Break Them Up
Why breaking up dominant firms is often the right (but not the only) thing to do.
Welcome to The Counterbalance, the newsletter of the Balanced Economy Project.
** Announcement: we’re supporting an anti-monopoly event for international civil society organisations in London on March 25th. Click here for details.**
Fourteen reasons to break up dominant firms
The Financial Times recently published a long article entitled How Germany became Europe’s leading Big Tech trust buster, lauding the country for its proactive regulation of tech giants. The EU’s incoming European Digital Markets Act (DMA,) the article continues, is potent but rather softer and slower - while recent political turmoil in the UK has held up progress (but things are speeding up again now).
Still, some doubt that even Germany is going remotely far enough.
“We will have to see whether on the basis of behavioural obligations we can tame the gatekeepers and neutralise their competitive weapons,” said Thomas Höppner, a partner at law firm Hausfeld, who spoke at our recent event in Brussels. “If these don’t do the job, the next step would be to break them up.”
Yet regulators around the world haven’t been breaking up dominant firms: more often they have acted like facilitators or even cheerleaders for mergers and acquisitions. As we have noted before, less than 0.5 percent of all mergers notified to the European Commission since 1990 were blocked; the biggest five Tech firms have bought out over 1,000 companies in the past two decades, and until last year no regulator, anywhere, had blocked or unwound any of them.
Meanwhile, European competition laws should protect against "abuse of dominant position" - rather than against "dominant position" - waiting for abuse is like shutting a stable door after a horse's exit. While the incoming DMA is more forward-looking and does envisage breakups - it only allows them "as a last resort." Why?
The backdrop to all this is a pro-monopoly, pro-merger world view that has coalesced inside a dominant technocratic "competition bubble," replete with academics and pundits ready to explain why each merger is good and breakups are bad. This accompanies broad ideological currents that demonise government regulation, and which argues that private actors are efficient so we should let them merge if they want to.
Box: what is monopoly?
We avoid dictionary definitions of monopoly, which tend to talk about a single seller in a market. We are interested in all situations of excessive market power, including when there is more than one seller, or buyer, of goods or services. People use horrible technical terms like "monopsony" or (ugh!) "oligopsony." We prefer, as some regulators do, to use "monopoly" as a blanket term meaning excessive and durable market power.
Section 1: The pundits are (mostly) missing - or ducking - the big points
Benedict Evans, an influential tech commentator with 175,000 newsletter subscribers, wrote a provocative post recently, entitled "Would breaking up ‘big tech’ work? What would?”
"Most of the things we worry about aren't actually competition problems," Evans writes, "but even where they are, breakups are unlikely to be effective."
His and others’ key arguments, simplified, go like this.
First, "we didn’t make cars safer by breaking up GM or Ford, but by writing rules about how you can make a car," as Evans put it. So don’t break them up: focus on Big Tech's bad behaviour and actions.
Second, that big firms are 'efficient,' reaping economies of scale and scope, and that breakups destroy value.
Third, that there are "natural monopolies" (for example, you wouldn't build two competing water supply networks side by side) and "network effects" (for example, you use Facebook because your friends and family use Facebook) - and breakups can't fix this: regulate instead.
Fourth, that waves of Schumpeterian creative destruction will push Big Tech aside to make way for the Next Big Things, so let’s wait for this to take its course.
Fifth, as France's competition boss feebly claimed recently, it is “too late” to overturn the dominant positions of Big Tech firms.
The competition-bubble pundits and academics, many of whom consult for big tech firms, tend to use carefully selected examples, wield straw-man arguments, address narrow questions, and above all tiptoe around the elephant in the room: power. Economic and political power.
This article shows why the pundits are flat wrong.
We should stress, at the outset, that big isn't always bad. A manufacturer of large passenger aircraft, for instance, needs to be far bigger than a bookseller does. Nor are all mergers necessarily harmful: some can create healthy new economic recipes from separate ingredients. If a merger builds excessive concentrated power, then it is a problem.
So we aren't a hammer, for which every problem looks like a nail. We don't want to break everything up. What we want to break is the establishment taboo over breakups, because some are urgently needed, to protect our democracies and economies.
The next section provides 14 reasons why breakups are an essential part of any democratic toolkit. (We’ll develop these in a longer report in due course.)
Section 2: Fourteen reasons
Reason 1: Monopoly is political
Monopoly power is political.
When Hitler became German Chancellor, the Nazis convened industrial leaders to extort donations and consolidate power. U.S. President Franklin Delano Roosevelt observed in 1938:
“the liberty of a democracy is not safe if the people tolerate the growth of private power to a point where it becomes stronger than their democratic state itself.”
This view influenced post-war reconstruction efforts in Europe.
Today, social polarisation is fueled not just by algorithms and fake news, but by a wider sense of powerlessness in the face of corporate power. The competition bubble rarely touches this central question of politics (Evans dismissed it as "handwaving.")
Reason 2: Monopolies are corrupt
In competitive markets, surplus profits get 'competed away,' as surpluses are invested to innovate and stay ahead of competitors, or to cut prices, or to attract employees. Monopolists, by contrast, generate large surplus profits by squeezing suppliers, workers, or consumers - which they are then free to spend on share buybacks, donations to pet causes, and of course lobbying.
So monopolists tend to have vastly more cash for lobbying. What is more, they don’t have to worry about sharing rewards from their “investments” in lobbying with competitors.
So monopolists have both greater resources, and stronger incentives, to lobby and be corrupt. [See also Zephyr Teachout, on why anti-monopoly and anti-corruption are intimately bound together, and why monopoly power not only corrupts, but is corrupt, from first principles.]
Reason 3: Monopolies are conflicted
Monopolies also corrupt markets.
Amazon, for instance, hosts sellers on its platform but also sells its own stuff on the platform, competing with those sellers, while also harvesting data to weaponise against them. It also lets sellers bear the initial risk of developing new products, monitors their performance, then swoops in with its own products to replicate the successful ones and capture the rewards from the risk-takers. It is referee and player, all in one.
How do we deal with such conflicts?
Think of each conflict as a hungry wolf in a children's playground.
You might post a sentry on every swing, and smack Wolfie on the nose if she tries to snatch a kid, or even build fences inside the playground.
But there's a more elegant solution: remove the wolf from the playground. In Amazon’s case, as we argued in a joint recent submission to the European Commission, separate its retail operations from its marketplace. Remove the conflict, for good.
Unlike fines, breakups tackle root causes, not symptoms. Breaking up monopolies "predictably - not just hopefully" - makes regulation work.
Reason 4: Monopoly manipulates and discriminates
Look at the market share of these search engines.
If Google pushes a business from page one to page six of its search ranking, that could destroy it overnight. That gives Google, but not DuckDuckGo or Ecosia, immense power to discriminate against, or punish, whoever it likes, and for whatever reason it likes.
This king-like arbitrary power to discriminate and manipulate is generic to monopoly power, and has been for centuries. Give people real options, and you lift boots off their necks.
Reason 4: Monopoly accelerates, monopoly is contagious
Jeff Bezos saw Amazon as a flywheel: the more buyers flock to Amazon, the more sellers need to be on there, and vice versa. Speed up one node, and you speed all the others.
Monopolists' super-profits (plus superior access to finance, and superior lobbying power) also grease expansion into new sectors, and predatory pricing to kill incumbents there.
So monopoly power expands naturally, automatically. Meanwhile, smaller firms on an increasingly tilted playing field feel they have no choice but to bulk up and merge, defensively.
So monopoly is contagious too. Breakups stop the cycle.
Reason 5: Behaviour is local, breakups are global
Although only nine percent of Facebook users speak English, an estimated 87 percent of Facebook’s spend on combatting misinformation is devoted to English speakers.
If the UK decided to "fix" Meta's behaviour, without breaking it up, that wouldn’t help Tanzania or Brazil.
By contrast, a breakup would have immediate global impact, curbing misinformation everywhere.
And, as demonstrated by the UK’s unwinding of Facebook’s Giphy acquisition, countries have more power to force global breakups than is commonly supposed.
Reason 6: Monopolies are more powerful than the sum of their parts
Successful monopolies are more powerful than the sum of their parts. For a monopolist, that is the point of many acquisitions. They can leverage conflicts, surplus profits, data, and chokepoints to amass power.
So a single merged entity with distorting gravitational power poses a bigger economic (and political) threat than its separated components collectively would.
Reason 7: "Disruption" won't fix it
Pundits always say disruption is just around the corner: Big Tech will be replaced with The Next Big Thing within 15 years - so why bother breaking them up? TikTok is breathing down Facebook’s or Youtube’s necks. And so on.
Such predictions led the US to wave through Google’s 2009 acquisition of its most serious mobile advertising competitor, AdMob, on the mistaken belief that Apple would soon disrupt the market. Oops.
This view kind of implies that anything bad that happens as a result of monopoly power during those 15 years doesn't matter.
But of course, it can matter quite a lot. Perhaps the current crop of Big Tech firms will fade in a decade, but in the meantime they have spurred US insurrection, and boosted the authoritarian forest-killer Jaír Bolsonaro - just for example.
Reason 8: Natural monopolies exist, but needn’t sit inside a single corporate form
Yes, Facebook enjoys 'natural monopoly' network effects. But there is no reason why multiple networks should sit in a single corporate form. Meta, which owns Facebook and Instagram and Whatsapp and much else, is not “a platform”: it is a conglomerate.
Why should Meta own Facebook and Instagram and Whatsapp? Why should the Amazon retail platform be owned by the same firm that sells cloud services? Some parts of these conglomerates don’t even depend on the network effects said to make them inseparable.
Breaking them up can be like cutting a diamond along its natural fault lines.
Reason 9: Monopolies privatise law, regulation and tax
When one newspaper bought its own advertising inventory, it found that up to 70 percent of the money spent leaks out, leaving as little as 30 percent for the publisher. This amounts to a crooked, unaccountable private tax system.
Monopolies sweep interactions between previously separate firms away from accountable public regulation and into their spheres of private regulation (e.g., onerous app store rules), where it is far harder to tackle bad behaviour.
Breakups open up private tax and law systems, and let the sunlight and public accountability back in.
Reason 10: On “alternative” approaches, like nationalisation
Many commentators deploy the "Either-Or Fallacy" as a strawman argument. Either break them up, or do something else.
But nobody is arguing for breakups alone.
The “New Brandeis” anti-monopolists in the United States explain that "network effects" mean that some parts of Big Tech conglomerates are platforms that may be regulated as public utilities. Meta could be split into separate platforms, while separate “inter-operability” rules could let users leave Facebook for other networks, still ensuring they can communicate with friends who stay.
Europe also has a long tradition of nationalisation and public ownership. As Zephyr Teachout notes in her superb book Break ‘Em Up, it isn’t an either/or question of nationalisation versus breakups - there is a role for both, such as with state media alternatives like the BBC, Radio France Internationale, or Deutsche Welle; or European public healthcare systems. Breakups can supercharge this too:
“Driving down [e.g. healthcare] prices is much harder when the government is negotiating with a powerful monopoly than when it is negotiating in a competitive market.”
Reason 11: Breakups are routine
Skeptics mutter that breakups are like unscrambling eggs: impossible!
But companies routinely split off parts of their business: shareholders often demand it. For example, AT&T spun off Warner Brothers, undoing a merger regarded as a "mistake." EY plans to split its consulting and auditing functions. Enforcers wouldn’t need to start from scratch to re-learn their lost art: fleets of experts are on hand to help.
John Kwoka and Tommaso Valletti, in their “Unscrambling the eggs” paper, identify many examples of natural fault lines for splitting Big Tech into separate functions or business lines, while keeping certain unitary pieces intact.
Reason 12: Breakups cut bureaucracy
Think of the effort and vigilance permanently required to police the outputs of Big Tech firms running dynamic black-box algorithms, or hosting subservient sellers. Even with ‘ring fences’ in place.
This is exhausting for outgunned regulators. And results are patchy. When the European Union ordered Google to create a “neutral” shopping site (after billions in fines failed), its tweaks still favored itself. Apple is similarly “walking backwards slowly.”
Ring fences weaken as memories of the scandals that built them fade. Britain is now watering down bank ring-fences installed after the last financial crisis.
Breakups are far simpler, with longer-lasting results.
Reason 13: Breakups can build resilience and security
Diversity and balance promotes economic and political resilience. Monopolies are single points of failure - economic chokepoints controlled by private actors accountable to shareholders, who generally don’t represent the public interest. They may even be agents of hostile actors, like Russia's Gazprom.
Monopolists have outsourced production to single countries or even single factories, stripping out redundancies in the name of ‘efficiency’, which meant that disruptions associated with the pandemic and Russia's invasion of Ukraine, for instance, often could not rapidly be fixed because monopolisation had killed the alternatives and backups.
Diverse local circulatory systems matter. Local businesses used to buy from each other, creating a virtuous cycle of local prosperity—and resilience. Now, “local circulatory systems for money [in agricultural areas have been] replaced by one-way conveyor belts shipping rural wealth out” to financial firms and monopolies in metropolitan centers, overseas and offshore. The result has been “increasingly extreme chokepointing within most industrial systems.” And public anger.
Reason 14: Competition is often (but not always) a force for good.
We left this till last, because it’s the most obvious. Mergers tend to reduce competition, breakups tend to increase it. When competition is healthy it can offer more alternatives not only to consumers, but also to workers, citziens, small businesses, and others. It can thus promote individual freedom from domination and manipulation, or foster revolutionary inventions and creative works.
Take Google’s Alphabet, for example, or Facebook’s Meta. They are digital advertising conglomerates whose business model incentivises keeping you online, to harvest more real-time data to sell. Polarising content drives engagement, so they prioritise that.
The first big 'social media' platform was not Facebook, but MySpace. It became notorious for poor privacy that facilitated sexual assaults and other harms. Facebook initially presented itself as a privacy-centric alternative, and rapidly won market share on that basis. But once it had seen off its competitors, it was able to lock users into a devil's bargain: for users, keeping social ties meant accepting horrible privacy conditions—and inflammatory algorithms.
Financiers talk of "kill zones" around monopolists, where minnows (like startups that safeguard privacy) are quickly absorbed or snuffed out. Breakups kill the kill zones.
Competition isn't always beneficial, of course. It can go wrong when regulators are blinded by bad ideologies, or when companies compete on externalities (such as greater willingness to pollute, or to take financial risks at others’ expense.) But the answer to these problems isn’t to monopolise.
Competition in the right places, in the right ways, is a force for good.
Section 3: Conclusion
We are at the start of a new era of fightback against monopoly power. New alliances are amplifying the voices of long-overlooked stakeholders. Regulators are waking up. Legislators are developing new toolkits to tackle power. A new era is taking shape.
It is time to bring back breakups: the most powerful single tool to meet the scale of our monopoly crisis. And we need to transform merger rules to stop the tide of consolidation that led to all the bickering about breakups in the first place.
Nicholas Shaxson is Acting Executive Director of the Balanced Economy Project.
Laurel Kilgour is a startup lawyer based in Dublin, and an adviser to the Balanced Economy Project.
This article has been slightly amended from the original, to fix broken links and add minor details.