good piece, #3 seems most likely to me. not AI productivity but a probably temporary shift in composition. however if I had to bet (and technically I am betting by not selling VTSAX) I think we will see AI productivity gains in 2026

How many of you got both of these right on the NYT faces of 2025 quiz?

"It would also take a few years to convince me that we were already enjoying an A.I.-fueled productivity boom and not just a volatile, temporary blip." - Jason Furman NYT

One reason I emphasize fog over a clear story for the GDP boom and jobs bust is that the story was exactly the opposite in the first part of the year. Looking at the data in the spring and early summer the puzzle was why was GDP collapsing while job growth was surprisingly high.

New piece in @nytopinion.nytimes.com trying to explain the GDP boom combined with job bust. I don't have a definitive answer--in fact I'm skeptical of grand unified theories given how the data is often revised and noisy. As always, only update a little on this news. www.nytimes.com/2025/12/23/o...

Depending on how you look at it growth in Q3 was very very strong or very strong or just possibly merely strong. Annual rates: GDP: 4.3% Real final sales to domestic purchasers: 2.9% Average of GDP & GDI: 3.4% GDI: 2.4%

It is with a heavy heart that I tell that the Michigan Econ community learned yesterday about the death of our chair John Leahy. He was such a bright economist and cheerful human being. I personally owe him a lot. RIP, John, the world today is darker without you. 🖤

Tragic death of John Leahy. An amazing economist, wonderful adviser, consistently modest and kind, he will be greatly missed. May his memory be a blessing.

Fascinating econ book recs from @jasonfurman.bsky.social Grabbing a couple for my high school. #TLSky #BookSky 💡📚 The Best Economics Books of 2025 - Five Books Expert Recommendations share.google/nXVLdvM2pupd...

The deficit is more daunting than it was than it was when Romney ran for President in 2012. I'm glad to see him updating, appreciating that taxes should be part of the solution. And not just minor tax "loopholes," he wants to seal the "tax caverns." www.nytimes.com/2025/12/19/o...

An interview on AI, bubbles and more with the excellent @shiringhaffary.bsky.social. www.bloomberg.com/news/newslet...

It appear that BLS made a big judgment error in its shelter calculation (effectively assuming 0 in October), leading to inflation understated. It is, however, very unlikely this error was political. If anything the opposite: they stuck to algorithm rather than using judgment.

Surprisingly tame CPI report. But should discount some due to shutdown-induced price distortions. But only discount some. Core CPI annual rates: 3 month: 1.6% (lowest since Feb-2021) 6 month: 2.6% (lowest since July -2025) 12 months: 2.6% (lowest since Mar-2021)

I can't wait for the stories about why everyone was all wrong about the economists being all wrong about the tariffs.

New @nytopinion.nytimes.com. I argue dissents are growing at the Fed, could grow more under the next Chair, & that's good--reduces groupthink & protects independence. But Fed will need to change how it does business, less forward guidance, more actual rate changes. www.nytimes.com/2025/12/16/o...

The headline jobs numbers are almost uniformly weak. I would discount some due to shutdown and Federal retirements. But only some. 41K jobs added in Oct & Nov 4.6% unemployment rate in Nov, up from 4.4% in Sep Private wage growth slowed Labor force participation rose

One problem with antitrust thinking and enforcement is how little updating there is based on experience and evidence. Much less than you see in macro policy--and to be clear there is not enough of it in macro policy either. Take the case of Roomba. A 🧵 .

A majority of CEOs say AI has *increased* headcount for senior leadership, mid career and entry level jobs. A majority of large institutional investors say it should do so going forward. Per a Teneo survey: www.teneo.com/app/uploads/...

AI is truly charming.

How some geopolitically relevant economies are doing relative to their pre-COVID trends.

A weakish jobs report but, as usual, not completely clear. Surprisingly strong 119K jobs added but 3-month average is only 62K. Unemployment rate up 0.12pp to 4.4% but LFPR and employment rate up and U-6 (broadest measure) down. Average hours flat.

This is not how I would describe myself.

Jobs numbers, per ADP. Back in the black but not by much--and 3-month moving average basically zero.

If average work hours per person fall dramatically in the future due to AI is that more likely to be because: 1. AI substitutes for labor, market-clearing wage falls below zero, people not hired. 2. AI complements labor, wages rise, income effects so people don't work as much.

I am also ranking Burhan Azeem as my #1 choice. I agree with everything Josh says. And I've also personally been really impressed with Burhan's amazing combination of commitment to Cambridge, deep thoughtfulness about the issues it faces and pragmatism and inclusiveness in his approach to them.

I am also ranking Burhan Azeem as my #1 choice. I agree with everything Josh says. And I've also personally been really impressed with Burhan's amazing combination of commitment to Cambridge, deep thoughtfulness about the issues it faces and pragmatism and inclusiveness in his approach to them.

People post too many memes.

Being an Anglo-Saxon country not great for borrowing costs right now.

This graph may or may not say a lot about the world over the last decade.

CEO pay has stagnated for the last quarter century.

Fixed by John Cambpell & Tarun Ramadorai is an important book--drawing on up-to-date research, grounded in experience in the US, UK and India, argues persuasively for more consumer finance regulation--an issue that has almost entirely faded from the debate. www.goodreads.com/review/show/...

The CPI-based ecumenical underlying inflation measure was 2.7% in September, up 0.1pp from August (and the same as the PCE-based measure for August). This is the median 7 concepts over 3, 6 and 12 months. My own judgment is closer to 2.5%, maybe a tiny bit below.

The Federal minimum wage was established in 1938. It was in effect for about 85 years. It has now, for better or worse, been effectively abolished.

If you call a bubble early you're not prescient, you're wrong. If you bought the stock market on the day Alan Greenspan announced "irrational exuberance" you would have made money no matter what day you ended up selling. (Don't @ me about inflation adjustment.)

The government made the reasonable decision to release CPI data because needed to calculate Social Security COLAs. Quick summary, core CPI annual rate: 1 month: 2.8% 3 months: 3.6% 6 months: 3.0% 12 months: 3.0%

The stock market is not the economy, Part 387 Productivity growth is often lagged, Part 129 Productivity growth before and after the dot.com bust

I discuss bubbles, the macroeconomy and more with Ross Douthat. www.nytimes.com/2025/10/23/o...

CBO is as good as it gets. But the world is hard to predict.

If you're not surprised & puzzled about Treasuries being <4% then you're either a seer, incurious or have better things to wonder about. Large deficits (albeit smaller than expected), capital demand, Fed independence risk, persistent inflation, uncertainty, all go the other way.

Three months ago I was wondering why jobs were so strong while GDP was so weak. Now with additional data and revisions I'm wondering the opposite. As always would place more weight on data messiness than any structural explanation (e.g., productivity, frozen labor market, etc)

According to official government data the unemployment rate was unchanged while the economy added 60,000 jobs in September. In Canada.

The other day a student asked me about the prevalence of insider trading in prediction markets. I now have an answer.

Core PCE inflation rising lately, running ~3% annual rate. But core PCE inflation ex portfolio services slowing lately, running just above 2% (when remeaned to reflect it usually runs low). Some of the reinflation we've seen is rising stock prices counting as higher inflation.

In 2000 & 2016 the Dems won the Presidential popular vote but lost the electoral college. Just chance the opposite hasn't happened. Eg 2004: Kerry came closer in the EC than the popular vote. This is the pop vote vote margin Dems would have needed in past elections to win EC.

Do you prefer Option 1 or Option 2? And why?

"In one respect, Breakneck is a classic work of a hedgehog. Dan Wang knows one big thing: China is run by engineers, and the United States is run by lawyers. He then applies that idea to many topics..." My review of the excellent Breakneck. www.goodreads.com/review/show/...

The United States ended FY 2025 with $30.287 trillion of debt. That is likely 99.9% of GDP, up from 98.2% last year. Note CBO's January forecast was 99.9% of GDP. In June CEA projected that if OBBBA passed it would fall to 96.2% of GDP this year.

Office hours today: Student 1: Last summer I got to visit the New York Fed. Student 2: I've visited eleven of the twelve regional banks--but I still haven't been to Dallas. (And this is in a principles of micro class. Oh, and student 1 added she had been to Cleveland too)

If you ask small businesses about job creation plans, job openings, inventory plans, earnings or capital expenditure plans they're much more dour than they were a few years ago. But they express much more optimistic about the future than before. www.nfib.com/wp-content/u...