💬 Discussion
Two Greek figures inspect urban infill housing, renovation scaffolding and an approval scroll.

Paying For Zoning Reform

The federal government cannot tell a city what it may build, so the housing law in force since 11 July 2026 pulls the two levers it has [1][6]. It eases environmental review of federally assisted housing and pays places that loosen their rules. Both changes came from the bipartisan package led by Senators Tim Scott and Elizabeth Warren and Representatives French Hill and Maxine Waters.

  • 50 million dollars a year already went to a federal program that rewards places for building more [6]
  • 23 percent lower rents than in a comparison were measured in Auckland eight years after it upzoned most of its residential land [10]

The proposal

Infill and small rehabilitation projects with federal assistance skip the full environmental assessment, and places that loosen their building rules compete for 200 million dollars a year in grants.

The analysis

Paying For Zoning Reform

53% of the weight is for, 47% against.

An argument's weight combines three questions: how big the effect is, how certain it is and how much the good at stake matters. The larger the slice, the more the argument weighs.

For

Against

Whether the review exemption and the zoning grants are a good measure, the analysis leaves open; it is a small law, and on average it comes out slightly ahead.

Unclear

Unclear: Under the assumptions used, the proposal brings more than it costs in 53 of 100 model runs.

 Overall effect: small

How strongly the proposal works overall, in five steps from very small to very large, comparable across all debates. The ring is always full, even when the overall effect is small; this step says how large it is.

The analysis leaves the result open because the two halves of the law cancel out: the review exemption saves about as much as the grants cost. The exemption is the solid half, since it works without anybody changing their mind. The grant half rests on two steps that nobody has measured together, whether money changes what a council decides and how far rents fall when homes are added. If rents answered as strongly as in Auckland, the law would come out clearly ahead. If the grants change nothing, it stands on the review change alone, against the cost of federal money in a local decision.

In detail

All arguments

Arguments for3 arguments

Reviews that do not happen

5.6points

Filling a gap in an existing block with federal assistance has required a full environmental assessment written for highways and dams. The law moves those projects to the short checklist category, which removes most of the bill and part of the delay.

Value 5 · Enforcement costImpact 1.9Plausibility 6
▸ Show reasoning & sources ▾ Hide reasoning & sources

Value

The stream is two things a review consumes: the money paid to consultants and staff to produce it, and the months in which a finished plan waits for it. Both are priced at the middle of the scale, the level this site uses for the cost of running a rule, and the delay is converted through the housing that stands empty rather than through anybody's time. It is a real cost rather than a transfer, because the hours and the fees are consumed and the months do not come back. Nothing is counted for the environmental protection given up, which is the argument opposite. Nothing is counted for the projects that never start because of the review, which would be larger and is not estimated. The value is the middle of the scale, and the delay is priced through the homes that stand unbuilt rather than through anybody's waiting time.

Impact

The statute moves infill housing, new construction of five to fifteen units and rehabilitation of the same size into the categorical exclusion that remains subject to the related federal laws — floodplain, contamination, historic preservation, noise — so what disappears is the environmental assessment, not the checklist that sits beneath it [6]. Something like four thousand assisted housing actions a year are affected, a setting: the housing department does not publish the count, and the range runs from 2,000 to 8,000. The difference between an assessment and the checklist is put at about 17,000 dollars of consultant and staff time each, in a range from 5,800 to 46,000, which is about 70 million dollars a year. The delay is the softer half: about 1,200 infill and rehabilitation projects a year of roughly forty units each are held for something like three months less, in a range from one to nine, and forty units unoccupied for three months at about 460 dollars of net housing value a month is about 67 million dollars a year. Together that is about 137 million dollars. The new categories apply only to money appropriated after the department has written its rules, so about eight of the ten years of this evaluation carry the effect, which brings the average to about 109 million dollars a year. The Impact is the largest here and it is the only figure in this evaluation that does not depend on anybody changing their behaviour.

▸ Show calculation ▾ Hide calculation
Assisted housing actions a year moved from a full assessment to the checklist category Setting, range 2,000 to 8,000: the housing department publishes no count by review category [2][6] 4,000 4,000 actions a year
× Consultant and staff cost of the assessment beyond the checklist that remains Setting, range 5,000 to 40,000 euro-equivalent (about 5,800 to 46,000 dollars): the floodplain, contamination, historic and noise checks still have to be done 15,000 euro-equivalent (about 17,000 dollars at 1.16 dollars to the euro) 60 million euro-equivalent a year
+ Plus the homes standing unbuilt while the assessment runs Setting, range 1 to 9 months: a project held by review is usually held by financing and permits as well [1] 1,200 projects × 40 units × 3 months × 400 euro-equivalent (about 460 dollars) 118 million euro-equivalent a year
× Years of the ten in which the new categories apply the statute applies them only to money appropriated after the department's rules are in force [6] 8 of 10 94 million euro-equivalent a year
× Weight of a dollar in public and project budgets the standard weight this site uses for money spent on running a rule 1.0 94 million euro-equivalent a year
÷ Normalised Impact scale of this evaluation 50 million euro-equivalent a point 1.88
Score 1.88 Impact × 5 Value × 6 Plausibility ÷ 10 = 5.6 of 100

Plausibility

The counterfactual is the same projects under the review categories as they stood before July 2026. The reclassification is written into the statute, so whether the assessment is dropped is not in question; what is estimated is how much it was costing and how many projects it reached. Consultant fees for environmental assessments are commercial prices and appear in project budgets, but the housing department publishes no count of the actions it processes by category, so the volume is a setting rather than an observation. The confounder that matters is that a project delayed by review is often delayed by several things at once — financing, local permits, contractor availability — so dropping one step may not shorten the wait by the full three months. That is named and unresolved, and it is why the delay range runs down to one month. Reverse causation does not arise. The Plausibility is above the middle: the reclassification is statutory, the price of an assessment is a market price, and the number of projects it reached is nobody's published figure.

Evidence: Precedent · plausibility at most 8 Evidence of effect: Definitional

Rents that ease

3points

The grants pay places that allow more density, shorten their permitting or expand supply. A home sold at the market price is worth what it costs to build, so what the law adds is not the homes themselves but the small amount they take off everybody else's rent.

Value 5 · Household budgetsImpact 1.7Plausibility 3.5
▸ Show reasoning & sources ▾ Hide reasoning & sources

Value

The stream is rent that does not have to be paid, priced at the middle of the scale like all money. Since the sixteenth of September this site prices an ordinary home sold at the market price the same on both sides: what it is worth to the people who live in it is what they are willing to pay for it, and that is also what it costs to build and keep. A home that comes into existence is therefore neither a gain nor a loss in itself, and counting it as one would count the same thing twice. What survives is what the extra supply does to the price everybody else pays, which is money moving from landlords to tenants; what it is worth turns on the difference between what a dollar is worth in a household spending a third of its income on rent and what it is worth to the owner receiving it. Nothing is counted for the construction jobs, which are labour moved from one site to another. The value is the middle of the scale, and what is priced is rent that moves rather than homes that appear.

Impact

The programme is 200 million dollars a year for the fiscal years 2027 to 2031, next to a 50 million dollar programme that already existed and continues either way [6]. What that buys in permission is the open question, and the figure used is 8,000 additional homes a year nationally, in a range from 2,000 to 25,000. Over a ten-year horizon an average of five annual cohorts are standing at any moment, so 40,000 homes, which is under a tenth of a percent of the roughly 44 million rented homes in the country. That count assumes the homes keep coming at the same rate after the grants end in 2031: what a grant buys is a rule change, and a rule stays in force when the money stops, while its first homes arrive a year or more after the decision, so a flow that starts late and runs on is stood in for by one that starts at once and runs on. If the extra homes stopped with the last grant year instead, an average of 3.75 cohorts would stand, 30,000 homes, and everything below would be a quarter smaller. The Federal Reserve's estimate is that a one percent increase in supply takes about 0.19 percent off average rents [9], which on that share and a national rent bill of some 850 billion dollars is about 145 million dollars a year that tenants keep and owners do not receive. A dollar is worth about 1.4 in the renting households and about 0.7 to the owners, so what the movement is worth is the difference, a Normalised Impact of about 1.74 on the scale of this evaluation. The range is wide and it is the elasticity rather than the home count that drives it: Auckland upzoned three quarters of its residential land in 2016 and rents there sat about 23 percent below the synthetic control eight years later [10], which is nearer a full unit elasticity and would put this at a Normalised Impact of about 9; at the Federal Reserve figure with only 2,000 homes a year it is about 0.4, and 0.34 if those homes also stop after 2031. The homes themselves carry nothing, because what they are worth and what they cost are the same number [7][8]. The Impact is a little smaller than the review saving, and it is a price effect on a very large base rather than a count of homes.

▸ Show calculation ▾ Hide calculation
Additional homes permitted because of the grants Setting, range 2,000 to 25,000: the grants pay for planning and process rather than construction, and no study links a grant to a zoning decision [1][6] 8,000 a year 8,000 homes a year
× Cohorts standing at any moment inside a ten-year horizon Assumption: the rule changes the grants buy stay in force after 2031, so homes keep coming at the same rate to the end of the horizon; if they stopped with the last grant year, 3.75 cohorts would stand (30,000 homes) and the result would be a quarter smaller 5 on average (0.5, 1.5, … 9.5 within the year) 40,000 homes
÷ Share of the rented stock the homes themselves carry nothing: what a market home is worth and what it costs to provide are the same number ÷ 44 million rented homes 0.09 percent of the stock
× Rent response to extra supply Federal Reserve estimate; Auckland, where three quarters of the residential land was upzoned, sat 23 percent below its synthetic control after eight years, which is nearer a unit elasticity [9][10] 0.19 percent of rents for each percent of supply 0.02 percent off rents
× National rent bill 730 billion euro-equivalent a year (about 850 billion dollars at 1.16 dollars to the euro) 125 million euro-equivalent a year moving from owners to tenants
× Difference between what a dollar is worth to the two sides renting households against the owners receiving the rent; only the difference is worth anything, the payment itself is a movement 1.4 minus 0.7 87 million euro-equivalent a year
÷ Normalised Impact scale of this evaluation 50 million euro-equivalent a point 1.74
Score 1.74 Impact × 5 Value × 3.5 Plausibility ÷ 10 = 3 of 100

Plausibility

The counterfactual is the same municipalities without the grant, and nothing constructs it. What upzoning does where it happens is measured, and measured well: Auckland's 2016 rezoning is followed by a large and sustained rise in construction against a synthetic control and by rents well below it [3][10], and California's statewide rules on backyard flats produced tens of thousands of units in cities that had permitted almost none. Those are findings about rule changes, not about grants. The link this argument needs — that a competitive grant of a few hundred thousand dollars changes what a city council decides — has no evidence at all, and the mechanism runs against it: the political cost of allowing density falls on the council, and the money does not compensate for it. The confounder that matters is selection: the places that apply for such grants are the places already minded to reform, so any correlation between grants and reform would be self-selection rather than effect. That is named and unresolved, and it is the weak link, not the step from supply to rents, which is measured. Reverse causation is not merely possible here, it is the likely explanation. The Plausibility is well below the middle: what extra supply does to rents is measured, whether money persuades anybody to allow it is untested, and selection would produce the same pattern with no effect at all.

Evidence: Mechanism · plausibility at most 6 Evidence of effect: Mechanistic · at most 6 Placement: Chain open · plausibility 3–3.5

Open: The grant programme runs a competition with applicants who win and applicants who lose. Comparing the rule changes and permit volumes of near-miss applicants against winners would identify the effect directly and could carry this above 5.

Rehabilitation that pencils

0.86points

The exemption reaches rehabilitation as well as new building, and a review that costs more than the work itself is what stops small repairs to assisted housing from happening at all.

Value 6 · Housing supplyImpact 0.3Plausibility 5.5
▸ Show reasoning & sources ▾ Hide reasoning & sources

Value

The stream is homes kept in use that would otherwise have been left to deteriorate or closed: the roof replaced, the boiler renewed, the block brought back up to code. A home kept in use is not a home newly built, and the difference matters since the sixteenth of September: a new market home is worth what it costs, so it carries nothing, while keeping an existing one standing costs a repair and preserves a home worth far more than the repair. What is priced is that difference, the housing service that continues net of the work that keeps it going, and these are assisted homes let below the market, where what the occupants get is more than what they pay. Nothing is counted for the tenants' avoided moves, which are real and small. The value sits in the middle-upper part of the scale, because what is kept is a home and what it costs is a repair.

Impact

Small rehabilitation projects are the ones a fixed review cost stops, because a 35,000-dollar assessment on a 230,000-dollar repair does not pencil while the same assessment on a 12-million-dollar development does. The statute helps the smallest of them most: repairs to buildings of one to four units drop out of the checklist category altogether, while repairs to buildings of five to fifteen units were in most cases already in it and gain little [6]. The figure used is 800 assisted homes a year kept in use that would otherwise have gone out, in a range from 200 to 3,000, at about 4,600 dollars a year of housing value each once five annual cohorts are standing; that is about 19 million dollars a year, and 15 million once the eight of ten years in which the new categories apply are averaged in. The number is small because most rehabilitation goes ahead anyway, because the exemption reaches only projects with federal assistance in them, and because the larger repairs were already in the lighter category. It is separated from the review saving above because that argument counts the cost of a review that happened and this one counts the work that did not. A home kept in use is a real good and carries no income weight. The 4,600 dollars is what the home is worth beyond what it costs to keep standing and beyond what its occupants pay, not the gross worth of the home: the resource at stake is a repair on a building that already exists, not a new building, which is why something survives here that does not survive for a newly permitted market home [7][8]. The Impact is the smallest on this side, a seventh of the review saving, and it reaches only the projects small enough for a fixed cost to be decisive.

▸ Show calculation ▾ Hide calculation
Assisted homes kept in use that would otherwise have gone out Setting, range 200 to 3,000: repairs to one-to-four-unit buildings leave the checklist category, repairs to five-to-fifteen-unit buildings were mostly in it already [2][6] 800 a year 800 homes a year
× Cohorts standing at any moment inside a ten-year horizon 5 4,000 homes
× Housing value each a year below the figure used for a newly permitted home, because a rehabilitated assisted unit is already subsidised 4,000 euro-equivalent (about 4,600 dollars at 1.16 dollars to the euro) 16 million euro-equivalent a year
× Years of the ten in which the new categories apply the statute applies them only to money appropriated after the department's rules are in force [6] 8 of 10 12.8 million euro-equivalent a year
÷ Normalised Impact scale of this evaluation 50 million euro-equivalent a point 0.26
Score 0.26 Impact × 6 Value × 5.5 Plausibility ÷ 10 = 0.9 of 100

Plausibility

The counterfactual is the same buildings under the previous review requirement. The chain is complete and its shape is ordinary: a fixed cost falls hardest on the smallest project, so removing it changes the arithmetic only for the small ones, which is exactly the group the statute names. What is not measured is how many projects that arithmetic was actually stopping — no one collects data on repairs that did not happen. The confounder that would lower this is that a small owner deterred by review costs is usually short of capital for other reasons too, in which case removing the review does not produce the repair; that is named and unresolved. Reverse causation does not arise. Because the mechanism is standard and nothing measures the size, the number rests on a complete chain rather than a finding. The Plausibility is at the middle: the arithmetic of a fixed cost on a small project is not in doubt and the number of projects it was stopping has never been counted.

Evidence: Mechanism · plausibility at most 6 Evidence of effect: Mechanistic · at most 6

Arguments against3 arguments

Two hundred million a year

5.6points

The grant programme is 200 million dollars a year for five fiscal years, on top of a 50 million dollar programme that already existed. It is small by federal standards and it is money.

Value 5 · Public financesImpact 1.7Plausibility 6.5
▸ Show reasoning & sources ▾ Hide reasoning & sources

Value

The stream is federal money spent, priced at the middle of the scale as public money always is. It stands opposite the homes the grants are meant to produce, which appear on the other side as housing rather than as money. Nothing is added for the administrative cost of running a competitive grant round, which is inside the appropriation. Nothing is deducted for the local money the grants may leverage, which is other people's and would be counted at the same weight if it were counted at all. The review exemption costs nothing and appears only on the other side. The value is the middle of the scale, and the outlay stands here while what it is meant to buy stands opposite.

Impact

The law authorises 200 million dollars a year for the fiscal years 2027 to 2031 in competitive grants [6]. The 50 million dollar programme that already ran for the same purpose is not added, because it exists with or without this law and the comparison is with the world in which the law was not passed. Five funded years inside the ten of this evaluation give an average of 100 million dollars a year, at the standard weight for public money. One thing would change the figure: an authorisation is not an appropriation, so Congress may fund less than the law permits, which points only downward. The Impact is the largest cost here and it is a little smaller than the review saving it stands against.

▸ Show calculation ▾ Hide calculation
Competitive grants authorised by the law [6] 200 million dollars a year for the fiscal years 2027 to 2031 200 million dollars a year
× Funded years inside the ten of this evaluation the existing 50 million dollar programme continues either way and is not counted 5 of 10 100 million dollars a year
÷ In euro-equivalent, at the standard weight for public money exchange rate used throughout this evaluation 1.16 dollars to the euro, weight 1.0 86.2 million euro-equivalent a year
÷ Normalised Impact scale of this evaluation 50 million euro-equivalent a point 1.72
Score 1.72 Impact × 5 Value × 6.5 Plausibility ÷ 10 = 5.6 of 100

Plausibility

The counterfactual is the previous position, with the 50 million dollar programme alone. The amount and the five fiscal years are written into the statute, which is as firm as a cost figure gets. What is not certain is whether Congress appropriates the full authorisation each year, which is a political question rather than an estimate. There is no behavioural link carrying the quantity: the money is spent or it is not. Reverse causation does not arise. The plausibility is held below the top because an authorisation is a ceiling rather than a commitment. The Plausibility is above the middle: the amount and the years are statutory and whether the money is appropriated in full is not.

Evidence: Precedent · plausibility at most 8 Evidence of effect: Definitional

Federal money in a local decision

2.2points

Zoning is the clearest example of a decision Americans have left to the smallest unit of government. Paying communities to decide differently is not a command, and it is not nothing either.

Value 6 · Local discretionImpact 0.7Plausibility 5
▸ Show reasoning & sources ▾ Hide reasoning & sources

Value

The stream is a decision made under an inducement rather than on its merits: a council that allows more density because a grant is available rather than because its residents wanted it. This site places that with the working order of the political system, because what is affected is where a decision is taken rather than whether a right exists. It is priced lightly and for a specific reason: nothing here compels anybody, a community that wants to keep its rules keeps them and loses only a grant it never had. Nothing is counted for the communities that do not apply, who are unaffected. The value sits in the middle-upper part of the scale, and it is priced lightly because nothing here compels anybody.

Impact

About 180 million people live in places that make their own zoning decisions and would be eligible to compete for these grants. The inducement is priced as if it cost each resident about 23 cents a year, in a range from six cents to 1.20 dollars — far below the figure, about two and a half times as much, that this site uses in the neighbouring evaluation for a legislature actually overridden, because a grant competition leaves the decision where it was. That is a Normalised Impact of 0.72 on the scale of this evaluation. The figure would be larger if the grants were conditioned on rule changes as a matter of eligibility rather than scored competitively, and larger again if they were tied to existing federal transport or housing money, which some proposals in this area would do and this law does not. The Impact is the smallest in this evaluation and it is set an order of magnitude below the equivalent figure for a measure that actually overrides a state.

▸ Show calculation ▾ Hide calculation
People in communities eligible to compete for the grants [5] 180 million residents
× Value of the inducement, per resident per year Setting, range 0.05 to 1 euro-equivalent (about 6 cents to 1.20 dollars): far below the figure this site uses where a legislature is actually overridden, because a grant competition leaves the decision where it was 0.20 euro-equivalent (about 23 cents) 36 million euro-equivalent a year
÷ Normalised Impact scale of this evaluation 50 million euro-equivalent a point 0.72
Score 0.72 Impact × 6 Value × 5 Plausibility ÷ 10 = 2.2 of 100

Plausibility

The inducement is definitional — the law offers money for specified rule changes — and what it is worth is a valuation. The counterfactual is the same communities without the grant programme. The chain is short and complete. What has never been measured is what a community loses when a decision it would have taken on the merits is taken under an inducement, and this site has one anchor for a popular vote set aside and nothing at all for a grant condition, so the figure is derived by analogy and by judgement about how much weaker the second is. The confounder that matters is that a grant may simply pay for what a community wanted anyway, in which case nothing is distorted and this argument is zero; that is named, unresolved, and is the same selection problem that limits the benefit argument on the other side. Reverse causation does not arise. The Plausibility is at the middle: the inducement is certain and the price put on it is derived by analogy from an anchor for a different thing.

Evidence: Plausibility · plausibility at most 5 Evidence of effect: Definitional

Reviews that were doing something

0.6points

Environmental review is not only paperwork. The checks for floodplains, contamination and historic sites stay, but the wider assessment — cumulative effects, alternatives, public comment — is what the law drops, and occasionally it was the thing that changed a project.

Value 7 · Environment and natureImpact 0.2Plausibility 4.5
▸ Show reasoning & sources ▾ Hide reasoning & sources

Value

The stream is what an unexamined question does to the people who end up living with the answer: traffic and air quality on a block that takes several projects at once, an alternative site nobody weighed, a neighbour whose objection was never heard. This site places that with the environment and broad participation, one class below life and health, because most of what an assessment catches is a risk managed rather than a harm suffered. What is priced is the harm that occurs because nobody looked, not the value of looking. Nothing is counted for the floodplain, contamination, historic and noise checks, which the statute keeps for every project it moves. Nothing is counted for litigation, which is procedural. The value sits in the middle-upper part of the scale, and what is priced is the harm that follows from not looking rather than the loss of the procedure itself.

Impact

The projects the law moves land in the exclusion category that remains subject to the related federal laws, so the floodplain, contamination, historic preservation and noise checks continue [6]. What stops is the environmental assessment above them: cumulative effects, alternatives, a public comment period. Of the roughly four thousand actions a year, a small share of assessments changed the project for a reason the checklist would not have caught. The figure used is one in two hundred — twenty projects a year, in a range from four to eighty — with an average harm priced as if it cost about 700,000 dollars each, in a range from 120,000 to 3.5 million. That is a Normalised Impact of 0.24, and 0.19 once the eight of ten years in which the new categories apply are averaged in. The reason it is small is structural: the exemption is written for infill and rehabilitation, which by definition sit on land already built on and examined, and the hazards an assessment most often finds are the ones the surviving checklist still asks about. The Impact is a tenth of the review saving it stands against, which is the honest ratio when a procedure is mostly cost and occasionally decisive.

▸ Show calculation ▾ Hide calculation
Assisted housing actions a year the reclassification reaches [2][6] 4,000 actions a year
× Share where the assessment would have changed the project for a reason the checklist misses Setting, range 1 in 1,000 to 1 in 50: the floodplain, contamination, historic and noise checks continue for every project the law moves [6] 1 in 200 20 projects a year
× Harm each where nobody looks Setting, range 100,000 to 3 million euro-equivalent (about 120,000 to 3.5 million dollars): cumulative traffic and air quality, a worse site, health effects on residents 600,000 euro-equivalent (about 700,000 dollars) 12 million euro-equivalent a year
× Years of the ten in which the new categories apply the statute applies them only to money appropriated after the department's rules are in force [6] 8 of 10 9.6 million euro-equivalent a year
÷ Normalised Impact scale of this evaluation 50 million euro-equivalent a point 0.19
Score 0.19 Impact × 7 Value × 4.5 Plausibility ÷ 10 = 0.6 of 100

Plausibility

The counterfactual is the same projects with the assessment performed. The chain is complete: an assessment sometimes finds something the checklist does not, a finding sometimes changes the project, and an unchanged project sometimes harms someone. What is missing is any measurement of the first two frequencies for this class of project — the housing department publishes how many reviews it conducts and not how many changed anything. The confounder that matters cuts in the direction of a smaller figure: infill and rehabilitation sites are usually already developed and documented, and the statute keeps the checks that catch the common hazards, so what is lost is rarer than it looks. That is named and unresolved. Reverse causation does not arise. Because nothing measures the frequency and a genuine counter-mechanism sits against it, the number rests on a complete chain rather than a finding. The Plausibility is below the middle: assessments do find things, and how often they find something the checklist misses on the sites this law covers has never been counted.

Evidence: Mechanism · plausibility at most 6 Evidence of effect: Mechanistic · at most 6 Placement: Chain closed, unevidenced · plausibility 4–5

Open: The housing department records the outcome of every environmental action it processes. Publishing how many resulted in mitigation or a change of site, split by project type, would put a number on this directly.

How robust the result is

Under the assumptions used, the proposal brings more than it costs in 53 of 100 model runs.

gains more than it costscosts more than it gains

Each dot stands for 200 of the 20,000 model runs.

Everything in this law, for and against, is small, and one convention decides how large the housing side can be. Sold at the market price, a home is worth what it costs to build and keep, so the home itself counts for nothing on either side. The measure earns its result through what it changes: how soon homes arrive, what they cost to build, what they do to everybody else's rent. The grant half therefore rests on a price effect: 40,000 homes standing is under a tenth of a percent of the rented stock, which takes about 0.02 percent off rents [9]. Only the difference in what a dollar is worth to tenants and to owners counts. On that reading the law comes out ahead in a good five of ten runs. Read with Auckland's eight-year record, which points to a far stronger rent response, that part is five times larger and the law comes out ahead in eight of ten runs [10]. Read with the Federal Reserve figure and two thousand homes, it nearly disappears and the law stands on the review change alone [9].

Show the distribution of the model runs

20,000 simulated futures

Unclear · 53 %

In 53 of 100 simulated futures the proposal brings more than it costs.

No effect is known exactly; every argument has a range. The calculation draws a possible size for each argument from its range 20,000 times and counts how often the benefits exceed the costs.

Net effect in points per future. Left of zero the drawbacks outweigh the benefits, right of zero the benefits win. Eight in ten futures fall between −6.1 and +9.6 points.
Scale and assessment

Scale of this evaluation: Normalised Impact, unitless and calibrated to this topic. For comparison, one point here corresponds to about 50 million euro-equivalent a year.

Size class:very small

Effect small (+0.058) · 53 % of the effect is positive

Areas show individual argument weights. The positive share of the effect nets out money transfers between the two sides; the probability of a positive net effect is determined separately.

How we score →

Review and sources

AI evaluation · not yet reviewed by a human

This evaluation was produced and sourced by an AI model; a human review is still pending. Figures and conclusions may still change. The review log is at the foot of the page.How review works →

Last reviewed by Claude Opus 5.5 · October 29, 2026 · 8× AI, not yet reviewed by a human
  1. October 29, 2026AI reviewClaude Opus 5.5record updated

    Page texts for the new page layout (draft G) with quotes from primary sources; numbers unchanged.

  2. October 21, 2026AI reviewClaude Fable 5.1record updated

    Money amounts in argument texts stated in dollars per style-en.md; scores, chains and balance unchanged

  3. October 20, 2026AI reviewClaude Fable 5.1record updated

    Summary and scale note rewritten to the style of 20 Oct 2026; numbers unchanged

  4. October 16, 2026AI reviewClaude Opus 5.5record updated

    Ranges added from the settings in quadrature for five of six arguments; the appropriation risk on the grant cost has no stated range and keeps the fallback width.

  5. October 14, 2026AI reviewClaude Opus 5.5record updated

    Prüfbericht 11.10. (SO-08): Fortwirkung der Regeländerungen nach 2031 als Annahme offengelegt, Variante mit Zubau-Ende (3,75 Kohorten, ein Viertel kleiner) in Text und Rechenweg.

  6. September 16, 2026AI reviewClaude Opus 5record updated

    Buchungsregel wohnbau-kosten/1.0 an den US-Setzungen geprüft: die 5.000/4.000 € je Wohnung sind bereits Überschuss über die Bereitstellung (rund 12.000 €/Jahr bei 190 $/sq ft und 6,76 % Zins) — Kostenseite beziffert, Quellen 7/8 ergänzt, Ankerfrage im Maßstab ausgewiesen; Zahlen unveränder

  7. September 14, 2026AI reviewClaude Opus 5re-scored

    Gesamtprüfung 08.09.: Alt-50-Mio (F0) raus, 5 statt 10 Förderjahre (con-1 i 0,215→0,086), NEPA-Kategorie laut PL 119-101 §206 (Checks bleiben; pro-1 0,235→0,094, con-2 0,048→0,0096), w auf realen Gütern entfernt; r 0,61→0,64.

  8. September 6, 2026AI reviewClaude Opus 5First evaluation

    First evaluation: the environmental review exemption priced from published fees and volumes, the grant effect on zoning decisions placed on an open chain.

Evaluations are produced with AI support and reviewed on a schedule for new developments; human passes are marked separately.How we review →

Show all 10 sourcesHide sources
  1. Congress.gov: H.R. 6644, 21st Century ROAD to Housing Act. congress.gov
  2. U.S. Department of Housing and Urban Development: Pathways to Removing Obstacles to Housing, and environmental review requirements. hud.gov
  3. Greenaway-McGrevy and Phillips, Economics of Transition and Institutional Change: The impact of upzoning on housing construction in Auckland. sciencedirect.com
  4. National Low Income Housing Coalition: The 21st Century ROAD to Housing Act becomes law. nlihc.org
  5. Congressional Research Service: Estimates of a housing shortage. congress.gov
  6. Public Law 119-101, 21st Century ROAD to Housing Act: Section 206 (environmental review categories) and Section 208 (Innovation Fund, fiscal years 2027 to 2031). congress.gov
  7. U.S. Census Bureau: Characteristics of New Housing — new single-family homes sold 2025: median price 417,400 dollars, median floor area 2,194 square feet. census.gov
  8. Freddie Mac: Primary Mortgage Market Survey — 30-year fixed-rate mortgage averaged 6.76 percent, 10 September 2026. freddiemac.com
  9. Board of Governors of the Federal Reserve System, Finance and Economics Discussion Series 2020-044: Housing Supply and Affordability: Evidence from Rents, Housing Consumption and Household Location — a one percent increase in supply lowers average rents by about 0.19 percent. federalreserve.gov
  10. Greenaway-McGrevy, Economic Inquiry: Can zoning reform reduce housing costs? Evidence from rents in Auckland — rents about 23 percent below the synthetic control eight years after the 2016 upzoning. onlinelibrary.wiley.com

More debates in this area

Likely worse · 15 %
Size: small
Investor Home-Buying Ban

Bar investors that control 350 or more single-family homes from buying more existing homes for fifteen years, while leaving new construction, build-to-rent communities and the homes they already own untouched.
Likely better · 89 %
Size: large
The Housing Credit Expansion

Raise each state's allocation of the tax credit that finances subsidised housing by twelve percent permanently, and halve the bond threshold a project must meet to qualify for the smaller credit.
All debates in Housing and Land Use →