Could Progressive Taxation Solve our Monopoly Problem?
Big companies are creepy, but so is antitrust action. Could more progressive taxation solve things better?
Anti-trust is cool again. After 40 years without the US government breaking up a large monopolistic company (Bell in 1982) the US government now has recently pursued (or publicly considered pursuing) antitrust action against Google, Meta, and Amazon, with action against other companies likely.
There are good reasons for this. Corporate consolidation has increased dramatically over the last four decades with a growing number of industries controlled by a small number of companies. This is bad in all sorts of ways. It increases inequality, reduces economic dynamism, increases economic fragility, and gives a small number of people a scary amount of power over everyone else in a way that is deeply creepy.
But antitrust is kind of creepy too.
It’s creepy when a big company has huge amounts of power, but it’s also creepy when big companies have strong incentives to act in the interest of the people with political power because they are afraid of politically motivated antitrust action.
This arguably wasn’t that big a deal when the companies being punished were Standard Oil, American Tobacco, DuPont, Alcoa, or even AT&T. But it’s a massive deal when the issue is companies like Meta and Google who have a huge ability to influence the information people consume and thus what people think is happening in the world.
It’s a big enough deal that it might be worth finding a better way of limiting corporate power than antitrust.
In theory the federal process for deciding whether a company should be broken up apolitical. The case is put together by the FTC or the DOJ and decided by a federal judge. In theory all these people make decisions in an apolitical way by following the letter of the law.
The real world is more messy. While nominally independent, The FTC and DOJ are led by political appointees. Similarly, federal judges are also political appointees. Moreover the legal cases at the heart of anti-trust typically depend on complex subjective decisions about what constitutes a market and what constitutes an abuse of power, providing lots of room for political bias.
This isn’t to impugn the moral virtue of the people involved. I’m a big fan of FTC-head Lina Khan and personally donated a non-trivial amount of money to her friends at the American Economic Liberties Project. I just think the whole system is messed up.
Even if everyone involved acts in an entirely neutral way, the mere fact that regulators might be biased towards the interests of the president is going to give any company under the threat of anti-trust action a strong incentive to do whatever might be in the interests of whoever has power.
Even if you think this is fine because you like the current president and their appointees and think they would never abuse their power, maybe it won’t be fine in the future when a president you loathe has ceased the levers of power.
Back when I was at Google, I once found myself interviewing someone who had already aced all the questions I’d rehearsed, so I decided to make up a new question on the spot and asked him “how many robots would Google need in order to take over the world?”.
Without pausing, his immediate answer was “how do I know they haven’t already done so?”. And he had a good point. He got his news through a Google web browser, using a Google Search engine, and a Google Phone OS. Sure, maybe he had an iPhone, or watched some TV news, or read Facebook, but maybe Apple, CNN, and Google are working together.
People used to say that if you wanted to do a coup, the first thing you needed to do was to take over the TV stations so that you could announce that the coup had already succeeded. The modern equivalent of that is to control Facebook and Google.
Arguably the ideal situation is that nobody should have as much power over information as Facebook and Google. It’s creepy if the government can control public discourse, but it’s also creepy that a few billionaires can control public discourse.
While corporations may not actually be people, there are definitely some similarities.
One similarity is that, just as it’s creepy when a company gets too powerful, it’s also creepy when an individual becomes too powerful. Since power-begets power, a person with too much wealth is likely to use it to manipulate the system in ways that give themselves even more wealth. And a country where all the power is in the hands of a few people is likely to go bad in lots of different ways.
But we don’t have antitrust action for people. There is no government agency that selectively identifies people who have too much power and brings court cases against them to reduce their power.
Instead, the main tool we have to prevent individuals becoming too rich and powerful is progressive taxation. Typically an individual pays no tax on the first chunk of their income, then each additional chunk of income is taxed at an increasing rate. In this way, people still have an incentive to work harder to earn more money, but the nasty feedback loop where wealth begets wealth is at least somewhat dampened.
At times progressive taxation has been very aggressive. The top tax rate was 91% in the 1950s, and that was a significant driver of the lower income inequality of the time. Many people would like the US to return to higher top tax rates, but even our current top rate of 37% has a significant effect.
Progressive taxation of corporate income is pretty common, but the top income threshold is usually far too low to distinguish between a normal sized business and a monopolist. Prior to 2017 the US had 4 income tax brackets, but the highest bracket was for income over $10m. Google isn’t realistically going to break itself up into so many companies that each of them makes less than $10m.
One important way that corporations are different from people is that companies can break themselves apart into multiple smaller companies.
If a human is earning so much money that they are stuck in a high tax bracket then they don’t have the option to break themselves up into multiple smaller people each of whom pays a lower tax rate, but a company could decide to break itself into multiple smaller companies.
One of the problems with progressive taxation on individuals is that it discourages economic activity, because the individual might choose to do less productive work if 90% of the extra money they earn would get taken away. But if a company found itself in the top tax bracket, it would also have the option of breaking itself into multiple smaller companies that are each in a lower tax bracket.
Rather than having the government using the courts to force big companies to break themselves up it would be the shareholders putting pressure on the company to break itself up, and the executives constantly thinking about ways they could break their company into multiple smaller companies without reducing their ability to do good work.
How big is too big? It depends.
Some companies just need to operate at a certain scale in order to do the thing they do. Google needs to be a big company in order to make a search engine, because indexing the whole web and running the machine learning models needed to understand costs a lot of money. Similarly, Apple needs to be a big company in order to do the expensive design work and manufacturing work needed to design an iPhone.
But maybe Google could split themselves into different companies for Search, Android, Chrome, and Docs. Or maybe Apple could split themselves into separate companies for iPhone, Watch, iOS, and Mac?
It’s hard for someone outside a company to know what things need to be done by a big company, or what things could plausibly be spun out into a separate company. However if a company has an economic incentive to split themselves up into smaller companies then maybe they will work out their optimal size by themselves.
Every company should be as big as necessary, but no bigger.
So if higher corporate tax brackets are such an obvious simple way to discourage companies from getting too big, why aren’t they a widely applied approach?
The standard answer is the same reason the US lowered its top corporate tax rate from a top rate of 35% to 21% - when tax rates are high, big companies have an incentive to find ways to dodge tax, or to move more of their operations to low tax countries.
Similarly, if big companies in your country face a higher tax burden than big companies in another country, then companies in your country will attract less international investment, because they will lose more of their profits to tax, putting your domestic industries at a disadvantage.
But these problems apply to antitrust lawsuits too. If companies in the US are more vulnerable to antitrust lawsuits than companies elsewhere, then companies have an incentive to locate themselves in countries without antitrust laws and investors are likely to price in the risk of antitrust action when choosing whether to invest in companies in that country.
In the end, the problem of tax avoidance by large multinational companies is probably a problem we just need to solve anyway. Indeed there is a growing global movement to reform international taxation to reduce the ability of multinational (and particularly Internet) companies to take advantage of low tax jurisdictions, including the growth of Digital Service Taxes, and the OECD’s BEPS (Base Erosion and Profit Sharing) project.
The other obvious way a company could dodge a progressive income tax is by breaking itself into lots of little companies that still behave like one company. They might be subsidiaries of a common parent, or owned by the same shareholders, or have exclusive contracts that tie them so closely together that they behave as if they are one company. Indeed this is often done for reasons of risk management, regulatory compliance, or to make it easier to work across multiple countries.
To be effective any progressive taxation system would have to tax such companies as if they were one entity, taxing them at a rate determined by their combined income. In practice the legal decision about whether to treat such companies as one entity for tax reasons is likely to get subjective in some cases, but it would likely still be less subjective than our current antitrust system.
But is money the right metric for deciding when a company is a dangerous monopoly?
You can imagine a world in which Facebook was a non profit, but still had monopoly control over information flow due to network effects. Similarly, you can imagine a world where all the big tech companies had such razor-thin margins that they barely paid any corporate income tax.
You can imagine alternative solutions that tax companies based on how many users they have, or how much time people spend using the product, or some other kind of metric that determines whether they have too much influence on how information flows in society. However these kind of metrics get pretty messy, particularly for products that don’t require users to verify their identity.
However the situation right now is that the companies that have a lot of power tend to also make a lot of money, so using money as the metric seems like it would solve the problem.
Betteridge’s law of headlines says that if the title of an article is a question, then the answer to the question is “no”. The title of this essay is “Could Progressive Taxation Solve our Monopoly Problem?”. Maybe the answer is “no”, but I think the answer might at least be “somewhat”.
Even if this approach has issues, it’s possible it has less issues than the alternatives of letting big companies stay big, or governments using antitrust law to selectively break them up.
This would be pretty effective at keeping businesses from getting too big, but I think the main goal of antitrust policy should be anti-anti-competitiveness, not anti-bigness.
From Yglesias:
> I want to draw a distinction here between anti-bigness and antitrust policy because there’s a set of influential figures who’ve been working to muddy the waters here and it’s confusing.
>Antitrust is about competition. If Nike tried to buy Reebok there would be a serious antitrust question about monopolization of the sneaker market. The issue isn’t that the combined Nike/Reebok entity would be “too big” (there are lots of bigger companies than the two of them combined); it’s a specific concern that we want companies to compete with each other rather than merge and raise prices. In theory you could have an anti-competitive cartel between three very mid-sized companies, if they happened to jointly control a market in something obscure like a particular kind of gasket or what have you.
> By contrast, although McDonald’s is a company that’s about as big as Nike in market cap, there’s really no conceivable McDonald’s merger that would raise major antitrust concerns. If McDonald’s merged with Starbucks, the combined entity would be much bigger than NikeReebok, but “restaurants” or even “fast food” would continue to be an extremely competitive market.
You can say that bigness is a useful proxy for anti-competitiveness, but the Yglesias article has lots more on the many large companies that provide more social value than smaller companies. He views bigness as a fairly poor proxy for anti-competitiveness, and would say we'd lose a lot from an anti-bigness tax.
https://www.slowboring.com/p/small-business-is-not-the-answer