Don't believe the Moody's number.

We have two distributional consumption measures:

- BLS Consumer Expenditure Survey: 23%. The direct household survey of spending.

- BLS Distributional PCE: 26%. A BLS prototype designed to repair underreporting.

https://twitter.com/LeylaKuni/status/2059380155416739947

First, Moody's isn't measuring spending. It's bascally just income.

They pull top-10% income shares from one survey, apply them to BEA disposable income, subtract a modeled saving estimate, call what's left "outlays."

Antoine walked through

https://x.com/LevyAntoine/status/1985127826207772920?s=20

"Abduction and the Demand Curve"

A new paper with @EconTraina

The demand curve is the most basic object in economics. Hold everything else fixed, change the price, see what happens. Ceteris paribus. Day one stuff. Okay, maybe day 3.

But what does "everything else" include? Unobserved quality, local tastes, recent advertising. Things the econometrician doesn't see. A market's demand curve holds those fixed.

Now suppose you run a randomized experiment. Set a price, observe quantity, repeat. You've eliminated confounding. You have a causal effect. We love experiments. Perfect. Right? Right?

Are these the same things? This maybe isn't well-known outside of IO, but the answer is no.

When they aren't, what are you supposed to do? This paper connects the experimental literature with the structural IO demand estimation literature to make clear the interplay .

In the experiment, you've averaged over all those unobserved conditions. You know what happens on average across markets when you set a price.

You don't know what happens in THIS market, with THIS unobserved quality, at that price. The experiment gives the average demand response. Policy happens in a specific market.

Two markets produce the same quantity at the same price. An experiment can't tell them apart. But at any other price, they diverge. The demand curve is a market-specific object. So what bridges the gap? Good ole' Berry (1994_ inversion.

You observe a market's shares, prices, and characteristics. Inversion recovers the unobserved demand index, the δ*, that rationalizes what you see. It pins down WHERE on the demand function this particular market sits.

Prior work treats this as a computational convenience. Berry (1994, p. 249) compares it to "taking logarithms of observed data." Berry and Haile (2021, p. 40) call it a "trick." They leave as an open question what happens when invertibility fails, "perhaps involving partial identification."

We answer. Without inversion, even price-only counterfactuals are set-identified. The trick is not optional. Inversion is not just sufficient but necessary for recovering market-specific counterfactuals.

But when exactly do you need it? Berry and Haile (2021) say experiments "generally" don't identify demand. Angrist, Graddy, and Imbens (2000) showed that when demand differs across markets beyond an additive shift, IVs identify a weighted average of derivatives, not any single market's response. Imbens even reiterates the point in his Nobel lecture.

We first make "generally" exact beyond the linear case of AGI (2000). We characterize precisely when the experimental average price response equals every market's demand slope (if and only if additive separability holds, a knife-edge that every standard discrete-choice model violates).

So outside of that case, what are we to do? That hasn't stopped IO economists. Are they just making stuff up? No! Berry inversion baby!

Along the way, we can make a few more connections. @yudapearl asked whether ceteris paribus demand can even be formally defined in counterfactual language. We do that.

The demand curve is the unit-level counterfactual Q_p(u) for a market with realized conditions held fixed.

We also show the connection to Pearl's causal hierarchy. Experiments give Rung 2 (causal). The demand curve is a Rung 3 object (counterfactual). There's generically a gap between them. Berry inversion is what is called abduction in SCM to move between those rungs.

The econometrics and CS frameworks are saying the same thing, and the demand curve is the natural, well-developed setting to see it.
@instrumenthull @borusyak @lihua_lei_stat @asheshrambachan @causalinf @steventberry @jamesbrandecon @lewbel

"The cost of living crisis"

I'm completely annoyed by this phrase.

It sounds like it means something but doesn't. Of course we want things to be affordable. No one debates that.

But what are we willing to trade-off? And how can we lower costs? Those are the questions When politicians talk vaguely about "tackling the cost of living," they can justify almost any policy.

Price controls? Affordability. Subsidies? Affordability. More spending? Affordability.

It literally can mean anything. That's fine for politics but not for understanding.

I'm generally quite confused what decision Tim Wu read.

He doesn't accurately characterize the opinion or the state of antitrust law.

A few examples:

Right out the gate, Wu is wrong. That is not what the ruling said! Boasberg did not rule on whether Meta had a monopoly is personal social networking.

He rejected PSN as a separate market. That's a big difference.

🚨 HUGE: The court just dismissed the FTC's case against Meta (related to Instagram/WhatsApp)🚨

It is mostly a market definition decision (boring), but the Judge looked to do a good job on the economics behind market definition.

Let's go through it 🧵 The vibe is set in line 1: Heraclitus and never stepping in the same rivier twice.

Whatever you think of the initial decisions to not block acquiring WhatsApp, we're dealing with the world as it exists today.

Social networking overlaps with social media today.

Economists hate this idea (price controls) for good reason!

The policy fails to achieve its own stated ends and wreaks havoc along the way.

A few thoughts on this particular defense of price controls 🧵 Piece here nytimes.com/2025/11/16/opi…

Good response from @JohnHCochrane already grumpy-economist.com/p/price-contro…

I'll add/expand on a few points

Semiconductors are clearly one of the most important industries in the world. They run everything: AI, cars, the digital economy.

Yet people haven't focused on market competition in this industry.

New paper out with @geoffmanne, David Teece, and @MZunigaP remedies that Link to paper:

When you start studying semiconductor manufacturing, two things immediately pop out.

First: Moore's Law.

Tech people may see this as some inevitable technical progression. laweconcenter.org/resources/from…

🚨 Updated paper alert 🚨

"Market microstructure and informational complexity" with @GuthmannR

Why do we have organized markets with intermediaries like NYSE or Amazon?

Our answer: they drastically reduce the information burden on everyone else.

Here's how we get there 🧵 Hayek taught us markets require minimal information. Traders only need to know prices, not everyone's preferences or endowments.

This generated formal proofs, particularly from Jordan (1982)

The competitive market is uniquely informationally efficient, or minimally complex.

Two macro trends over 40 years:

- Rising markups
- Falling business dynamism

Are these related? Many economists say so.

@UpdatedPriors and I have an updated paper where we take an IO approach and look sector by sector.

Spoiler: The industry data tells a different story 🧵
The economic logic makes some sense:

Higher markups → firms have market power → barriers keep out competitors → less entry/dynamism

This story is so widely accepted. It's in the background of all antitrust discourse.

But what happens when we test it at the industry level?

Anthropic just released policy proposals.

There are actively bad ideas, generally good but unrelated to AI, and others that miss the forest for the trees.

I read the overarching framing as "how do we minimize harm?" not "how do we harness benefits and navigate trade-offs?"

https://twitter.com/AnthropicAI/status/1978157815911891094

AI is bad, so tax it! Okay?

It's not exactly clear what the externality is, but we don't just tax everything an some economist calls an externality.

🚨 2025 Nobel Prize in Economics goes to Mokyr, Aghion and Howitt 🚨

"for having explained innovation-driven economic growht"

The best prize in years! For most of human history, living standards barely changed. Then something shifted.

This is THE question. The hockey stick. Why did it happen? What happened?

Modern economics get critiqued for being meaursable stuff that doesn't matter. Not these guys

Matthew Lynn's WaPo piece claims economists were wrong about tariffs. Economists are a diverse group so you can definitely find mistakes.

But overall it's a bad piece.

Cherry-picking data, timeline confusion, and misunderstanding basic tax data and incidence.

https://twitter.com/BrianCAlbrecht/status/1976674351366971487

It's a bit of shadowboxing against I'm not sure who.

2 private economists who made predictions mid-April?

Okay. Lynn says "six months on" as if tariffs have been stable since April, so these are the right predictions to judge.

The Google Search remedies decision just dropped. The court rejected breakups, payment bans, and choice screens.

But it still ordered data sharing and syndication duties.

This reasoning will shape antitrust for years. My thoughts 🧵

https://x.com/nytimes/status/1962982232013906263

First, what the court actually ordered:

- No exclusive contracts for Search/Chrome/Assistant/Gemini
- Must share some search index metadata (one-time)
- Must offer syndication to rivals
- 6-year term

What it DIDN'T order matters just as much.

🧵 of EJ Antoni completely not understanding economic statistics, being partisan hack, or both

https://twitter.com/josephpolitano/status/1955041060197114136

x.com/garywinslett/s… x.com/garywinslett/s…

The first statement is true for any particular tax.

But it tells us nothing about comparisons across taxes. You can’t compare so easily.

Two simple examples to see why: Suppose a good is in fixed supply (perfectly inelastic). Increasing that tax generates no deadweight loss.

Are you going to compare to another tax?

Lesson 1: Elasticities matter still.

Planet Money had an interesting podcast tying together the BLS story with the broader questions about the role of economists and economics.

Some thoughts:
npr.org/2025/08/01/125… I think there's just a huge disagreement about what economists do. Everyone seems to want pure prediction.

There's the usual stuff about not predicting the Great Recession and Financial Crisis. I'm left wondering, compared to who? Which group of people saw it coming?

Highly misleading data from Lina Khan. This is all about new businesses, not small businesses.

Those new, growing businesses that will end up making jobs and innovations are not filing RPA lawsuits. They aren't doing anything on "more local control." The thing about small businesses is... They suck, in terms of job creation, productivity, etc.

The only good ones are a very few of the new ones.

And those ones don't stay small for long! Then they become big and then Khan would try to kill them

Michael Pettis, again in FT, ignoring how prices adjust.

Pro-tip: Write down a model. Check your equilibrium conditions. 🧵 of what happens when you don't Don't be confused by the currency/captial/trade stuff. What is he actually saying?

"country’s investment is constrained not by scarce saving but rather by inadequate domestic demand"

He is saying quantity is constrained, not by supply, but by demand. It's both!

The next stage of the Google AdTech case is remedies, with a hearing this Friday and initial proposals due May 5.

September 22nd is the other date to remember.

A little background as we move into this next stage. The liability phase is in the books.

Judge Brinkema found Google monopolized the open-web publisher ad-server market (DFP) and the ad-exchange market (AdX), plus an unlawful tying of the two products. Count III (advertiser networks) was tossed. static01.nyt.com/newsgraphics/d…

Cracking open the 115‑page ruling.

Going in, two Supreme‑Court cases oomed large:

- Trinko (refusal‑to‑deal) and
- AmEx (two‑sided platforms)

I want to focus on those in this first pass 🧵

https://twitter.com/antitrust_bot/status/1912876835869974873

Judge says: Google’s sin is tying customers, not refusing rivals—so Trinko safe‑harbor ≠ available.

The FTC's trial against Meta is underway.

So far, it seems the FTC (and Zuckerberg in 2012) fundamentally misunderstands network effects and how they interact with competition.

A short thread on the economics of network effects as relevant to this case: First, network *effects* are not inherently anti-competitive or bad.

That could just mean that the value to users grows as more users sign up. Then we want more users to sign up!

I explained this in a 2-pager about scale. But just read "network" laweconcenter.org/resources/scal…