Control of Human Actions Through Land

Ask why a given American city sorts land into "single-family residential," "commercial," and "industrial," and the honest answer is rarely a planning theory. It's history, mostly forgotten, doing the work a theory is usually credited with.

Where the categories actually came from

New York City's 1916 Zoning Resolution — the first comprehensive zoning code in the country — split the city into three use districts: residential, commercial, and "unrestricted," the last one meant for the ports, utilities, and heavy factories nobody wanted next door. It wasn't derived from a general theory of good land use. It was Edward Bassett's commission answering two very specific New York fights: Fifth Avenue retailers trying to keep garment factories, and the immigrant workforce that came with them, out of their shopping district, and the public backlash over the Equitable Building — a 42-story tower built straight up from the lot line that plunged the surrounding streets into permanent shadow. The categories solved that problem, in that city, in 1916.

Then the categories went national, as a template rather than a rediscovery. In 1921, Secretary of Commerce Herbert Hoover appointed an advisory committee to draft a model law any state legislature could adopt wholesale — the Standard State Zoning Enabling Act, first issued in 1922, with more than 55,000 copies distributed to towns that had never had a garment district, a Fifth Avenue, or an Equitable Building problem in their lives. Town after town adopted essentially New York's answer to New York's fight, because a model act made adoption easy and inventing a locally-derived alternative was work nobody had time for.

The categories hardened into their present shape for a reason that should complicate anyone treating them as neutral. In 1917, the Supreme Court struck down explicit racial zoning in Buchanan v. Warley — ordinances that assigned blocks by the race of who could legally live there. Cities didn't drop the goal; they changed the instrument. Nine years later, Village of Euclid v. Ambler Realty Co. (1926) upheld use-based zoning as a legitimate exercise of municipal police power, and cities pivoted almost immediately to sorting people by "character of use" rather than race directly — with single-family districts doing, more quietly, much of what racial districts had done in the open. Eight cities had zoning ordinances when Buchanan was decided in 1917. Within twenty years, more than 1,200 did. Richard Rothstein's The Color of Law (2017) traces how explicitly that pivot was discussed at the time, in language that treated apartment dwellers and undesirable "character" as close cousins of the racial categories the Court had just closed off.

None of that makes every zoning rule illegitimate on its face. It does mean the categories now enforced as though they were self-evident — a taxonomy of what a neighborhood naturally is — were assembled once, for a specific fight, exported as a template, and then reshaped by a court case that rewarded whichever tool could do exclusionary work without saying so. The Rule of the Unquestioned named this pattern in a school-subject roster; zoning is the same reification with a deed attached.

What the categories cost now

The economic case against the strictest version of that inheritance is not fringe. Roughly three-quarters of residential land in many American cities is zoned for single-family homes only, and minimum lot sizes, parking mandates, height limits, and outright bans on multifamily or accessory units constrain supply hardest in exactly the places demand is highest. Economists Edward Glaeser and Joseph Gyourko have spent two decades documenting the resulting "zoning tax" — the gap between what a unit costs to build and what it sells for once land-use restriction scarcity is priced in, a gap that runs into six figures in the country's most restricted coastal metros. Chang-Tai Hsieh and Enrico Moretti's widely cited 2019 study in the American Economic Journal: Macroeconomics estimated that tighter land-use restrictions in a handful of high-productivity cities — chiefly New York, San Francisco, and San Jose — measurably dragged down aggregate U.S. economic growth over four decades, by keeping workers out of the places their labor was worth the most.

Zoning also shapes who lives where more than its neutral language admits. Strict low-density zoning correlates with sharper income and racial segregation in the metros that kept it tightest, which is close to the least surprising finding available given the history above — a tool built to survive the end of explicit racial exclusion doesn't stop sorting people just because the sorting is now legally silent about why.

The one American city that tried the alternative

Houston is the largest U.S. city without a Euclidean zoning code, and it's worth taking seriously rather than treating as a curiosity. Land use there is governed instead by private deed restrictions — covering a substantial share of the city's residential land — layered with subdivision rules, parking and setback codes, and ordinary nuisance law. It's not deregulation in the libertarian imagination's cleanest form; Houston still regulates plenty, and it has its own real problems with sprawl and infrastructure. But it's the closest thing the country has to a controlled comparison, and it suggests the choice was never really "zoning versus chaos." It was always zoning versus some other institution doing the sorting — a private covenant, a nuisance suit, a subdivision rule — with different owners of the decision and a different distribution of who gets a say.

Transit-oriented upzoning: real leverage, real limits

Where reformers have had the most traction recently is narrower than abolishing zoning outright: upzoning specifically near transit — raising density caps, cutting parking minimums, allowing mixed use within a quarter- or half-mile of a station. The logic is straightforward and reasonably well supported: residential density near frequent transit correlates strongly with higher ridership, and modeling of new rail investment finds the welfare gains run several times larger when height and floor-area limits are relaxed around stations than when the same transit gets built under restrictive zoning left in place. Los Angeles's Transit Oriented Communities program, adopted by ballot measure in 2016, is the clearest recent test case: it substantially increased development proposals and shifted their mix toward income-restricted units, though the net addition to the overall housing stock over several years was more modest than the program's capacity studies implied it could be.

That gap between theoretical capacity and realized production is the honest caveat. Upzoning is frequently necessary and rarely sufficient on its own — construction costs, interest rates, environmental review, and ordinary discretionary permitting all throttle the supply response independently of what the zoning map allows. Large shares of land even within easy walking distance of major stations often remain zoned single-family regardless of the transit sitting a few blocks away, which is less a flaw in the theory than a reminder that the zoning fight described above is still being fought station by station.

The question underneath the question

All of that assumes the premise most transit-oriented zoning reform takes for granted: that concentrating housing around fixed-route transit is the right thing to be optimizing land use for in the first place. It's worth asking directly, because the answer isn't obvious everywhere the policy gets applied.

Fixed-route rail and frequent bus service perform well under fairly specific conditions — dense, largely monocentric travel patterns, with most trips converging on a dominant core, and service frequent enough to compete with driving on door-to-door time. Parts of New York, the inner Bay Area, and much of urban Europe and East Asia meet those conditions and get correspondingly high ridership for it. Most American metros, sprawling and polycentric by comparison, don't, and the national data reflects it: transit ridership per capita has trended down for decades despite rising subsidy, farebox recovery on many systems runs well under a fifth of operating cost, and rail projects specifically have a long, well-documented history of large capital cost overruns — a pattern the infrastructure economist Bent Flyvbjerg has tracked across decades of megaprojects worldwide, rail chief among them.

None of that is an argument for pure car dependence, and it isn't an argument against density on its own merits — plenty of people want to live somewhere walkable regardless of what a train schedule says. It is an argument that privileging transit-adjacent land use as the organizing principle for zoning reform, in a metro whose underlying travel patterns don't actually support high-frequency fixed-route transit, risks reproducing the same mistake as the original 1916 template: treating one place's answer as though it were a general theory, and then building the next century of policy around it before checking whether the premise transferred.

What's actually being argued about

Splitting the Atom of Sovereignty described California's newer housing statutes — SB 9, SB 35, the builder's remedy — as a state increasingly willing to strip a noncompliant city's standing to object at all, not just win the argument against it. Whether that's the right amount of state force to bring against a category this arbitrary in its origins is a genuinely open question, and the history above cuts in more than one direction at once. A category invented for one city's 1916 turf war and then used for decades to sort people by race in all but name is not a category that gets to claim sanctity by tradition. But a state deciding, by statute, exactly how much density every city must allow near every transit stop — on the assumption that transit-oriented density is the obviously correct target everywhere — risks trading one unexamined template for another, just enforced from a higher floor.

The honest position isn't that zoning should be abolished, or that every inch of it should be defended. It's that the categories now treated as descriptions of what a neighborhood naturally is were never that. They were one commission's answer to one city's fight, exported by a model act, hardened by a Court that rewarded whichever tool could exclude quietly, and they remain exactly as arguable now as they were assumed to be settled a century ago.

---

Sources: Bassett, Edward M., and the New York City Commission on Building Districts and Restrictions, 1916 Zoning Resolution; Buchanan v. Warley, 245 U.S. 60 (1917); Village of Euclid, Ohio v. Ambler Realty Co., 272 U.S. 365 (1926); U.S. Department of Commerce, Advisory Committee on Zoning (Sec. Herbert Hoover), A Standard State Zoning Enabling Act (1922, rev. 1926); Richard Rothstein, The Color of Law: A Forgotten History of How Our Government Segregated America (2017); Edward Glaeser and Joseph Gyourko, research on land-use regulation and the "zoning tax," including Rethinking Federal Housing Policy (2008) and subsequent work; Chang-Tai Hsieh and Enrico Moretti, "Housing Constraints and Spatial Misallocation," American Economic Journal: Macroeconomics 11, no. 2 (2019); Los Angeles Transit Oriented Communities (TOC) program, adopted via Measure JJJ (2016); Bent Flyvbjerg, research on megaproject cost overruns, including Megaprojects and Risk: An Anatomy of Ambition (2003) and later studies.