Romer on Urban Growth

Here’s one bit from an excellent interview of Paul Romer on urban development:

Q. How are economics and planning and development of cities related each other?

Urban Expansion is an exception to the usual rule that an economy does not need a plan. Creating new built urban area requires a plan for the public space that will be used for mobility (sidewalks, bus lanes, bike lanes, auto lanes …) and for parks. At a minimum, this plan should provide for a network of arteries big enough to allow bus travel and dense enough that no location is more than 0.5 km from such an artery. This is the only thing that needs to be planned up front for land that is not yet developed. Everything else can wait. But if informal development comes first, it is too late. The area will never have enough public space to allow successful urban development.

I think Romer is correct. What surprised me most when studying Gurgaon in India was that despite strong demand, there wasn’t a lot of common infrastructure being built. The transaction costs of ex-post planning were simply too high.

In addition to transport arteries, I would also mention the importance of setting aside space and access points for sewage, electricity, and information arteries. It’s not even necessary that government provide these services or even the plan itself (private planning of large urban areas is also possible) but a plan has to be made. By reserving space for services in advance of development, developers and residents can greatly improve coordination and maximize the value of a city.

A simple, minimal urban plan is analogous to the rules of the game.


Every economist's home turf is an exception to the usual rule that an economy does not need a plan?

Paul Romer's incentives are aligned with a plan being needed?


Planners plan for the need for their plan.

It's not really an exception. Transaction costs are very high, so planning tends to happen inside an entity rather than among a lot of different entities. Just like firms. If you get an innovation or a framework (like existing infrastructure) that results in lower transaction costs, then it's easier to have market coordination.

My great uncle was an architect, specializing in landscape architecture. His passion was municipal design (he wrote a book on the subject long considered a classic). He was from the mid-west and educated in the east (Harvard) but spent his career in Sacramento, where there are many green spaces (gardens) he designed (some now named in his honor). What's surprising is that this occurred almost 100 years ago! Why have we gone backwards when it comes to municipal design. I think the answer is obvious.

"I think the answer is obvious." Jolly good; why not tell us what it is, then?

Vaccinations and GMO foodstuffs.

Your great uncle wasn't planning on large groups of net tax consumers in prime urban real estate, or a million foreigners moving in a year, needing places to live and throw away their trash.

This is why we can no longer have nice things.

That is a little harsh. Our cities declined between the 1940s and the 1990s for reasons having to do with the legacy of slavery, the popularity of the automobile, and poor governance. You want net tax consumers out of prime urban real estate? Wait for prices to go up, they'll leave. Privatize urban schools. This is happening and we're getting nice things again.

Our cities did not decline. Core municipalities declined. Prior to 1924, municipal annexation was bog standard and the municipal ambo grew with the settlement. It was eliminated first in New York and New Jersey and then various other loci. The core municipalities declined with the decay of public order and real estate development driven by affluence. Space is a superior good for most people and families could also afford two cars, which even professional-managerial bourgeois seldom had prior to the war, so some degree of sprawl would have taken place even without the enhanced investment in limited-access highways and increasing disorder in inner-city schools and neighborhoods.

A hundred years a go we had lots of immigrants needing places to live.

Nothing has changed in 100 years?

Again, we are not lacking open space for tract development. Commercial and residential development in cities is a small share of total available land.

Reading Caro's book on Robert Moses, NY planning czar for many years. At one point around WWII Moses built more state parks in NY than existed (says Caro) in the entire USA, and that would include your great uncle's works in Sacramento, California.

AlexT would not like Manila (30M people in an area roughly the size of the beltway in DC area), with little or no planning for infrastructure. To pick just one example: people build illegally so close to public roads that they typically appropriate the right-most lane, on both sides of the street, thus cutting the road down in size, hence making congestion even worse. Yet poor rice farmers and their offspring continue to flood into Manila, hoping to make the $11 a day minimum wage (a good wage compared to the $5 a day found in the provinces). Pathetic. After a year in Manila, and after finding my girl (less than half my age), we moved the heck out of that Dan Brown-inspired hell on earth (see the Sci-Fi movie "Elysium" for a preview of Manila).

Moses' ways cannot and should not be revived. There is too much conceded to obstructive veto groups, but no appointive state official should be able to destroy neighborhoods and engineer wanton evictions the way he was allowed to do (especially in the Bronx). Also, the man loathed public transportation, a perverse preference around New York and northern New Jersey.

Your observations about Gurgaon can be explained by India being inherently a low trust society.

Funding & maintaining "common" infrastructure is a recurrent problem in a low trust society. So many profitable Indian firms operate out of air conditioned, designer swanky office suites located in buildings whose stairways & passageways were dank, smelly, crumbly hellholes.

True. Similarly, many Indians in urban areas leave in clean homes which are surrounded by filthy stairwells, sidewalks, parks, and pretty much all space outside their homes. Not only do they not care, they actively contribute to the creation of the filth.

What makes India a low trust society? Is it possible that those things could change so that trust rises? Or is it simply stuck that way?

Al, I have a hypothesis about this, based on pure conjecture and no data! (grin) But I have wondered about this for a long time, so here's my guess.

1. Assume humans on average have a finite amount of trust. That is, they cannot (on average) simultaneously trust government, the media, their neighbors, their family, the church, and The Colbert Report, etc. all at a 100% level all the time.
2. If this is true, a society which places more of its trust "reserve" on one thing will have less trust for another.
3. I observe -- freed from the prison of actual facts! -- that the Indian FAMILIES I know have immense amounts of cohesion, inter-dependency, and mutual respect and trust.
4. If 3 is true and 1 is true and 2 follows from 1, then societies with very high levels of intra-family trust must have less trust elsewhere... e.g. in the state.

How's that for a theory? Pretty weak I know, but it's my guess.


I think this theory has been discussed before, but it deserves to be repeated. (BTW, Fukuyama wrote a book on societal trust.)


I'm sure these things can change and do change. I don't think the level of trust in a society is static. But I'm not sure how these things come about or can be hastened.

Hmm, isn't any bunch of people looking to spend serious money on some venture a "low trust society"? That is, if you are to invest millions of your own bucks, then will expect even nice Scandinavian business partners to sqeeze as much profit from the deal as they can and will write your contracts accordingly.

In India, I guess the crunch comes when you try to enforce that contract. The cultural factors don't work directly, rather they create a government that doesn't do it's job, and everyone knows this.

After reading these comments, I'm wondering if a slight terminology adjustment might be useful.

I don't know about you, but when I tag a freeway sign with my crew's initials, urinate on someone's front door, throw flaming garbage out my monster truck window at a pedestrian, or break all of the senior citizen center's windows with a stolen aluminum baseball bat, I do it because I have no respect for others and feel no common interest with anyone else in my community, society, city, state, country, etc.

Plus, it's fun!

Maybe instead of a "low trust" society, we should be talking about a "no respect" society or a "high fun" society.

India might be a high fun society, but not in the way you are talking about.

India is low trust, i.e. people are not antisocial, they just don't see much value in dealing honestly with strangers. At least by the standards of honesty we now have in relatively corruption free rich democracies.

Trust dynamics work one lever deeper than what you suggest. People have an incentive to deal honestly with strangers if and only if they have a reasonable assurance that such honesty will be reciprocated. But there's a vicious cycle here. Once a person has an unpleasant dealing with another, it not just reduces trust in that person but in everyone else as well. And that's the beginning of a slippery slope leading to a low trust society. What's the chicken and what's the egg here? Assuming the chicken came first, I'd say a lack of trust is the chicken, which engenders dishonest behavior. In India's case, lack of trust could have its origins in foreign invasions, which destroyed native societies, bred collaborators from within the native society, etc.

The novelist Rana Dasgupta wrote a book about present-day Delhi society, and one of the quotes (not translated perfectly) in the book is an old Punjabi saying: Whatever is in your stomach belongs to you; everything else belongs to the invader. (The actual Punjabi quote I believe refers to Ahmed Shah Durrani, the Perso-Afghan invader of NW India in the 18th century.)

So this is definitely related to trust, but if a society keeps getting pummeled century upon century by invaders, the concept of building and maintaining public infrastructure goes for a toss. This is very anecdotal and relative to the Indian mean, but I find trust and care for public spaces to be least in the north-west of India, and they gradually increase as you travel south and east. The Jamshedpur I grew up in during the 80s was moderate trust (we didn't lock our doors at night, our kids could wander halfway across town and be brought back home by conscientious strangers.) The Delhi (Punjabi-dominated) I reside in today has to be among the least-trust places on the planet.

Green field development is one thing, but to me the more difficult and interesting problem is dealing with an expanding urban core that is spreading out into a built semi-urban or suburban environment. The costs are incredible. Boston for example can't even build a relatively simple extension to its Green Line: The costs have ballooned to $400 million / km - no tunneling involved and it would be built over existing railroad beds. The project which has been discussed for over 20 years is about to be canceled because the costs are so out of control. If this line can't be built then they might as well just give up.

Isn't eminent domain a crucial issue? At least in India my observation has been that plots that might have sold for $20,000 last year will expect 100% more during acquisition because of the potential windfall payoff to the hold-out owner.

Eminent domain in the US is a very different proceeding than in India.

It can be expensive, but isn't the deal killer that it is India.

Delhi's metro two competitive advantages: low cost loans and better eminent domain.

Well, true. Irrespective of the beauty of the legal principle India has a glacially slow legal machinery.

No matter how good or bad the laws, if the case has to wait 20 years to be heard (and that isn't uncommon at all) then, yes, you do end up with a logjam during acquisitions.

If Walt Disney had only lived another 20 years....

“This is the only thing that needs to be planned up front for land that is not yet developed. Everything else can wait. But if informal development comes first, it is too late. The area will never have enough public space to allow successful urban development.”

Maybe, but probably not. More unicorns are unneeded.

Easterly and Desoto’s observations of what really happens regarding third world “planning” (which is highly related to “economic development” planning authorities in the first world) shows the outcome of urban “planning” to be highly undesirable.

The planning becomes political dupery and nitwitery of the first order as explained by public choice theory inquiry.

I wish there was a credible way for the majority to coordinate & pay off the problematic minority.

e.g. There's an arterial road in Bombay with a perennial traffic snarl due to the two lanes occupied by parked tractor trailers. Parking is illegal but the drivers bribe the cops.

The irony is that far numerous & far wealthier people waste hours every day crawling down that same route. It'd be a worthy & insignificant investment to bribe the cops to do their job, far exceeding what the truck drivers could ever pay.

But never happens!

And you can think of countries that have working, "clean" governments as ones where the good burghers have successfuly organised to bribe their public servants and keep them bribed. Egads! I think that's what Friedrich Engels said.

I do agree that the basic transportation corridor layout is certainly the most important to lay out, but stormwater management is a fairly close second. It's easier to run new drinking water pipes, which are generally smaller in size, than digging stormwater removal tunnels after the fact. Hence why we see the common flooding in newly built Chinese cities, because "not dealing with stormwater" in purely urban environments simply isn't an option. Running new pipe under areas where it's not too deep, and not super dense is a nightmare, and the adding it after the fact rather than before development, leads to either massive cost, or the project never getting done.(See the recent attempts to remedy the flooding in the "near NE" part of DC, which even today sometimes floods.

Most everybody favors 'planning' versus no planning. Hardly an insightful newsflash.

Gut issue, of course, is WHO makes the plan and choices from the myriad of options & costs. And WHO implements/enforces the always subjective plan.

Who are the masters and who are the servants in modern urban planning?

Cities certainly need plans concerning the placement of infrastructure, but the nature of those plans depends on how use of the infrastructure will be priced. Less space will be needed for parking and automobile traffic if congestion is priced.

AT, I'm confused. Have your views changed?

"If we think about, what are the best cities in the United States, particularly for the poor, it’s places like Houston, which have no zoning and which have very easy regulatory systems in which you can build. You can get a permit to build within a matter of days, compared to New York where you’ve got to go through a dozen different permitting processes and you have to hire specialized people whose only job is really to stand in line to help you get through the process….So, people of modest means can still buy a house in Houston. And they can’t do that in many other places in the United States because of zoning and not-in-my-backyard rules, a kind of secession of the rich, not in terms of gated communities but in terms of adding on rules and restrictions on how large your lot has to be in order to build a house, how many people can live in the house etc. All of these things have made it extremely expensive to buy in any of these cities, which use more top down planning." - See more at:

"The Voluntary City: Choice, Community and Civil Society, edited by David Beito, Peter Gordon, and Alex Tabarrok. It's very much anti-zoning. Here you can look inside the book. I am genuinely puzzled by this progressive meme that GMU economists don't speak out adequately against zoning and in favor of more diverse, more vital, and more practical urban environments." - See more at:

"Eliminating zoning altogether and relying on covenants and other private solutions, as does Houston, is probably best." - See more at:

My views have changed but not in the way you suggest. Going into the paper, I thought Gurgaon was a mess but a mess that was due mostly to the speed of the (magnificent!) private development that has occurred. Thus, I hypothesized that over time the issues of infrastructure would be worked out by the large developers. Writing the paper, however, I didn't see that happening hence my greater appreciation for the size of ex-post transaction costs and the importance of an early plan which would coordinate action. (Note that in the paper we also discuss private urban planning, as for example, in Reston, VA or Jamshedpur, India.)

I remain opposed, however, to zoning! Zoning is not what Romer or I are talking about. Note that Romer is talking about set asides for *public* space, by which he means a plan for where the major roads and sewage lines will go that everyone knows about *before* there is development. Indeed, the point of the plan is to increase the value of private development.

Zoning is an interference with private development which neither Romer nor I want. That's why I analogized the simple, minimal plan to the rules of the game. It's good to have let's say well recognized contract law but not rules governing what people contract about. I want freedom of contract and part of what makes freedom of contract work is a set of rules governing contract. Similarly, I want freedom of urban development and part of what makes freedom of urban development work is that everyone knows where the major roads are and how to connect to the sewage system and where electricity can be tapped into and where the parks will be located and so forth.

Hope that clarifies.

Yes, it does clarify. Thank you.

Sounds like AlexT sold out to The Man. Good OP, RPLong. AlexT's ex post parsings sounds like the weasel words that they are.

In fact, you don't need rules of the road, as any Third World shanty town will show you. And according to Hernando de Soto, its those shantytowns that provide a good portion of any city's GDP (black market). if you provide them with legal title, ex post, you can increase their value even more.

AlexT has stated he favored something similar to Houston's approach. Which is not a "Third World shanty town" by any stretch of the imagination.

When all Ray's got is a hammer, everything looks like a nail half his age.

"...looks like a nail half his age."

I laughed.

"When all Ray’s got is a hammer,"

And then I laughed some more.

But does Ray even have a nail?

And who was it who said if he owned both hell and Texas, he'd live in hell and rent out Texas?

"And who was it who said if he owned both hell and Texas, he’d live in hell and rent out Texas?"

Union General Philip Henry Sheridan. More infamously known for being the first to initiate scorched Earth tactics during the US Civil war. Resulting in the "Burning of the Shenandoah Valley".

Zoning is an interference with private development which neither Romer nor I want.

You won't want that factory or warehouse they put up across the street from you, but you're gonna get it good and hard.

Romer oversimplifies the process of planning a well-functioning city.

As Los Angeles demonstrates, there are plenty of opportunities to go astray even after a city has built a monstrous freeway system and planned a grid of wide boulevards and significant urban park spaces.

The population can swell and the number of cars can still overwhelm sizable streets and freeways. Buses are caught in traffic jams and attempts to build surface light rail lines still wind up constrained by the unstoppable crush of the car population.

A hell of a lot of middle class people choose not to use public transportation because they feel unsafe and uncomfortable with other members of the general public and dislike the unpleasantness of the buses and trains. It often takes longer and they push a lot of annoying advertisements at you on these conveyances too.

Good management is essential.

It seems to me LA is one place where a certain amount of additional transportation regulation and, if possible, a well managed freeway access fee system _might_ actually improve the economy, not destroy it.

"exception to the usual rule that an economy does not need a plan"

No its not.

The world's cities developed quite nicely without a government plan.

The problem with a government plan is the government employees doing the plan.

I'm sure that Londoners just LOVED their city being full of sewage and disease before the government starting planning about these kinds of things.

London was already developed before the started the sewer system. The sewers were installed where the population already was.

Sewers may have been used in Rome but modern systems did not develop until after the Industrial Revolution began.

The dense settlement around London has a population nearly 3x what it did a century ago. Somehow I doubt catch as catch can would have been the most prudent policy.

The world’s cities developed quite nicely without a government plan.

Sayeth a man who does not while in America have to contend with open sewers nor (except on rare occasions) flash floods.

But, there is also the matter of political (& social) controls.
Take a look at the Shogunate use of controls by precincts - right down to streets, in the organization of urban centers and all the activities in them, including the human flows.

We might also remember the renovation of Paris to accommodate the use of artillery in crowd control.

Sounds like Napeoleon's "Whiff of Grapeshot" quote (which apparently was invented by Carlyle years later).

"Elected president of the Republic of France in 1848, Napoleon's nephew became emperor on December 2, 1852, one year after his coup. Under his new name, Napoléon III decided to modernise Paris after seeing London, a city transformed by the Industrial Revolution, which offered large public parks and a complete sewer system. Inspired by Rambuteau's ideas, and aware of social issues, he wished to improve the housing conditions of the lower class; in some neighbourhoods, the population density reached numbers of 250,000 people per square mile in conditions of very poor sanitation. The goal was also for public authority to better control a capital where several regimes had been overthrown since 1789. Some real-estate owners demanded large, straight avenues to help troops manoeuvre.[2]"

And this:

"There were also political reasons for the renovation. From 1790 to 1850 France had transitioned between Royal dynasties, Republics, and Empires six different times! At this point revolutions seemed to be the fate of every French Government a fate that Emperor Napoleon III was keen to avoid. Paris was the focal point of revolutionary movements and by this point Parisian barricades had become synonymous with revolution. Napoleon III realized that the narrow streets of Paris were easy to barricade and the winding disorganized layout made it difficult for troops to quell rebellions quickly. Therefore it was hoped that by with the new layout providing quick access for troops and broad streets discouraging barricades, future revolutions could be avoided."

I think we are selectively taking a shallow reading of Romer: We seem to be focusing only on the main roads and sewer lines. Note he also adds "sidewalks, bus lanes, bike lanes". It is these finer levels of planning that are most controversial and where "planning" really matters.

Major roads: The entire U.S. contains a skeletal layout of major roads. From hereon out, it is just adding to them. Moreover, in today's U.S., major road construction and development happen together. It is not like planners plan the major roads and then developers move. This is a very different world from pre-1970s.

Even if we accept that there are some (very) major roads to be built like new interstates, I assure you that these will be built in areas that developers have already identified as prime land for development.

Sewers: Actually this is one of the easiest infrastructure pieces to build. They are underground and it is not like digging a tunnel for a train. They can run underground the front or backyard. Easements simplify the process. Not to mention the large number of houses that have been built on septic systems (although I suspect that trend is coming to an end).

Parks: Ok, these do need some planning, but again in today's world they go hand-in-hand with development. Nobody is going to plan a park, unless they know that developers are already eyeing the area. Planning for the park will be done simultaneously with the wishes of developers in mind.

Strangely, I do not see planning for water, perhaps the most critical piece of infrastructure in today's world.

Only in the USA, and to a lesser extent Canada, would there be controversy about sidewalks, bus lanes and bike lanes.

There was a similar article in last week's Economist about the success of Cambridge, England. The takeaway seemed to be that local government should concentrate on the core public goods provision and stop needlessly interfering with entrepreneurs and university business/technology incubators.

Depending on just how "planned" these plans are supposed to be we should ask what the ex post transaction cost induced by the plans might be under various scenarios of urban planning. For instance, if the urban plan sets aside certain areas for the infrastructure access (various right of way limits prior to the development) that assumes a high density commercial activity but we find that what the general market demand has become is that of lower density residential use what happened? Lots of wasted space? In adequate/inefficient infrastructure paths? Or maybe just second choice locations as the plan has effectively zoned the area for a particular usage and even officially "rezoning" it won't change the reality of the design.

Seems like the claim must be that on the whole a non-optimal urban plan has to be better than the evolved urban outcome for a large majority of the cases (and cultures)..

This actually argues for more set-asides not less. In a civilised country it is a lot cheaper and easier to sell unneeded easements etc to developers than to buy needed ones back from property owners.

access points for sewage, electricity

Every now and the I try to imagine what sewage system planning and operations must be like in Shenzhen, Beijing or Shanghai. Talk about a dirty job!

Now brace yourself and cast your imagination to Lagos, Mumbai, and Bangkok

People aren't going to click on links just because you put them there. You have to describe them a bit.


You state that private planning of large urban spaces is possible.

Theoretically. However, assembling the land is quite difficult and expensive. Then the developer, who probably borrowed money, has to wait while sufficient people and businesses move onto the land. This will not happen overnight. The developer has to carry the cost of the interest on the loan and the taxes on the vacant land. This can lead to inadequate returns of capital and disaster in a downturn.

For example, Columbia, Maryland, which was developed by the Rouse corporation as a planned city and I have been told is extremely pleasant, was very nearly the financial death of the company in the 60's.

Robert E. Simon had to sell his Reston also.

Sewers are the foundation of civilization. That is the key to any serious metropolitianization.

What is amusing is that privately owned gated communities tend to be far more restrictive regarding what people can do within them that are far worse than the most restrictive municipal zoning regulations.

The former Reston resident in my circle offers that the reason is sociological: the community board attracts and retains the world's officious nuisances. He was told by one member thereupon he'd be sued by the community board if he did not repeat a repair using a different shade of paint. He pointed out to her that the shade he used was the one specified in documentation when he bought the place, that if it did not match it was because the surrounding coat had weathered, and "I'd like to see you sue me over this". He never heard from her again on that particular issue. She was a serial harasser.

Robin Hanson thinks sewers did nothing for improving humanity.

Our urban planning is run by the National Association of Realtors.. profits first, planning later. Every politician in every city and suburb is on the take. Chicago and suburbs are a perfect example. I could tell you stories.... just check out newspaper archives....

There's a discussion of the problems of urban planning in an issue of Mars Hill Audio Journal and also in Mark Hinshaw's True Urbanism. The problem identified by both authors was that planning grew exceedingly stereotyped after the war, taking no account of local features. Hinshaw said you'll see local building codes just photocopied from model codes which make little contextual sense.

living in Tj makes me think a city need only allow sufficient taxi supply and gas taxation, to achieve quick and nifty transport

both of those seem hard to pull off tho

Comments for this post are closed