“Source: our own minds”

Mackenzie Crook’s magical suburban folk tale, #SmallProphets published by #PenguinBooks and #PuffinBooks down the years. A 🧵
1/

The AI Becker problem Who will train the next generation?

As @lugaricano.bsky.social puts it in this v smart piece from last summer

www.siliconcontinent.com/p/the-ai-bec...

Looks like software firms are developing an AI Becker problem

Well that was a good bit of analysis wasn’t it Duggee?

A-woof!

Hundreds of thousands of Muslims fought in the world wars. The armies that fought+won the second world war against fascism had an ethnic + faith mix that matches the Britain of 2025 more than that of 1939-45, so arrival of Windrush is the story of black RAF servicemen returning to Britain in 1948

Forthcoming in AEJ: Applied Economics “University as a Melting Pot: Long-term Effects of Internationalization”

Debates on international students often focus on capacity, funding, or competition.

My paper shows the main effects on natives are not academic or economic, but social.

👀👀👀

Gulp

found something rather baffling when researching my column this week…

I wanted to see if there was any evidence that AI tools were helping economists to make their research more readable. So I analysed the text of NBER working paper abstracts…

BBagel on Howland St. £1.50, reasonable for a bagel

It’s very good

A hot cross bagel

Delicious Judeo-Christian heritage

2) primarily. Left it vulnerable to normal interest rates and small shocks, let alone a pandemic. First PE owner does the sale-and-lease-back, then 3i split the service bits off, finally Macquarie do the leveraged buy out to load debt onto what are basically some upward-only leases. Sell those on.

Trying to work out whether this is

1) another piece of the post-COVID city slotting into place, or
2) just a company being loaded up with debt and then going bust.

Changes to UK working habits push car park group into administration National Car Parks hit by falling demand since the pandemic as people altered how they shopped and worked

fascinating to see the collapse of NCP as people’s driving/working/shopping habits have changed so much

(and possibly another ominous sign of town centre decline?)

www.ft.com/content/2e80...

How to design a tube line Plague pits, secret teams at TfL and a massive Bedfordshire warehouse: Making a new underground line is about a lot more than digging a big tunnel

Urban infrastructure is fantastically complex. Excellent long read by @londonermag.bsky.social

www.the-londoner.co.uk/how-to-desig...

Full house for today’s UCL Economic Geography Network seminar with Zara Shabrina - a superb talk on modelling Jakarta as a sinking city

Full house for today’s UCL Economic Geography Network seminar with Zara Shabrina - a superb talk on modelling Jakarta as a sinking city

Sounds uncomfortable

cc @sarahoconnorft.ft.com, @jburnmurdoch.ft.com ICYMI

This is a must-read. Acemoglu, Autor and Johnson on Pro-Worker AI 👇

Just been reading this. It’s very good.

Congratulations! Looks fantastic.

A couple of thoughts about AI and academia (in economics at least): you can’t look at Claude Code without concluding that it’s a complete game changer in the production of research. 1/

Delighted to announce UCL’s new Professor of Pudding Studies

It must be very hard to publish null results Publication practices in the social sciences act as a filter that favors statistically significant results over null findings. While the problem of selection on significance (SoS) is well-known in theory, it has been difficult to measure its scope empirically, and it has been challenging to determine how selection varies across contexts. In this article, we use large language models to extract granular and validated data on about 100,000 articles published in over 150 political science journals from 2010 to 2024. We show that fewer than 2% of articles that rely on statistical methods report null-only findings in their abstracts, while over 90% of papers highlight significant results. To put these findings in perspective, we develop and calibrate a simple model of publication bias. Across a range of plausible assumptions, we find that statistically significant results are estimated to be one to two orders of magnitude more likely to enter the published record than null results. Leveraging metadata extracted from individual articles, we show that the pattern of strong SoS holds across subfields, journals, methods, and time periods. However, a few factors such as pre-registration and randomized experiments correlate with greater acceptance of null results. We conclude by discussing implications for the field and the potential of our new dataset for investigating other questions about political science.

I have a new paper. We look at ~all stats articles in political science post-2010 & show that 94% have abstracts that claim to reject a null. Only 2% present only null results. This is hard to explain unless the research process has a filter that only lets rejections through.

No shade intended! In the UK present setup we is pretty clearly central government? Either directly (eg budgets for policy X) or indirectly (eg funding research on X, funding WWCs on X type things).

Most public policy is more experimental than we’d like to admit - we often don’t know for sure what works (where, for whom), so we have to try things out.

That’s difficult and risky! But everyone benefits from knowing the answers. *National Govts need to support and invest in that.

Evaluation is a public good, and we* should fund it properly - this is VG from @ioramashvili.bsky.social

Absolutely tremendous work here, no notes.

(Apart from to say that this proposal is by the actual president of RIBA)