Archie Hall

Agreed, DMP work was fascinating

Synthetic controls with a heavy weight on the US always worry me a little, but this is a careful bit of work from Nick Bloom and co with (as I see it) two big implications

1. They put the Brexit hit to GDP at 6-8%..!
2. Their measure of the damage hasn't bottomed out yet

www.nber.org/papers/w34459

Why the developing world needs wider streets Roads are the arteries of cities, making the whole more than the sum of its parts. Cities of the developing world need more of them.

Great piece: www.worksinprogress.news/p/why-the-de...

www.economist.com/leaders/200...

From an Economist leader in August 2000

We're fortunate, then, that the shutdown looks like it's ending and that the attacks on the independence of the BLS have diminished. No substitute for government data quite yet.

Here's the link to the full piece again: www.economist.com/finance-and...

That chart above is also a comparatively generous perspective. Zoom in on the one-month level (i.e. release-to-release) and initial ADP prints are pretty much entirely uncorrelated with initial private payrolls prints (and somewhat, but not vastly, correlated with final prints).

Another issue is that private-sector data tends to benchmark itself heavily to the official stuff.

That means even the better sources, like ADP's payrolls measure, look a lot worse when you look at the actual initial releases, not the final revisions.

A bit more on one paper that does a more rigorous job than my chart in teasing that all out
x.com/ArchieHall/...

One simply proxy for whether government data still matters is whether it moves markets. Over the past few years, government data has been more market-moving that ever.

The trouble is that even in places with plenty of solid alternatives (like payrolls), they rarely agree.

Further down, a very quick 🧵 with a few more highlights.

As the shutdown grinds to an end, I took stock for @economist.com on how well private-sector data has done guiding America through the past few months' data drought.

www.economist.com/finance-and...

A quick take from me ahead of tomorrow’s Supreme Court arguments on IEEPA tariffs
— There are lots of non-IEEPA authorities Trump can use rebuild his tariff wall
— But, losing IEEPA will make the tariff-setting process more lumbering and chaotic. That could be a problem

economist.com/finance-and...

Nigel Farage’s economic plans are a disaster Three choices: fiscal implosion, deep austerity or a hasty U-turn

Farage's jettisoning of the 2024 manifesto in favour of what promises to be a much more fiscally-conservative approach is a significant moment. My colleague @archiehall.bsky.social ran the rule over the policybook as late as May. It wasn't pretty. www.economist.com/britain/2025...

Striking, though, that you don’t see this in America (at least in the survey data, anecdotally you get lots of the same complaints) — or in 2023/4 when inflation expectations seemed to be coming down

You don't see a similar gap in the US, at least if you use the NY Fed inflation expectations data rather than the UMich series. (Even in the jumpier UMich series, the cause is pretty clearly tariffs, which is a fairly clear story.)

No pretty chart but see that roughly below:

Here's something I don't have a good explanation for— British households' long-run inflation expectations are pretty severely de-anchored, but markets are totally relaxed.

Have data centres raised power bills? Not really, our analysis @economist.com finds—at least so far. Adding data centres to a state-level model of electricity prices shows no significant effect

But backlash is brewing. My latest, on the politics of the data centre: www.economist.com/united-state...

And a dispatch from @dlknowles.bsky.social from the ICE raids in Chicago www.economist.com/united-stat...

11/ So at least for now, Zero Migration America is here to stay. Best start getting used to it.

Here's the piece again: www.economist.com/finance-and...

10/ But: even unpopular Trump policies have a habit of sticking around. And there's little indication the administration has much desire to slow down.

Business complaints, even in industries with effective lobbying operations, so far haven't done much.

9/ The $100k H1-B fee also polls terribly: 57% of Americans opposite it, only 26% support it. Even among Republicans it's barely above water.

8/ Will the policy stick?

Public opinion on migration has rebounded sharply upward this year. 79% of Americans now thing immigration is good for the country, the highest on record.

7/ ... its issues with an aging workforce, especially if what we're seeing today sticks.

6/ Zero Migration also worsens America's deficit problems, and...

5/ The hit to growth from a serious attempt to squeeze skilled migration would be pretty stark.

4/ But the more consequential story is long-term.

The H1-B fee mess showed a fairly clear direction of travel on the Trump admin's desire to pull down skilled migration. The next place to watch is OPT, the post-grad work status that gets most talented foreigners into the US.

3/ Short-term, this will clearly be vastly disruptive.
— Plenty of industries like construction and farming lean very heavily on immigrant workers.
— High migration helped enable the 'soft landing' in 2023/4, by absorbing outsize labour demand. We could now see the reverse.

2/ What's happened? Crossings at the southern border have crashed after Trump's asylum ban.

Deportations are up and skilled-worker visa are already creeping down (even before talk of the $100k H1-B fee).

That is a stark reversal: net migration was ~2-3m in 2023 and 2024.

For the first time in about 70 years, net immigration to America could be zero. Beneath the noise of tariff and budget fights, migration may well be the biggest economic story of 2025.

My latest for @economist.com: Welcome to Zero Migration America

Link: www.economist.com/finance-and...

🧵 below