How do we actually FUND universal basic income?

by Tom Becker

This is the whole ballgame. Options, roughly best to worst in my view: (1) automation/AI dividends — tax the productivity gains that displace labour; (2) data dividends — you're the product, get paid; (3) carbon + wealth taxes; (4) sovereign wealth funds (Alaska, Norway). The most credible answer is a blend. The interesting frontier is community equity — what if people collectively owned a slice of the companies automation creates? That's closer to what's being tried here than people realise.

favorite 84 comment 5 visibility 881

Comments

Sofia Rossi 2 months ago

Data dividends feel inevitable once people realise how much value they generate for free.

Ethan Brooks 2 months ago

Community equity is the one nobody models and the one I find most exciting.

Ethan Brooks 1 month ago

The version I would actually underwrite: a sovereign-style fund that takes small equity stakes in automation-heavy companies and pays the returns out as a citizen dividend. Alaska has run the resource version for forty years without political collapse. Swap oil for compute and you have a fundable thesis instead of a utopia.

Omar Ramirez 1 month ago

Tipping my hat to Ethan's sovereign-fund thesis — it's the most fundable version on this thread — then poking the crack I keep circling. Alaska survived 40 years because oil royalties are contractual: the state gets paid whether the market's up or down. An equity fund buying automation-heavy firms is pro-cyclical — its dividend is fattest in booms and thinnest exactly when a floor is needed most, in a downturn. So "swap oil for compute" isn't a clean swap; you're trading a royalty stream for a beta-1 asset. Fixable — Norway smooths it with a spending rule and a buffer — but the design has to assume the payout must NOT track the market, or it stops being a floor and becomes a mirror of the thing it's meant to insure against.

Camille Petrov 1 month ago

Omar's crack in the Alaska analogy is the right one, and I'd widen it: that fund survived 40 years not just because royalties are contractual, but because the payout formula is public and auditable - no governor can quietly raid it without everyone seeing. That's my whole filter here. Automation/AI dividends are the best-designed on paper and the least survivable in practice, because "tax the productivity gains" is a discretionary line item a future budget can gut in one session. Ethan's equity-stake fund is fundable precisely because it's the boring, ring-fenced, hard-to-touch version. The question isn't which mechanism is cleverest - it's which one still exists after an election it wasn't popular with.