How Economists Corrupted the Internet
FTC documents show that antitrust economics establishment and Obama-era corruption led to a monopoly takeover of our economy. Joe Biden can undo the damage.
|Matt Stoller||Mar 22||39||24|
Welcome to BIG, a newsletter about the politics of monopoly and finance. If you’d like to sign up, you can do so here. Or just read on…
Last week, documents leaked showing that the Obama administration nearly brought antitrust charges against Google in 2012. I’m going to write about why they didn’t, the damage that decision caused, and why Biden will forge a different path.
Also short pieces on:
How Gmail Quietly Controls a Vital Channel for Political Speech
The Slow Collapse of Corporate Republicans
The Monopoly Behind the Nuclear Weapons Lobby
The Coming Merger Boom
Why Golf Clubs Are Getting Worse
Before the main event, some house-keeping. I have a piece in the New York Times on the Arizona state legislative fight against the app store monopolies of Google and Apple. Also, I was recently on Marketplace to talk about Google. Finally, reporter Alec McGillis has an important book out on Amazon and the tearing apart of American society. It’s called Fulfillment: Winning and Losing in One-Click America, and I’ll be writing more about it shortly.
Bad Search Engine Results Kill People
Americans expect Google to deliver the most relevant and best results for any particular query. But Google has an edge case problem. When an unsophisticated or desperate user really needs information about something important, and marketers are trying to lie or defraud the user, Google may deliver results that are not only bad, but actively harmful.
For instance, in 2017, reporters Cat Ferguson and Dave Dayen showed that Google’s poor search results had become a useful tool for con artists trying to entice addicts and alcoholics to sham rehab facilities. Google’s marketing tools often worked, helping shoddy treatment center firms cheat addicts, some of whom no doubt relapsed.
Offering poor quality rehab facilities is wrong, and Google didn’t cheat the addicts directly. But what made this line of business profitable was among other things the easy access to customers enabled by using Google. Indeed, as Ferguson noted, these companies were “united by their dependence on Google.” Embarrassed by the publicity, Google eventually made some effort at addressing the problem, but never really figured out how to stop con artists from using its service to harm these desperate people.
Similarly, in 2019, the Wall Street Journal reported on millions of fake listings on Google Maps, which con artists used to cheat customers and blackmail honest small businesses. Users were screwed. But for businesses, the only recourse was to spend more ad money on Google; complaining got you nowhere, or worse. Said one businessman, “It’s less harmful to piss off the government than piss off Google. The government will hit me with a fine. But if Google suspends my listings, I’m out of a job. Google could make me homeless.”
In other words, while generally speaking you will get good results from Google, in edge cases you may get results that are extremely harmful, like a repairman who cheats you, a bad doctor, or someone who wants to steal your money in the guise of helping you recover from addiction. Since most people expect to get credible results from Google, it’s a form of mass deception. And those who rely on Google to convey information to customers, like small businesses, are often on a knife’s edge, existing at the whim of a search monopolist that does not notice them.
These quality problems are a result of Google’s monopoly; poor quality is a classic symptom of monopoly power. How Google seems to offer good results on the whole, but sometimes undermines quality at the edges, is a somewhat subtle story.
Why Does Google Help Kill Addicts Seeking Recovery Services?
Google’s main search engine is what is called a ‘general search engine,’ meaning it provides general results based on indexing most of the web.
There are other types of search engines. Yelp and Expedia, for instance, are known as ‘vertical search engines’ who focus on a much narrower topic, like local businesses and travel. You can’t ask Yelp generalized questions about research or culture, but it is likely better (though not perfect) at removing local restaurant listing spam than Google, because that is its entire business.
Of course Google isn’t just a general search engine. It has vertical search lines of business as well. It competes with Yelp, Expedia, etc, listing restaurants, health providers, travel information, etc, and has user reviews. But the incentives are different for Google. If Google Maps stopped listing every restaurant in New York City, the lost revenue literally wouldn’t show up on Google’s income statement. Yelp, however, would see it as a crisis for its business.
The CEO of Yelp no doubt spends a lot more time thinking about removing fake listings of restaurants than Google CEO Sundar Pichai, just because Pichai has nine products with more than a billion users. Maybe Google is better at building stuff than most companies, but it’s not so much better that its executives can spend no time on a search problem and, all things being equal, still outperform a specialized search vertical. In other words, the reason Google isn’t very good at finding the right health care provider or local business is because that’s not really what its executives think about.
All of this is a way of saying that vertical search engines are sometimes better at finding certain kinds of information than Google. In its original form, from 1998-2007, Google helped blend the world of general and specialized search; it simply chose the best results, sending people to the right place on the web or to the right vertical search engines that had the best results. As Google co-founder Larry Page once put it, “We want to get you out of Google and to the right place as fast as possible.” People built businesses around an open web. Yelp was founded in 2004, back when you could still found firms adjacent to Google; Yelp got a lot of traffic from Google because it had the best local results.
But in 2007, Google stopped trying to send users to the most relevant place to answer their query, and started to try and keep people on Google properties. It began transforming itself from a general search engine into a walled garden, and it arranged its business strategy to exclude competitors, both vertical and general search, from the market, especially as people started to use their mobile phones to find things. At first this change was subtle, but Google gradually expanded its walled garden, encompassing more and more content. In doing so, it directed ad revenue to itself, eventually strangling not only vertical search competitors, but also publishers, online video and mapping competitors, and advertising technology firms.
Today, Google is the key gatekeeper to the web for users and advertisers, and venture capitalists will not invest in firms adjacent to it. Google’s dominance is also why the web in 2021 is increasingly a mess, a place for scam artists and disinformation. Today, if there were a vibrant competitive market for search, this rehab clinic fiasco might not be a problem; a health-based vertical search engine might be able to solve the problem that Google cannot. But in Google’s walled garden internet, that’s no longer a possibility. And as there really is no distinction between the web and the offline world, Google’s absentee landlord relationship to problems involving credible information is one reason scam artists and disinformation are proliferating globally.
It didn’t have to be this way. And in fact, in 2012, the Federal Trade Commission, which is our antitrust enforcer, nearly filed a case that would have stopped Google from corrupting our information commons.
Fumbling the Future
And this brings me to Leah Nylen’s story last week titled “How Washington Fumbled the Future”, looking back on the Obama administration’s policy vis-a-vis Google. She got her hands on a series of allegations the FTC had in 2012, in documents kept secret for nearly a decade. Recently, there have been multiple antitrust suits launched against Google, two by states and one at the Federal level. What is astonishing is how the FTC in 2012 had the evidence to bring most of the suits in court today.
Those of us who follow this area didn’t think that the 2012 FTC documents would be that interesting. The vote to close the Google investigation was unanimous, 5-0, with both Republican and Democratic commissioners letting Google skate. We figured that the FTC just didn’t see the problem clearly, as technology markets tend to morph quickly. Back in 2011 when the investigation started, who would have imagined that Google would become this powerful and dominant?
And yet, it turns out that the FTC had evidence of Google’s behavior, and just chose not to act. Yelp CEO Jeremy Stoppelman called these released documents a “smoking gun” showing how “Google methodically destroyed the web.” Stoppelman competes with Google, but other more neutral observers agree with him. William Kovacic, a Republican ex-FTC member, said, “I always assumed the staff memo was not so specific, direct and clear about the path ahead. A lot of the DOJ case is in there. It’s really breathtaking.” Kovacic, who voted to open an investigation in 2011, left the FTC before the complaint came up for a vote, so he hadn’t read it until this week.
These documents revealed many things, one of which was a battle inside the institution. The FTC’s lawyers, while meek, were evidence-based, and did want to bring charges. The FTC economists, by contrast, fought them every step of the way, with predictions that today - and frankly even then - were jaw-dropping in their wrongness and stupidity.
So what were the allegations?
The FTC’s lawyers made two observations about Google. First, Google intentionally harmed and killed vertical search rivals by putting its own product first, which is known as ‘self-preferencing.’ “Google,” said the complaint, “routinely, and prominently, displays its own vertical properties, while simultaneously demoting properties that are identical to its own, but for the fact that the latter are competing vertical websites.” Google also would scape from its rivals and present their user-generated content as its own. Second, Google used exclusivity agreements and its own market power to deprive its rivals of user traffic, advertising revenue, data, and advertisers.
In doing so, Google drained the vitality out of the web, so much so that venture capitalists began using the term ‘kill zone’ to describe the un-investable spaces adjacent to Google.
If They Had the Evidence, Why Didn’t the FTC Bring a Suit?
If there was such strong evidence, what happened?
There were three main reasons for the reluctance to bring a case. The first is simple: corruption. Google was extremely close with the Obama administration, with its lobbyists averaging a meeting a week at the White House. Today, four out of the five FTC Commissioners who voted on the case, as well as staffers involved like Howard Shelanski, receives money directly or indirectly from one of the five big tech giants. (The only former FTC Commissioner who voted on the case and does not receive big tech money is deceased.)
Even more tellingly, the resolution of the case happened a little over a month after the reelection of President Obama, a campaign in which then-Google CEO Eric Schmidt was on election night, as The Wall Street Journal reported, “personally overseeing a voter-turnout software system for Mr. Obama.” Today this sounds bad, but at the time, the attitude in the White House was that Google, Facebook, and Amazon were making the world a better place. Republicans loved them because the GOP loves business, and Democrats loved them because they were culturally progressive and good for consumers. (For example, in 2013, well-known Democratic pundit Matt Yglesias wrote about Amazon, “Amazon, as far as I can tell, is a charitable organization being run by elements of the investment community for the benefit of consumers.”)
The second main reason was ideological. Even the FTC lawyers were under the spell of the ‘consumer welfare’ standard, and felt that Google had the right to monopolize as long as consumers weren’t harmed. The reason, for instance, that the lawyers didn’t want to bring a charge for Google’s self-preferencing to kill vertical search engines wasn’t that they didn’t know the goal. They realized it was a scheme to monopolize. But they felt that Google had presented enough evidence showing that consumers might like it if Google’s search engine prioritized Google’s own properties. This ideology made the lawyers meeker, though they still wanted to bring suit.
The third and most significant reason is that the commission’s antitrust economists made a very strong, and entirely wrong, argument against the case, which in retrospect rested on a set of laughably inaccurate predictions. While the FTC lawyers at the Bureau of Competition wanted to bring some charges, the economists at the Bureau of Economics said there was no case. They believed the online ad and search markets were competitive, consumers were benefitting, and there was little evidence of misbehavior.
Here were some of the assumptions and predictions from the Bureau of Economics in 2012.
Consumers will continue to rely on desktop computers to search, not smartphones or tablets
Surveillance advertising tracking users across the web have only a “limited potential for growth”
Google is not a monopoly
Search engine quality isn’t primarily driven by data (“The premise that more click data is the primary and critical determinant of competitive position and quality… is contrary to the history of the general search market.”)
Google was not a particularly significant source of traffic for vertical search engines like Yelp
Google choosing to downgrade more relevant content in favor of its own content is good for consumers
Defaults don’t matter to consumers
Some of these assumptions are wrong, but a few are downright crazy. Take the assertion that that search engine quality isn’t primarily driven by data, which is what FTC economists argued. That notion was contradicted by testimony the FTC got from Google official Udi Manber. Manber told the FTC, “The bottom line is this. If Microsoft had the same traffic we have their quality will improve *significantly*, and if we had the same traffic they have, ours will drop significantly. That's a fact." And who was Manber? He was Google’s former chief of search quality. The importance of data is not a controversial assumption. It’s just one that FTC economists chose to disbelieve.
Most of the other assumptions were similarly outlandish. Why did these economists, who are well-trained and smart, rely on such silly assumptions? It’s not as simple as financial corruption, though that plays a role. It’s that antitrust economics, writ broadly, has nothing to do with understanding markets or monopoly power. At the same time, for instance, as FTC economists were dismissing the idea of Google as a monopoly, the economists at the other antitrust enforcement agency, the Department of Justice Antitrust Division, were being similarly ridiculous in justifying why they didn’t bring a single case against a monopolist.
Carl Shapiro, the head economist at DOJ, argued that there just weren’t any monopolies to go after. I’m not kidding. Here’s a direct quote.
First, I can say from personal experience that when I was the chief economist at the DOJ during 2009-2011, the Antitrust Division was genuinely interested in developing meritorious Section 2 cases, and we were prepared to devote the resources necessary to investigate complaints and other leads, but we found precious few cases that warranted an enforcement action based on the facts and the case law.
It’s not just that money from dominant firms offered to the antitrust economics world is endemic; Shapiro, for instance, is now a consultant for Google. It’s that antitrust economics is designed purely as a language for excluding ordinary people from debates over political economy. If it were designed around some sort of scientific validity, the FTC Bureau of Economics failure over Google would be understood as so egregious that it would justify firing much of the economic staff. But that didn’t happen, in fact what happened is nearly ten more years of aggressive self-congratulations. During the Trump era, the only part of the FTC to receive boosts in funding were the economists.
That’s how deep-rooted the problem is.
Will Joe Biden Clean Up Obama’s Mess?
So is Biden going to break with Obama on antitrust? The short is yes. The only question is how aggressively he does so. Back in October, I did a bunch of interviews with people around Biden-world for a piece on how a President Biden would take on corporate power. I noted that Obama loved Silicon Valley, even musing that after he left office he might become a venture capitalist, while Biden by contrast called big tech CEOs "little creeps." Obama adored technocrats, but Biden himself when he was in the Senate had little patience for economists.
Mostly I’m pretty happy with my predictions, in which I said Biden would spend a lot of money and be more populist than Obama. But Biden has out-performed my expectations. He hired antitrust expert Tim Wu for the White House, and is reportedly appointing antitrust expert Lina Khan to the FTC. And Biden himself when he was in the Senate had little patience for economists. Meanwhile, Republicans, who were just as in thrall to Google, have changed as well. Trump brought a Google antitrust suit, the conversation on Capitol Hill has undergone a stark change on the right, and the awesome Texas case against Google is heavy on evidence of monopoly power and very light on theoretical economics.
So both parties have broken from the antitrust status quo. That said, the question now is whether political leaders writ large can ditch their affection for economists. These debates are more than theoretical. When someone uses Google to find help for their addiction, that person is not accessing a search engine, but a manipulation machine designed to serve the interests of whoever pays Google the most. It didn’t have to be this way. And if we downgrade economics, it doesn’t have to be this way anymore.
How Gmail Quietly Controls a Vital Channel for Political Speech: One point of frustration for newsletter writers is that Gmail shifts certain email that people have requested into the Promotion tab, and sticks ads that look like emails on top of the Promotions list to get Gmail users to click on more ads. A contact told me about how this practice hits political organizers and political leaders particularly hard, and is now causing a crisis in political communications. Politicians and organizers are randomly are blocked from communicating with volunteers and small dollar donors.
“People have spent the last 5 years fretting about the role Facebook might play in augmenting election results — but Gmail can completely subvert an candidate’s electoral chances and has gotten minimal scrutiny.”
The Political Collapse of Corporate Republicans: Last week, the House Antitrust Subcommittee had a hearing on suggested changes to the antitrust laws, part of a series of hearings that will likely result in a significant overhaul in how the state and corporate power relate to one another. The hearing illustrated the problems that the GOP coalition is having meshing its increasingly anti-monopolist conservative base with its big tech supportive corporate wing. This tension became evident watching Republican FTC Commissioner Noah Phillips get grilled, not so much by Democrats, but by Republicans.
An increasingly significant concern for the right is big tech firms using their power to censor conservatives, as they did when they removed Donald Trump from most major platforms after the Capitol Hill riot. Conservative Republican leader Jim Jordan, who is the highest ranking member on the full Judiciary Committee, is extremely deferential to corporate power, but he is first and foremost a conservative movement leader. So on the one hand, he generally wants to help promote his vision of a free market, but on the other, he cannot abide dominant firms censoring his political faction. In this hearing, he was seeking some sort of reconciliation of these two goals, and asked his Republican ally, Noah Phillips, for help.
Phillips is a corporate Republican, and in this case, he simply had nothing to offer. In a back and forth, he told Jordan that censorship was not a monopoly concern. He also gave no ground on Section 230 of the Communications Decency Act, which gives a liability shield to tech platforms, and is often offered by Republicans as a way to address censorship concerns without making antitrust laws more powerful.
Jordan asked, “If we're not going to use antitrust to deal with censorship and you are reluctant to talk about Section 230, what's the answer?”
Phillips responded, "I'm afraid I don't have a good answer."
Phillips couldn’t respond with anything, because the only way he could have discussed addressing big tech censorship was by acknowledging that the state should be structuring markets deliberately on behalf of some sort of common good. And that concession would crush his ability to fight against state power and on behalf of dominant firms. Jordan needed something Phillips couldn’t offer.
Movements don’t end when their ideas prove to be wrong, they end when solving one political problem causes another, bigger one for their coalition. And that’s why libertarianism is falling apart.
A Nuclear Missile Monopoly Lobby: I’ve written a bit on the Northrop monopoly over the nuclear missile business, the result of a merger between Northrop and rocket engine maker Orbital ATK. As it turns out, monopolist Northrop is a big donor to the group of Senators who support spending more money on these missiles, informally known as the ICBM Coalition.
The efforts of the ICBM Coalition have been supplemented by lobbying and campaign contributions from ICBM contractors, led by Northrop Grumman, which has received a sole source, $13.3 billion contract to build a new ICBM, known formally as the Ground-Based Strategic Deterrent, or GBSD. Current estimates indicate that building and operating the GBSD and related warheads will cost $264 billion over the life of the program, which would provide a steady flow of revenue to Northrop Grumman and associated companies for years to come.
A quarter of a trillion dollar sole source contract? Not bad, Northrop.
The Coming Merger Boom? With the increasing drumbeat of merger announcements, most recently the $25 billion railroad merger of Canadian Pacific and Kansas City Southern, it’s clear there’s a merger boom afoot. As AELP analyst Krista Brown noted, “the FTC received 1130 preliminary merger filings in the last 4 months. That is nearly double the 683 filings received in the same time period last year... The distressed market is causing a merger boom that will only continue to grow in the coming months.” Stay tuned.
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.
P.S. BIG reader Ross Baird emailed me the following.
I play golf as the occasional hobby, I like being outside and walking for hours. I bought a club about 20 years ago, an Adams Tight Lies 3 Wood, that I have hit for 20 years and loved. My club unfortunately recently broke (it was durable and lasted 20 years!) and I went shopping for a replacement. Not to get too in the weeds, but the design and style of Adams golf clubs was definitely superior - it was a small startup golf club success that hit the big time.
Unfortunately I could not find a golf club like the one I bought 20 years go, today's version of Adams Tight Lies 3 wood is definitely inferior to the version 20 years ago and super expensive ($300+!) So i figured there's a monopoly story here. Adams Golf was started in Plano, TX and went super successful off the back of the Tight Lies Club, which was definitely better than any other product. Yet in 2012 Adams was acquired by multinational conglomerate Adidas, which turned Adams clubs into carbon copies of the other (inferior) brands they owned (corporate efficiency!). In 2017, after the roll up merger failed, Adidas sold off Adams and a couple of other brands to private equity firm KPS partners.
So I should not have been surprised that 20 years ago I bought a product that is clearly superior to the same product today, after a monopolistic merger and private equity spin-off, even though the latter is more expensive.