Build the Next American Places

Summary

America should build the next places people will need, but it should not confuse ambition with starting from scratch. In some regions the best answer will be restoring the capacity of a legacy city that already has streets, water mains, hospitals, and universities. In others it will be a new district connected to an existing town, campus, rail line, or employment center. A smaller number of locations may justify genuinely new cities. The national mission is not to produce a predetermined number of cities. It is to create affordable, resilient, beautiful communities wherever the evidence shows they can work best.

That is a change in how this brief once read. An earlier version treated new construction on open land as the headline and legacy renewal as the responsible first course before it. Both framings smuggled in an answer that should be earned. A National Resilient Communities and Corridors Program should establish three eligible development tracks, hold them to one published standard, and let the federal government select the strongest projects rather than deciding in advance that one development model must always win.

Every funded project, whichever track it enters through, meets the same floor: daily needs within a walk or a short transit trip, transit running from first occupancy, at least 30 percent permanently affordable homes, nature built in as working infrastructure of the kind Medellin built into 18 roads and 12 waterways,¹ and construction that tests regionally adapted building systems on measured performance. Funding moves in phases, and later phases depend on verified results from earlier ones.

Build the receiving capacity before the emergency

A country that cannot build new places is a country that has decided its map is finished. Ours is not. Climate change is already redrawing it, and the only question is whether the redrawing is planned or panicked: whether families leave flood-prone and fire-prone regions on their own terms, toward places prepared to receive them, or as refugees after the disaster, toward places that never made room. The principle is to build the receiving capacity before the emergency, in regions with water and comparative safety, connected by rail to the country’s economic life, and designed from the first street for the century they will actually face.

Nowhere ready to receive the people climate will move

Three failures converge here. The housing crisis is worst exactly where adding housing is hardest: the regulatory gauntlet this pillar’s lead brief documents makes building in high-demand metros a decade-long, parcel-by-parcel war, while the vetocracy brief shows why. Climate displacement is arriving without receiving capacity, and the Great Lakes is now the subject of a dedicated effort to prepare anticipated receiving cities, on the premise that housing and services have to be planned before migration accelerates rather than after.² And the country has largely stopped making new places at all. The nation that built railroad towns, land-grant communities, the New Deal greenbelt towns, and the TVA’s purpose-built cities now treats founding a place as something other countries do.

Some of that capacity is not missing so much as abandoned. Cleveland, St. Louis, and Buffalo have each lost more than half their 1950 population, roughly 60 percent between them, leaving street grids, water systems, schools, and hospitals built for far more people than live there now.³ Whether that inheritance is an asset or a liability is a question of condition rather than sentiment, which is exactly what this program is built to test.

Other countries have kept building, and the record is instructive in both directions. China developed more than 3,800 new towns over four decades, which came to house over 150 million people, proof that governments can build at civilizational scale. The same program produced the ghost cities, the speculative debt, and the empty towers, which is proof of what happens when construction runs ahead of demand. The scholarly literature treats both halves as contested, which is the honest way to borrow the lesson.⁴ Saudi Arabia’s Line is the cautionary tale at the other extreme, a monument designed around a rendering rather than around residents. The lesson is neither “never build” nor “build anything.” It is that new places succeed when demand anchors come first, when phases stay small enough to stop, and when the design serves the people who will live there. Meanwhile the American population math counsels focus: the Census Bureau projects the national population peaking around 370 million near 2080 before declining.⁵ This program is not justified by national growth. It is justified by redistribution, the millions who will move as climate pressure builds and housing costs push, and by the shortage of good places for them to land.

One program, three tracks, one standard

The National Resilient Communities and Corridors Program establishes three eligible tracks. Regional partnerships propose projects, every proposal is scored against the same published criteria, and the best projects win regardless of which track they entered through. The federal role is infrastructure, credit, land stewardship, and standards.

Legacy City Renewal. Restoring housing, infrastructure, neighborhoods, and population capacity in existing cities and towns. This track should be prioritized where a community has vacant or underused land, infrastructure that can be economically rehabilitated, existing hospitals, universities, schools, cultural institutions or employers, viable rail connections, local support for substantial growth, manageable long-term climate risks, and a credible plan to benefit existing residents and prevent displacement. The critical discipline is not assuming existing infrastructure is automatically usable. A century-old combined sewer, a water system sized for a population that left, or a rail spur that has not carried freight in thirty years may be an asset or a liability, and the difference is a question of fact. Every legacy proposal requires an assessment of infrastructure condition, remaining capacity, rehabilitation cost, environmental hazards including lead and industrial contamination, and long-term maintenance burden. Reuse is often the cheapest path. It is not automatically the cheapest path, and the scoring should be able to tell the difference.

New Urban Districts. Comprehensively planned, rail-connected districts built beside existing communities. This is the likely center of gravity of the program, for a straightforward reason: a large district attached to an existing city can preserve the freedom to design a coherent street, transit, utility, housing and public-space system while gaining immediate access to an established labor market, hospitals, schools, and civic institutions. It captures most of what makes new construction attractive without paying the full cost of building a society from nothing. Suitable sites may include former military installations, underused rail yards, obsolete commercial areas, brownfields and abandoned industrial land, large areas of surface parking, publicly owned land, and carefully selected land immediately outside an existing urban boundary. What matters is that these districts function as additions to real communities, knitted into the existing street grid, transit network, and civic life, rather than isolated technology campuses that happen to have housing. Calling this the likely center of gravity is a prediction about where the strongest proposals will come from, not a quota. If legacy renewal proposals consistently outscore district proposals, the portfolio should tilt that way.

New City Pilots. Genuinely new cities remain a legitimate but limited option. A proposal in this track carries a heavier burden and should have to show a strategic regional location, favorable water and climate conditions, a viable transportation corridor, committed anchor employers or institutions, ecologically responsible land assembly, a financially manageable first phase, a specific reason the same capacity cannot be created more effectively in an existing community, and competitive lifecycle costs compared with rehabilitation or expansion elsewhere. The most common error in new-city advocacy is treating cheap land as cheap development. Inexpensive land does not mean inexpensive development. A new city must finance water and sewer, roads, power, schools, public safety, transportation, healthcare access, broadband, parks, and municipal administration before it has a tax base large enough to carry any of it. The land is the small line. Everything downstream of the land is the real number, and a proposal that cannot show it honestly has not made its case.

How projects are chosen

Every proposal, in every track, is evaluated against the same standard: public cost per new resident; public cost per completed home; housing costs relative to regional incomes; infrastructure construction and lifecycle costs; existing infrastructure that can genuinely be reused; water supply and future climate exposure; operational and embodied emissions; land consumption and habitat disruption; access to employment, healthcare, education and daily needs; rail and transit viability; displacement and speculation risks; expected population demand; public financial return; local democratic support; and the benefits received by existing residents.

Publishing one scoring rubric across three tracks is what makes the comparison honest. It also makes the program’s central empirical question answerable: nobody currently knows, with confidence, whether a dollar of federal money creates more durable housing capacity in Toledo, or in a planned district outside Madison, or in a new community on a Midwestern rail corridor. Running all three against one rubric is how that gets settled.

Projects advance in phases, and later funding depends on verified results from earlier phases. Demand validation comes before heavy infrastructure, never after: land assembly and planning first, then a validation phase in which actual pre-commitments from employers, institutions and households must clear published thresholds, and only then rail extensions, schools and major public works. If validation fails, what exists is land and plans, which unwind cleanly. Empty stations do not.

Building with the land

Natural, vernacular, and Indigenous approaches belong here as a serious building and land-management strategy, not as an architectural theme applied to a conventional building. The three terms are distinct and worth keeping distinct. Natural construction uses minimally processed materials: earth, straw, timber, hemp, lime, stone. Vernacular construction reflects methods developed in response to a particular climate, landscape and culture, which is why a Gulf Coast porch and a High Plains windbreak are solving different problems. Indigenous knowledge belongs to distinct Indigenous nations. It is not a free collection of ideas available for government or corporate extraction, and treating it as one is both a wrong and a good way to get the engineering incomplete.

What the program should support is modern hybrid construction: engineered timber structures, prefabricated straw or agricultural-fiber insulation, hemp-lime infill, compressed-earth blocks or engineered rammed earth, lime and earthen finishes, passive solar orientation, and climate-responsive building forms, combined with modern ventilation, filtration, moisture control, fire protection, accessibility and structural engineering. Alongside the buildings, nature-based infrastructure does real work: wetlands, bioswales, native planting and watershed restoration for cooling, flood control and habitat.

No material earns a place here because it is natural. Every system is evaluated on health, safety, durability, repairability, ecological effects, lifecycle cost, embodied carbon, local availability and resident experience, measured in occupied buildings rather than in demonstration structures nobody lives in. Straw-bale walls that mold in a humid climate are a failure regardless of their embodied carbon. Rammed earth that cannot be repaired by local trades is a liability in the fortieth year even if it is a triumph in the first. The approach has to be regional: the Great Lakes, the Plains, the Southwest and the Southeast face different combinations of heat, cold, humidity, fire, flooding, drought and wind, and have different materials at hand. A single national building palette would fail most of the country.

Partnership with Indigenous nations

Indigenous participation in this program means shared authority, consent, and material benefit. It does not mean a review meeting after the site has been chosen. Concretely: government-to-government consultation beginning during site selection; protection of treaty rights, ancestral lands, cultural sites and sensitive knowledge; tribal authority to decline participation outright or to restrict how knowledge is used; paid leadership positions in planning, research, construction and ecological management; Indigenous data-sovereignty and intellectual-property protections; contracting opportunities for tribal enterprises; revenue sharing or ownership where Indigenous knowledge contributes to a commercial product; and direct support for projects conceived and governed by tribal nations themselves.

This is also where the platform’s permitting commitments are load-bearing rather than decorative. The Voice-Not-Veto Test already holds that tribal consultation and treaty obligations are a floor no streamlining touches, because a nation-to-nation obligation is not a procedural delay. That principle governs here without modification.

Technology as a tool, not the purpose

These communities should test technologies, but they are places to live, not permanent demonstrations. Nobody should be asked to live inside an exhibit. Appropriate pilots include district energy and microgrids, water reuse and stormwater systems, prefabricated and modular construction, low-carbon materials, passive design, renewable energy, modern public transportation, controlled-environment agriculture where it performs well, and digital infrastructure governed by strong privacy and cybersecurity rules.

Two honesty requirements attach to all of it. The first concerns cost curves. Sustained public demand drove dramatic cost declines in solar, and that experience is often generalized into a promise that anything deployed at scale gets cheap. It does not follow. Government research, procurement, standardization and volume can reduce costs for manufacturable, repeatable products. Buildings, rail, water systems and site-specific infrastructure are substantially assembled in place under conditions that vary by site, and they do not necessarily follow the same curve. The program should fund what competes on measured lifecycle cost and stop describing every technology as one deployment away from cheap.

The second concerns food. Vertical and controlled-environment agriculture is a possible supplement for fresh produce and local resilience, not a path to food independence; its energy costs and crop range are real constraints. The right frame throughout is resilient interdependence rather than self-sufficiency. These communities should maintain essential backup capacity, water, power, food, medical supply, while remaining connected to regional and national systems. A community that tries to supply everything itself is usually more fragile than one that can survive an interruption and then reconnect.

Public value, protected by design

Public investment must not become a transfer of land, infrastructure and risk to private developers. That outcome is the default unless the terms prevent it, so the terms are part of the policy rather than an afterthought. The program requires continued public ownership or long-term control of strategically important land, with ground leases where appropriate; permanent affordable and social housing; community land trusts, cooperatives and resident ownership; open contracts and financial records; competitive procurement and periodic rebidding; open technical standards; public access to performance data; clawbacks when private partners fail to deliver; profit-sharing or other public return when public investment creates substantial private value; anti-speculation rules; strong protections against displacement; and a public-sector comparison completed before any public-private partnership is used.

Land acquisition ahead of route announcements is the mechanism that makes most of this work: the appreciation that rail and infrastructure create returns to the public that created it, instead of to whoever bought ahead of the news. Acquisition happens at published independent appraisal, and parcel purchases are disclosed on a public register on a statutory schedule. Private companies may construct buildings and operate contracted services. They must not own the government or monopolize essential public systems.

Permitting by risk, not by category

The platform’s Voice-Not-Veto principle applies here in full, including its requirement that speed come from process rather than from weakening what a project must meet. Applied to this program, that means calibrating review to risk across three levels: permit-by-rule for low-risk, repeatable projects using certified designs, the housing typologies, street sections and utility details a program like this should be standardizing anyway; consolidated accelerated review for larger projects using established mitigation methods, with one lead agency and a firm schedule; and full review for projects carrying major environmental, water, safety, displacement, cumulative-pollution or tribal implications.

One rule matters more than the tiering. A high-impact project is not automatically approved because an agency missed a deadline. Deadline-triggered approval sounds like accountability and functions as a loophole: it converts an agency’s understaffing into a permit. A missed deadline should instead trigger escalation to a named senior decision-maker, added review capacity, a public explanation of the delay, and where appropriate a refund of the applicant’s fees. The applicant is entitled to a timely decision. Nobody is entitled to a decision made by default.

Absolute limits, not just efficiency

Efficiency alone is not enough. A development can improve every per-unit metric and still increase total emissions, total water withdrawal and total land converted. Each project therefore tracks absolute figures: total operational emissions; embodied carbon; water withdrawals and consumption; land disturbed; habitat effects; waste generation; material use per resident; household transportation demand; energy costs; infrastructure maintenance obligations; and resident health and satisfaction. The objective is improving human well-being while reducing total environmental burdens, not increasing consumption under a green label.

Measure before scaling

Each project publishes independently verified results on occupancy and population retention; employment and business formation; housing affordability; construction time and cost; public subsidy per resident; infrastructure reliability; transportation use; energy and water consumption; emissions and ecological performance; indoor environmental quality; health and safety; resident satisfaction; maintenance and repair costs; and public financial return.

Expansion depends on performance. Methods that do not work are revised or discontinued without threatening the whole national program, which is the main structural argument for running three tracks at once rather than betting the program on one. A portfolio can lose a project. A single-model program that fails, fails entirely.

What we know, and what we are testing

Honest policy distinguishes what the evidence supports from what it merely permits. Relatively well supported: compact, mixed-use development can reduce automobile dependence; transit performs best when connected to real housing and employment demand; green space and nature-based infrastructure can contribute measurably to cooling, health and resilience; public research and procurement can accelerate some technologies; building low-emission infrastructure at the outset can avoid expensive retrofits later; and existing infrastructure can deliver substantial value where its condition and capacity justify reuse.

Explicitly hypotheses, requiring testing: that new cities will generally be cheaper than existing-city renewal; that large numbers of people will voluntarily relocate; that every proposed rail hub will attract sufficient employment; that natural construction will reduce total costs in every region; that vertical farming will materially improve overall food security; that urban design alone will produce major population-level mental-health improvements; that all selected technologies will get cheaper through deployment; and that a new development will avoid corporate capture without binding governance rules.

The program’s structure follows from that division. The well-supported findings set the design floor every project must meet. The hypotheses are what the phase gates exist to test.

Cost: not yet scored

This program is not yet scored, and this brief does not invent a number; the enabling legislation gets scored like everything else. What the design fixes in advance is the direction of the flows. Each project’s total federal exposure is capped in the authorizing statute before ground breaks, with the cap set by reference-class forecasting against the documented actual costs of comparable projects, including their overruns rather than their brochures. If projected costs reach 125 percent of the authorized ceiling, work pauses and nothing continues without public re-authorization. Land-value capture and ground-lease revenue service the infrastructure debt, and the affordability share is financed by the land economics rather than bolted on afterward.

Every phase gate also carries a published failure disposition, written before groundbreaking. If validation fails, assembled land is disposed of at market over a decade with proceeds retiring the bonds, with right of first refusal to adjacent communities and to the legacy renewal track, and with any affordability covenants surviving the sale. If a later gate fails, completed infrastructure conveys to the county or utility district at written-down value rather than sitting in escrow. Publishing the unwind plan is not pessimism. It is the difference between a pilot and a boondoggle.

The objections, answered

Some will say this is central planning. The design is close to the opposite: government builds enabling infrastructure and sets public-interest rules, and communities then grow, diverge and govern themselves, with funding gated on results rather than faith. The centrally planned version of this policy would be the one that decides in advance how many cities to build and where. This one publishes a rubric and lets proposals compete.

Some will say fix the cities we have instead. Often that will be the right answer, and legacy renewal is a full track precisely so it can win on the merits, repeatedly, and in most regions, if the evidence goes that way. What this brief will not do is assume the answer before the assessment. Existing infrastructure is sometimes a bargain and sometimes a sunk liability, and only a real condition assessment distinguishes them.

Some will say rail stops cannot conjure economies. Correct, which is why corridors must be viable without the new stop and why anchor commitments must exist before ground breaks. Some will say it is corporate welfare in green paint; that is the failure mode the public-value terms exist to prevent, and they are precisely the terms a giveaway would omit. And some will say people will not move. Some will not, and nothing here compels anyone. Families in repeatedly damaged places should have a genuine choice: relocation assistance at replacement cost rather than depressed post-disaster value, open to renters as well as owners, with continued adaptation investment for those who stay. Managed retreat and forced displacement are opposites, and the statute should keep them so.

What Would Change Our Position

If new city pilots consistently show higher public cost per resident and worse outcomes than legacy renewal or district projects, the new city track shrinks to whatever niche the data supports, and the money follows the evidence into existing communities. If district projects underperform legacy renewal on the same rubric, the expected center of gravity moves and this brief is rewritten to say so. If pilot districts miss occupancy and employment benchmarks, expansion stops, and that is a commitment rather than a forecast. If relocation assistance shows coercive drift in practice, the program pauses before it scales. If hybrid building systems lose to conventional construction on measured lifecycle cost and durability, they remain research rather than policy. And if land-value capture fails to finance the infrastructure as modeled, the fiscal design gets rebuilt before another site is announced.

The plain-language version

America used to know how to make places. Railroad towns, land-grant colleges with towns around them, whole communities built on purpose. Then we mostly stopped, and now housing costs too much where the jobs are, the regions climate change will empty have no plan, and cities built for far more people than live in them sit half-used.

So build again, but let the evidence pick the method. Sometimes that means rebuilding a city that already has the bones. Usually it will mean a well-planned new district attached to a town that already exists. Occasionally it may mean a genuinely new city, where the water, the rail line and the employers all line up and no existing place can do the job. Cooled by trees and restored streams as well as air conditioning. Walkable. Connected by rail. A real share of homes permanently affordable. Built with methods suited to the region, in partnership with the nations who have been reading this land the longest, and measured honestly enough that the failures get stopped and the successes get repeated.

Not a monument, and not a rendering. Places where an ordinary family can afford a good life.

Footnotes

1. UN Environment Programme on Medellin’s Green Corridors The temperature figure is Medellin’s mayor as quoted by UNEP, not an independent UNEP measurement; UNEP’s Colombia head is quoted in the same piece saying monitoring would be key to demonstrating the benefits over time. ↩

2. New America, Great Lakes Receiving Network An initiative working with anticipated receiving cities across the Midwest on housing and infrastructure planning ahead of climate migration. ↩

3. U.S. Census Bureau, Population of the 100 Largest Cities and Other Urban Places in the United States: 1790 to 1990 (Gibson, Working Paper No. 27) 1950 counts: Cleveland 914,808; St. Louis 856,796; Buffalo 580,132. Current counts from the 2020 Census and subsequent city population estimates. ↩

4. Boqian Xu, China’s New Towns in Controversy: A Literature Review, Journal of Planning Literature (2022) “In the past forty years, more than 3,800 new towns emerged and accommodated over 150 million urban inhabitants in China,” alongside the polarised scholarly dispute over the ghost cities. ↩

5. U.S. Census Bureau, 2023 national population projections ↩