Economics: Weekly Summary (October 20-26, 2025)

I spent the week skimming a stack of short, sharp blog posts about economics. They come at the same topic from a bunch of angles — trade fights, AI spending binges, money itself, wages and mobility, housing, and a few oddball pieces that are quietly important. I would describe them as a messy, slightly loud conversation. To me, it feels like folks are arguing in a kitchen while a pot of something complicated simmers on the stove. You hear snippets, you smell the spice, and you try to figure out if the dish is going to be good or over-salted.

The AI gold rush and the bill that might arrive later

There were more than a few pieces tussling over whether this whole AI boom is a genuine growth engine or a version of the familiar market froth. Read side-by-side, the voices split into two camps: the builders and the bean counters. The builders — the ones sketching out datacenter blueprints and chip roadmaps — sound excited and a little breathless. The bean counters — the folks worrying about returns and price power — sound quietly suspicious.

Start with Dwarkesh Patel. He pokes at Sam Altman’s watchword of massive weekly AI infra buildouts and asks practical questions: where will the power come from, who will staff rapid datacenter growth, and how much is real demand versus headline theater? He’d say the bottlenecks aren’t just chips. They’re power plants, docks, construction crews. It’s like planning to open a dozen new restaurants and assuming the city already has enough cooks and water mains.

Then there’s Dave Friedman, who turns the logic toward economics. He argues two things, both plain and unsettling: first, AI CapEx might not pay for itself. Investment is piling up, but the users — the firms and people who should pay for AI — may not hand over the kind of money vendors assume. Second, as agents and automation cut transaction costs, they don’t erase frictions entirely. New bottlenecks form; platforms will charge to gatekeep — and fees will follow. I would describe his tone as a bit like someone at a farmers market warning you that the pretty stall will charge for bagging your tomatoes.

Paul Kedrosky and thezviwordpresscom add color. Paul notes that AI CapEx sits awkwardly in GDP and trade measures. Nvidia and its GPUs are often painted as scarce, heroic things, but he says the shortage stories sometimes have a whiff of myth. Thezvi wonders whether the market’s current love affair with AI stocks is a bubble at all — or just a normal re-rating for a new tech wave. I’d say these posts feel like watching people at an auction. Some are convinced the painting is worth millions; others suspect the gavel will tell a different story.

A small but important echo came from Naked Capitalism about the financial strain inside the AI ecosystem, and from another piece that worried about “brain rot” — how tools that help can also dull thinking when used poorly. This links back to Dave’s point: even if we build huge capability, the human and institutional ways we use it will determine how much economic value actually lands in pockets.

There’s an everyday way to imagine this: imagine a small town invests in a huge new shopping mall because a celebrity store says it will open there. The mall gets built. The shoppers don’t come in the numbers promised. Rents are high. The contractor’s paid, but little else ripples through the town. Same with AI CapEx — the buildings can exist without the long-term demand to justify their cost.

Tariffs, trade tantrums, and the China conversation

Trade and tariffs keep bubbling to the surface. The tone ranges from academic skepticism about populist myths to real-world retail worries.

Richard Hanania took on the ‘China Shock’ literature and how it’s been used by protectionists. He’s blunt: trade with China did hurt some communities, no doubt, but the net benefits to the U.S. economy are being used clumsily by people who want political cover to stop creative destruction. I would describe his critique as a mix of history lesson and cold slap: trade forces change, and change is messy.

On the other hand, Mike "Mish" Shedlock and others looked at the practical fallout from tariff policy. Goldman Sachs’ breakdown — highlighted by Mish — says consumers will shoulder most of Trump's tariffs. That’s subtle but important: tariffs are often pitched as protecting domestic workers, but the map of pain shows ordinary buyers paying more for goods. Mish also flagged the Fed losing ADP data access, and how fragile information channels are during political gridlock — a reminder that policy debates aren’t only theoretical.

Then there’s the small-real-world voice: John H. Cochrane relays a CEO’s take on tariff timing. Companies often absorb costs for a while until old, cheaper inventory runs out; then prices jump. It’s the sort of lagged effect people miss in political speeches. I’d say the tariff debate here feels like someone complaining that you can’t just suddenly turn the thermostat up and expect the house to warm evenly.

The political theater was sprinkled through other posts too. Dougald Lamont and others used the Canada–U.S. spat over a Reagan ad as a case study in how trade rhetoric and politics can flare into real negotiations being shelved. And in a very small corner, AmericanCitizen ties China’s curbs on rare mineral exports to the messy political blame game: policy choices ripple into supply chains and jobs, and people point fingers.

The recurring note is practical: trade policy isn’t abstract. It lands in prices, inventories, and whether a factory can keep a roof on. If you like the meat of this, go read the authors' full notes; there’s a lot of juicy detail packed in the little posts.

Wages, work, and the slow-moving human side of markets

A few posts circle labor markets, mobility, and what happens when workers can’t — or won’t — move.

Mish flagged that only 11% of Americans moved in 2024 — the lowest in decades. That’s not just a stat. It suggests people are stuck for reasons that mix high housing costs, remote work, and economic fear. Maxwell Tabarrok then pushed back into theory, asking whether monopsony explains minimum wage debates. His read: monopsony can justify minimum wages in some special markets, but low-skilled labor largely acts like a competitive market. I would describe his argument as the sort that nudges you to stop repeating the simple textbook lines and think about where power really sits.

Jamie Lord brought the Brexit angle into view: politicians admit the damage Brexit did to labor supply, yet they keep tightening immigration rules. That’s the kind of policy incoherence that makes you grind your teeth. It’s like promising to build a house, then refusing to let the carpenters in. The housing tracker by Kevin Erdmann also loops back here — rents aren’t a simple supply story; demand swings and the 2008 mortgage crackdown left persistent scars.

These posts together sketch a theme: labor is sticky. People don’t move like chess pieces, and policy that assumes they will is often naive. The personal tensions matter. Folks want security. They weigh moving house against kids’ schools, job risk, and social ties. That's not a footnote; it’s the main line in how markets actually behave.

Money, banks, debt — it’s mostly information

One of the fuller, quietly important threads was about what money really is. Dougald Lamont wrote a short, nice piece arguing that money is not a static commodity but a system of information and agreements. Banks create money when they lend. Simple sentence, big implications.

This pops up in other places too: Political Calculations gave a snapshot of who owns U.S. national debt and how things shifted in September 2025. Figures like $37.6 trillion and changes in who the creditors are are useful for the headline, but they’re also a reminder: the national debt is a network of claims, not just a scary number. Minh Quang Duong talked about Gen Z and credit, which shows how the system creates winners and losers in access to that information and trust.

I’d say that when writers remind us money is information, they’re asking us to rethink a lot of default policies. If money is a ledger, then debt relief or restructuring looks like changing records, not breaking physics. That makes big policy options feel a little more possible — and also easier to botch if you don’t understand the plumbing.

Housing, food, and the price squeeze

People care about the kitchen bill, and the posts this week didn’t forget that. Naked Capitalism dug into food inflation — beef, coffee, eggs — and how volatile those categories are. Their point was small but human: food prices bite and they bite fast, especially for poorer households.

Housing got a spotlight too. Kevin Erdmann argues rents primarily drive home prices and that supply-demand hasn’t been in equilibrium since 2008. That’s a neat, if annoying, way to summarize a decade of policy and finance choices. If rent goes up, house prices follow — and that pushes people into mobility decisions, into wage fights, or into just staying put.

Pair that with the low mobility data and you begin to see a personal map: people paying more for food and rent, moving less, and feeling pinned down. This is the kind of slow bleed policy discussions often miss because they prefer quick, shiny graphs.

Small business, market tweaks, and plain human puzzles

A handful of posts look at economics at the micro level — what a musician, a restaurateur, or a hobbyist actually faces.

Branko Milanovic told a little story about a night in New York, a conversation about whether to nationalize small restaurants, and how people argued across Russian and Western terms. It’s charming because it’s small and real. He ends relieved that someone agreed with his anti-nationalization tilt. I’d describe his note as a reminder: abstract policy talk has to pass the smell test of real lives and places.

Arthur Turrell offered a neat little solution to ticket resellers: charge market price but rebate at the venue to punish scalpers. It’s an elegant, slightly cheeky policy patch. No Dumb Ideas proposed a marketplace for whole hobbies — imagine selling your guitar and the teacher, the sheet music, the group you jam with. It’s a tidy little insight about transaction value: selling relationships and knowledge is usually worth more than selling stuff alone.

These micro-posts matter because markets are made of small, stubborn human transactions. Cheap policy wisdom often forgets that.

Historic echoes: cities, scale, and moral hazard

Not all posts were about the now. A few looked back and sideways.

Bram Hubbell wrote about trade cities before 1450, using Constantinople as an example. It made me think about cycles: trade hubs rise, shrink, and rise again. The pattern repeats. Naked Capitalism also ran a piece reflecting on E.F. Schumacher’s idea that bigger is not always better. That dovetails oddly well with today's AI building mania: bigger data centers might be impressive, but bigger isn’t always better for people.

And there was a concise, useful piece on moral hazard by bookofjoe with a neurosurgeon example. It’s little, but it’s exactly the sort of anecdote that sticks. You see how incentives nudge behavior. That’s the real fabric of economics — little incentives shaping big outcomes.

Politics, geopolitics, and a pinch of culture

Several posts tied economics to politics and geopolitics. Philoinvestor collected Europe-themed essays about the Eurozone’s structural tensions and the need for strategic autonomy. Razib Khan interviewed Noah Smith about Japanese politics and how both Japan and the U.S. are shifting culturally and economically. It reminded me that economic debates are always shadowed by political stories.

And then there are the toss-in pieces that make you look twice: Shawn K compared modern AI to alchemy — the idea that a little cleverness can transmute near-nothing into value. It’s a neat analogy. Imagine turning code into revenue like an alchemist turning lead into gold. Some folks in the blog stack are thrilled about that potential; others shrug and ask who pays for the furnace.

The recurring hooks and little frictions

If I was to pin a few repeat notes across the posts, they would be:

• Money, data, and information matter more than people often admit. Several posts circle back to this. When you see someone arguing as if money is a pile of coins, that’s a simplification — and the knotty parts of policy live in the information flows.

• Infrastructure and capacity are underrated. Whether it’s chips, power, or skilled labor, the limiting factor often isn’t demand; it’s the ability to supply reliably.

• Policy is noisy and slow. Whether it’s tariffs, migration rules after Brexit, or housing supply lags, politics often contradicts the technical fixes it claims to seek.

• Small things add up. Food price spikes, ticket scalpers, hobby marketplaces — these tiny markets tell the story of distributional pain and opportunity. It’s not glamorous, but it matters.

• There’s a tension between optimism and realism. Many writers are excited about technological change. Many are also skeptical about who captures the gains and who ends up paying the bill.

You see these threads again and again. It’s almost like listening to a radio station where the same two songs play on repeat but with different covers. At times it gets repetitive. At times it feels clarifying.

A couple of stray but useful notes

• If you care about the nuts and bolts of tariffs and who pays, look at the Goldman Sachs breakdown mentioned by Mike "Mish" Shedlock. It’s the kind of piece that moves talk into tax incidence.

• For a tidy read on how money and banks actually work, Dougald Lamont has a short piece that cuts through classroom metaphors. It’s helpful if you’re tired of hearing money described like some magic commodity.

• For those who chase markets and charts: Derek Thompson flagged that uneasy chart linking stock highs and declining job openings. It’s an eyebrow-raiser and worth your time if you like connecting financial markets to labor reality.

• If you like humans-in-the-small, Branko Milanovic and Arthur Turrell wrote pieces that are small but land heavy. They remind you that the big policy lines are made of lots of ordinary choices.

I’d say the week’s blog crop is useful in part because it doesn’t converge on one tidy truth. That can be maddening. But it also means there’s a lot to nudge at: power, incentives, data flows, and the small market frictions that add up into big outcomes. If you want more meat than these hints, follow the links and read the full posts — the authors keep the goods in their pieces. There’s more to chew on there, and I suspect the next week will bring more arguing, and more reasons to care.