Eminent domain limits for private development

The proposal

In Kelo v. City of New London (2005), a 5–4 Court held that taking unblighted homes for a private redevelopment plan was a “public use” under the Fifth Amendment, reading “public use” as “public purpose.”¹ Because the Takings Clause binds the states, that reading is the floor for every government.² The amendment as commonly framed would define “public use” in the Fifth Amendment itself: no government may take private property to transfer it to another private party for economic development. State-record variants: a flat ban on private-to-private transfer (New Hampshire); a public-use definition with a health-and-safety blight exception (South Carolina, Texas); heightened compensation (Michigan).³

The evidence

The only independent national poll with a party breakdown is from the autumn after the decision: 81% of Americans opposed it — 85% of Republicans, 79% of Democrats, 83% of independents (Saint Index, 2005 — a land-use series fielded by the University of Massachusetts Lowell Center for Economic and Civic Opinion, sponsored by a consulting firm). Opposition was as flat across ideology as across party — 77% of liberals, 81% of moderates, 82% of conservatives — and 63% opposed strongly.⁴ The next year the same series found 71% supporting a ban on taking private property for private development, 43% strongly (Saint Index, 2006; no party breakdown published).⁵

The record otherwise agrees:

• Twelve states amended their constitutions by popular vote, 2006–2012; every clean measure — a private-transfer limit and nothing else — passed by wide margins, from Louisiana’s 54.8% to South Carolina’s 86%.⁶ Every failed measure (23.9% to 47.6%, 2006–2008) was bundled with regulatory-takings compensation or rent-control repeal.⁷
• The House passed the Private Property Rights Protection Act 376–38 in 2005 — Republicans 218–2, Democrats 157–36 — and three more times through 2018; the Senate never did.⁸⁹
• The newest independent party split: 78% of Iowans opposed companies using eminent domain for carbon-capture pipelines — 72% of Republicans, 82% of Democrats, 79% of independents (Selzer & Co. for the Des Moines Register, 2023) — a proxy, since most Kelo reforms exempt common carriers.¹⁰
• An advocacy-commissioned poll adds a second party split: 95% disagreed with the ruling — 92% of Republicans, 94% of Democrats, 99% of independents (Zogby International for the American Farm Bureau Federation, 2005). That wording drops the public-benefit rationale the ruling turned on and reads fourteen points above the Saint Index. A six-state Midwestern survey found 81% opposed (Embold Research for Bold Alliance, 2024; no party breakdown published).¹¹

We searched nine national pollsters and found no poll on this question since 2014 with a party breakdown.

Where it stands

The bipartisan floor is 79% of Democrats (Saint Index / UMass Lowell, 2005): Strong, with a six-point gap. It is two decades old; staleness is the test at risk. Nothing since suggests it moved — the ballot record ran to 2012 without decay — but nothing measures it.

The “everyone agrees because no one is thinking about it” test is now measured, and it cuts both ways. Two years after the decision, only 21% of Americans could correctly say whether their own state had passed a reform law, and only 13% both knew and could judge whether it worked (Saint Index, 2007; no party breakdown published).¹² Attention is that shallow. The opinion under it is not: 63% opposed the ruling strongly in 2005, 43% backed a ban strongly in 2006, and taking “low value homes … to build a shopping center” was acceptable to 4% of New Jerseyans (Monmouth University, 2005; no party breakdown published).¹³ Depth under named exceptions and costs is untested nationally. Framing dependency is well measured: bundled measures ran 15 to 60 points behind clean ones.⁷ The only partisan split is on the federal remedy: House Democrats voted 39–157 on a 2005 funding-cutoff rider and 157–36 on the standalone bill.¹⁴

The symmetry test is met by construction — a public-use definition binds every government and every future majority — and reforms passed under legislatures of both parties: over a Democratic governor’s veto in Iowa, 90–8 and 41–8, with members of both parties on both sides, and over a Republican governor’s opposition in Mississippi.¹⁵ The why-an-amendment test has a precise answer: “public use” is constitutional text, so only the Court or an amendment can change its meaning, and the statutory substitute — a spending condition — has passed the House four times and never the Senate.⁹

The existential-problem test is open. Economic-development takings are rare; forty-three to forty-five states have reformed;¹⁶ the federal government binds itself by executive order;¹⁷ two sitting justices have asked the Court to reconsider Kelo (2021).¹⁸ Against that: the decision remains the federal floor; of the 36 legislatures that acted, 22 passed laws one published assessment grades largely symbolic, and the eleven states that reformed by popular vote produced the stronger ones; one state has never reformed and one rolled back;¹⁶ and this would restore a clause already in the Bill of Rights. Whether that is a threat to the country’s future or a policy preference is the question.

Framings we tested

The clean ban clears in every form asked: “prohibit government from taking private property for transfer to another private individual or business for economic development or increasing tax revenue” won 80.1% in Michigan (2006); public use with a health-and-safety blight exception won 86% in South Carolina (2006) and 81.0% in Texas (2009); owner-occupied homes only won 62.0% in California (2008).⁶ Ballots carry no party breakdown, but the states span red, blue, and purple. Two framings do not clear: the ban bundled with regulatory-takings compensation or rent-control repeal, rejected five times,⁷ and the redevelopment framing — “rebuild an area” — a 47%–39% plurality (Monmouth, 2005).¹³

What would move it

In: a national poll with party breakdowns on forward-looking, mechanism-specific wording — “no government — federal, state, or local — may take a person’s home, business, farm, or land in order to transfer it to another private owner for economic development, even if the owner is paid fair market value” — clearing 60% in both parties with exceptions (blight, common carriers) and costs (forgone redevelopment) named. That settles staleness; the existential-problem test would remain.

Out: a floor under 60% with exceptions and costs named (depth); a result turning on “rebuild an area” versus “homes for a shopping center” (framing dependency); or a finding that a problem this rare, and remedied this widely by the states, is a policy preference (the existential-problem test).

← All candidates