, markedly increasing their market power. These acquisitions, many of which have the potential to become “killer mergers“, have gone largely unchallenged by antitrust enforcers, who are usually wary of challenging smaller deals that have a limited effect on concentration in the short term.

“Of course, you cannot extrapolate from these results how to rewrite the antirust rulebook,” says Pellegrino. “But one thing that we do know is that to the extent that market power is driven by startup acquisitions—which seems consistent with some of the trends we’re seeing—the current antitrust toolbox is not well-equipped to deal with the problem.”

Photo: “Big Fish Eat Little Fish.” Designed by Pieter Bruegel the Elder, engraved by Pieter van der Heyden and published by Hieronymus Cock in 1557.

The ProMarket blog is dedicated to discussing how competition tends to be subverted by special interests. The posts represent the opinions of their writers, not necessarily those of the University of Chicago, the Booth School of Business, or its faculty. For more information, please visit ProMarket Blog Policy