Has Big Business Gotten Too Big?

Historical Trends in Concentration

Let’s start with historical trends. Elaine Tan, in a study of share of the economy by large firms in the United States from 1931 to 2000 found, “When big business is defined as the largest 200 or 500 non-financial corporations, its share of assets was never as high as it was during World War II, and was declining or stable after the merger wave of the late-1960s.”⁵ Likewise, the percentage of manufacturing industries in which the top four firms accounted for more than 50 percent of shipments increased only slightly from 1952 to 2007, from 35 percent to 39 percent, hardly evidence of rampant monopolization.⁶ The revenue of the largest 200 US corporations as a percentage of total business revenue did grew, but only from 24 percent in 1954 to 29 percent in 2008.⁷ These figures include overseas sales, and since large firms are more likely to sell overseas, and since overseas sales have grown faster than domestic sales over the last half century, this increase doesn’t necessarily reflect domestic market share.⁸ Moreover, the number of mergers under the Hart-Rodino reporting requirement was lower in the 2000s (1,524 per year) than in the prior two decades (2,881 and 2,246, respectively).⁹

It’s also important to use the right measures. The Obama Council of Economic Advisers warned that “the majority of industries have seen increases in the revenue share enjoyed by the 50 largest firms between 1997 and 2012.”¹⁰ But the C50 ratio (the amount of any particular market captured by the largest fifty firms) is largely meaningless from an anti-trust perspective. If all fifty firms in an industry had equal market share, they would hold just 2 percent. Or the increase could have come from firms with relatively small market share (firms forty-one to fifty) gaining market share from firms fifty-one to one hundred. Moreover, the CEA looked at two-digit industries (e.g., the broadest classification of different industries, such as wholesale trade, finance and insurance, etc.) which are too broad to represent real markets where market power can be exercised. As Carl Shapiro, U.C. Berkeley economist and former member of the Obama Council of Economic Advisers, states, “I don’t know any Industrial Organization economist who thinks that’s very informative regarding market power. At some broad level, larger firms are having a larger share of economic activity—I think that’s true, but that doesn’t directly tell us about competition or concentration in markets where market power can be exercised.”¹¹

To really understand market power and competition it’s important to look beyond two and even three digit industries and beyond C50 and even C20 ratios. A more relevant indicator is change in the C4 and C8 ratios at the six-digit industry level. From 2002 to 2012, 59 percent of 792 six-digit-level industries (e.g., flour milling) saw an increase in the C4 ratio, 63 percent in the C8 ratio, and two-thirds in the C20 and C50 ratios.¹² This seems to support the neo-Brandeisians. But on closer look it’s not so clear.

First, at the three-digit level, many industries saw no change or even a decline: accommodation and food services (0 percent); arts, entertainment and recreation (0 percent); real estate and rental and leasing (−6 percent); and wholesale trade (−25 percent). Within the twenty-one three-digit manufacturing industries, almost half saw either no increase or a decrease. For example, despite some high-profile mergers, the beverage and tobacco manufacturing industry saw a decrease in the C4 ratio of 14 percent.¹³

Second, it’s important to look not just at the direction of change but also at the absolute concentration levels. If industries with low C4 and C8 ratios get slightly more concentrated, that is not a problem. For example, the C4 ratio for the administrative and support and waste management and remediation services industry increased 32 percent between 2002 and 2012. But the share held by the largest four firms increased from just 6 percent to 8 percent, hardly evidence of monopoly power. Likewise, retail C4 increased by 23 percent, but from just 11 percent to 14 percent. In fact, the majority of the C4 increases were in industries that were relatively unconcentrated. Just 16 percent of the industries that saw a rise in the C4 ratio had a C4 ratio higher than 40 percent in 2002, and just 19 percent had a C8 ratio of more than 50 percent.

In addition, increases in concentration often benefit consumers. For example, the fifty largest retail firms increased their market share by 11.2 percentage points. But this meant that more Americans were shopping at efficient, lower priced retailers. As Ryan Decker and coauthors write,

Most of the labor productivity growth in this sector has been attributed to net entry. In many cases, existing firms improve productivity in retail trade primarily through adding new, more-productive retail locations rather than expanding existing establishments. Moreover, much of the exit of low-productivity retail establishments in the US economy has been dominated by the exit of “mom and pop” single-establishment firms.¹⁴

We see a similar picture in other industries. Between 2002 and 2012, about half of manufacturing industries got more concentrated and half less. But even the concern with the ones that got more concentrated overstates the problem because the issue is what levels of concentration were reached. Between 2002 and 2012, of 86 four-digit manufacturing industries (such as communications equipment manufacturing), 33 saw declines in their C4 ratio, while the average firm saw its C4 ratio increase by just 8 percent, to 30 percent.¹⁵ Moreover, sixty-one either had a four-firm concentration of 30 percent or less in 2012 or the change from 2002 to 2012 was negative. Thirty-eight saw declines in the eight-firm concentration ratio, with the average ratio going up just 6 percent. But these modest increases in concentration were likely pro-growth as a number of studies find that plants acquired by other firms in the same manufacturing industry experience above-average productivity gains after they are purchased.¹⁶ Robert McGuckin and Sang Nguyin found that plants that were acquired increased their productivity, writing: “acquisition suggests that synergy is a dominant motive for takeovers during the period under study.”¹⁷

The neo-Brandeisian case is in part that concentration leads to higher profits, something the scholarly literature generally finds to be true. One review found that although some studies found that firms with higher market share were less profitable, most studies found they were more profitable.¹⁸ The article estimated that a 1 percent increase in market share led on average to a 0.14 percent increase in profits measured as return on sales. But the key question is whether these profits come from oligopolistic rent seeking or from superior productivity and performance. If it’s the latter, then the firms are simply being rewarded with greater sales for superior performance that benefits the economy.

Importantly, the review finds that increases in industry concentration do not lead to an increase in pricing power. The authors write, “The meta-analysis also fails to support market power theory. Price and industry concentration (which can be proxies for market power) do not moderate the impact of market share on profits.”¹⁹ In other words, firms with greater market share enjoy higher profits but this does not seem to be a result of market power. For example, a firm that has 3 percent of a market would be more profitable if it had a 6 percent share, even though at 6 percent it would normally still have no market power. The authors suggest that one reason increased size leads to higher profits is because size makes it easier for firms to boost product quality. Moreover, multiple studies have found that mergers increase the combined firm productivity. For example, a report by the US Bureau of Labor Statistics noted: “Mergers are found to have a positive impact upon TFP [total factor productivity] growth, accounting for 0.36 percentage points of total factor productivity growth between census years.”²⁰

One recent study that has been receiving considerable attention is from economists Jan De Loecker and Jan Eeckhout who attempt to measure the price levels firms charge above the marginal cost of producing each good or service.²¹ They found that this markup has gone up over the last 37 years from 17 percent to 67 percent and postulate that much of this increase is owing to increases in market power. But a closer look suggests that something other than market power is at work here. First, they compare markups with overall growth of the market value of firms in the United States, assuming that reflects total net present value of profits. The total amount of profits is not the right measure, however: the rate of profit is. As we note, the rate of corporate profits is essentially the same today as it was in the 1960s when markups were much lower. If markups increased almost four times during this period, surely the rate of corporate profits would have gone up more than a few percent.

Second, they find that markups tend to be higher in smaller firms, which, by definition, have less market power.²² Third, they find similar markup patterns among industries that have dramatically different concentration ratios. Markups went up considerably in industries such as agriculture, real estate, and arts, entertainment and recreation, all of which have extremely low concentration ratios (the C4 ratio in real estate is just 6.1 percent, 5.4 percent in arts and entertainment, and even lower in agriculture). If market power were really driving this change, markups should not have increased very much, if at all, in these unconcentrated industries. Finally, they release data a few individual firms. It is striking to note that Apple, the most profitable firm in the world, had lower, not higher, markups in 2014 than it did in 1980, while Walmart, the world’s largest retailer, had essentially the same levels. While the markup for General Electric increased from 1.45 in 1990 to 1.71 in 2014, its operating margin fell from 22.3 percent to 14.1 percent.²³ A more logical explanation for this finding of increased markups, leaving aside the possibility of faulty methodology, is that the ratio of fixed costs to marginal costs has increased in most industries, particularly as investments in intangible capital (e.g., marketing, software, R&D) have increased significantly.

If the neo-Brandeisians are right that concentration is increasing and enabling more firms to increase profits through market power, then profit rates should be growing faster for large companies than for small since on average large firms have greater market share and potentially market power. In fact, from 1994 to 2013 profits (defined as net income as a share of total receipts) grew 5 percent faster for firms with less than $5 million in revenue than for firms with more than $5 million in revenue and 10 percent faster than for the largest firms, with incomes of more than $50 million.²⁴ In 2013, corporations with receipts of less than $500,000 enjoyed net income as a share of receipts of 7.1 percent, while the largest corporations, those with receipts of $250 million or more, had a net income of just 6.8 percent. When just the largest corporations, those with more than $250 million in sales, are compared to all the rest, their profits were only slightly higher, 6.8 percent versus 5.6 percent.²⁵

But maybe the problem is just with a small share of large firms. Abernathy and coworkers claim that “in 2014 the rate of returns for corporations in the top 10th percentile was five times that of median firms; in 1990, the ratio was two to one. In theory, innovation or improved productivity could be the cause—but the companies capturing greater profits tend to be older, suggesting that the culprit may be a monopoly advantage.”²⁶ In fact, this divergence is likely to be related more to differential growth in productivity than to growth in market power. The OECD found that the most of the leading firms in various industries have continued robust productivity growth since 2000, while the other 90 percent of firms have seen lagging productivity growth.²⁷ As the OECD report notes, “A striking fact to emerge is that the productivity growth of the globally most productive firms remained robust in the 21st century but the gap between those high productivity firms and the rest has risen.”²⁸ A study conducted under the aegis of the McKinsey Global Institute reports a similar phenomenon.²⁹ Higher productivity in these leading firms naturally translates into higher profitability. Given this large divergence in productivity performance the real question is why haven’t concentration ratios grown even more as the less productive firms lose market share to the global leaders? Clearly this would often be a positive outcome as it would mean higher global productivity.

When it comes to particular industries, the Brandeisian case that higher concentration leads to higher prices often simply falls apart. For example, the industry that has shown one of the highest increases in prices, health care and assistance, saw declining industry concentration over the fifteen years of 1997 to 2012, with the share held by the largest fifty firms declining 1.6 percentage points.³⁰ Conversely, retail pharmacy concentration has increased while profits have fallen. Yet this doesn’t stop Brandeisians from holding up the retail pharmacy industry as an example of the negative effects of concentration. The Institute for Local Self-Reliance praises a North Dakota law essentially banning big pharmacy chain stores, asserting this leads to better consumer outcomes.³¹ But if the outcomes are really better, why is there a need for a law banning large pharmacies? Wouldn’t consumers just naturally choose the smaller ones? Moreover, if concentration drives up profits, why has the retail pharmacy industry’s profitability (i.e., industry average return on assets) dropped 50 percent since the early 1980?³² Moreover, pharmacy industry productivity grew faster than US productivity growth from 2000 to 2009 than from 1987 to 2000, when the annual value of mergers and acquisitions in the industry was six times higher in the earlier period, setting up the industry for a period of robust productivity growth as larger and more efficient chains gained market share from less efficient smaller companies.³³

Lynn and Longman complain that Americans face a beer duopoly, writing that “more than 80 percent of all beers in America—are controlled by two companies, Anheuser-Busch Inbev and MillerCoors.”³⁴ Any visit to a liquor store makes it clear that the choice of beers in America has never been higher, even more than in the last half of the 1800s before the emergence of national brewers. The explosion of independent microbreweries, over 1,500 in 2016, gives the lie to the notion of lack of consumer choice.

Robert Reich asserts that, “Antitrust laws have been relaxed for corporations with significant market power, such as big food companies, cable companies facing little or no broadband competition, big airlines, and the largest Wall Street banks. As a result, Americans pay more for broadband Internet, food, airline tickets, and banking services than the citizens of any other advanced nation.”³⁵

Many readers of Reich might believe his claims. But a cursory examination of the evidence shows he is wrong. US broadband prices are higher than those of some nations but lower than those of at least eight other OECD nations, including the Netherlands and France;³⁶ no mean feat, insofar as the United States is the second least densely populated nation, which makes deploying broadband wires a lot more expansive than in densely populated nations such as South Korea and Japan.³⁷ The United States does not even make the list of the top ten countries with the highest cost of food.³⁸ In part because of the restrictions on large farms in France, the average French consumer pays $336 a month for food, compared to $267 in the United States. Of seventy-five nations, the United States had the seventh cheapest air travel, behind mostly developing nations such as India and Algeria, which have low labor costs.³⁹ And when it comes to banking, the consulting firm CapGemini finds that average prices in the United States for core banking services are lower than the global average while another study finds that of 11 major developed nations, that US banking costs for consumers are second lowest.⁴⁰

Barry Lynn also argues that big business and mergers lead to higher prices and presumably higher profits.⁴¹ He criticizes a litany of mergers, including Hertz car rental (merging with Dollar and Thrifty) and Safeway (merging with Albertsons), Kraft Foods (purchasing Nabisco), and Procter & Gamble (purchasing Gillette). If he is right that these mergers gave these firms market power, it would seem logical that their profit rates must be exorbitant. In fact, in 2015 their net profit margins were actually lower than the Dow Jones average. The Dow average profit was 9.6 percent, but Hertz (6.6 percent), Safeway (1.6 percent), Kraft (5.7 percent), and P&G (9.2 percent) were all lower.

The Case for Concentration

The neo-Brandeisian argument has morphed into absurdity when even the Economist writes, “Slower growth encourages companies to buy their rivals and squeeze out costs.”⁴² But squeezing out costs is usually about raising productivity, the key to higher prosperity. Indeed, there is a strong case for even more concentration, at least in some industries, because capital-intensive industries, high-wage industries, and industries that do a lot of R&D all have larger firms.⁴³ Yet neo-Brandeisians ignore important structural differences that can influence competition, including scale economies, network effects, innovation, and global market competition.

Scale Industries

In some industries, firms are big because of economies of scale. Yet, neo-Brandeisians go out of their way to deny the very existence of scale economies because they know that this reality, more than any other, undercuts their claim that breaking up big companies would be good for the economy. Matt Stoller, reflects that view when he tweets, “I’m increasingly convinced the biggest con in business history is the notion of ‘economies of scale.”⁴⁴

Stoller is implying that hundreds if not thousands of economists, operations management scholars, and economic historians are not only inept, but intentionally fraudulent in finding economies of scale in production. Yet, the evidence is overwhelming that Stoller is wrong. In one of the earliest studies on this, Perspectives on Experience, the Boston Consulting Group found that “costs appear to go down on value added at about 20 to 30% every time product experience doubles.”⁴⁵ This means that “fragmentation of production among many competitors places an extremely high penalty on consumers.”⁴⁶ More currently, the Obama Council of Economic Advisers’ issue brief, “Benefits of Competition and Indicators of Market Power,” acknowledges scale efficiencies as one reason for a possible increase in concentration.⁴⁷ If marginal costs go down the larger a firm gets, it becomes efficient for the firm to grow in size.

The US government has long recognized the existence and importance of scale economies. A 1980 FTC report on the electric light industry concluded that production technology:

would preclude an atomistic industry. A minimum efficient high volume plant would produce from 7 to over 60 percent of the total out-put, depending upon the type of lamp produced. Therefore, one would have to expect four-firm concentration levels ranging from 28 to 100 percent at the product level at least for the products in the widest use.⁴⁸

The report goes on to note that “the importance of plant scale economies precludes the possibility of breaking up the industry into an atomistically competitive market. To produce at minimum cost, only a few firms could operate.”⁴⁹ In other words if the most efficient electric light factory has to produce at least 10,000,000 bulbs a year, then a fragmented and competitive market composed of firms producing 500,000 bulbs a year each would lead to higher costs and prices. More recent studies have also found scale efficiencies in a number of industries. One study found that most plant and firm acquisitions increase productivity, concluding “the market for corporate assets facilitates the redeployment of assets from firms with a lower ability to exploit them to firms with higher ability.”⁵⁰

The banking industry is an example of an industry that benefits from scale economies. Yet neo-Brandeisians will have none of this. Barry Lynn warns, “The total number of banks in America has fallen by some 60 percent since 1981, even as the population has grown substantially.”⁵¹ In fact, despite the supposed the rise of banks “too big to fail,” the C4 and C8 concentration ratios for commercial banks actually fell between 2002 and 2012, from 29.5 to 25.6 and from 41.0 to 35.8, respectively.⁵² Moreover, if one were to create a US banking system from scratch, it wouldn’t look like the current one, with its more than 5,000 banks (though that number is down from over 12,000 in 1980.)⁵³ As discussed in chapter 2, this massive number was a reflection of local bank protectionism, which led states to erect unit banking laws barring banks from opening branches across state lines. As the states relaxed these archaic laws in the 1980s and Congress passed legislation in 1994 eliminating most of these restrictions, smaller banks were bought up by larger, more efficient ones to, take advantage of economies of scale, as the Federal Reserve Bank shows.⁵⁴

But because most other nations never had American-style unit banking laws, they have always had significantly fewer banks per capita. In 1998 Japan had just 170 banks, or one bank for every 747,000 people. Canada, widely viewed as having the safest banking system in the world, had one bank for every 1.16 million residents. The United States has one bank for every 58,000 people. And in 1999 the share of deposits and assets of the five largest US banks was just 27 percent, compared to 77 percent in Canada, 70 percent in France, and 57.8 percent in Switzerland.⁵⁵

But even with the number of banks falling by more than half in the last few decades bank economies of scale have still not been exhausted and the United States still suffers from too many banks. As the Federal Reserve has found, even the largest banks face increasing returns to scale in terms of cost, meaning as they get larger, their cost per customer and dollar deposited goes down.⁵⁶ The Fed found, “Our results suggest that capping banks’ size would incur opportunity costs in terms of foregone advantages from IRS [increasing returns to scale] in terms of cost.”⁵⁷ Other studies have found similar results.⁵⁸

Even if neo-Brandeisians were to acknowledge that fewer banks would mean higher productivity, they argue this would mean more restrictive lending policies with big banks being less likely to lend to local, small businesses. But as one study of bank size and lending found, “The declining trend in the importance of small banks has, if anything, increased overall bank lending rather than reduced it.”⁵⁹ Likewise, another study found that “greater market share by [large, multimarket banks] is associated with increased competition in small business lending.”⁶⁰

Retail also benefits from economies of scale. As one study finds, “much of the increased competitive pressure on small retailers is due to the fact that growing chains face decreasing marginal cost.”⁶¹ In other words, when a large store gets larger its costs go down because of scale economies. No wonder small retailers are losing market share; they are less efficient and stock fewer products. Because of its size, Amazon can afford to deploy highly automated fulfillment centers with increasing levels of robotization. Walmart can afford one of the world’s most sophisticated software systems for inventory management. And consumers benefit.

Moreover, advances in information technology (IT) are expanding the number of industries that benefit from scale economies. Ten years ago few would have thought that the taxi industry exhibited anything more than modest economies of scale, insofar as most taxi companies were small and local. But the emergence of companies like Uber and Lyft, empowered by software and GPS-enabled smart phones, enabled at least some of the functions in the taxi industry, such as hailing and payment, to benefit from scale economies.

Innovation Industries

Concentration is also a driver of consumer welfare in industries that depend on innovation: regularly bringing to the market new products, services, or business models. Because marginal costs are significantly below average costs in innovation industries, many firms tend to be big and many industries concentrated. In the software industry, for example, it can cost hundreds of millions of dollars to produce the first copy but nothing to produce additional copies.

Firms producing innovative physical products can have also declining marginal costs. For example, it took Boeing almost eight years of development work and an expenditure of more than $15 billion before a single 787 Dreamliner, the first carbon fiber jet airplane, was sold.⁶² That $15 billion has to be built into the overhead of every 787 sale. If the market were curtailed for the 787, for example, through unfair government subsidies going to Europe’s Airbus and China’s Comac, then Boeing would be less able to invest in the next innovative jet airplane. Economists refer to this as increasing returns to scale. While virtually all high-tech industries have this, most low-tech industries do not. A study of more than 1,000 European companies found increasing returns to scale for high-tech firms but, past a certain size, decreasing returns to scale for low-tech ones.⁶³ This means that in innovation industries, increased firm size and industry concentration mean lower industry-wide costs. Having ten, or frankly even three, aviation firms each investing $15 billion to develop a 787-like jet would be a waste of societal resources, as would having ten firms producing PC operating systems software since all would have to invest considerable amounts in software programming but each would have on average one tenth of the sales compared to just one firm.

Increased firm size and industry concentration also make higher profits possible, in part because sales are higher relative to fixed costs than they would be with many more competitors. But rather than being anti-consumer, these higher returns are a boon to consumers because most innovation companies have to reinvest these profits into the next round of risky innovation if they are stay alive;⁶⁴ the next 787 plane, or the next version of the operating system. This is why William Baumol, a leading scholar of innovation economics, writes: “In markets without too much difficulty of entry, an increase in concentration in the longer run may not be ascribable to attempts by firms to achieve monopoly power but, rather, to innovation and the resulting technological changes that make it efficient for output to be provided by firms that are larger than previously was the case.”⁶⁵

Neo-Brandeisians regularly see the large market shares and high profits of some technology-based firms as evidence of monopolistic exploitation. But as Baumol points out, “Prices above marginal costs and price discrimination become the norm rather than the exception because … without such deviations from behavior in the perfectly competitive model, innovation outlays and other unavoidable and repeated sunk outlays cannot be recouped.”⁶⁶ Indeed, numerous studies of innovation industries have found that increased sales means more R&D.⁶⁷ A study of European firms found that for high-tech firms, “Their capacity for increasing the level of technological knowledge over time is dependent on their size: the larger the R&D investor, the higher its rate of technical progress.”⁶⁸

We see this same dynamic in the pharmaceutical industry. As the former Congressional Office of Technology concluded, “Pharmaceutical R&D is a risky investment; therefore, high financial returns are necessary to induce companies to invest in researching new chemical entities.”⁶⁹ Likewise, the Harvard economist F. M. Scherer writes, “Had the returns to pharmaceutical R&D investment not been attractive, it seems implausible that drug-makers would have expanded their R&D so much more rapidly than their industrial peers.”⁷⁰ This is why the Organisation for Economic Co-operation and Development (OECD) writes that, “There exists a high degree of correlation between pharmaceutical sales revenues and R&D expenditures.”⁷¹

But even though higher profits are the very source of continued innovation, at least for the firms lucky enough to be successful innovators, (the unlucky ones have negative profits) neo-Brandeisians worry about innovation monopolies. But firms in innovation industries are more likely to compete through innovation, making market leadership highly contestable. In other words, firms pursue innovation not just to gain a small share of a stable market but to fundamentally disrupt it: the process of creative destruction. As Joseph Farrell and Michael Katz write, “In network markets subject to technological progress, competition may take the form of a succession of ‘temporary monopolists’ who displace one another through innovation. Such competition is often called Schumpeterian rivalry.”⁷² As Schumpeter wrote:

As soon as quality competition and sales effort are admitted into the sacred precincts of theory, the price variable is ousted from its dominant position. … But in capitalist reality as distinguished from its textbook picture, it is not that kind of competition which counts but the competition from the new commodity, the new technology … competition which commands a decisive cost or quality advantage and which strikes not at the margins of the profits and the outputs of the existing firms but at their foundations and their very lives. This kind of competition is as much more effective than the other as a bombardment is in comparison with forcing a door.⁷³

Schumpeter could very well have had in mind the 1932 Supreme Court case New State Ice Co. v. Liebmann. The case revolved around the fact that Oklahoma passed a statute requiring that ice producers be licensed as public utilities. The court rightly overturned the law. But Brandeis defended the statue, seeing ice as a social necessity and writing that the “business of supplying to others, for compensation, any article or service whatsoever may become a matter of public concern.” But the development and spread of the refrigerator (with built-in freezers) quickly made this issue moot, as refrigerators went from less than 15 percent of American homes in 1932 to over 80 percent by the early 1950s.⁷⁴

This is why competition in innovation industries is usually more about innovation than about price. According to Baumol:

Oligopolistic competition among large, high-tech, business firms, with innovation as a prime competitive weapon, ensures continued innovative activities, and very plausibly, their growth. In this market form, in which a few giant firms dominate a particular market, innovation has replaced price as the name of the game in a number of important industries. The computer industry is only the most obvious example, whose new and improved models appear constantly, each manufacturer battling to stay ahead of its rivals.⁷⁵

Network Industries

A third kind of industry in which higher concentration often increases welfare are network-based industries. These are also industries in which fixed costs are high but also where the value of investments increases as the size of the network increases. Some examples are transportation (e.g., air travel, railroads), utilities (e.g., electricity, gas, water), and information (e.g., broadband and Internet applications).

Despite the benefits of scale in network industries, neo-Brandeisians decry increased firm size in them. As Abernathy and coworkers write, “Perhaps most alarming is the tech sector, where a combination of network effects, outdated laws, and permissive regulation has enabled a handful of companies to consolidate vast control over key internet services.”⁷⁶

But increased concentration in network industries usually works in the direction of innovation and consumer welfare. A case in point is residential broadband, where most US consumers can choose from at least two wireline competitors, cable and telephone company (as well as a satellite provider and multiple wireless providers, though at least with current technology, with either higher prices or lower quality). Neo-Brandeisians decry this duopoly and advocate for more competition, including having government fund and operate the construction of a third wireline network. But using government policy to add more competitors would make consumers and the economy worse off. This is because total costs would by definition increase as three separate networks would have to be built and operated, but each network owner would now capture on average at most one-third of the customers instead of half. So costs would go up and revenue down, making it all but certain that prices would rise even if profits (which are already around the US business average levels) went down. Unfortunately, in part because of federal “universal service” policies that subsidize small inefficient telecommunication and broadband providers, many of these companies are small and have not exhausted scale economy potential. In 2012, there were 3,520 wired telecommunications carriers in the United States, with average employment of just 215 workers.⁷⁷

Another network industry neo-Brandeisians criticize is airlines. Barry Lynn and Phillip Longman write, “In the late 1970s, the Carter administration repealed this body of law, in the name of ‘deregulation.’ In the years since, airlines have been allowed to consolidate to a degree unknown even to the railroad barons. Today four super-carriers control 80 percent of traffic, and enjoy outright monopoly on many routes.”⁷⁸ But this overlooks the fact that this more concentrated industry structure provides real value to consumers in the form of more direct flights and better connections. Indeed, the four major domestic airlines coordinate their network of flights to best manage connections. In addition, airline prices increased only about half as fast as the rate of inflation from 1995 to 2015, 29 percent compared to 55 percent.⁷⁹ In addition, airline consolidation has had major positive impacts on airline productivity. Between 1997 and 2014, private, nonfarm multifactor productivity increased 19 percent. But airline productivity, in an era of mergers, increased a staggering 74 percent.⁸⁰

In this case as in so many others, the Brandeisians’ main objection relates to equity rather than efficiency. As Lynn and Longman write, now airlines “discriminate among people who live in different cities, cutting service and hiking fares to places like St. Louis, Memphis, and Minneapolis, in ways that make it harder to attract and keep business.”⁸¹ But in fact, all three of the cities saw airline ticket price increases at or below the national average. (To be fair, since 1995 the average price of airline tickets has increased faster on some routes than others. Fares to or from places like Hawaii and Alaska more than doubled, while fares to or from places like Denver, Milwaukee, and Richmond, Virginia, grew by less than 10 percent, even as inflation increased 55 percent. With market-based pricing and routes and larger airlines there have been winners and losers, but mostly winners.⁸²

Railroads are another industry with network effects. Building two rail lines to and from the same place would be a significant waste of societal resources. And when government tried to force the industry into a particular structure while also price regulating it, the US rail industry almost went bankrupt. It was not until the Staggers Act of 1980, when Congress deregulated the industry, that it became healthy. For example, in the almost ten years after the act was passed, industry multifactor productivity more than tripled compared to the period before the act.⁸³ And as Craig Pirrong writes, “Some segments of the rail market have likely seen increased market power, but most segments are subject to competition from non-rail transport (e.g., trucking, ocean shipping, or even pipelines that permit natural gas to compete with coal).”⁸⁴

We also see industries characterized by network effects on the user side, where the benefit from the product or service is magnified if more people use it. As an Obama administration Council of Economic Advisers’ report notes, “Some newer technology markets are also characterized by network effects, with large positive spillovers from having many consumers use the same product. Markets in which network effects are important, such as social media sites, may come to be dominated by one firm.”⁸⁵

A good way to think about this is whether people would really like it if the government broke up a company like Facebook, splitting it into two companies, Facebook and Headbook. Half of your friends would be on Facebook and the other half on Headbook. So every time you wanted to post a picture of your kid’s birthday party you would have to do it twice. In other words, there is a reason why there is one major social networking firm (Facebook), one microblogging site (Twitter), one major professional networking site (Linkedin), and so on: consumers get much more value by being able to communicate efficiently with a lot of people. Besides, we shouldn’t really worry about current concentration levels in Internet-based network industries that provide their services for free because the relevant market from a competition perspective is not the social network or microblog network, it’s the advertising market. All these firms compete for advertising dollars and, notwithstanding their size, have little market power in the ad market.

We see the same dynamic in software. Neo-Brandeisians wanted the federal government to break up Microsoft after the Justice Department brought suit against the company in 1998 and claimed that by separating the operating system business from the “Office” market, competitive operating systems would emerge. But breaking up Microsoft would have had negative value for consumers. Who would want two different operating systems, making it difficult for users to share files between people or organizations using different systems? Indeed, the very reason Microsoft was not broken up was because the benefits of network effects for operating system and application programs (e.g., word processing) were clear. This is not to say that competition authorities should not address monopolistic behavior. For example, as part of the settlement with the US government, Microsoft was required to enable competitors in related products (e.g., browsers) to better interface with the Windows operating system. But that is very different from the neo-Brandeisian solution of breaking up digital monopolists and oligopolists.

Industries in Global Competition

A fourth set of industries in which scale improves economic welfare are those facing global competition. Since 1980 global oligopolies have emerged in dozens of industries, usually based in Europe, Japan, and the United States. But as we noted in chapter 2, this is actually the second wave of global consolidation. The first took place in 1918–1939, mainly in the form of international cartels that amounted to 30–40 percent of global trade (about the same as interfirm transfers today). Mergers and cartels at least in increasing returns industries are alternative methods of creating scale economies, although the former is superior. The choice of method reflects national laws. The postwar trend toward mergers rather than cartels reflected the influence of US hostility to cartels and lenience toward mergers on competition policy in Europe and elsewhere. Nevertheless, the tendency toward efficient oligopoly in an industrial economy has been remarkably pronounced, in both eras of industrial age globalization.

Unfortunately, neo-Brandeisians seem oblivious not only to the fact that the United States is now in intense global competition but also to the fact that it is losing this competition, as evidenced by its $500 billion-plus annual trade deficit. The Economist writes, “America in particular has got into the habit of giving the benefit of the doubt to big business. This made some sense in the 1980s and 1990s when giant companies such as General Motors and IBM were being threatened by foreign rivals or domestic upstarts. It is less defensible now that superstar firms are gaining control of entire markets and finding new ways to entrench themselves.”⁸⁶

Really? US firms are no longer threatened by foreign rivals? In the 1980s and 1990s, the US trade deficit averaged 40 percent less as a share of GDP compared to the period 2000–2015 (1.5 percent vs. 3.8 percent). And then the United States did not face competition from state-directed innovation mercantilist powerhouses such as China, where the government subsidizes and protects national champions, while attacking foreign competitors with unfair domestic practices (e.g., forced technology transfer, intellectual property theft, etc.). In this new global environment, firms in many industries need to bulk up to be able to have the resources to compete effectively at global scale.

Surprisingly, it’s the neo-Marxist scholars who present a more accurate picture of competition, recognizing the significant increase in global competition and the concomitant rise of large firms to effectively compete. One article in the socialist Monthly Review states:

The giant corporations that had arisen in the monopoly stage of capitalism operated increasingly as multinational corporations on the plane of the global economy as a whole—to the point that they confronted each other with greater or lesser success in their own domestic markets as well in the global economy. The result was that the direct competitive pressures experienced by corporate giants went up.⁸⁷

It goes on to note, “John Kenneth Galbraith’s world of The New Industrial State, where a relatively small group of corporations ruled imperiously over the market based on their own ‘planning system,’ was clearly impaired.”⁸⁸

Neo-Brandeisians reject this argument. Lynn and Longman write that “the idea entirely ignores all historical evidence. Under the system [Thurman] Arnold pioneered, the American economy prevailed over and ultimately vanquished two rival economic systems, those of National Socialism and, later, Soviet Communism. America became the ‘Arsenal of Democracy’ during World War II even as the Justice Department was busy slapping domestic monopolies with antitrust suits.”⁸⁹ The attempt to attribute American economic success during and after World War II to the legacy of Thurman Arnold’s Antitrust Division is creative but unconvincing. Let’s start with the fact that, as discussed in chapter 10, these neo-Brandeisian policies actually set the stage for the loss of US global competitiveness in a number of industries, including consumer electronics, copiers, and computing.

Second, at that time, far from winning on a competitive playing field, most of the US economic competition was flat on its back, devastated by war in Europe and Asia. And the rigid, state planning system in the Soviet Union could never produce viable competitors. Third, there was relatively little globalization from the 1930s to the 1970s. In the early 1950s exports and imports accounted for just 7 percent of US GDP, compared to 30 percent from 2012 to 2015. Even more important, over the last fifteen years an array of nations, led by China, has embraced an array of aggressive mercantilist policies designed to take global market share away from firms in the United States that go largely unchallenged in the World Trade Organization System.⁹⁰

Take the semiconductor industry, for example, an industry that the United States pioneered and still leads, at least for the present. However, the Chinese government has set a goal to depose the United States from that position and become a world-class player in all major segments of the semiconductor industry by 2030. One Chinese official states that the government intends to have “the visible hand of government join with the invisible hand of the market.”⁹¹ The most visible manifestation of that hand comes in the form of government subsidies, specifically National and Regional IC [integrated circuit] Funds, which have already accrued more than $100 billion in assets, with most of the funds channeled from the government through state-owned enterprises (SOEs) into a “private” equity firm so that China can claim that the funds will support “market-based” transactions in accordance with World Trade Organization principles.

A substantial portion of those funds is being used to acquire foreign competitors in the semiconductor industry; in fact, since June 2014 Chinese entities have made seventeen acquisitions across different levels of the semiconductor industry value chain, with the most notable effort being China’s Tsinghua Unigroup’s $23 billion failed bid for Micron Technologies in July 2015.⁹² In this environment, the size of the US market leader, Intel, is one of America’s few saving graces, for if any firm has the resources to compete with state-backed firms that can sell below cost for many years and that refuse to follow other global trade rules, it is a firm the size of Intel.

State-backed competition occurs not only in the semiconductor industry. China’s 121 biggest SOEs increased their total assets from $360 billion in 2002 to $2.9 trillion in 2010, in part because during the recent financial crisis approximately 85 percent of China’s $1.4 trillion in bank loans went to state companies.⁹³ SOEs account for more than 40 percent of Chinese GDP and 70 percent of China’s offshore foreign direct investment (OFDI) activity.⁹⁴ In fact, total Chinese OFDI stock grew from $4 billion (USD) in 1990 to $298 billion in 2010 to $1.3 trillion in 2016.⁹⁵ As Yasheng Huang notes, China’s OFDI is “state-driven and centralized,” and it is “probably historically unprecedented for the SOEs to invest on such a massive scale.”⁹⁶ Nor is China alone in practicing heavy-handed mercantilism designed to challenge US commercial leadership. Brazil, India, Indonesia, Russia, and many other nations are savvy practitioners of this. In this environment, characterized by flaccid prosecution of trade enforcement and a weak, even nonexistent US national competitiveness strategy, including extremely high corporate tax rates, firm size provides at least one potential defense against both fair and unfair foreign advantages.

The Case for Monopsony

While neo-Brandeisians focus mostly on seller power (monopoly), they also worry about buyer power (monopsony). They worry that big companies will unfairly use their market power to hurt business suppliers. As one neo-Brandeisian, Sabeel Rahman, writes, “Despite its low prices, Wal-Mart, for example, has power as a platform: like Amazon, it can leverage its huge consumer base to pressure producers who want their goods on the shelves.”⁹⁷ Rahman writes that Paul Krugman agrees and states, “Amazon is a different kind of monopoly. It does not extract rents from consumers but rather operates as a monopsony, a company whose buying power allows it to discriminate against suppliers.”⁹⁸ Rahman concurs:

Amazon is a critical hub through which almost any bookseller or buyer must pass. As a result, Amazon can use its position to unfairly discriminate between publishers, wielding its access to its vast user base as a weapon. It did so with Hachette, refusing to accept pre-orders for the publisher’s books because Hachette had demanded the ability to set prices for its e-books.⁹⁹

Amazon was pressuring a publisher, Hachette, so that it, Amazon, could lower prices. It appears that the new call from neo-Brandeisians is: “Side with use and we will make sure that businesses don’t lower the prices of things you buy!” Liberal economist Dean Baker gets it right when he writes, “I also don’t have much sympathy for the publishers and authors who fear declining incomes due to Amazon’s pressure.”¹⁰⁰ To be clear, this is not to say that Amazon could not use its market position to act unfairly, but using it to lower prices for consumers is not one of these cases.

While large buyers may be able to pressure suppliers, the result is usually beneficial because it forces them to become more innovative and competitive. Strong buyers are in a better position to require continued cost cutting and innovation on the part of their suppliers, both of which benefit consumers.

More broadly, the goal of competition policy should not be to protect business suppliers but rather to encourage efficiency and competitiveness and to benefit consumers. Yet the Obama Council of Economic Advisers privileged businesses over consumers when it stated: “If an entrepreneur sells its products to downstream firms rather than to end-users, it would benefit from there being a greater number of downstream firms to which it can sell products.”¹⁰¹ In other words, if the government breaks up big retailers like Walmart so that they have less bargaining power, small business suppliers will get to charge higher prices. The only way this cannot have a negative impact on consumers is if neo-Brandeisians believe that the large retailers have complete market power and pass along none of the cost savings from lower supplier prices to consumers, a proposition for which there is no evidence. Moreover, businesses, especially small ones, want an efficient distribution system, not a system that maximizes the number of distributors, because this keeps prices low, which increases consumer demand.

Small Businesses Can Have Market Power

For those in the tradition of Louis Brandeis, market power comes from size. Only big firms can act in anticompetitive and anticonsumer ways. But in fact, some of the most egregious “monopolists” are small firms that collude either through professional rules and guild-like restrictions or with the help of government. By banding together and limiting competition, these small-business-dominated industries hurt consumers and innovation.

A case in point is the optometry industry, which has a long and checkered history of colluding to keep contact lens customers from being able to buy lenses from other, often cheaper sources, such as Walmart and online seller 1800Contacts. By conventional definition, optometrists don’t possess market power, since most are small, local practices. Moreover, they don’t appear to make supranormal profits that a monopolist might. Nonetheless, the industry has long engaged in anticompetitive behavior to limit the ability of contact lens purchasers to buy contact lenses elsewhere. The industry has this power because it plays a gatekeeper role: customers can’t purchase contact lens without a prescription.¹⁰² The industry has used that power to collectively pressure contact lens manufacturers to not sell lenses to lower-cost distributers (particularly online sellers of lenses), making it clear to the manufacturers not toeing the line that optometrists will refuse to prescribe their brand. The collusion was based on professional norms, repeated in blogs and trade journals and at professional conferences, but it had the same effect as a coordinated boycott.

We see this kind of anticompetitive behavior from many industries dominated by small firms: realtors, wine wholesalers, lawyers, car dealers, and others. Car dealers have protected their profits by getting laws passed making it illegal in all fifty states for car manufacturers to sell cars directly to the consumer. Wine wholesalers benefit from laws they support making it illegal for vineyards to sell their wine directly to liquor stores. Lawyers have fought the provision of legal services software. If neo-Brandeisians really want to go after abuse, they should start here, with these unfair practices that benefit relatively well-off professionals and small business owners, for not only do these practices harm consumers directly, they also artificially limit the market share of larger, more efficient firms.

None of this is to say that there are not cases where market power has led to higher prices and consumer harm. Several studies have found that to be the case with US hospitals, which have undergone a consolidation wave in the wake of the rise of managed care. This is why in December 2015 the US Federal Trade Commission announced that it planned to block the combination of two large Chicago-area hospital systems, Advocate Health Care and NorthShore University Health System.”¹⁰³ But these cases represent the exceptions, and in most cases existing competition authorities provide adequate means for constraining them. The argument of the neo-Brandeisian school that many or most economic problems in the United States stem from rising monopoly—itself alleged to be a result of lax antitrust enforcement since the Reagan years—does not stand up to scrutiny and going down this path would surely lead to negative consequences for US economic competitiveness and growth as well as for consumer welfare and per-capita incomes.

11 Has Big Business Gotten Too Big?