Paul Milgrom, a Nobel and Emmy winner who became Google’s antitrust witness
semafor.com
semafor.com
His assertions in this case are anodyne. Second-price sealed-bid auctions are a well-studied auction format [1][2]. The judge's question about sealed-bid auctions was reasonable; so is Milgrom's answer.
Google isn't abusing its monopoly power because it's using a proven auction format. How the auction is run is orthogonal to what the auctioneer charges. This article is well written. But the title is catnip to those who have a vague political but not fundamental understanding of the term.
[1] https://en.wikipedia.org/wiki/Vickrey_auction#Weaknesses
[2] https://scholar.harvard.edu/files/athey/files/comparingforma...
https://ia800602.us.archive.org/6/items/gov.uscourts.dcd.223...
It's abusing its monopoly power through "Pricing knobs" (see page 83)
A YouTube millennial-aged podcaster, who apparently read this opinion, summarised the abuse as "inflat[ing] the predicted click-through rate of the second highest bidder" and "introduc[ing] a degree of randomness in determining the winner", i.e., "if bids from the top two contenders are close the runner-up may randomly be credited with the win".
This "makes the outcome less predictable causing the winner and runner-up to raise their bids in future auctions" and "both parties end up bidding more than they otherwise would."
"Over the years, he also helped shape the online advertising industry by working with dot-com companies to build the marketplaces that eventually became the flourishing ad tech industry."
Either way, the OP's conclusion doesn't change much in a duopoly.
Any half decent intellectual actively avoids earning the monicker
It's not surprising. Economists are known for constructing theories that sound plausible to the typical person, but which are subtly geared to concentrate wealth at the top. Did we not learn from "trickle-down economics"? Economists core moral value is maximum efficiency, but I think overall efficiency is better at a happy medium rather than a maximum. Like Ivan Illich has talked about, sometimes it's better to have a speed limit on progress...
> During the Google antitrust trial, Milgrom picked apart the government’s assertions that the company had abused its power
The entire problem is that Google has too much absolute power. We need a new version of antitrust that doesn't just target percentage power, but absolute as well.
Do you know which of his books mentions that idea?
It's just a general rule of existing in this universe that fortune favors the fortunate and misfortune favors the misfortunate.
That doesn't mean we should embrace that, but it shouldn't be surprising economists talk about it and study it a lot. People just seem to get upset when economists have to be the ones to tell them that making it a law that iron is equal in value to gold will not make everyone rich.
> It's just a general rule of existing in this universe that fortune favors the fortunate and misfortune favors the misfortunate.
But it is not a rule of the universe that society must allow those fortunate who make billions to keep or get that wealth.
This is a total non sequitur from the comment you're responding to. Descriptive versus prescriptive [1].
[1] https://en.wikipedia.org/wiki/Positive_and_normative_economi...
But this isn't true. The wealthy in our society get more wealthy because those are the system we have. There's no law of the universe that says we can't increase taxes more, or spend more resources to ensure those at the bottom don't starve.
"when economists have to be the ones to tell them that making it a law that iron is equal in value to gold will not make everyone rich."
Economists are the same people hand-wringing and saying that we couldn't possibly tax those who have too much because ... profits.
If you have wealth, that means you have tools at your disposal to protect yourself and concentrate more wealth. If you do so.
If you don’t have wealth, you don’t.
If society likes it or not just means others are actively trying to take it away (or not) and might still overwhelm whatever tools you have/are using. It doesn’t mean you aren’t still advantaged by having them. All you need to see that is by looking at what happens when society tries to hurt those without wealth, and how much easier that goes.
Examples include everyone from warlords to drug kingpins to the ‘strongest lion’.
It happens literally without societies existing, to within things like the USSR.
Karl Marx was an economist. This is like saying engineering is fundamentally about building surveillance systems and serving ads.
No, it's not. But it's important to understand the fundamental forces at work that create such a system, because if you push too hard against those forces, they will break whatever system you create. The same way physics breaks things that push against it's forces.
History is littered almost to the point of being a landfill with examples of societies trying to go against economic forces and ending up worse off because of it.
You know, I used to think something along these lines but reading `Debt: The First 5000 Years` really changed my mind. We're not min-maxer's homo economicus. And I know the last 30 years of economics are trying to inject more psychology into the field (thanks Daniel Kahneman) but when it applies to human interactions at its core economics is a math solution to imaginary problems. The macroeconomics field is cool tough.
That doesnt mean the problems are imaginary, they are real, with real consequences.
The economic question of how many people will starve if we impose a new tax on food is a real one. The attempts to model the situation and derive an answer are imperfect. That doesnt mean they cant be useful, or even provide accurate answers. There are domains where spherical cows are entirely appropriate.
> a situation where resources are allocated efficiently and no further improvements can be made to society's well-being.
Of course the unstated rule that the AI left out is that this is the best you can do if you're not allowed to redistribute to the poor from the wealthy. Apparently that is by definition "inefficient".
Vilfredo Pareto of course is credited by some as a precursor to Fascism, acknowledged as an inspiration to Mussolini turning to Facsism, and somone who publicly welcomed Mussolini's takeover as step towards a "minimal state". But some people get really annoyed if you say he's a fascist.
This is ad hominem. That isn't just a criticism of the comment, but pointing out that it's literally illogical.
Heisenberg and von Braun worked for and sympathised with the Naxis. That doesn't make quantum mechanics and rocketry fascist.
I would definitely want to read an alternative history where the Nazis built their entire ethos on quantum mechanics and rocketry, but that's not the reality we live in.
We live in the one were someone who was anti-democratic because he thought the workers and the speculators would use it to take power from the aristocrats and the business owners inspired Facsism.
Milgrom wrote the book on auction theory, a subcategory of game theory, a domain of mathematics. Your criticism is with a consulting firm started by an economist. Not economics per se let alone mathematics.
Even then, your critique is unsustained. Milgrom is saying Google ran fair auctions. I don't believe anyone has put forward evidence to suggest it didn't. The point isn't that he's wrong; it's that he's irrelevant. Google's antitrust problem isn't in it running unfair auctions, but being the only auction house.
> "trickle-down economics"
Not economics [1]!
> Economists core moral value is maximum efficiency
Karl Marx was an economist.
[1] https://en.wikipedia.org/wiki/Trickle-down_economics#History
...and the idea that running fair auctions are subtly geared to concentrate wealth at the top is silly. Yes, there is a well understood principle that fair auctions will out-compete unfair auctions in the long run, but whether that concentrates wealth at the top is a consequence of a lot of other factors that are entirely external to the auction dynamics.
In what sense? If anything, the market failure of non-competitive markets almost implies the reverse: power prefers unfair auctions.
> whether that concentrates wealth at the top is a consequence of a lot of other factors
Correct. (And as an aside, sealed-bid auctions are particularly good for small bidders or bidders in an adversarial market, the latter for the same reasons a secret ballot is better for small parties.)
I do not follow your reasoning here. What do you mean by "the market failure of non-competitive markets"?
Seriously, this stuff is econ 101. Ultimately auctions are about bringing together as many buyers and sellers as possible to produce as much liquidity as possible. Fair auctions naturally attract more buyers and sellers than unfair auctions. If the auction is unfair to say, buyers, then you'll have fewer buyers using them than auctions that are fair or favour buyers, which in turn means that sellers will get access to fewer buyers, so they'll go somewhere else.
You’re going off about econ 101 but do not recognise the term market failure [1]?
> Fair auctions naturally attract more buyers and sellers than unfair auctions
Citation needed. (Plenty of examples of open-bid ascending-price auctions beating better systems due to familiarity, for instance.)
More fundamentally, auctions are subject to network effects. They benefit from anchor participants, i.e. repeat participants of substantial size. Those anchors tend to want perks for participating. The entire domain of market microstructure and ECNs rebating market makers is a story of broad participants’ desires for a fair system competing with narrow participants’ desires for a biased one. This effect is exacerbated when one side of the market is repeat or big while the other is ephemeral or small.
> Plenty of examples of open-bid ascending-price auctions beating better systems due to familiarity, for instance.
Yeah, I think we're talking past each other. I wasn't making a statement about "better" systems, and there are certainly other factors that can influence outcomes. I was only speaking of the principle that it is advantageous for an auction to be fair, and open-bid ascending-price auctions can be fair. I am quite certain that if you were to murder anyone who participates in a fair auction, then unfair auctions will be more successful, but that's outside the scope of the principle I'm speaking to.
What is the metric used to quantify this? Should any group be subject to this criteria? Do governments also fall under this?
(An aside: If an intelligent alien were to visit Earth, they will observe that solely profit chasing companies are bad but power chasing governments have done worse. Both are bad and ideally no entity should have a lot of absolute or relative power.)
Capitalism fully captured the discipline of modern economics - which are the priests of capitalism
Anything that isn’t capitalist is “heterodox” and not taken seriously or seen as adversarial to Capitalism (true)
And this:
> which are the priests of capitalism
Is emotional manipulation without substance, and not appropriate for HN.
Capitalism is a philosophy that believes chiefly property or capital is a real objectively measured thing, and that property ownership and labor inputs and the rightful owner of value increases are separate, but interconnected and managed by rational actors with increasingly perfect information
Specifically that there should be, not just that there is, but that the natural state of the world is that there is a distinction between capital ownership and labor inputs with respect to assignment of value an ownership
More specifically capitalism‘s core claim is that value creation is independent of labor inputs and as a result, the capital owner is the expected and rightful claimant to whatever portions of the value increase they can demand. A modern capitalist economist would point at the “failure” of the labor theory of value to actually be put into use in a economic system.
Further, it assumes that labor has perfect economic flexibility and mobility, and that the flexibility and mobility of capital is simply limited by trade agreements and lack of intermediaries in price setting across all markets
So the final end result of the process of capitalism is what the Mises Institute economists, and all of the Chicago school would effectively finally land at which is this pure economic liberalism otherwise known as anarcho-capitalism where every single individual is its own independent legal entity with perfect information making perfect rational decisions disintermediated from governments or groups.
No. Capitalism is an economic system based on private ownership of the means of production and their operation for private profit.
> interconnected and managed by rational actors
Rational actors are a simplifying assumption that toy models make, like frictionless surfaces. Both are taught in 101 classes. The difference is physics rarely has political implications while economics commonly does, so you get a lot of people who think their 101 class (or less) qualifies them on economics in general.
(Another point of defining rational systems is to observe how they deviate from reality. Sometimes, this yields useful insight into how humans behave, like hyperbolic discounting. Sometimes, it reveals inefficiency and corruption.)
> with increasingly perfect information
This article is about Milgrom, who wrote the book on auction theory. Auction theory is all about describing actors with imperfect and incomplete information.
> capitalism‘s core claim is that value creation is independent of labor inputs and as a result, the capital owner is the expected and rightful claimant to whatever portions of the value increase they can demand
No. Every macroeconomic model of note involves capital, labor and productivity to create value. Capital and labor compete for their claims to value production.
> pure economic liberalism otherwise known as anarcho-capitalism
These are vastly different concepts and at the fringe of both politics and economics.
You ignore the assumptions that define the system as implied and non-differentiated
So, not all economic systems define private property as a valid concept- yet you want to smuggle it in as though it’s a homogenous assumption across all disciplines
Capitalism is distinct in this and simply collapsing into “privately held means of production” is specifically a perspective long the Marx - smith axis ignoring the anarchist (Proudhon) and mutualist-indigenous hunter gather tradition
You need to explicate all assumptions in order to make a definition
Your definition is already corrupted by assuming property is a valid concept
I literally said “capitalism is an economic system based on private ownership.” That makes zero universal assumptions.
If you don’t have private ownership in your economic system, it isn’t capitalist. (There is a limited universality to the concept of means of production, though you can absolutely define an economic system without it. It will simply struggle to describe an industrial society. Similar to how a classic capitalist model struggles to describe the complexity of informal social interaction.)
> simply collapsing into “privately held means of production” is specifically a perspective long the Marx - smith axis
The existence of a spectrum doesn’t imply the non-existence of another. Like, if I say red light is about 700 nm, that doesn’t say nor even imply that I don’t recognise neutrinos.
> definition is already corrupted by assuming property is a valid concept
It’s an internal definition. Property is a capitalist concept.
Arithmetics is commonly defined as “an elementary branch of mathematics that studies numerical operations like addition, subtraction, multiplication, and division” [1]. The fact that addition is a term that, in this context, entirely derives its meaning from within arithmetics is both circular and okay and does not corrupt the definition. (We can do the same thing with natural sciences. Entomologists study insects, with the term insect being internally defined by entomology and/or biology.)
I've never heard anyone ever say that they knew economists for this. It seems likely you just made this up.
> Economists core moral value is maximum efficiency
Citation needed. I've never met or seen a single encomist who believes that efficiency is a moral imperative.
> We need a new version of antitrust that doesn't just target percentage power, but absolute as well.
Who is "we" and what empirical evidence suggests that we "need" this?
I bet you haven't. It's like asking a programmer if designing weapons is their primary imperative - nearly everyone will say no. But when you examine the number of people that make contributions that could be weaponized or is abused as a weapon by their superiors, the statistics tell a different story. Nobody wants to say they contribute to a surveillance state, but you also can't really deny that technology is being promoted so it can reach that stage.
Economists don't want you to know what sausage it is they're making. But the most successful and influential ones are certainly being used as adversaries to pro-consumer economic policy because that's why these people are paid in the first place.
> Who is "we" and what empirical evidence suggests that we "need" this?
"we" are people that import American software products, and the evidence is the crumbling and neglected shit-heap of smartphone regulation. It is obviating serious privacy laws, creating stagnant markets that do not benefit users, and forcing developers to cover first-party costs when it should be platform-holders and customers picking up the bill.
What do you think the economists at the FTC are doing?
What?
Very few. Non-competitive markets are a well known market failure [1] and a condition that an unchecked political economy tends towards.
Just because one's work can be used as a weapon doesn't mean that designing weapons is one's primary imperative. "primary imperative" isn't about outcomes, it's about objectives. I'm pretty sure the primary imperative of the Wright Brothers was to get something into the air, not to design a weapon.
The most influential economists are paid to.. make rich people richer? Who? Which? Economists are largely academics, which are paid by universities to teach and research without editorial control. If you have damning evidence of explicit corruption, please cite that versus some generality. It would also be good to know (with sources) the prevalence versus a one off.
If you look at NBER’s latest research (https://www.nber.org), which is a top economist think tank / affiliated with Harvard, I’m not seeing anything that (to me) supports anything you’re saying.
I have dated a powerful economist and hung around in those circles, and never have I heard anything remotely like what you’re describing. But it is true that I’m not an economist myself nor am deeply embedded to have certainty.
In a trial like this, a "star witness" was almost certainly paid for their testimony. Not to suggest that their opinions were altered by the payment but that businesses generally need an incentive for someone like this to go on the stand.
It's not an accident that this person said a bunch of things which were positive for Google, because they would have specifically looked for someone who would say such things.
This is shifting the goalposts. The parent commenter is talking about the "most influential" and it's likely that those who want to be influential will make public appearances and statements. Those are the sort in a field who are paid by large companies to publicly agree with them.
Google continuing to exist or blowing up won’t fundamentally change the economic outlook of the US at a broad population level. Changing taxation, regulation, and import/export policies will.
This seems to be focusing too much on the fact that it was initially phrased as "the most successful and influential" -- kinda suggesting "all of the most ..." -- and that's not really a claim I would support. I don't think that's really what the other commenter meant, either. Apologies for the confusion.
> Google continuing to exist or blowing up won’t fundamentally change the economic outlook of the US at a broad population level. Changing taxation, regulation, and import/export policies will.
Right, this is what the FTC is attempting to do with this lawsuit. A person who testified against them in court is an economist who has won both an Emmy and a Nobel Prize. That seems like at least one who's at or near the top in their field seemingly attempting to influence policy at a high level.
I was doing some work for the other side, and asked one of their lawyers why that professor would testify to something so ridiculous. Wasn't he worried that this would damage his reputation?
The lawyer told me that all the people in academics and industry know about the expert witness gig. If they see that the professor emeritus said something ridiculous at a trial they will assume that he wanted a new RV or boat or to remodel his house or something like that, and was offered enough money to pay for that in exchange for reading the claims in the case, spending maybe a few days writing up a report in support of the side that hired him, and spending maybe 30 minutes to an hour on the stand testifying about that report.
I then asked one of their expert witnesses what he was doing with his pay for his report and hour on the stand. He told me he was buying a Tesla Roadster which at the time was launching soon and sold for around $100k, and it was being paid for entirely from that one expert witness gig.
I imagine that's caused by bad news-reporting, but I don't actually have evidence of this.
Wealth tends to concentrate. So any action that creates wealth will also increase wealth concentration (in most cases). So unless economists, or anyone for that matter, try to construct theories that compensate for that externality, then most wealth creative theories will also have the effect of wealth concentration.
>Economists core moral value is maximum efficiency, but I think overall efficiency is better at a happy medium rather than a maximum.
You are talking about pareto efficiency vs pure efficiency and economists are well aware of this tradeoff. Vilfredo Pareto was himself an economist. A situation is pareto efficient when you can't make anyone better without making anyone else worse (your "happy medium"). Your claim is that "economists" don't care about pareto efficiency and only care about efficiency, which is absurd, they even defined the term. A good economist will lay out both the benefits and problems of their theory, and will propose mitigating measures for any negative externalities that it may cause.