David Munro

random shower thought, but it seems like it would be feasible to use these weird redactions on like citations/excel code etc. to get a pretty good handle on what is being redacted, lol. Especially given the volume of material.

Must have been citing Epstein Zin preferences paper!

dude, what issss this? Why is it the Epstein files?? Some references in the biblio are redacted! lol.

Never underestimate the quest for knowledge no matter how irrelevant it may seem in the moment

If I hadn’t gone to a public university, I honestly wouldn’t have ever bothered to watch a basketball game. But now? “Boxing out” is a top ten skills for boarding public transportation

Public debate links worsening job prospects for AI-exposed occupations to the release of ChatGPT in late 2022. Using monthly U.S. unemployment insurance records, we measure occupation- and location-specific unemployment risk and find that risk rose in AI-exposed occupations beginning in early 2022, months before ChatGPT. Analyzing millions of LinkedIn profiles, we show that graduate cohorts from 2021 onward entered AI-exposed jobs at lower rates than earlier cohorts, with gaps opening before late 2022. Finally, from millions of university syllabi, we find that graduates taking more AI-exposed curricula had higher first-job pay and shorter job searches after ChatGPT. Together, these results point to forces pre-dating generative AI and to the ongoing value of LLM-relevant education.

arXiv📈🤖
AI-exposed jobs deteriorated before ChatGPT
By Frank, Sabet, Simon et al

We added some more analysis in a revised version of the paper, and we do see some evidence of softening of investor demand. Thought you might be interested.
bsky.app/profile/dmun...

We also add a lot of other analysis and robustness checks. E.g. some folks seemed quite worried about post George Floyd stuff (which was in the first draft). We don't really see much here in terms of crime. Minn. still looks unique house price wise relative to other "defund" the police cities, etc.

Of course, it's hard to know why investor demand weakened. Could be due to perceptions of a more elastic housing supply, and less room for capital appreciation. Could also be due to regulatory uncertainty (e.g. legal challenges to 2040 plan that occurred). But an interesting result nonetheless.

On the demand side, we leverage a Redfin dataset that tracks the share of new purchases from "investors" (e.g. buyer had LLC in their name), and find that the share of purchases from investors was lower in Minneapolis (relative to other metros) in the post-reform period.

On affordable housing, the 2040 plan incentivized affordable housing creation via tax incentives and financing. We see stronger affordable housing creation, notably in the most affordable category (<31% Area Median Income), relative to other cities/townships in the region.

Finished up some new revisions on this paper. Mostly bringing a more prominent role of a DiD analysis alongside the syn. control (very similar results) and more on mechanisms. Notable findings, mechanism-wise, include affordable housing creation (supply-side), and a weakening of investor demand.

Here is Sharyn Alfonsi’s email to her ‘60 Minutes’ colleagues in full:

Ha, well then I'm extremely unhelpful!!

If I'm remembering correctly his work/point was highlighting this is a common thing to do (regressing on these estimates), that is inherently problematic... but I don't recall the details.

I don't know if I have much to add here. I recall Glenn Harrison having some work of a similar flavor (recollection from one of his talks I watched) but I don't recall which specific paper that was, if any (could have just been a point he was making in passing). Sorry I'm not more helpful!

Thanks for the update! Weird that the numbers between you and Cawley are so different. Some single postings seem to be posting for multiple position... but I would be surprised if this accounts for the difference (maybe save, e.g., central banks hiring many?) Here are two examples:

FRONTLINE | "The Warning" - Preview | Season 2009 | Episode 14 Long before the meltdown, one woman tried to warn about a threat to the financial system.

It's wild to me that the profession, macro in particular, essentially ignored Larry's role in laying the foundations for the 2008 crisis. Their dismissal of Brooksley Born's warning is also dripping with sexism, so par for the course. @akhilrao.bsky.social www.pbs.org/video/frontl...

And the Insured Unemployment Rate ticked up slightly to 1.26%. Overall unemployment picture during the shutdown points towards a gradually weakening labor market. Hopefully signing off! 🫡 @josephpolitano.bsky.social @estruby.bsky.social

Initial Claims (week ending Nov. 8th) up 12k or 5.6% (SA) and up 22k or 10% (NSA).

Since we don't yet have new official claims data, my last (?) rogue report (I missed last week!), for week ending Nov. 1st. CC increased 20k (SA) or 7k (NSA) last week.

US drew $900mn from IMF account as Argentina debt payment loomed Buenos Aires’s reserves rose by the same margin as Washington seeks to bolster Javier Milei’s government

The @financialtimes.com suggests US drew $900 million worth of SDRs from its (~$170 billion) account at the IMF and transferred them to Argentina so Argentina could make a scheduled debt repayment to the IMF. on.ft.com/4oBeQMm

Every month for the last 77 years, the Bureau of Labor Statistics has estimated Americans' unemployment rate

This streak dies tomorrow

Govt staff could not survey Americans about whether they're working or looking for work in Oct, the 1st time in 934 months

This blind spot will remain forever

OMG, now I'm just thinking about all the future headaches this causes in terms of having missing data in a time series...

And the Insured Unemployment Rate ticked up slightly: 1.24 to 1.26% over the week. @josephpolitano.bsky.social

Initial claims down slightly: 3k or 1.4% (NSA), 3k or 1.6% (SA). Thus, the rise in insured unemployment likely driven from lower job finding (people hanging out in the claims pool longer) not new layoffs.

Week #5 Rogue Unemployment Report, Halloween Edition 🎃 (hoping we won't have a Christmas Edition...). Data up to week ending Oct. 18th. Insured Unemployment ticked up slightly: +34k or 2.2% (NSA), +28k or 1.5% (SA)

New this week, Insured Unemployment Rate (CC/cov. emp.) (worth noting cov. emp. is, by construction, a pretty lagging measure of employment). Stayed relatively constant week-to-week (around ~1.25%), but has been trending upwards over the past year. @josephpolitano.bsky.social @mchinn.bsky.social

Initial Claims also pretty steady: -5k or -2.7% (NSA). -8.5k or -4% (SA).

Week #4 Rogue Unemployment Report (compiled from state data). Insured Unemployment relatively steady (+18k or 1.2% (NSA). -6k or -0.3% (SA). (excl. CO, MA, TN, not reported yet)

Government of the scammer, by the scammer, for the scammer