In a court filing on October 9, 2024, the US Department of Justice (DOJ) let it be known that it was considering a break-up of Alphabet, with the addendum that it would also be pushing for the company to share the data it collects across its multiple platforms with competitors. There is many a slip between the cup and the lip, and it is entirely possible that these are threats designed to extract more concessions from the company, but the break-up talk is a continuation of a debate about the power accumulated by big tech companies, in general, and with Microsoft, Amazon, Apple, Alphabet and Meta, in particular, and what should be done about that power. With politicians, economists and lawyers all in the mix, offering widely divergent solutions, I look at the evolution of anti-trust law in the United States, and whether that law can or should be used to counter big tech. In doing so, I will start with the disclosure that I am not a lawyer, and have no desire to be one, but the problem, in this case, may be that there are too many lawyers involved, and too little business sense.
The Law in Spirit and Letter
In the latter part of the nineteenth century, as the United States was transitioning from an emerging market to a global economic power, its growth was powered by three industries - steel, railroads and oil - all requiring large investments in infrastructure. In each one of these businesses, powerful men earned their "robber baron" standing by squashing competition and building dominant companies that aspired for pricing power. In oil, it was John D. Rockefeller, who started Standard Oil and built a sprawling empire across the nation, acquiring other players in the still nascent oil business. With Carnegie Steel as his vehicle, Andrew Carnegie took control of the growing steel market, before selling his business to J.P. Morgan, who took it public as US Steel. In railroads, a network of tycoons controlled swathes of the country, with Cornelius Vanderbilt, Jay Gold and Leland Stanford all playing starring roles, as heroes and villains. Along the way, they created the trust structure, organizations of companies which controlled production and prices, effectively monopolizing the businesses .
As these companies laid waste to competition, exploited labor and overcharged customers, a political and economic backlash ensued, manifesting in the Sherman Anti-trust Act of 1890 and the election of a Teddy Roosevelt, campaigning as a trust buster. The Sherman Act used the constitutional power of Congress to regulate interstate commerce to authorize the federal government to break up the trusts and "restore competition", with the latter words vaguely defined. While the law outlawed "every contract, combination, or conspiracy in restraint of trade," and any "monopolization, attempted monopolization, or conspiracy or combination to monopolize", the Supreme Court added the constraint that the law only forbade competitive restraints that were "unreasonable". That vagueness initially worked against the government, in its enforcement of the act, with the Supreme Court ruling against it in its attempt to break down the American Sugar Refining Company, in 1896, but the kinks were worked out in the next decade. In 1911, President Taft used the act to break up Standard Oil into multiple oil businesses, and the entrails of that breakup can be found in many of the largest oil companies of today.
In 1914, Congress passed the Clayton Act to clarify and augment the Sherman Act, and expanded its reach to cover a whole host of activities that it classified as anti-competitive, including some mergers, predatory pricing and sales ties. It also barred individuals from sitting on boards of competing companies and created the Federal Trade Commission (FTC) as an institution to provide the specifics on what constitutes unfair competition and to work with the Department of Justice, to enforce these rules. In subsequent years, Congress returned to add provisions and modify the act, including the Robinson-Patman Act in 1936, which reinforced the laws against price discrimination, the Celler-Kefauver Act of 1950, which filled in gaps on the merger provisions, and the Hart-Scott-Rodino Act of 1976, which introduced the need for any company planning an acquisition that exceeded a transaction value threshold (reset at regular intervals) to file a pre-merger notification with the Justice Department and to wait at least thirty days before consummating the acquisition.
Enforcement Ebbs and Flows
The effectiveness of laws at dealing with the problems that they purport to solve depends in large part on how they are enforced. In fact, one reason that the Clayton Act created the Federal Trade Commission in 1914 was to enforce the anti-trust laws, and the FTC states its mission as protecting "the public from deceptive or unfair business practices and from unfair methods of competition through law enforcement, advocacy, research and education." In carrying out this mission, the FTC often relies on the Department of Justice (DOJ), where an antitrust division was created specifically for this purpose, in 1919.
Through the history of anti-trust laws in the United States, the enforcement has ebbed and flowed, partly as a result of changing administrations bringing in very different idealogical perspectives on its need, partly in response to Court judgments in its favor or against it, but mostly because of questions about whether the central objective of the laws is to enhance competition or to protect consumers. The divide between enhanced competition and consumers played out in competing viewpoints, with one school, led by Robert Bork, arguing that the original intent of the law is consumer protection, and the other pushing back that the end game of the law is to stop cartels and monopolies, i.e., enhancing competition. That tension continues to underlie much of the debate of the law today, in both political and economic circles, and will come into play if the DOJ pushes ahead trying for a big tech breakup.
It is undeniable that for most of the last few decades, the consumer protection argument has resonated more strongly with courts, and has played out as a restraint on what actions the FTC can take, and how far it can go in its enforcement of antitrust law. It is this context that Joe Biden's choice of Lina Khan as the youngest person to head the FTC was viewed a signal of change in focus, since Ms. Khan's most well-read treatise, Amazon's Antitrust Paradox, written while she was still a student at Yale, argued that the company's increasing power was hurting both competitors and consumers. In that paper, she posited that platform-based companies prioritized growth over profits, using their platform size to decimate competition, and that antitrust laws would have to be retooled to rein in these companies. The central part of her argument is that while Amazon’s consumers benefit in the short term, because of lower prices and better service, they would lose out in the long term because less competition leads to less innovation and fewer choices. While her appointment led many to expect a sea change in antitrust enforcement, the effects have been modest, at least in terms of activity:
That graph, though, does obscure the fact that the government has been more aggressive about challenging high profile mergers, and publicly proclaiming its intent to do so, in others. The results have been mixed, with wins in a few cases coming with losses in several others, with the failure to stop Microsoft's acquisition of Activision representing one of it s highest profile losses. In short, while Ms. Khan's argument for use of antitrust laws to restrain platforms may have found a receptive audience among some legal thinkers and politicians, it has not won over the courts (at least as of now).
The Remedies: Sticks and Stones!
No matter where you fall on the consumer versus competitor protection debate, the remedies available to the government fall into three groups, ranging from its power to stop (require) activity that it believes will stymie (advance) competition to breaking up companies, with the possibility, albeit rarely used, of allowing a company to establish monopoly power, but with pricing power restraints.
1. Operating restraints and changes
The anti-trust laws give the government the power to affect how a company operates by stopping it from acting (by acquiring another company, introducing a new product or entering a new market) or changing its behavior (in terms of pricing it products and operating its business), in the interests of increased competitiveness. In doing so, though, the courts require the government to make the case that the actions that it is stopping or the behavior it is altering are unreasonable and that it meets the "rule-of-reason" threshold, i.e., that there are anticompetitive effects that exceed any pro-competitive effects.
a. Merger Challenges
Corporate mergers in the United States, where the transaction value exceeded $111.3 million in 2023, required the acquiring company to file a pre-merger notification with the Justice department, with consummation of the merger happening only after approval. In its most recent update to requirements on pre-merger notifications, the DOJ expanded its information disclosure requirements to include transaction-related documents from deal teams and more complete information about both the products and services offered by the companies, as well as about corporate governance. As we noted in the last section, the degree to which the government uses it power to challenge mergers has waxed and waned over time, and even if challenged, the last word rests with the courts. In a report that it is required to file under the Hart-Scott-Rodino Act for the 2023 fiscal year, the DOJ listed out the number of merger challenges for the year (16), breaking them down into wins (1), consent agreements (4), ongoing litigation (1) and abandonments/restructured complaints (10). The report also lists out the industries that were targeted the most, in terms of merger challenges:
|
Hart-Scott-Rodino Annual Report for 2023 (DOJ) |
Again, note that notwithstanding Ms. Khan's high profile thesis on the need for antitrust enforcement against technology companies, the bulk of the challenges have been directed at more traditional businesses.
b. Operating Changes
In some settlements, the government extracts concessions from a targeted company that it believes will improve the competitive standing of the business. These can range the spectrum, and I will use some of the 2023 settlements to illustrate:
- Forced divestitures: As part of a settlement allowing a proposed merger of Vistra Corporation to acquire nuclear plants owned by Energy Harbor Corporation, where the FTC raised concerns about less competition and higher energy prices for consumers, Vistra agreed to divest its power plant in Ohio. In its challenge of Intercontinental Exchange's acquisition of Black Knight, it required Blue Knight to divest some of its businesses, as a condition for the merger to go through.
- Product bundling/Pricing: As a condition for allowing Amgen to move forward on its acquisition of Horizon Therapeutics, where the FTC feared that Amgen would use its large drug portfolio to pressure pharmacies to push Horizon's two monopoly products, the FTC secured a consent order where Amgen agreed not to condition any of its product pricing or rebates on whether Horizon drugs were prescribed.
- Corporate governance: In EQT's acquisition of Quantum, the FTC's concern was that as these companies were direct competitors, giving EQT a seat on the board and a large shareholding in Quantum would reduce competition. Consequently, EQT was forced to divest its EQT shares and was prohibited from having a board seat.
In most of these cases, the government used the threat of more extreme punishment to extract concessions from the targeted companies.
c. Pricing Oversight
If it is price fixing by a company that has drawn the attention of the antitrust enforcers, it is possible that the remedies sought will reflect changes in the way a company prices its products and services. In 1996, Archer Daniels Midland (ADM) pleaded guilty to fixing prices for Lysine, an animal feed, in collaboration with Japanese and Korean companies. The company, in addition to paying a large fine and having top executives face jail time, was also required to change its pricing processes. In 2024, the FTC published a warning that the use of algorithms by multiple competitors in the same business, to set prices, can violate antitrust laws, and sued RealPage, a property management software, for allegedly allowing landlords to use its algorithms to drive up rental prices. As AI makes algorithmic pricing more of a norm in other businesses, the FTC will undoubtedly be challenging more businesses on pricing practices.
2. Break ups
The most extreme action that the DOJ can take against a company in response to what it views as anti-competitive behavior is to break up the company. Since their effects on the company in question are so wrenching, they are rarely pursued and even more rarely court-approved, but when they do occur, they are memorable. Here are three that stand out:
The
Standard Oil break up, in 1911, was not just the first big break up in history, but given that it targeted what was then one of the largest companies in the United States, it had major consequences. At the time of the breakup, Standard Oil effectively controlled the entire oil business and it was forced to break itself up into thirty four companies:
The eight major companies that emerged from that breakup have morphed over time, and remain dominant players in the oil business, albeit in modified form. - The other big breakup of the twentieth century happened closer to the end, when AT&T, then the monopoly phone company in the United States, was broken up into a long distance company (AT&T) and seven baby Bells, based upon geography: A few decades later, the business has not only changed dramatically, but it has reconsolidated itself into four ventures, with AT&T and Verizon remaining the biggest players.
- The third breakup, albeit one that did not go through, targeted Microsoft in 2000, where the DOJ sought to break up the company, separating its operating system (Windows) from its application software and browsing businesses (Office and Internet Explorer). The courts initially found in the government's favor, but that ruling was subsequently set aside. Eventually, the company settled, agreeing to share some of its application programming interface with third-party company, but avoided major restructuring.
While each of these breakup (including the potential Microsoft one), got significant attention at the time that they happened, the net effects on competition, consumers and the companies themselves are still being debated, and we will return to examine the trade offs in the next section.
3. Regulated Monopolies
The phone business was still in its nascency, when the Willis Graham Act was passed in 1921, arguing that "(t)here are monopolies which ought to exist in the interest of economy and good service in the public welfare, monopolies which must be promoted instead of being forbidden. The telephone business is one of these. Legitimate consolidation will promote economy. It will promote service. It is foolish to talk about competition in the transmission of intelligence by telephone. It is silly to believe that there can be real competition either in service or in charges… The thing that the American Congress ought to do is to.. regulate those monopolies so as to get reasonable prices and good service for the people…" That act allowed AT&T, then the leading phone company in the United States, to acquire its mostly troubled competitors to create a monopoly, with a catch. That catch was that the company's pricing power would be regulated to deliver a reasonable rate of return for its investors, thus creating the basis for regulated monopolies.
The notion of a natural monopoly was not restricted to just telecommunications, and was used for other utilities, such as water and power, with the only difference being that most of the companies offering those utilities obtained local monopolies rather than national ones. Arguably, the decision delivered benefits for customers, as the services were extended to almost every part o the country, albeit at the cost of innovation. As a side benefit, these regulated monopolies, protected from competition, had the capacity use their surplus funds to support activities that sometimes generated societal benefits, that they would not have in a competitive marketplace. With AT&T, that was the case with with Bell Labs, AT&T's in-house research laboratories, where some of the greatest inventions of the twentieth century were made.
The End Game
I mentioned at the start of this post that I am not a lawyer, and I understand that antitrust is full of shades of gray, where absolutism can lead to poor outcomes. Thus, I do get Robert Bork's point that the ultimate endgame in antitrust law is not promoting competition, for the sake of competition, but only if delivers net benefits to consumers. At the same time, I don't think we can dismiss Lina Khan's arguments that large tech companies, using the networking benefits and access to data from their immense platforms, can obtain monopolistic power that may work against consumer interests in the long term, not only by stymying innovation, but also potential increasing prices for consumers down the road, once they reach dominance.
At the risk of adding to an already complex trade off, I believe that three other factors have to come into play in assessing the right action forward:
- Business economics: The notion that increased competition increases innovation and delivers more consumer surplus is deeply set, at least as taught in basic economics courses, but there are businesses where that is not true. In these businesses, the business may be more efficiently run and customers better served, with fewer competitors, rather than more, and to illustrate, consider two examples. The first is the airline business, an absolute mess, where none of the stakeholders (investors, employees, customers, managers or regulators) feels well served, as we lurch from boom to bust. Forty seven years after the business was deregulated, a strong case can be made that the business will be better served with consolidation and allowing more of the weakest players to fail. It is worth noting that the most activity in the Lina Khan DOJ stint have come against airlines (JetBlue and Spirit, a withdrawn challenge to Alaska and Hawaiian), with consumer protection as the rationale, but with no serious assessment of business viability. The second is the streaming business, where Netflix has broken the entertainment business, but it has not been replaced with a viable business model. In fact, as you sort through a dozen streaming choices, it is quite clear that most of these services cannot subsist on their own, with the only pathway to viable business models being a consolidation into three or four streaming services. Forcing competition in businesses where consolidation is the better path to efficiency will create more unstable businesses, more unhealthy competitors and more unhappy customers, i.e., there will be no winners.
- Investors: Implicit in antitrust law and enforcement is the belief that investors in the errant companies are the beneficiaries of anti-competitive actions, but is that true? In the case of trusts, it was quite clear that by clearing the competition and exploiting their monopoly power, investors in the trusts benefited. There are anticompetitive actions, however, where it can be argued that investors see little in benefits from the actions, in the short or the long term, even though managers may rationalize them as beneficial. Thus, if the argument is that a company is using a cash cow business to subsidize its entry into other businesses, investors and regulators may be on the same side on the question of shutting down that subsidization. Ultimately, anti-trust actions are more likely to find investors as allies, if the company being targeted is mistrusted by investors and has a track record of wasting money on long shots.
- Economy and Markets: It is also worth emphasizing that as government regulators, the antitrust enforcers have to consider how their actions against companies, on antitrust grounds, play out in the nation's economy and its markets. If, by allowing a company or companies to reach a dominant position in the market, you are increasing their competitive advantages against foreign competitors or adding to the aggregate payoff to investing in stocks in markets, should you put those gains at risk by handicapping those companies? It is worth remembering that the Chinese government decided to crack down on its tech giants (Alibaba, Tencent, JD) in 2019, motivated more by control than by any consumer or competitive interests, and in the process not only set them back in the global markets by a significant amount, but hurt the Chinese economy and markets.
If you bring these all into the mix, you will be making the work of antitrust enforcers even more difficult, but you will be considering the effects of your actions more fully:
If your job as an antitrust enforcer is to balance competing interests, and do what is right only if there is a net plus to your action, you should be considering the effects of antitrust activity on all four dimensions. That said, if you have blinders on, and view only one of these dimensions (consumers, competition, company or the economy) as critical, it is entirely possible that the actions you take can have net negative consequences, in sum. Using this framework to assess the AT&T break up in 1981, the break up into seven regional phone companies and a long distance one was initially praised as an action that would promote innovation and new thinking, but history suggests otherwise. The regional phone companies continued to behave like the old Ma Bell, investing little in new technologies, and continuing with the high debt and high dividend policies of the original. Much of the innovation in telecommunications came from outsiders entering the business, and the business itself has reconsolidated suggesting that the economics cannot support a dozen or more players. And just as a bonus, Bell Labs was renamed Lucent Technologies, and after an initial burst of enthusiasm about promise and potential, sank under its contradictions.
The Big Tech Dilemma
This post was precipitated by the Justice department’s targeting of Alphabet, with threats of a break up and requiring the company to share its data. While neither threat has been made explicit, it is worthwhile thinking about how the big tech companies measure on the competitiveness scale, and whether antitrust law can or should be used to cut them down to size. The challenge, as we will see, is that we all agree that big tech has become perhaps too big, but the question of how it got that big has to be answered before we respond to the bigness.
The Rise of Big Tech
Looking at the DOJ's arguments for breaking up Alphabet, it is clear that the same arguments can be used against some of the other big tech companies. In this section, we will look at Alphabet, Amazon, Apple, Microsoft and Meta (bundled together as the Fearsome Five), all of which have been rumored, at times, to be in the crosshairs of antitrust enforcers, and the reason for their targeting, which is that they are all big, perhaps even "too big", and that can be backed up with multiple metrics:
a. Market Capitalization: If the companies that we have listed look like they belong together, it is because they were bundled as the FANGAM stocks in the last decade and as part of the Mag Seven in this one. In each case, that bundling was used to illustrate how dependent the US equity markets have become on just a few stocks, to deliver overall equity returns. In the graph below, we look at the rise of these companies, in terms of market capitalization, since 2010, and how much of the aggregated market cap at all US stocks has come from just these companies:
As you can see, these five companies, in the aggregate, increased their dollar market capitalization from $716 billion at the end of 2009 billion to $12.1 trillion on October 16, 2024, accounting for 23.16% of the increase in market capitalization across all US equities over that period. On October 16, 2024, these five companies accounted for 20.22% of the market capitalization of all 6132 US equities, and in sum, they had a market capitalization that was greater than that of any other equity market in the world.
b. Revenues and Earnings: The rise in market capitalization did not just come from vibe or momentum shifts and was backed up increases in revenues and income over that period that were truly extraordinary, given the scale of these companies:
These companies increased revenues 18.8% a year between 2009 and 2024, while preserving enviable profit margins - gross, operating and net margins stayed relatively stable. In sum, these companies have delivered a combination of revenue growth and operating profitability that is unmatched, given the size of these companies, in history.
c. Day-to-day life: There is a final component on which you can measure how big these companies have become, and that is to look at how much of our time and lives is spent on one or more of their platforms. In a New York Times article from 2020, the writer talked about trying to live without big tech for six weeks, and how difficult she found the consequences to be. During the same year, I chronicled in a post how much time I spent each day on the platforms on one or more of the big tech companies, essentially concluding that I was in their grip for all but fifteen minutes of the day. As a thought experiment, consider what your day at work or at home will look like today, if all five of the Fearsome Five decided to make you persona non grata. Mine would be a grind, with this post not being written (it is on a Google Blog), the graphs not showing up (they are in Microsoft Excel) and my computer not responding (it is a Mac).
In short, I don't there is any debate that the big tech companies have become big on every dimension, and become central players not just in the economy and markets, but in our personal lives. It is therefore no surprise that when Lina Khan and others argue that these companies have become too big, and need to be restrained, they find a receptive audience.
Pathways to Bigness
While, for some, bigness alone is a sin that needs to be punished, the pathways that these companies took to get to where they are now needs to be examined for a simple reason. If those pathways were cleared by legitimate business actions and choices, it would not only be unfair to punish them for their success in foiling competitors and establishing dominance, but it would also make the legal challenge of using antitrust laws to restrain them much more daunting. In this section, we will look at what these companies did (and are doing) that explains their success.
- Core Business Dominance: Looking at the fearsome five (Amazon, Apple, Meta, Alphabet and Microsoft), each one, with the possible exception of Microsoft, has a core business in which it dominates, driving the bulk of it revenues, with Microsoft perhaps being the exception. For Alphabet and Meta, that core business is online advertising, with Apple, it is the iPhone, and Amazon's revenue base is in the retail business. Microsoft's dependence on its software business has waned over the last decade, and while Windows and Office continuing to deliver as cash cows, the company has increasingly become a cloud and business services company.
- Shaky Side Businesses (with a cloud exception): Largely funded by cashflows from their core businesses, the big tech companies have tried to enter new businesses, mostly with little to show for their investments. Alphabet has been most open about its ambitions to be in multiple businesses and its renaming was largely a signal of that intent. Amazon's ambitions to be a disruption machine have been widely documented, with forays into logistics, entertainment and even health care. Apple has been more restrained, but it too has tried its hand at entertainment and other businesses. Meta, after facing market backlash for its badly framed entry into the Metaverse, has retooled itself and is trying for success in AI and virtual reality. For the most part, these side businesses have been cash drains, and added little in value, with one exception. For three of these companies, Amazon, Alphabet and Microsoft, the cloud business has become not only a large part of their revenue base, but also an even bigger contributor to their profitability. With Apple, the services business is offering promise in terms of growth and is a gold mine when it comes to profitability, but it draws much of its value from the iPhone franchise.
- Consumer subsidies: These companies have also created subsidy mechanisms for consumers, offering them products and services that are "free" or "bargains", at least on the surface. Amazon Prime remains one of the best deals in the world for consumers, since for an annual fee of $139, you get free shipping, entertainment and a host of other services. In fact, Amazon makes explicit the cost of the shipping subsidy in its annual reports each year, and it has spent tens of billion each year for the last decade, supporting that service. Alphabet offers a whole range of products, from Google Docs to Google maps, at no explicit cost, and there are hundreds of millions that use WhatsApp around the world, with no monthly charges or fees. Apple and Microsoft, befitting their standing as the elder statesmen in this group, have been more stingy about providing free add ons, but they too have sweeteners that they offer, usually in exchange for data from users.
The question then becomes whether any of this is "unfair", and the answer is debatable. Listening to those most critical of these companies, there are five arguments that I have heard to back up the "uneven playing field" argument:
- Subsidize their product offerings: One of the critiques of tech companies is that they use the massive profits they generate from their businesses, core and cloud, to subsidize their product offerings to customers. By doing so, critics argue, they make it more difficult, if not impossible for competitors, to succeed in these subsidized businesses. That is probably true, but cross product subsidization, by itself, is neither uncommon, nor illegal, and consumers are the beneficiaries.
- Networking benefits: Most of these companies have large platforms, and in the businesses that they operate in, that can work in their favor. In online advertising, Alphabet and Meta have a significant advantage over competitors, because advertisers want to go where people gather, and they are more likely to find that on larger versus smaller platforms. That said, those networking benefits are inherent in online advertising, and punishing the companies that were able to climb the competitive ladder most competently does not seem fair.
- Use of private data: When users spend their time on the tech company platforms, they are providing data to these companies that can be used to their benefit. Staying with the online advertising giants, Google and Meta, is clear that the information that they collect from user interactions on their platform is being used to target advertising better, making them an even more attractive destination for advertisers. While conceding these points, it is worth noting that advertisers should have no complaints about better targeted ads, users share private data voluntarily, in return for conveniences, leaving competitors again as the only complainants.
- Squashing competing technologies: When your platforms become ubiquitous, your competitors might need your permission to play on these platforms, and the big tech companies often make it either more difficult to play or claim a large chunk of revenues. Apple, for instance, has faced pushback because it charges a 30% fee for third-party apps that go through its app platform, and Google has also received criticism for restricting third party app stores on Google play and Android. Here, the argument can be made that in addition to competitors being hurt, consumers are being denied choice and paying higher prices for third party offerings.
- Not paying fair price for content: Many of the big tech platforms allow users to access content for free, with the content developers feeling shortchanged. The big tech companies benefit from this content access, because that access increases platform usage and their revenues (from advertising, device sales etc.), but in a fair system, they should be sharing this revenue with the content developers and providers. It is at the heart of the tussle that is ongoing between media companies (newspapers, magazines) and the big tech companies, and while the former are becoming more savvy, they are operating at a disadvantage.
I am sure that all of these issues will be litigated, but I do think that governments (and antitrust enforcers) are on far stronger ground, on the last two, than on the first three. More generally, if you were to look big tech sins, there are two general conclusions:
- Hurt competitors, subsidize consumers: As you look at the critique of big tech, it is clear that the damage from big tech company behavior has been felt mostly by competitors. In fact, consumers for the most part have benefited from the subsidies that they have received, and if they are aggrieved about the use of the data that they have shared with the companies, it is unclear how much they have been hurt by that sharing.
- Current versus Prospective sins: Extending the first point, even the most severe critics of big tech argue that the costs of allowing them to dominate will be in the future, Lina Khan's criticism of Amazon is that while customers benefit right now from Amazon Prime and other freebies, there will be costs they bear in the future that will outweigh the benefits. In particular, she argues that there will be less choice and innovation, because of Amazon's dominance, and that Amazon will eventually become powerful enough to raise prices, and consumers will have nowhere to go. The problem that Ms. Khan and others in her camp will face is that there is nothing in the company's behavior currently that would lead us to extrapolate to those dire endings.
Ultimately, anti-trust actions are as much about politics as they are about economics, and they work only if they carry public approval. On economic grounds, that is why pushing strong anti-trust actions against big tech will be a much more difficult sell than against other dominant businesses in the past. After all, how do you convince customers that they paying more for Amazon Prime and being charged for Google Maps will make them better off, because there may be more innovation and choice in the futures with more competition?
The Choices
The DOJ court filing suggests that the die has been cast, and that Alphabet will be the target of the anti-trust enforcers in the near future, with success or failure in that endeavor perhaps resulting in expanded action against the other big tech companies. Using the framework from the last section in assessing the costs and benefits to consumers, competitors, investors and the economy, we can evaluate the choices.
1. Break up
Can the government break up Alphabet, just like it did AT&T and Standard Oil, in the last century? It can push for it, but to understand why it will be difficult, and even if plausible, unwise, here are some considerations:
- While you can think of the multiple platforms that Alphabet operates as separate, the truth is that the core business is advertising, and whether you are on the Google search box, YouTube or on Android, that business derives its value from keeping users in the Google ecosystem, rather than on independent platforms. With Facebook, that problem is magnified, since Facebook, Instagram and WhatsApp are all part of the same ecosystem, with the end game keeping you in it. In short, the platforms, separated, would both be unable to survive as stand alone businesses as well as less attractive destinations for users.
- There is an added reason why breaking either Alphabet or Facebook into individual platforms makes no economic sense. Online advertising is a business with networking benefits, and any solution that pushes you away from consolidation, may create more competition, but will worsen business efficiency and health. In fact, assuming that you were able to break both Alphabet and Facebook into individual platforms, it is not clear to me who will benefit. Consumers will no longer have access to their subsidized products, online advertising will be less targeted and effective for advertisers and even the competitors who may be helped in the near term will find those benefits fade quickly.
- As we noted in the last section, the big tech companies have generally not been able to deliver value in their side ventures, with the exception of their cloud businesses, for Alphabet, Amazon and Microsoft, and the services business. You can demand that Alphabet be forced to divest itself of all of it non-ad related bets, but very few of these businesses can stand alone. It is true that the cloud businesses have the capacity to stand alone, but what is the argument that you would use for forcing divestiture? After all, in the three companies that have significant cloud businesses - Alphabet, Microsoft and Amazon, their success in the cloud had little or nothing to do with core business domination and divestitures make it less likely that consumers gets subsidized products, which will make them worse off. In addition, divesting these businesses will do nothing to break the dominance that these companies have in their core business, since that dominance comes from networking benefits and private data. In fact, the only company where an argument can be made for a break up is Apple, where the services business draws its value from the Apple stranglehold on the smartphone business.
Summarizing, breaking up any of the big tech companies risks the worst of all outcomes. It will make the companies (and their investors) worse off, but not by as much as critics think, but it will also have negative effects that ripple across the economy and markets, while making the businesses that they operate in less efficient. Competitors will derive short term benefits from the breakup, but those benefits are unlikely to last, if the business economics still point towards consolidation. Finally, consumers will be left off worse off, in the short term, with only promises of a better tomorrow filling the void.
2. Regulated Monopoly
The second pathway that has been suggested is that the government big tech companies as regulated utilities, just as they did phone, power and other utility companies in the last century. While that would give the government power over how these companies price products and services, and make them less profitable, the flaws in the argument are large and potentially fatal:
- The regulated monopolies of the last century agreed to the pricing restriction quid quo pro because the government gave them monopoly power in the first place. With tech companies, what exactly would the government be offering these companies in return for the loss of pricing power? With Alphabet and Meta, the online advertising market is not the government's to give away, and with smartphone (Apple) and online retail (Amazon), it becomes an even bigger reach.
- If, in fact, the government did get control of pricing power at these companies, who would be the beneficiaries? With online ads, the benefits would flow to the advertisers, a transfer of wealth from one set of companies (the Big Tech advertising companies) to another set of businesses (the many companies that advertise on the tech platforms), and that is neither fair not equitable.
- If the end game is innovation, and with technology, it is the lubricant for success, creating regulated monopolies and requiring them to earn their cost of capital will not only destroy incentives to innovate, but leave these companies exposed to disruptors from other markets.
In short, there is no pathway that works to make any of the big tech companies look like Ma Bell, and even if that pathway existed, how would that benefit consumers, markets or the economy?
3. Targeted changes
Given how much of a reach it would be to break up the big tech companies or bring them under the regulated monopoly umbrella, the pathway, if the government is intent on sending a signal will take the form of constraints on and changes to operating practices. I will start with a list of changes, where I think that the government has a better chance of prevailing, because the laws and public opinion will be on their side:
- Platform access: If you own a platform where users congregate, you cannot make the roadblocks to third parties being on the platform so onerous that they are put at an almost insurmountable disadvantage. I think that Apple and Alphabet will be pushed to make their platforms more accessible (technically and economically) than they are right now.
- Paying for content: As AI looms larger, the fight over content ownership will get more intense, since AI can not only be a monstrously large consumer of content, but can do so with little heed to where the content comes from, or who owns it. Content owners and developed may need an assist from the government, when they fight to reclaim the content that belongs to them.
- Customer and User Recourse: The power dynamics when you use a tech platform are imbalanced, and as a user or customer, you often have no power against the company operating the platform, if it chooses to act against you. As someone who has kept my blog on Google Blogger and my videos on YouTube, there is almost nothing I can do if Alphabet decides to shut them both down, other than appeal to the company and hope to get a fair hearing. Governments may push more formal appeals processes, with independent arbiters, to provide for more balance.
There are three other changes, where the government is less likely to succeed, and deservedly so:
- Share data with competitors: It is possible that the government will try to get tech companies to share the data they collect, but I believe that this is neither fair nor a competitive plus. While having the data gives them an advantage over their competitors, that can be said about competitive advantages in many other businesses, and companies in those businesses are not asked to do the equivalent. Does Coca Cola have to share its syrup makeup with competitors because it has the most valuable brand name in the beverage business? Should Novo Nordisk be asked to share its patent rights for Ozempic and Wegovy with other pharmaceutical companies, because having these rights gives it a leg up in the weight loss business? If your answer is no, why would you use a different set of rules for big tech companies. Of course, if your answer is yes, your problem is not with big tech but with capitalism, and that is an argument for a different time and setting.
- No cross subsidization: It is also possible that the government will take a stand on cross business subsidies, arguing that the money that big tech companies make in one business should not be used to establish advantages in other markets. The problem is that cross subsidization is part of almost every large company, where successful, cash-rich portions of the company subsidize cash-poor portions, perhaps with growth potential. Those subsidies can sometimes hurt shareholders of the company, but it is not the DOJ's job to provide them with protection. In fact, the big tech companies have not been immune from investor backlash, as Meta found out, when it pushed its Metaverse investing plans forward with no clear pathway to monetization.
- Device Compatability: Big tech companies are often criticized for making it difficult for other company devices to play on their platforms. Thus, the Apple platform works much better with Apple devices (iPhones, iPads and Mac computers, Apple iPods) than with Android devices. Much as this may frustrate us, as consumers, no company should be obligated to make it easier for competitors to take business away, and government attempts to suggest otherwise will be heavy handed and ineffective.
4. Do nothing
There is a final option, and it will not be appealing to many anti-trust enforcer who came into their professions wanting to push for change. That is to do nothing! That sounds defeatist, but at least in technology, it may be the best choice, given the following:
- Tech life cycles are short: As many of you may be aware, I believe that companies, like human beings, go through a life cycle, evolving from start-ups (baby) to mature (middle age) to decline (old age). That said, there is also evidence that tech companies age in dog years, scaling up much faster, not lasting at the top as long and declining much more quickly than non-tech companies.
That, in turn, reduces the need for governments to intervene on behalf of competitors or consumers, since tech companies that look unassailable and dominant today can quickly find themselves under threat in a few years. - The Innovation Trade off: As an extension of the first point, if innovation costs money, and life cycles are short, companies have to be allowed to make money during their brief stints at the top, to justify innovation. In short, if you make the lucrative years for a tech company less so, by taking away pricing power and capping profitability, it will reduce the incentive to start and grow new technology companies. I don't think it is coincidence that the EU, where rule makers take a dim view of excess profits and market power, has no great tech companies.
- Disruption is always imminent: To the extent that big tech companies are tempted to play it safe, cutting back on innovation and using their market power to increase prices on customers, i.e., the Lina Khan doomsday scenario, they expose themselves to disruption far more than manufacturing or consumer product companies do. Blackberry's failure to adapt left them exposed to the smartphone disruption, and Yahoo! lost its search engine dominance to Google in the blink of an eye. I would wager that the big tech companies are acutely aware of that threat, and I don't blame them for creating as safety buffers.
You may have guessed already, but I do believe that doing nothing is, in fact, the most sensible option, with big tech companies. Are there risks in adopting this path? Absolutely! The big tech companies may have found ways to extend life cycles and they may buy out disruptive innovation, just to squash it, and we may all be worse off, as a consequence. I have seen no evidence of any of that behavior so far, but that fear remains, and I will remain vigilant.
Conclusion
I do not see eye to eye with Lina Khan, but I will start with the presumption that she has good intentions and that her argument is deeply thought through. My concerns with her big tech views are two fold. The first is that she is a lawyer, and law schools around the world do an awful job on teaching their graduates about business, which is one reason that laws tend to be one-size-fits-all. Just to illustrate, competition is good in some businesses, but consolidation works in others, and a law or lawyer that does not discriminate between the two will do more damage than good. The second is that she is a true believer, and if you start with the view that big tech companies are evil, you will undoubtedly find good reasons to cut them down to size.
I do recognize that there are non-economic considerations at play, and that you may fear the effect that big tech platforms are having on our politics and social discourse. I share that concern, but I am not sure that there is an economic solution to that problem. If you think that breaking up Google and Meta will lead to more polite discourse on social media and a return to the cultural norms of yesteryear, you are being naive, since the problem lies not in Twitter, Facebook or Reddit, but in ourselves insofar as participating on social media seem to bring out the worst in us. I am afraid that we have opened Pandora's box, and there is no shutting it now!
Comments 0