Economics: Weekly Summary (September 29 - October 05, 2025)

This week’s economics blog chatter reads like a neighbourhood full of people shouting from their porches. Some are warning about a fire; some are counting the cost of the storm last month; others are arguing about who left the gate open. I’d describe the voices as anxious, combative, and oddly repetitive — not in a boring way, but like a song that keeps changing the chorus. To me, it feels like everyone’s trying to name what’s broken before we agree on how to fix it. Read the original posts if you want the full grit — I’m only nudging at the highlights and the bits that stuck in my head.

The AI bubble: loud spending, hollow promises, and who pays the price

If there was a dominant theme this week it was AI. Multiple writers kept circling the same idea: big, hungry spending on AI infrastructure, and a very real chance that this will blow up into something nasty for ordinary people.

Start with the picture-machinery notes. Edgar E Peters pokes at a pattern: tech firms spending aggressively on AI looks a lot like past bubbles. He even compares the scale of corporate AI capex to massive public works. The picture he paints is almost absurd — like a town building a new stadium for a team that hasn’t shown up yet.

That simmering worry shows up again and again. Derek Thompson and Will Lockett each run down the dot-com analogies. Their point isn’t novel, but they make the risks feel concrete: if the AI circus collapses, the spillover could be deep. I’d say Thompson’s framing — that capital concentration in AI can crowd out other areas of the economy — works like a bright warning light. It’s the kind of thing you might shrug at until it’s your small business that can’t get a loan.

There’s a sharper, more human-angled take from Michael Spencer. He treats AI not as a tech curiosity but as an economic force shaping jobs, wages, and future generations. He’s worried about deskilling, fewer entry-level openings, and an ever-widening gap between the top 20% and the rest. To me, his coverage felt grim — like watching a factory slowly automate away the jobs of your cousin. He ties this to demographics too (fewer young people in the workforce), which makes the headline feel more structural than cyclical.

Naked Capitalism and its contributors add a healthy dose of skepticism. Their notes — and the piece by Servaas Storm in the same vein — flag GenAI pilot failures and the reality that scaling models doesn’t equal customer value. The claim that 95% of GenAI pilots don’t grow revenue is the kind of stat that should make boards pause. It’s like someone promising you a fancy espresso machine that only makes cold coffee.

Two practical tangents keep turning up. One: the venture math is flimsy. ARR multiples and headline valuations hide lousy unit economics, says Credistick. Two: investors and corporates may be suffering from FOMO — they’re pouring cash into AI because everyone else is, not because the market needs the product. That’s a classic stampede setup.

Read any one of these and you’ll get a similar mood: high spending, thin returns, and a social cost that falls hardest on people who already have the least.

Jobs, youth, and the academic squeeze — who’s getting squeezed

The labour narrative was messy. On one hand you have signs of resilience; on the other hand you have worrying pockets of weakness.

Mike “Mish” Shedlock highlights a sobering data point: since April 2023, the unemployment uptick has been concentrated among the young (16–24) and Black workers. That matters. Young people are supposed to be the flexible part of the labour force; they’re supposed to be the ones trying out gigs and learning on the job. When that pipeline chokes, it’s not just short-term pain. It’s lost experience, lost pensions, lost buying power. I’d say it’s like watching a couple of rows in a schoolyard become empty while the rest keep playing.

Then there’s the slow collapse of the econ PhD job market. Christopher Brunet walks through the numbers: a steep fall in tenure-track openings and a placement rate that would make any grad student swallow hard. This is the academic equivalent of a neighbourhood diner suddenly closing. It’s not just about jobs; it’s about future teaching, research, and the intellectual supply chain. Brunet notes causes like falling undergrad enrollment, the demographic cliff, AI, and reputational damage to the field. The proposed escape hatch? Decentralized outlets like Substack. Sounds a little bit like moving the church service from the big cathedral into people’s living rooms.

The political scene makes the jobs story worse. Dean Blundell and others point to private sector job cuts and the immediate fallout from government standoffs. Shutdowns, furloughs, and delays ripple through local firms and contractors. Mish also breaks down the furlough risk: hundreds of thousands of federal employees could be idled if the funding fights go south. That kind of disruption is not academic. It’s real bills, closed daycares, and postponed mortgage payments.

So, the labour picture is uneven. Some sectors and metrics look okay. Others — the young, early-career academics, certain minority groups — feel like they’re getting the short end over and over.

Housing: it’s not all mortgage rates — supply, rents, and stubborn prices

Housing commentary this week pushed against the simple headline that ‘mortgage rates ruined everything.’ Two pieces from Kevin Erdmann argue that supply matters more than rates. He’d say prices are mostly about land rent and shortages of homes. That’s a useful tweak. If prices are being driven by tight supply, then lowering rates alone is like giving someone a bigger spoon while their bowl is still empty.

Political Calculations tracked new-home affordability. The median new home price and the share of income consumed by mortgage payments stayed high; new homes remain just above conventional affordability thresholds. That matters to real families. Even a small move in rates won’t fix deep supply shortages. It’s like blaming traffic on the rain when half the lanes are closed for repairs.

Erdmann’s second note teases apart mortgage vs rent. If renters can afford higher rents, then buying prices won’t magically drop just because rates tick up. In some cities mortgage/rent ratios vary wildly. The point I kept circling back to: don’t treat mortgage rate changes as the only lever. Zoning, land use, and construction pipelines matter more for long-term affordability.

Money, misinformation, and the politics of “how money works”

There were a few pieces pushing against common myths about government finances. One of the louder notes came from Dougald Lamont, who took issue with Canada’s Parliamentary Budget Officer for treating government finance like household finance. Lamont emphasizes that a sovereign currency issuer is not a household — an old but still spicy point in public debate. He brings out historical examples and warns that bad advice could steer policy toward unnecessary austerity.

Lamont’s follow-up on money creation in Canada reads like a mini-lecture on how banks and the central bank interact. He frames money not as a thing you hoard under the mattress, but as social obligation and balance-sheet entries. If that sounds academic, the real implication is political: governments can, under some conditions, create the money they need to meet social commitments. Whether they should, and how that affects inflation and distribution, is a fight.

Richard Hanania threw another flag into this field with his call for economists to fight misinformation. He argues economists haven’t been as active in countering false beliefs as climate scientists or public health folks. That resonated with the Lamont pieces because a lot of the confusion about deficits and money is exactly the kind of “false belief” he wants economists to challenge.

I would describe the debate as loud and messy. It’s a little like a cook-off where everyone’s using a different recipe for stock. Some folks want more transparency and correction. Others warn that simplified messages can mislead as much as they illuminate. Read Hanania’s call and Lamont’s replies if you like the smell of actual policy fights.

Politics, tariffs, and old bailout patterns

Politics kept spilling into economics this week. Aaron Rupar writes the kind of piece that feels knee-deep in policy history: farm bailouts, tariffs, and how winners and losers get carved out. He argues that tariffs under the Trump administration worsened conditions for farmers, and that bailouts are coming because policy created the problem in the first place. It’s a classic political economy loop: policy promises to protect X, but protection invites retaliation and then taxpayers pick up the bill.

Naked Capitalism had a number of posts tying political choices to economic risk — from threats of military action to domestic policy missteps. The message here is that political instability isn’t just rhetorical. It moves markets, trade flows, and investor confidence. That’s not new, but the writing this week made the connection feel sharper: policy choices have direct economic consequences, sometimes slow and sometimes immediate.

Healthcare, time-use, and peeling back finance to basics

A few posts weren’t about markets or bubbles but about deeper systems.

Santi Ruiz hosted Anup Malani explaining how to bring down healthcare costs. It’s practical, not ideological. Malani talks about how CMS and economists can tackle fraud, improve efficiency, and use data. Healthcare is a giant, creaky machine, and small, smart fixes can cut costs without tearing the system down. That idea is worth a few deep reads if you care about policy that actually helps people afford medicine.

Arthur Turrell wrote on time-use and the potential of using wearables to get better data. It’s a neat pivot away from headline economics. Time is an economic resource, he argues, and we’re pretty bad at measuring it. Better data could inform policy on everything from transport to work hours. It felt like a tidy, hopeful piece in a week full of alarmist angles.

And then there’s Brett Scott with the “Ten Layers of Finance” — a reminder that finance sits on top of things like food, transport, ecosystems. He’s basically saying: don’t treat finance as the economy. It’s the icing, not the cake. That metaphor stuck with me. It’s easy to be dazzled by the shine on top and forget the loaf underneath.

A small digression: Thomas Klaffke in “Rabbit Holes” pushes a different kind of fix — cultural and community resilience. There’s a little note there about smoking being romanticized as a social anchor in a disembodied digital life. Odd tangent, I know, but it ties back: when institutions fray, people look for rituals to stay human.

Startups, venture math, and the long tail of failed promises

A couple of posts pulled back the curtain on venture metrics. Credistick argues that simple ARR multiples are a lazy shorthand. They mask churn, unit economics, and how much cash a company actually needs. That’s a practical point for investors and founders. It’s like using the odometer on a car to guess how many repairs it’s had — possible, but risky.

Alex Wilhelm worried about AI revenue narratives and IPO receipts. There’s buoyancy in the market, but also a nervousness about where actual revenues are coming from. The interplay between big public exits and private burn rates is a place to watch. If you like the theatre of startups, these posts read like a rehearsal for a farce.

The noise machine: media, watchdogs, and the week’s sentiment

Several entries were about media itself. Naked Capitalism got several shout-outs and fundraising pleas. Their supporters — including essays like the ones praising it as a “bullshit detector” — pitch the site as a counterweight to mainstream spin. It’s comforting in a way. Like a trusted neighbour who keeps an eye on the street. The tone of these pieces tends to be sharp and partisan, but the simple fact is they’re satisfying a hunger for scrutiny.

There were also a few regional and cultural notes that softened the week’s otherwise technical tone. İsmail Şevik wrote a short, pointed piece about emotional scarcity in his society — less economic analysis, more social diagnosis. It was raw and a bit unexpected among the policy papers. That emotional thread keeps popping back: economics is numbers, sure, but it’s also about how people feel and how communities hold together.

Threads that keep appearing — the patterns I couldn’t ignore

Some patterns came up so often I started to hear them like a background hum.

• Fear of concentrated capital and what it does to the rest of the economy (AI spending crowding out other investments).
• Worry about labour-market scarring for young people and those already disadvantaged.
• Arguments that housing shortages, not just interest rates, are the main driver of long-term unaffordability.
• A recurring pushback against simplistic views of government finances; several pieces wanted to change the story about what governments can and cannot do.
• A taste for practical policy fixes in healthcare and data collection, sitting next to big, sweeping warnings about bubbles and empire.

I’d say this week’s conversation was less about neat answers and more about naming the right questions.

If you like the smell of real argument rather than soothing consensus, click through to the posts. Michael Spencer lays out the social costs of AI hype. Dougald Lamont and Canada’s debate over money make you rethink that old deficit metaphor. Christopher Brunet will make any aspiring academic check the exits. And if you want something that reads like a mix of tech history and cautionary tale, Derek Thompson and Edgar E Peters both give you good narratives to chew on.

I’m not trying to be the map here. Think of this as a weather report. On some streets the sky is clear. On others the clouds are thick and thunderous. There’s thunder about AI and money. There’s a slow drizzle over housing and jobs. Every so often the sun peeks through with concrete policy ideas, like better time-use data or healthcare fixes.

If you want to keep poking, you’ll find the full textures on the authors’ pages. They do the heavy lifting. I’d say next week watch two things: who spends what on AI, and whether any of those health and housing fix proposals actually make it past the idea stage. It’ll tell you whether these are just loud warnings, or the start of something that changes how regular folks pay rent and clock into work.

Go read them — the posts are a mix of anger, data, and a kind of weariness. That weariness felt real to me. Like after a long community meeting when everyone goes home, but the problems are still on the table. Maybe the papers this week are the beginning of a longer conversation. Maybe they’re just noise. Either way, they’re worth following.