The Tech Monsters

Big Tech is under fire from both left and right. Even as they engross ever larger shares of overall corporate earnings, colossal tech firms like Amazon, Facebook, Google/Alphabet and others are coming under attack from all sides. Many progressives denounce them out of hatred of capitalism or hatred of big business of any kind. Many conservatives claim that social media companies like Facebook, Twitter, and Google-owned YouTube—controlled by liberal Democrats—are using their power to censor conservative and libertarian viewpoints.

Ideology aside, what about the economics? Here it is important to distinguish tech firms that have grown while pursuing their core businesses—Google in the search business, for example—and those that have expanded by branching into unrelated businesses, like Google’s spun-off holding company Alphabet.

Companies can grow on their own or they can grow by merging with other companies. There are three major kinds of mergers: vertical mergers, horizontal mergers, and conglomerate mergers. Vertical mergers incorporate part or all of the supply chain into a single corporate enterprise—say, Ford Motor Company owning the mines and the mills that produce steel for its cars. Horizontal mergers fuse several firms into a single larger firm.

In industries with increasing returns to scale, like manufacturing, or network effects, like telecommunications and infrastructure, large horizontally and/or vertically-integrated firms can be more efficient than small ones. That is why there are not hundreds or thousands of local mom-and-pop passenger jet manufacturers, and why cities do not have dozens of small, local utility grids. If efficient scale is accompanied by undesirable effects on society—say, predatory pricing by a large, efficient electrical utility—then the public policy solution should be regulation, like utility rate regulation, as opposed to breaking up a big municipal grid into lots of tiny ones, whose competition, unless checked, would soon once again produce one or a few giant winners anyway.

Conglomerate mergers are an entirely different matter. A conglomerate merger creates a single firm by combining formerly independent companies in unrelated industries.

The last wave of conglomerate mergers in the United States occurred in the 1960s and 1970s. It was motivated by the desire of companies to boost profits by adding streams of revenue from firms in multiple industries, rather than engaging in the hard work of boosting productivity or growing sales for their core businesses. The results were often grotesque. As Robert D. Atkinson and I observed in Big Is Beautiful: Debunking the Myth of Small Business (2018):

Between 1950 and 1978, Beatrice Foods made 290 acquisitions and W. R. Grace made 163. The latter, originally a chemical company, acquired Hostess Twinkie snack cakes, Mexican restaurants, sports teams, fire extinguisher makers, banks, and western wear makers, among other firms. RCA purchased Random House, a publishing house, Hertz Rent-a-Car company, Banquet Foods, which made frozen meals, a carpet company, Coronet, and a company that made golfing attire.

Writing in the Fordham Law Review in 1979, Michael Pertschuk and Kenneth M. Davidson observed:

Gulf & Western passed its corporate youth as a modest midwestern manufacturing firm with $8 million in annual sales and 500 employees. According to its most recent annual report, Gulf & Western has, over the past 20 years, gained control over 100 companies employing more than 100,000 workers and earning $4 billion in annual sales. Gulf & Western bought 64 advertising pages in Time magazine to boast that it owns Madison Square Garden, grows sugar cane in the Dominican Republic, publishes books under Simon & Schuster, makes cigars, weaves clothing, manufactures pulp, rolls steel, and lends money, to name a few of its businesses.

Pertschuk and Davidson write that Gulf & Western was joined by other conglomerates in the 1960s and ’70s:

Mobil Oil purchased Montgomery Ward, the seventh largest retailing firm in the United States: Kennecott Copper purchased the Carborundum Corporation for more than half a billion dollars; Philip Morris inhaled Miller Brewing; and Sun Oil acquired a controlling interest in Becton, Dickinson, a distinguished maker of fine surgical instruments.

Conglomerates may generate financial benefits for a small number of shareholders and managers, who now siphon off money into their pockets from several different business lines rather than one. But decades of research have failed to show any benefits to the economy from conglomerate mergers, in the form of increased productivity among the constituent units. Even John Kenneth Galbraith, a critic of those who idealize small business in general, told Congress in 1978: “There is not the slightest reason to believe that after being absorbed by the conglomerate, the small enterprise is more innovative, more efficient, more effective, or more profitable than before. If anything, the evidence is in the other direction.”

In the 1970s and ’80s, a reaction against conglomerate mergers set in, with some conglomerates spinning off divisions in unrelated lines of business to concentrate on their core competencies. Today, however, we are witnessing a second wave of conglomerate mergers, initiated this time by tech companies rather than the manufacturing and oil companies which were behind the first conglomerate merger wave half a century ago.

Alphabet, the holding company that includes Google, is the W.R. Grace or Gulf & Western of our time. Alphabet owns YouTube, the world’s largest video-sharing site; a smartphone division with Android and Pixel phones; Waymo, a self-driving car project; Project Wing, a commercial drone delivery service; Google Fiber, a high-speed internet, TV, and phone service that competes with cable companies; Google Cloud, a cloud-computing platform; G Suite, which includes Gmail, Calendar, and Hangouts; Verily, a health care company; Sidewalk Labs, an urban development company; Google Capital, a “growth equity investment fund”; DeepMind, which focuses on artificial intelligence (AI); Project Loon, which seeks to use hot-air balloons to expand global internet access; Jacquard, which makes smart fabric; Soli, which uses radar for touchless gesture control; and Spotlight Stories, which makes virtual reality films, among others.

There is no rational reason why a single firm should be engaged in all of these unrelated lines of work. Economies of scale? No such economies result from pairing Waymo with YouTube.

Economies of scope? There are almost certainly diseconomies of scope. It is impossible to imagine any management team, no matter how brilliant, that has the expertise to effectively oversee a hot-air balloon project, a real estate development enterprise, a health care entity and a maker of smart fabrics.

Are there “synergies” arising from the increasing importance of tech and wireless communications in all sectors? Any such defense of Alphabet and similar conglomerates created by tech companies would be silly. That logic would have justified the acquisition by General Electric, half a century ago, of every manufacturer of every appliance with an electric motor, including automobiles and airplanes, plus all electrical grids in the United States.

A holding company like Alphabet, to be sure, may have lots of spare cash sloshing around that can fund useful innovation. But that argument for conglomeration is not convincing, either. If most of these divisions and projects were free-standing companies, they could be funded directly by investors, without Alphabet’s investors and managers taking their cuts.

Does this mean the government should take a cleaver to every big tech firm? No.

Whether a line of business is related or unrelated to the core competencies of the original corporation is a question that can be settled only on a case-by-case basis. If Amazon is defined as a retail company, then it seems perfectly reasonable for it to acquire grocery store chains like Whole Foods and brick-and-mortar bookstores which, together with package delivery via the mail, are three complementary forms of delivery. The same rationale might justify Amazon’s shipping service and courier service.

But if Amazon is a retailer, then it has no business owning Twitch, a video game streaming site; Kuiper Systems, a satellite company; or Health Navigator, an online health information service.

What about platforms that offer their own product lines, in addition to displaying the wares of others? Retail chain stores like Safeway and CVS have long offered their own in-house brands of paper towels and the like. That practice did not kill off competition.

But Safeway doesn’t own YouTube and CVS doesn’t have its own car company and satellite company. It would seem wiser to err on the side of caution and completely ban internet platforms and social media companies like Amazon, Google, and Facebook from offering their own product lines, in addition to advertising or displaying the goods and services of independent firms.

Today’s wave of conglomerate mergers in tech, then, is as indefensible as the earlier conglomerate fad of 50 years ago. To believe otherwise is to believe that there would be no problem if Amazon, Facebook, Google and the other big tech companies merged to form Acme, Inc.—a single holding company that owned a firm in every industry in the United States.

There is growing interest in a “Glass-Steagall” for the tech sector. Just as the Glass-Steagall Act of 1933 (partly repealed by the Gramm-Leach-Bliley Act of 1999) separated commercial from investment banking, so a Glass-Steagall for tech would force companies to choose between being platforms or producers of goods and services. A Glass-Steagall for the internet era is a good idea.

Also justified would be Justice Department antitrust initiatives that force inefficient, unwieldy tech conglomerates to spin off unrelated businesses and focus on their core competencies. Using antitrust law surgically, to split incoherent conglomerates into individual firms in different business lines, makes much more sense than using antitrust litigation mindlessly and indiscriminately against all big firms.

Horizontal and vertical mergers that increase productivity should be allowed and even encouraged. But conglomerate mergers that merely increase parasitic financial rents for the investors and managers of octopoid holding companies are bad. Our guide in these matters should be the late business analyst Peter Drucker, who, in a September 1998 interview with Fortune, famously observed: “Securities analysts believe that companies make money. Companies make shoes!”