Why Did Congress Just Vote to Break Up Big Tech?

The Judiciary Committee voted 21-20 to split up Google, Amazon, Apple, and Facebook. What happened was unbelievable.


Welcome to BIG, a newsletter about the politics of monopoly. If you’d like to sign up, you can do so here. Or just read on…

It’s been an extraordinary week. On Monday, the Supreme Court ruled against the NCAA for abusing its monopoly over the terms and wages of student athletes. Yesterday, the House Judiciary Committee, which is the body that has jurisdiction over antitrust law, wrote and voted on legislation to break up big tech firms Apple, Google, Amazon, and Facebook. There are problems with the bills, and I’ll get into them. But the underlying content is less important than the political message, which is that breaking up big tech is looking increasingly inevitable.

Meanwhile, the new Federal Trade Commission Chair Lina Khan has taken full control of the agency. She’s hired acting directors to run the bureaus, and announced that the FTC will be holding open commission meetings where the public can watch the commissioners vote and debate things. Open meetings haven’t happened in decades, so this is something of a shock to the agency. There’s even a period where the public gets to talk back to the commissioners. Imagine that. The underlying agenda is pretty aggressive (though in the administrative weeds), and starts with finalizing a rule against lying about Made in USA labels. (If you want to sign up to speak, you can do so here.)

And now to the good, bad, and meaning of the break-up votes. Here’s Jerry Nadler, the Chair of the Judiciary Committee.

The Good

The Judiciary Committee wrote and passed six different bills, two of them being general purpose antitrust acts and four being big tech-specific ones. These bills are an outgrowth of the 16-month investigation into Apple, Google, Amazon, and Facebook, with an analysis of millions of documents and hundreds of witnesses.

This isn’t the end of the process, but it’s the first Congressional vote to actually restructure powerful firms since the 1996 Telecom Act, and the first real Congressional attack on concentrated corporate power since the Bank Holding Company Act Amendments of 1970. Congressman Mondaire Jones summed up the hearing where they voted to do so. “Unless we break these companies up,” he said, “they will continue to be above the law.’”

So what do these bills do?

The first two are relatively simple. The first increases the amount of money that our antitrust enforcers can use to bring cases and regulate markets. (The FTC’s budget is $351 million, this would boost it to $418 million, while the Department of Justice Antitrust Division would go from $188 million to $252 million.) I wasn’t so keen on this one for a long time, because the Federal Trade Commission and the Antitrust Division are terrible and asking for more resources was an excuse for bad legal strategy. But with Lina Khan at the FTC, I’m more optimistic that she can restore the agency’s legitimacy. Or at least, now I know there’s someone there who recognizes the task at hand.

The second is a bill that is very procedural, but antitrust is a weedy area, and it matters. One of the techniques that monopolists use to avoid scrutiny is to move cases brought by state attorneys general to courts that are friendlier to big corporations. California, for instance, is well-known for tech-friendly judges - Google tried to move one key antitrust case on adtech to its home state. But big pharma does it too. In 2016, 40 state attorneys general filed suit in Connecticut against 18 pharmaceutical companies alleging price-fixing and market allocation of 15 generic drugs. The pharmaceutical companies, most of which were headquartered in the Philadelphia-area, successfully transferred the case to the Eastern District of Pennsylvania. It still hasn’t gone to trial. The second bill stops this nonsense, and lets state AGs keep the cases in the district they choose to bring suit. (Jurisdictional fights have always been a problem - in my book I profiled a 1937 suit over the monopolist Alcoa, in which the firm got the suit moved to its home town of Pittsburgh, and Congress in response nearly passed a law making it easier to remove judges.)

These two bills might not seem like a big deal. However, if these two bills were all that passed, they would still comprise the single most important strengthening of Federal antitrust law in a generation. For decades, antitrust was just not important, and the Judiciary Committee didn’t bother to focus on it. So to have these markups, and pass these bills, is in itself meaningful.

The other four bills solved for problems specific to Google, Apple, Amazon, and Facebook, problems ostensibly laid out in the big tech report by the subcommittee last year. Here are the four bills and what they did.

1) The ACCESS Act mandates that big tech firms have to make their systems open to competitors and business rivals, in the same way that AT&T customers can talk to T-Mobile customers, or users of different email systems can communicate with one another.

2) The merger bill makes it harder for big tech firms to buy rivals.

3) The nondiscrimination bill is intended to ban the ability to big tech firms to preference their own products, the way Google substitutes its own reviews for Yelp reviews, even if Yelp’s reviews are better.

4) The break-up bill is supposed to split apart big tech firms by prohibiting platforms from owning any line of business that uses that platform.

All four passed the committee, which is extraordinary and unexpected. And not only did they pass, but they passed with both Republicans and Democrats working on them.

The Anti-Monopolists vs Friends of Eric

There were two coalitions at work, though these coalitions weren’t partisan in nature The first was the coalition of big tech, meaning Silicon Valley California Democrats led by Zoe Lofgren, as well as conservatives led by Republican Jim Jordan. Call them ‘Friends of Eric,’ after Google ex-CEO Eric Schmidt.

The second coalition opposing big tech consisted of Democrat David Cicilline, Republican Ken Buck, along with a host of progressives including Pramila Jayapal (who was the sponsor of the break-up bill) and Mondaire Jones, as well as right-winger Matt Gaetz.

Republican Congressman Gaetz praised Democratic counterpart Jayapal, saying Teddy Roosevelt would be proud, even as Republicans Jim Jordan and Darrell Issa lavished agreement on Silicon Valley Democrats Zoe Lofgren and Eric Swalwell. The bipartisanship was evident among partisans paying attention. Tucker Carlson praised the bills and laid into GOP leaders for opposing them, and conservative Republicans noticed Jim Jordan doing the bidding of Google.

The Bad

The bad is pretty simple. The tech-specific bills, as written, probably won’t deliver what their sponsors think they will, because they didn’t get all the specifics right. This isn’t intentional, it’s just that it is hard to write this kind of legislation. A friend once told me a good legal expression, ‘to write a good law, you have to think like a criminal.’ And that’s basically right. There are extremely well-paid lawyers who will spend their time exploiting the tiniest loophole, so good drafting means thinking about making statutes airtight. Competition law is complex and warped to make drafting full of legal minefields, so without extreme care in the language, the law will likely be subverted.

To understand why it’s so hard to get these laws right, it helps to start with the two basic problems with antitrust law. The first is that regulators and enforcers make key policy decisions, and have done a very bad job at it. A good example is they just decided to stop enforcing the anti-chain store Robinson-Patman Act, which prohibits certain forms of kickbacks, as well as prohibiting giving better prices to bigger customers. At some point in the 1970s and 1980s, the Department of Justice and FTC chose not to enforce the law anymore. And when they stopped doing so, Walmart and other chain stores, and eventually Amazon, exploded in size and power.

And then there are judges. Judges have been trained in a type of thinking in which antitrust is all about promoting a certain form of economics, known as ‘consumer welfare.’ Most of the things you and I would consider unfair, like paying kickbacks to someone to stop them from selling rival products, or selling below cost to drive your competitors out of business, or intentionally making your products incompatible to undermine smaller rivals, judges tend to see as ‘pro-competitive,’ which is to say, good and efficient. I’m not kidding. Yesterday, Obama-appointed judge Daniel Crabtree dismissed an antitrust case against Epipen maker Mylan, which was paying bribes to stop their competitor’s product from being available to consumers. To Crabtree, such bribes weren’t corrupt, they were efficient!

It’s not that judges are corrupt, it’s that there is now 40 years of case law saying that they must generally be hands-off and let firms do what they want. And to get judges to rule in your favor, plaintiffs must spend millions of dollars getting an economist to make up fancy models saying that intervening in a particular case creates more economic value than not intervening, and then hopefully the judge flips a coin and likes your expert more than Amazon’s expert. To put it differently, imagine if, say, you had to show in any robbery case not just that your money was stolen, but that you would spend your money more wisely than the person who took it. That’s basically what antitrust is like these days. This is called ‘consumer welfare’ but it is in fact just a corrupt and foolish way to understand law.

The way to address both problems - bad regulators and bad judges - is to write very specific and careful legislative text. Give clear instructions on which practices are and are not legal, and try to avoid corrupted words like competition, which only invite judges to opine on economic questions. Moreover, have a clear vision on what gets broken off from what. These bills don’t really do that. (Neither did Klobuchar’s bill introduced earlier this year; state Senator Michael Gianaris’s antitrust update in New York came closer.)

The break-up bill, for instance, centers on conflicts of interest between lines of business. But it never defines what it means by “line of business,” and this stuff gets very blurry. Big tech will use this ambiguity to its advantage. Right now, for instance, Amazon has a marketplace on Amazon.com. It also has a logistics business, Fulfillment by Amazon. If you try to split these two obviously different divisions apart, Amazon will claim that these are all one line of business, with Amazon Prime, Fulfillment by Amazon, and Marketplace all one thing. And then a judge gets to decide whether that’s true, because this bill doesn’t. Judges really don’t like to make what they perceive of as product design decisions, preferring to defer to monopolists. So yeah, that’s a problem.

This problem is pervasive across the legislative text. The merger bill, rather than a straightforward ban on big tech mergers, instead says big tech firms have to jump through a bunch of hoops showing that whoever they are buying doesn’t compete with them or potentially compete with them. That sounds fine, except that judges understand ‘competition’ to mean very expensive and unwieldy fights over how to define the market, according to fancy expensive economists. So basically enforcers will still have to unnecessarily spend massive resources to stop big tech mergers, though they will have more authority to do so. (Also, during the mark-up, big tech managed to punch a hole through this one, exempting mergers of less than $50 million. But that happens.)

And finally, the two other bills - one on self-preferencing and one on interoperability - each have their own problems. During the markup, Congress inserted a provision into the bill on interoperability mandating that users give consent before their data is shared. This intuitively makes sense, except that it basically allows big tech to block any data sharing they don’t want to deal with. Apple, for instance, now asks users if they want to share data with third parties, and just 5% of users opt-in. There are other concerns; standards expert Mark Nottingham pointed out that the interoperability bill makes the FTC effectively an internet regulator and breaks existing institutional arrangements. I don’t have enough knowledge to opine on this critique, but Nottingham’s point of view seems reasonable. And then on the self-preferencing bill, big tech managed to insert an amendment that made it subject to the consumer welfare standard, which guts the intent of the bill.

There are other problems with these four bills. All of them require regulators to make key policy decisions, such as designating which platforms are subject to these rules, or defining terms like data. These should all be done in legislative text, because regulators often just don’t follow through on Congressional intent, especially when that intent is vague. Ironically, this is something North Carolina Republican Dan Bishop - who opposed most of this legislation - tried to fix during the markup.

In other words, there’s a lot of legal workshopping that has to happen to make these bills workable.

The Meaning

But here’s the thing about the critiques I just offered. They aren’t really that big a deal.

Sure, the bills need to get fixed before they move to the next stage of the Congressional process. But that’s manageable. What matters, however, is that the Judiciary Committee found the votes to pass them. These members would have voted for well-structured bills, and writing well-structured bills is quite doable.

And that is truly stunning.

Does this mean these bills in a reasonable form will end up signed into law this session? Well the funding and jurisdiction bills, yes. The other four will take more time, because there’s clearly not yet a consensus. But the pressure to deal with monopolies across the board is only going up. So over the next five years, it’s hard to imagine these firms remain in one piece, and that our markets remain as concentrated as they are.

What has been happening since the financial crisis of 2008, and what continues today, is nothing short of a revolution in the way Americans think about political economy and monopoly power. We haven’t had such ferment over these issues since the 1890s or 1930s, during the Gilded Age and New Deal. Any one of these bills put forward would be the biggest strengthening of Federal antitrust enforcement in at least fifty years, so that all six passed out of committee - especially the bill that breaks up these firms by statute - is remarkable.

There is one other important aspect of what these bills mean. Put simply, they show representative politics can work. Big tech lobbied aggressively against these bills, and lost the debate. Yes, there are corruption problems, but ideas and investigations matter. Republicans and Democrats actually came together and decided, for once, to govern.

We will win some, and lose some. But this was a very good week.

Why a Pro-Monopoly Supreme Court Ruled Against the NCAA Monopoly: Earlier this week, a very conservative Supreme Court published its decision on whether the National Collegiate Athletic Association can continue to put rules in place to suppress the wages and terms of student athletes. In NCAA v. Alston, the court ruled 9-0 on narrow grounds for the student-athletes and against the NCAA, which is a clear labor-fixing cartel.

The court finally noted that suppressing competition in labor markets is an antitrust harm. This may seem like it’s already the case. But while courts have said that agreements not to hire the employees of competitors is a violation, they hadn’t yet ruled on using market power to directly lower wages. Now the court has done so. The court also narrowed certain cases that are favorable to monopolists. Most fundamentally, the case shows that the philosophy underpinning the dominance of monopolists - the ‘consumer welfare standard’ - is falling apart.

A Private Equity Giant Trained the Group that Assassinated Jamal Khashoggi Private equity is everywhere. Here’s the New York Times on the group that trained the Saudi men who assassinated Washington Post Jamal Khashoggi.

Four Saudis who participated in the 2018 killing of the Washington Post journalist Jamal Khashoggi received paramilitary training in the United States the previous year under a contract approved by the State Department, according to documents and people familiar with the arrangement…

The training was provided by the Arkansas-based security company Tier 1 Group, which is owned by the private equity firm Cerberus Capital Management. The company says the training — including “safe marksmanship” and “countering an attack” — was defensive in nature and devised to better protect Saudi leaders. One person familiar with the training said it also included work in surveillance and close-quarters battle.

Such a training firm could have been owned by a non-PE owner as easily as a Cerberus, since the problem here isn’t a result of financial engineering. But there is one element linking private equity to this scandal. Cerberus is politically connected. Trump nearly put the head of Cerberus, Stephen Feinberg, as a top official at the Office of the Director of National Intelligence. Ownership tends to bring political connections for portfolio firms. And this training program was approved by the State Department. Ergo…

Thanks for reading. Send me tips on weird monopolies, stories I’ve missed, or comments by clicking on the title of this newsletter. And if you liked this issue of BIG, you can sign up here for more issues of BIG, a newsletter on how to restore fair commerce, innovation and democracy. If you really liked it, read my book, Goliath: The 100-Year War Between Monopoly Power and Democracy.


Matt Stoller