Universality: Redistribution Without Categories

Owning Essential made a narrow claim about a California ballot measure: once a government grants expedited process and exemption from rules to a defined list of project categories, the list itself becomes the site where a value system gets imposed, quietly, under the cover of a word — "essential" — that sounds too neutral to argue with. This piece asks the same question about a much older and much larger category of government action: not who gets to build faster, but who gets a check.

Start by granting the premise a large share of readers already hold and live by in 2026 and that this piece has no interest in relitigating for now: that government has a legitimate role in redistribution. A society wealthy enough to prevent destitution and decline is making a choice, not avoiding one, and a social safety net funded through general taxation is, for most people across most of the political spectrum, a settled feature of what a government is for rather than a live argument. Take that as the floor for everything that follows.

It's worth naming where the word doing the heaviest lifting in that sentence actually comes from, because "essential" wasn't always this comfortable a label. Its modern currency dates to a specific, recent, and genuinely extraordinary period: the early months of the Covid-19 pandemic, when governments at every level had to decide, in a matter of days and under real emergency conditions, which workers had to keep showing up and which activities had to stop. "Essential services" and "essential workers" were born in that emergency moment, and the moment itself supplied the thing that made the category go down easy — a declared emergency is exactly the circumstance in which a population accepts a government's line-drawing without much appetite to challenge it, because the alternative to trusting the list on short notice looked like refusing to trust anyone during an actual crisis.

That was a defensible trade during an actual emergency. What's less often noticed is that the word survives the emergency that minted it, carrying the same reflexive deference along with it into contexts — a construction-permitting ballot measure, a redistribution program — that are not emergencies at all, and don't obviously deserve the benefit of the doubt a pandemic earned once, honestly, under conditions nothing now resembles.

Granting that much doesn't settle anything about how. And the how is where Owning Essential's actual argument transplants almost without modification: a government that has accepted the redistributor's role still faces the same design choice Proposition 45 faced with construction projects — does it redistribute through a list of who qualifies, or does it find a way to redistribute that never has to build a list at all?

The same word is doing the same work, right now, in disputes that have nothing to do with construction permits. Whether gender-affirming surgery counts as "medically essential" care is contested precisely at the level of that word — insurers, state Medicaid programs, and courts are being asked to decide not just whether a treatment works, but whether it clears a threshold that determines who gets it covered and who pays out of pocket, and the fight is conducted almost entirely on the word "essential" rather than on the underlying medical evidence, because winning the label wins the funding.

Stewardship of Loneliness already covered the other end of the same move: California didn't just fund loneliness counseling, it stood up an apparatus that classified loneliness itself as a condition requiring an official response, and that classification — essential enough to build a Social Health Council around, not essential enough to have existed as a funded category five years earlier — was itself the entire policy decision, dressed up as a diagnosis.

Neither example is raised here to argue the coverage question either way. They're raised because both disputes prove the same point this essay keeps returning to: "essential" isn't a finding, it's a verdict, and whoever gets to hand it down has just made the actual redistribution decision without ever having to defend it as one. Heavy state politicians like feeding and maintaining sacred cows because they're powerful.

The welfare state's own essential-project problem

Look at how redistribution actually gets built, in California or anywhere else, and it's lists all the way down. Disability benefits require proving a qualifying condition to an adjudicator. Unemployment insurance requires proving the job loss met specific conditions and that the recipient is actively searching for another one. Means-tested programs set an income cliff, sometimes several stacked cliffs, above which eligibility disappears entirely regardless of how close a household sits to the line. Family status, immigration status, age, work requirements, asset tests — every one of these is a category, drawn by a legislature or an agency, and every one of them does exactly what the nine categories in Proposition 45 do: decides, in advance and on someone else's behalf, who counts as deserving enough to receive what the state has decided to give.

That's not a criticism of any specific program's target population — there are real, substantive reasons a disability program tests for disability rather than mailing checks to the able-bodied. It's the same acknowledgment Owning Essential made about nuclear power and high-speed rail: every individual exclusion is defensible on its own terms.

What compounds across an entire welfare system built this way is the same thing that compounded across Proposition 45's list — a redistribution apparatus that runs entirely on eligibility categories becomes, over time, a machine of incredible complexity for encoding whichever coalition currently holds the pen's idea of who deserves help, contestable and re-drawable with each change of government, and expensive to administer precisely because someone has to be paid to verify who's really inside each line and who's gaming it from outside.

What UBI actually is, and what most of what gets called UBI actually isn't

This is where the distinction the question opened with does real work, because the term "universal basic income" gets applied loosely to two structurally different things.

A genuine UBI has a specific, narrow eligibility test: is the recipient a living human being within the jurisdiction. Not low-income enough. Not unemployed enough. Not disabled enough, elderly enough, a parent, a citizen of sufficiently long standing, or enrolled in a qualifying program. Just alive, and counted. Because that criterion is coextensive with the entire population the state has jurisdiction over in the first place, it functions less like a category that sorts people into winners and losers than like the absence of one — there's no adjacent group being excluded by the line, because the line runs around everybody at once.

Almost everything actually piloted under the UBI banner over the last decade is a different animal wearing the same name. City guaranteed-income pilots typically select a few hundred recipients by lottery from a targeted income band or zip code. Several state and national proposals phase the benefit out above a certain income, or restrict it to citizens, or attach work requirements after all — each a perfectly defensible design choice on its own terms, and each one a category, doing the same sorting work as a disability test or a Prop. 45 exclusion, just aimed at a more sympathetic-sounding population. Call it what it is: a means-tested guaranteed income, a real and possibly good policy, but a categorical BI rather than a UBI, and subject to the identical critique leveled at every other list-based redistribution scheme — somebody still has to draw the line, defend it, and live with everyone it leaves on the wrong side.

The test this actually passes

Run true UBI through the same test Owning Essential applied to Proposition 45's category list: does the eligibility rule relocate the state's discretion, or remove it? Every categorical program, redistributive or regulatory, relocates it — into the drafting of the list, the adjudicator's judgment call, the income threshold a legislature can move ten points in either direction with no debate about the underlying principle at all.

A UBI is close to the limiting case on the other side. There's no adjacent, worse-off category for a caseworker, a court, or a future legislature to quietly redraw, because the eligibility criterion was never fine enough to redraw in the first place. Nobody lobbies to get "living human being" expanded to include them. Nobody outside the definition exists to notice they were excluded.

That doesn't make UBI free of value judgments — the size of the payment, how it's funded, and whether it replaces or sits alongside existing categorical programs are all still choices a state has to make, and all still contestable. But it locates the values fight in a different place: not in who gets to receive at all, but in how much everyone receives together, which is a fight conducted in public, on a single number, rather than one buried in hundreds of pages of program-specific eligibility rules that most of the population will never read closely enough to notice what they encode.

Not just who receives, but what they're allowed to do with it

The same test applies a second time, one layer deeper, once the check has actually cleared. A categorical benefit rarely stops at deciding who qualifies — it usually goes on to decide what the money is allowed to become. Food assistance can buy groceries but not a bus pass. A housing voucher pays a landlord directly and specifies what counts as qualifying housing. A workforce grant pays for an approved training program, not whatever the recipient has actually judged will improve their own position.

Each of these earmarks is defensible for the same reason each Prop. 45 category was defensible on its own terms — a program built to fight hunger has a real reason to prefer that its dollars buy food. But stacked together, an earmarked benefit system does to the use of a transfer what a categorical eligibility system does to access to it: it substitutes the state's judgment about what a person's circumstances require for the judgment of the person actually living inside those circumstances, who has information about their own situation no caseworker or program designer will ever have.

Cash is what a redistribution system looks like when it declines to make that second substitution as well as the first. A government that hands someone an unrestricted UBI payment has done the giving and stopped there — it hasn't also decided, on the recipient's behalf, whether the better use of that month's money is rent, a car repair that keeps a job reachable, a child's school supplies, or something else the state never would have guessed.

It's the same shape as a park made of land conserved and opened rather than a daily leaf-blown recreation center with a posted schedule of approved activities: government supplies the basic resource without supplying the instructions for its use -- the version that stops at supplying the resource is the one that keeps its hand off the second decision as fully as a genuinely universal eligibility rule keeps its hand off the first.

Put the whole design in one line: government, in this model, supplies energy, not bias. It can add force to a system — money into an economy, water into a park, a floor under everyone's income — without also deciding which direction that force is supposed to push once it arrives. The moment a government starts specifying not just how much but toward what, or for whom, or on what condition, it has stopped supplying energy and started supplying a preference wearing energy's name, and every category this essay has traced back to a list is exactly that: force with a direction already built in, delivered as if the direction weren't a choice at all.

The part universality actually costs

None of this is a case that UBI is costless or obviously superior on every axis, and it's worth being straight about what a category-free system gives up. A flat payment to everyone spends real money on people who don't need it, which a means-tested program by design does not. It can't account for a wheelchair ramp, a dependent with a medical condition, or a city where rent runs three times the national median, without either leaving those people worse off than a targeted program would have, or bolting on a second-tier top-up — which immediately reintroduces a category, just a narrower and more defensible one than the ones it replaced. And funding a payment large enough to matter, sent to every adult rather than only the subset who need it, is a materially larger number to raise than a means-tested program covering the same need, which is a real fiscal argument against it and not merely an objection from people who'd rather keep the categories.

The honest accounting is that a purely universal transfer trades targeting precision for the one specific thing this essay has been arguing is worth trading for: a design that doesn't require the state to decide, case by case, who has earned the right to receive. Whether that trade is worth making is a genuine policy argument, weighed against real costs, not a foregone conclusion. What isn't a genuine argument, and what this piece has tried to name directly, is calling a means-tested, phased-out, category-bounded guaranteed income program "universal" and expecting the word to do the same silent work "essential" did for Proposition 45 — carrying a much narrower, much more contestable payload past the scrutiny the actual design would earn under its own name.

The funding side has to pass the same test

A universal payment is only as category-free as the revenue that funds it, and this is where a land value tax fits the design in a way most funding mechanisms don't. A tax on land value assessed uniformly — every parcel valued by the same formula, on the unimproved value of the ground itself rather than what's built on it or who owns it — is close to the funding-side mirror of a UBI's eligibility rule: the assessment doesn't ask what the land is used for, who's using it, or whether the use is one a legislature has decided to favor this decade. It just asks what the land under any given owner is worth, on the same terms everywhere. The moment that universality gets punctured — an agricultural exemption here, a redevelopment-zone abatement there, a lower assessment for whichever use a coalition currently wants to encourage — the tax base picks up exactly the disease this essay has spent its length describing on the spending side: a list, defensible parcel-exemption by parcel-exemption, that quietly re-imports the state's preferences into the one number that was supposed to be neutral. A universal basic income funded by a land value tax that's been carved into categories hasn't actually built the category-free system this essay is describing. It's just moved the list from the check to the assessment roll, and left it exactly as capable of picking winners as it ever was.

The vocabulary itself tracks the distinction, once it's pointed out. "Basic" names a floor — the same amount, under everyone, with nothing above it ranked against anything else. "Essential" names a ranking; the word cannot be spoken without some other thing, somewhere, having already been judged non-essential by the same act of speaking it. An income floor can honestly be called basic. It cannot honestly be called essential, because calling it that would mean a list already exists, and a list is the one thing a genuinely universal program was built not to need. Services can be called Basic. They can't be called Essential — not without quietly admitting that something, or someone, just got left off.