Grant Ellison

Everything has a second-order effect. Most people stop at the first one. Policy, economics, and incentive structures →

[Grant Ellison @secondorderfx]

New SOE piece out. The Reversibility Question, on European immigration.

6,000 words on what the data shows, why the policy persisted against voter preferences for two decades, and what can and can't be undone.

Link below.

[Grant Ellison @secondorderfx]

The 8% yield isn’t really representative of staples broadly. KO ~3%, PEP ~4%, PG ~2.5%. KHC is in the 6% range. The names actually yielding 8% (BTI, certain food names) are usually pricing in real concern, not paying out generously by accident.

Bonds vs staples isn’t a yield comp anyway. Different duration, different drawdown risk, different role in a portfolio.

[Grant Ellison @secondorderfx]

Worth adding that PFML caps at $135k of wages. Dual-earner $200k household pays more in it than a single-earner $400k household. So the “tax cut for working families” headline gets paired with a regressive payroll tax structure that protects high earners and hits the same working families harder. Different policy than the press release.

[Grant Ellison @secondorderfx]

@SahilBloom “I haven’t taken a vacation in three years.”

Reads as dedication. Actually means you’ve built a job or a team that can’t function without you, which is a failure of delegation, hiring, or system design.

[Grant Ellison @secondorderfx]

@ChucksNeTwerX @DrJackKruse Topical DMSO, 70%, twice a day will do wonders. Make sure nothing is contaminating your hand / your hand is clean when you put it on.

[Grant Ellison @secondorderfx]

https://t.co/rQ9wSUTZtR

[Grant Ellison @secondorderfx]

The argument in the piece isn't that immigration is bad. Native birth rates are below replacement. Some level of inward migration is necessary.

The argument is that the specific mix Europe chose, large volume + low selectivity + culturally distant source countries, produced predictable consequences and locked them in.

Link below.

[Grant Ellison @secondorderfx]

Social benefits are about 45% of federal spending. Most of that isn't the means-tested programs people picture. It's Social Security and Medicare, going to retirees and to medical providers.

The fight over "welfare" is mostly about a small slice of the actual number.

[Grant Ellison @secondorderfx]

The Casey audit on UK grooming gangs was commissioned by Starmer's Labour government in January 2025. Casey is a Labour appointee.

She concluded there had been a "collective failure to address questions about the ethnicity" of perpetrators.

The institutional class audited itself and named the problem.

[Grant Ellison @secondorderfx]

A system that requires you to ask permission to start a business, hire a worker, sell a product, fire someone, or change prices is a system optimized for the people running it.

Not for the people inside it. Not for the people it claims to serve.

[Grant Ellison @secondorderfx]

Of 153 identified perpetrators from the 2015-2016 Cologne New Year's Eve assaults, essentially none were native German citizens.

Two-thirds were originally from Morocco or Algeria. 44% were asylum-seekers. 12% were in Germany illegally.

These are BKA statistics, not contested numbers.

[Grant Ellison @secondorderfx]

There’s nothing in JOLTS that identifies AI as the driver of the openings increase. The series doesn’t break out openings by cause, only by sector.

“P&B services rose 668k” is the data. “AI is creating jobs” is a narrative layered on top. The chart can support the first claim but not the second.

Also JOLTS prior month prints get revised by 200-400k regularly, so single-month “9-sigma beats” against forecasts often soften considerably in the next release.

[Grant Ellison @secondorderfx]

There’s nothing in JOLTS that identifies AI as the driver of the openings increase. The series doesn’t break out openings by cause, only by sector.

“P&B services rose 668k” is the data. “AI is creating jobs” is a narrative layered on top. The chart can support the first claim but not the second.

Also JOLTS prior month prints get revised by 200-400k regularly, so single-month “9-sigma beats” against forecasts often soften considerably in the next release.

[Grant Ellison @secondorderfx]

MSRP conflates real inflation with mix shift. Truck/SUV/crossover share went from ~65% to ~80% over this period and EVs from ~2% to ~10%. Both push the average up without anything getting more expensive on a like-for-like basis.

The chart’s mostly showing what Americans bought rather than how prices changed.

[Grant Ellison @secondorderfx]

@Handre @TerribleSword Same experiment ran in reverse in China 1978-1984. Decollectivization, same farmers, same land. Grain output rose ~33% in six years. Lin (1992) attributes about half of that to the institutional change alone.

[Grant Ellison @secondorderfx]

@LizAnnSonders 1.03 is roughly the 2018-2019 norm. The 2021-2022 readings of 1.8-2.0 were the anomaly.

Also worth noting JOLTS response rates have dropped from ~65% pre-COVID to ~30% now, so monthly ticks carry more noise than they used to.

[Grant Ellison @secondorderfx]

https://t.co/rQ9wSUTZtR

[Grant Ellison @secondorderfx]

The thing about European immigration that breaks people's models is the demographic trajectory continues even if every flow stops tomorrow.

Differential birth rates do the work. Policy can change inputs, not stocks.

That's what the piece is really about. Link below.

[Grant Ellison @secondorderfx]

The top 1% earned 20.6% of income and paid 38.4% of federal income taxes in 2023. Top 50% paid 97%. Bottom 50% paid 3.3%.

Every conversation about "fair share" should start with those numbers and almost none of them do.

[Grant Ellison @secondorderfx]

The natural experiment in European immigration is Poland and Sweden under identical macro and technology conditions.

Different policy. Different outcomes.

Foreign-born share: Poland under 3%. Sweden 20%. Bombings per capita: not comparable.

[Grant Ellison @secondorderfx]

Coleman 2010 projected white British becoming a minority of the UK population by ~2066. He explicitly noted those projections assumed 2010 migration rates and would compress if rates rose.

UK net migration in the year to March 2023: 944,000. His timelines were conservative.