John Spry retweeted
Let's go!!!
50 years from now, people look back on current dentistry practices the way we look back at bloodletting.
John Spry retweeted
PS - if we got an additional 1pp per year, or hey, maybe even 2pp of growth over next decade, it would be a boomtime the likes of which none of us in the West have seen in our lifetimes! Not a disappointment!
John Spry retweeted
We have great org econ models of the challenges of shifting org architecture, diffusion/co-invention models of inefficiencies, labor transition models and empirics from history, etc. etc. etc. Economist's views here are not just bloviating. 8/8
John Spry retweeted
Bottlenecks are very hard to predict ex-ante. Eg, here is McKinsey 2019 (they weren't the only ones!): 25-40% displacement from automation by 2030, without even seeing LLM-based AI coming! Suffice to say, this isn't going to happen. 7/8
John Spry retweeted
AI World folks are thinking of things like "Cursor and agentic coding spread very quickly!" Well, kind of (most software is still not written this way!) and *productivity* in software is still hard to observe (has price collapses in *any* category?!). 6/8
John Spry retweeted
It gets worse: tasks are parts of jobs which cannot be broken apart (as in @lugaricano and friends' great "Messy Jobs"). Incentives, career paths, training are linked to existing pre-AI tech. End-user knowledge takes time to learn. 5/8
John Spry retweeted
The big benefit will be when architects and designers begin designing with this new robot in mind. With historic architectural invention, that has taken *decades*. Note here I am assuming no technological bottleneck and massive cost advantage! 4/8
John Spry retweeted
Regulatory barriers will limit use in many places. It will take time to produce and distribute the machines. Local masons will in many cases refuse to work with it. Co-invention will be needed for odd sites, or unusual brick layouts. 3/8
John Spry retweeted
3) Imagine that tomorrow we invent a $10k robot that can build brick walls. What does it do to growth? Very little! Doesn't affect brick prices (can even raise them in SR!) or "roll fee" for the truck that brings it to you. But it's worse. 2/8
John Spry retweeted
On bottlenecks: 1) economists have been exactly right about little extra growth despite very fast tech change. They thought a group of econs I organized were crazy >2 yrs ago at the AI labs when we said "1pp extra growth/yr through 2030". 2) bottlenecks *still* not understood 1/8
John Spry retweeted
That is striking. The U.S. already had a much shorter maturity on its entire portfolio of Treasuries than most governments in other advanced economies, and it's going even shorter. But the surpluses are very far in the distant future. Means that the U.S. federal government is, in fact, increasing the duration mismatch between its liabilities and its assets. And taxpayers are exposed to more and more interest rate risk.
John Spry retweeted
This is indeed a big deal. As my coauthors and I show (nber.org/system/files/workin…), the TCJA's increase in bonus depreciation raised investment by 4% and GDP by 1% in the US. And this is an even bigger reform!
That said, I'd love to see estimates of how this will affect the average effective tax rate (AETR) in addition to the METR.
The METR is the tax rate on the very last dollar of investment, which means that it governs the optimal scale of new investments conditional on those investments going forward. Importantly, in Canada that last dollar of investment is usually done by a large firm that has received some kind of bespoke subsidy from both the federal and provincial governments for that specific project.
The AETR takes into account all the fixed costs associated with big investment decisions. It governs the binary decision of whether to make a big investment in the first place at all. It also reflects the rates at which "typical" firms and projects - that don't receive those aforementioned bespoke subsidies - are taxed.
Canada has always been pretty competitive on the METR, but much less competitive on the AETR. In my view, this is one of the factors that explains our poor investment performance over the past decade.
BIG tax change announced today. A “productivity mega deduction” broadens how much is eligible for immediate expensing. That cuts the effective tax on new investment in half. More to do, but this is big.
Details: canada.ca/en/department-fina… #cdnecon
John Spry retweeted
"Gains from Trade Increase with Greater Differences in Opportunity Cost: A Classroom Experiment" by Kathryn L. Combs & John A. Spry (@JohnASpry) from @UofStThomasEcon.
doi.org/10.1057/s41302-026-0…
John Spry retweeted
A few months ago, Charles Manski of Northwestern University visited GMU for our Buchanan-Smith Annual Seminar. During our off-talks, he told me about his work on polling and non-response rates. Being generally dismissive of polls as "mehhhhh social science", I was surprised by the discussion going in that direction. Then he told me that it is not widely known within the wider public that political polls have a 2.5% response rate.
Bear that in mind: 2.5%!
You need to make truly insane assumptions about the distribution of opinions of non-responders relative to responders. Yes, polling firms know this problem and they try to deal with it. But its still pretty wild and more inaccurate.
Manski just published in the Journal of the American Statistical Association his paper on this. He notably proposes a new statistical tool (TME for Total Margin of Error) which seems to perform quite well.
Now, why does this matter? Because of the Hong and El-Sayed stuff of last week (one lost despite being predicted to win, the other won by far less than expected). Moreover, polling accuracy in recent elections in my native Quebec have fallen from election to election.
But here is why it matters even more -- because most of the pundits I hear who talk online and in the traditional media have NO statistical knowledge whatsoever. And they blurt stuff with such certainty that its infuriating.
Rant over.
John Spry retweeted
U of M Economist Christopher Phelan confirmed as Chair of the Council of Economic Advisers, continuing Minnesota's legacy of shaping national policy. cla.umn.edu/economics/news-e…
John Spry retweeted
For those Industrial Organization economists seeking to spice up their BLP lecture on demand systems with differentiated product prices being correlated with product unobserved (to the econometrician) attributes (such as wild raccoons), this is a noteworthy teaching example!!
John Spry retweeted
(BTW, if somebody wants to pursue a fully general APS-style analysis of equilibrium in private-monitoring games with discounting, I think generalizing this beautiful paper to infinite states is the most promising path out there.)
John Spry retweeted
You can compare across the different types of tax rates because as long as you know you're starting below optimal tariff rates, you know incremental tariff change to overall national welfare is positive, whereas it's highly negative for the income and capital taxes due to their efficiency costs and this would be true with any standard set of elasticities that yield a positive optimal tariff rate.
John Spry retweeted
That comports with standard economic wisdom that intermediate inputs should be untaxed, while final goods should. jstor.org/stable/1910538
John Spry retweeted
We are very happy that our paper “Means-Tested Transfers in the U.S.: Facts and Parametric Estimates,” joint with Nezih Guner @NezihGuner and Christopher Rauh, is now accepted for publication in the Review of Economic Dynamics.






















