What this covers

Paul Romer and Russ Roberts discuss how cities in the developing world can better serve their poorest residents, centering on one core claim: the binding constraint on human flourishing is not people's eagerness to migrate to cities but the legal supply of places where they can live. Rather than imposing top-down mandates around amenities or density, Romer argues that governments should focus on defining and protecting public space—the roads, sidewalks, and utility corridors that enable urban function—and creating genuine market for jurisdictions where people choose among competing cities. The conversation moves from theoretical economics to concrete governance mechanisms, drawing on examples from New York City's 1811 grid plan to Shenzhen's rapid growth after 1980.

The discussion examines why certain policy structures succeed and others fail. A central distinction emerges between reform zones and concession zones: temporary, favor-based arrangements predictably collapse, while permanent reforms that volunteers choose to enter can spread and endure. Romer illustrates how well-meaning regulations—minimum square footage, amenity standards—can backfire by making each new resident financially burdensome to cities, which then restrict growth. He traces why public space coordination consistently fails under private development alone and why even dense, transit-rich cities face political gridlock when transport capacity cannot grow fast enough. The conversation also probes tougher problems: how land value capture might fund infrastructure, why property rights hold reliably in some places and not others, and what governance legitimacy is necessary for reform zones to take root in fragile states.

Sharpest takeaway

Romer argues that the most powerful lever for human flourishing in the developing world is reforming how cities form and grow—especially by defining and protecting public space and giving people real choice between competing jurisdictions—rather than imposing well-intentioned mandates that suppress the supply of urban opportunity.

  • Reform zones (experiments meant to spread nationwide and last forever) succeed where concession zones (temporary favors) predictably fail
  • Most urban growth will come from expansion of existing cities, where cheaply planning and protecting public space (roads, sidewalks, utility corridors) unlocks development
  • There is massive excess demand for urban life; the binding constraint is supply—legal places to live—not people's willingness to move

The argument · threads8 threads · 32 claims
0.76

Poor people's barrier to improvement is lack of legal housing supply, not unwillingness to move; cities must compete to attract permanent residents.

5 pointscentrality 5/5
  • The barrier to poor people improving their lives by moving is not mobility—people already take enormous risks and costs to reach cities—but the absence of a legal spot where they can have a place to live; the real mission is creating the supply side: establishing locations that compete to attract migrants as permanent residents.

    “frankly it's not luggage. People are quite mobile... The problem is there's no place that's ready to give them, basically a legal spot where they can have a place to live.”

  • In early stages of urban expansion, laying out places where people can build modest accommodations—even with latrines, wells, or trucked-in water as New York had before sewers and the Croton Aqueduct after 1840—is worthwhile because if it is affordable, legal, and gives people a chance, they will pursue it and both mayors and constituents will favor it.

    “in the early stages of urban expansion, you might be laying out places where people could build very modest accommodations and in the initial stages that we may be dealing with latrines and maybe just trucked-in water. But if it's affordable enough and it's legal and it gives people a chance, they'll pursue that strategy.”

  • Rather than the Millennium Development Goals' moral suasion, the single goal worth pursuing is that every rural family should be able to choose among several cities that are willing and eager to attract them as permanent residents, because that kind of competitive choice is more likely to raise the quality of life of poor people.

    “I'd rather replace the lot of them, every single one of them, with a single goal, which is that every family should be able to choose between a number of different cities that are willing and eager to attract them as permanent residents.”

  • Roughly 5 billion more urban residents are expected over the next century, most coming from expansion of existing cities in the developing world, which is why advising those cities on expansion is the highest-leverage work even while waiting for opportunities to build entirely new cities.

    “roughly 5 million more residents in 100 years than we have right now, most of that is going to come from expansion of existing cities in the developing world.”

  • An attractive climate is a luxury good, so as average income in the United States grows there will be more demand for attractive environments such as the Mediterranean climate of the California coast, making such locations the natural sites for building competitive new cities.

    “I think one thing we have to keep in mind is an attractive climate is a luxury good.”

0.72

Reform zones let governments experiment with controversial governance changes on willing residents without coercion or imposition.

4 pointscentrality 5/5
  • Many zones around the world fail in predictable ways because they take the form of a 'concession zone'—giving tax holidays or favored treatment to particular parties for a finite time—rather than a 'reform zone,' so their ex ante probability of benefiting the country is very low.

    “they fail in ways that you could have predicted when you started them, because they took this form that I'm calling a 'concession zone.'”

  • A reform zone uses the decision to opt in to a new geographic area as an opportunity to implement reforms of any sort that would be controversial if imposed on people already in a location, allowing experimentation without coercion because residents choose voluntarily to live under the new regime.

    “you use the decision to opt in to a new geographic area as an opportunity to implement reforms of any sort, any type of reform, that might be controversial if you tried to implement it on a group of people who were already in a particular location”

  • A viable reform zone requires a government with at least some capability and some legitimacy so the reform will be durable and respected over time, but the best-run governments do not need a reform zone—creating a 'middle ground' where countries with bad governance yet a window of legitimate, democratically supported authority (as Honduras briefly had under President Lobo) are the candidates.

    “the governments that have the most of that don't actually need to implement a reform zone, because they are pretty well run. So you are looking for this middle ground.”

  • The test of whether something is a genuine reform is whether you would want it to extend to the rest of the country and want it to last forever; if not, it is merely a concession.

    “The test of whether something is a reform, or reform zone, is: Do you want it to extend to the rest of the country, and, do you want it to last forever?”

0.59

Urbanization and effective cities are the highest-leverage intervention for catch-up growth in developing countries far from the technological frontier.

1 pointcentrality 5/5
  • Most growth worldwide is catch-up growth, not innovation at the technological frontier, and for countries far behind the frontier more R&D and PhD training is irrelevant; what matters are gateways that open a country to an inflow of existing knowledge from the rest of the world, and effective cities are the key gateways—which is why urbanization is the most important place a government can intervene to enable spectacular catch-up growth.

    “most of the growth that we see is not growth at the technological frontier, but it's catch-up growth... How do they access to all the stuff that's already known some place on earth? And what I concluded was the speedup, catch-up growth--more innovation, more R&D, more college and Ph.D. training--all that's completely irrelevant.”

0.54

Public space protection and grid planning during urban expansion are nearly impossible to retrofit and directly enable city value and mobility.

3 pointscentrality 4/5
  • New York City's 1811 plan laid down a rectangular grid of streets and avenues on undeveloped farmland using a single surveyor over three years and a three-person commission, planning for a 7-fold expansion at very little cost—a model the Urban Expansion Initiative advises developing-world cities to copy, since defining and protecting public space costs almost nothing yet unleashes faster, more successful development.

    “they had a single surveyor who surveyed, you know, all the avenues and all the streets over about 3 years. And a 3-person commission who backed up his decisions. But sort of making the plan for the public space costs very little.”

  • Failing to protect public space during urban expansion is almost impossible to fix afterward; Bangkok expanded with affordable housing but, unlike New York City which protected its avenues and streets, did not protect public space and is now choked with traffic and lacks corridors for utilities like storm sewers.

    “they had a single surveyor who surveyed, you know, all the avenues and all the streets over about 3 years. And a 3-person commission who backed up his decisions. But sort of making the plan for the public space costs very little.”

  • The most important thing for cities expanding in the developing world is to define, before development takes place, what is public space versus private space—because public space (roads, sidewalks, utility corridors) provides the mobility enabling the interactions that make cities valuable—and this must inter-operate across private developments, something private initiative alone (as in Gurgaon) consistently fails to provide.

    “a distinction, before development takes place, of what's the public space as opposed to the private space. Because that public space is what you are going to use for the sidewalks and the road space, the mobility that facilitates all the interactions that make cities valuable.”

0.54

Management differences between cities, not just their similarity, produce vastly different outcomes in crime, governance, and urban thriving.

3 pointscentrality 4/5
  • Even when cities superficially look similar, they may implement very different management strategies—as discount retailers like Walmart, Kmart, and Costco do—and those management differences (for example in running a police force) can produce very different crime rates, causing cities to shrink or thrive.

    “even if superficially cities look similar, they may implement very different strategies for actually managing something like the police force. Those management strategies may lead to very different crime rates”

  • A competitive market for local jurisdictions—competition between governments, not just between firms—leads to better governance and better performance of city and state governments, an advantage the United States has exploited and one that big countries like China and India could leverage more than they currently do.

    “one of the advantages the United States has been able to exploit is that same kind of very competitive market for local jurisdictions that's led to better governance”

  • By analogy to the market for corporate control (takeovers, mergers, divestitures), the market for combining or splitting jurisdictions has been effectively shut down, and it should be made legally and politically easier to merge, annex, or divest jurisdictions—for example divesting parts of Detroit or merging too-small jurisdictions in St. Louis—so cities can operate at efficient scale and try different governance.

    “It's a little bit like the market for corporate control. Like takeovers, mergers, or occasionally divestitures. That market has been shut down for cities.”

0.45

Minimum housing mandates, though well-intentioned, backfire by making cities exclude residents and shut down migration to opportunity.

4 pointscentrality 3/5
  • Layering minimum-amenity mandates (minimum square footage, private toilets, indoor plumbing) onto cities makes each additional resident a large net expenditure exceeding the revenue they generate, so cities respond by trying to keep people out—meaning well-intentioned mandates can be counterproductive by shutting down the migration to opportunity they are meant to protect.

    “each additional 10,000 residents I get, I have a huge mandate of expenditure I have to meet somehow... So, all I'm going to do is try and keep people from coming to my city.”

  • When large numbers of people each hold a veto right through their property rights, it is almost impossible to make a collective redevelopment decision; switching people from individual property rights to shares in a corporate entity (requiring only a supermajority) can sometimes succeed, but holdouts who refuse to convert can still capture benefits and slow the whole process.

    “each one has a veto right through a property right, it's almost impossible to get a decision.”

  • Because the value of a subway system shows up largely in the value of the surrounding land rather than in fare revenue, you cannot finance a subway through fares alone; you must internalize the land-value increases the subway induces, and because subways are lumpy and large, even a big developer cannot finance one giving a neighborhood access to the whole city.

    “the value of the subway system will show up to a large extent in the value of the land. So that you can't just think about... try to finance it based on the fare revenue.”

  • Density and transport are closely linked in a reinforcing loop: residents oppose densification when new buildings would bring more congestion without added transport capacity, so the political opposition that prevents densification stems from inadequate transport, making the loop hard to start when streets already feel crowded.

    “it's problems advantaging[?] congestion and transport that partly induce the political opposition which prevents the densification.”

0.44

Evaluate governance proposals with institutional specificity; reframe moral mandates as potentially counterproductive to their stated goals.

7 pointscentrality 2/5
  • A general strategy for evaluating governance proposals is to get very specific rather than discussing things abstractly; for any new city proposal, one of the first questions to ask is who hires and fires the police chief, because that level of specificity is required to push a serious alternative forward.

    “one of the first questions I ask is: Who hires and fires the police chief?”

  • The standard economist suggestion of a congestion charge is a lose-lose proposition for people deterred from driving by it, because if the gains flow mostly to someone other than the deterred drivers, or there is no way to return those gains without undoing the deterrence, the policy will not work well politically despite being efficient.

    “this is a lose/lose proposition for the people who are deterred from driving by the congestion charge.”

  • An academic or intellectual contributes in two ways: suggesting possibilities that haven't occurred to people (since the market intuition that everything is already known is incomplete), and 'reframing'—taking options people filter out on moralistic grounds, like dormitory-style shared-bathroom housing, and bringing them into the set of practical alternatives, while also reframing morally appealing mandates as potentially counterproductive.

    “there are sometimes possibilities that just haven't occurred to people... The other is to do what I call reframing, which is to take something which people recognize as a possibility but they say, Oh, we couldn't do that because it's morally inappropriate.”

  • Charter city projects should be judged like startup firms: even with the right ex ante conditions the odds of success are not high, but probability of success is the wrong metric because failures cost very little while successes generate enormous benefits, so a high failure rate should be anticipated and tolerated.

    “even if 90% of startup firms fail, it still might be a good idea to tolerate or even encourage firms to start up. Because the failures cost very little and the ones that succeed can generate enormous benefits.”

  • Police forces run by corporate entities, including non-profits like universities, show a systematic tendency to discriminate in applying the law along lines beneficial to the entity—for example protecting athletes in schools with important sports programs from charges of sexual violence—so any organization with discretionary law-enforcement power must be held accountable to equal treatment under the law.

    “One of the things you see is a systematic tendency for those police forces to discriminate in their application of the law along lines that are beneficial to the entity.”

  • Cities should be thought of as intermediate entities between the nation and a business: they can borrow market-like mechanisms such as choice by users, but certain functions like policing cannot be privatized to a corporate accountability model and require political accountability.

    “I tell students to think of cities as intermediate entities between the nation and a business.”

  • The most important social value of autonomous vehicles may not be saving driving time but that they use far less of the scarce, valuable real estate in city streets, because human-driven cars require much greater following distances due to slower human reaction times.

    “the thing about a car driven by a human is it takes a lot more real estate than a car that's autonomously controlled, because humans have to allow a lot more distance between vehicles, because we have slower reaction times.”

0.40

Proof of concept through implemented examples spreads ideas far more effectively than theoretical papers and academic arguments.

3 pointscentrality 3/5
  • Stockholm made its congestion charge palatable by using the revenue to subsidize buses and running the arrangement for about seven months before removing both the charge and the bus subsidies, then holding a referendum which passed in the 50-55% range because 10-15% of the electorate learned through experience that the arrangement was beneficial—a margin that would not have existed in a vote taken beforehand.

    “10, 15% of the electorate through the experience discovered this was a beneficial arrangement. And that was enough to make it pass this 50% threshold in a referendum.”

  • One instance of showing how an idea works in practice is far more influential than many papers that describe it as a hypothetical, so implementing one successful charter city or one city copying the 1811 plan (as in Colombia, Ethiopia, Mexico, and India) will draw more attention from academics and mayors than n papers that merely tell.

    “there's a lot to be done with one instance of showing, compared to n papers that tell.”

  • Shenzhen, started essentially from scratch as a small fishing village around 1980 under central government control, grew to roughly 10 million people with cumulative growth rates of 15-25% per year after 1990, demonstrating how a controversial reform (opening to foreign investment) could be tried where people chose to go, succeed, and then spread, while avoiding the coercion that traumatized China during the cultural revolution.

    “the growth rates here are, like, in the 'teens or the 20s, cumulatively. So for two decades, you know, 15-25% growth per year.”