Amsterdam, painted by Jan Micker in 1652, 350 years before Google Earth.
Lately I have been provoked to reflect on the shapes and forms of urbanism past and future, about the nature of compact and dense urban places, and about what makes the next city, or any city, literally sustainable. Let me explain.
This last weekend we had a chance to make a last-minute visit to the National Gallery here, to see a sensational exhibition entitled “Pride of Place: Dutch Cityscapes of the Golden Age.” Featured were dozens of wonderful paintings, maps, and drawings of 17th century Dutch cities. Unfortunately for my readers, the show closed May 3rd. There is, however, a terrific catalog if you’re inclined.
This was a time of a powerful and compelling urbanism in the Dutch low country: Amsterdam, Delft, Haarlem, The Hague and other cities were burgeoning with wealth, commerce, and an explosion of what Daniel Burnham would much later call civic loyalty.
Jacob van Ruisdael, Haarlem, 1672.
17th century Dutch cities all shared a few common characteristics. They were watery – canals and barges are everywhere in the paintings. They were very dense and very compact. They had extraordinary public spaces, ranging from plazas in front of the City Hall or Church, to marketplaces, to tree lined quays along the banks of the canals (my favorite: the bier kade or beer quay, depicted by Jan van der Heyden in 1670 Amsterdam).
Jan van der Heyden, Amsterdam, 1670. The Bier Kade is at the left.
And interestingly, they were walled. They had very distinct edges. They were separated from the surrounding landscape for reasons of security, certainly. And to contain the water, perhaps. But I think something more may have been involved in maintaining this kind of very sudden and distinct threshold between the urban and the rural. Sustenance. Sustainability.
Cities have always been predicated on surplus. And Dutch cities enjoyed a surplus of foodstuff from the immediate and adjacent fields, as well as the bounties of their commerce and trade with distant markets.
In 1670 Amsterdam was a city of 200,000. This meant nearly a million pounds of food a day, every day, to keep everyone whole. And no Safeway, Costco, or Lean Cuisine in the freeze.
So agricultural land was a critical factor in making this urbanism possible. The dense and compact city could work because there was plenty of tillable land immediately adjacent to the community. Many of these paintings depict a completely rural setting with grazing beeves and rows of potatoes within a few feet of the city walls. Food was very close at hand.
And today, in our cities? Today, less than 2% of Americans are farmers. A typical supermarket stocks over 45,000 items, from all over the world. We have endless choice, all year round: asparagus in February, cherries in December. Food remains close at hand, thanks to an industrialized agriculture that reaches to every corner of the globe to put food on our tables.
This is a system that is, of course, completely untenable (read unsustainable), from any number of perspectives: pollution, security (when we can no longer sustainably feed ourselves and must rely on others, we can no longer rely on access to necessary quantity and quality), water scarcity, health. While it is true that grain production has increased three-fold since 1950, the real cost of this productivity is becoming clear as water disappears and aquifers dry out, streams and rivers are polluted, the air is filled with livestock methane and CO2, and increased demand drives prices up, both here and abroad. It now takes 10 calories of fuel to produce one calorie of food on Giant’s shelves. That’s five times more than 50 years ago.
So what about the next city? Maybe the Old Masters have something to teach us after all. Take a look at this:
10,000 feet above Fairfax, Virginia.
Here we are, 15 miles from the Washington Monument, in suburban Virginia. Cul-de-sacs and cars, strip centers, and folks spread out all over the place. You could pick most any American city and find the same patterns of settlement.
Now I am not advocating building walls around our cities. And I am going to continue to stay out of the “what-to-do-with-the-suburbs” debate. But I am struck by those 17th century paintings, that agriculture so close to the dense city. If we actually were to make the next city much more dense, as I think we eventually must, perhaps we too could sustain ourselves from nearby fields.
What, exactly, is a sustainable city? If a sustainable city is one that can support itself, and meet its needs, without doing so at the expense of others present or future, near or far, then perhaps the Dutch cities of the 17th century were indeed a Golden Age. Een taal is nooit genoeg.
Jacob von Ruisdael, view of Amsterdam, 1680.
I stumbled across this exhibit on a recent visit to the NGA and was really taken with it. The paintings were wonderful and they have me excited for our fall trip to Belgium and the Netherlands.
Great stuff – I actually purchased the catalog in order to hold on to the images. The only other similar movement that comes to mind, this urban in character, is the Ashcan School – Sloan et. al.
Thanks for posting the paintings!
I don’t think it’s coincidence that most of the best American cities are surrounded by water or other natural barriers to physical growth. Bringing the walls back might not be such a bad idea.
In the pre-auto era, one of the big social problems was site rent. Basically land was so expensive in the city center that very few people could afford to buy a home. This was creating huge social class differences as improvements in agricultural efficiency were driving people off the farm and into the city filling the tenements.
Site rent was one of the reasons behind the communist and socialist movements in the depression and one of the things that animated progressives like Henry George.
In the 30’s Frank Lloyd Wright was advocating something he called Broadacre City. He was advocating auto oriented development what detractors today would call sprawl.
He thought cities were too polluted and people were packed to tightly into small spaces. In a big country he argued, why should people be forced to live in such a small area?
One of the things about auto oriented development that appealed to Frank Lloyd Wright was that it would destroy site rent.
Enviromentally auto oriented development creates its share of problems, but socially it helped to create the middle class in this country after the war. As cities spread out, not everyone was competing for the very limited right to be near the city center. That means home ownership is expanded to a lot more people.
That makes housing more affordible for everyone. The appeal of Levittown was the ability to buy your home. Property ownership was no longer limited to the wealthy. Along with the GI Bill, auto oriented development was one of the big factors in the dramatic expansion of the middle class after the war.
One of the points that Ed Glaeser makes it that really tight land use policies kill housing affordibility.
http://www.cato.org/pubs/regulation/regv25n3/v25n3-7.pdf
Enviromentally, it probably would be a great idea if more cities looked like NYC. But socially, it would probably be a better idea if more cities functioned like Houston.
http://www.nysun.com/opinion/houston-new-york-has-a-problem/81989/
The reason cities like Houston and Atlanta are growing so fast right now is that socially they do a really good job of making home ownership affordible to the masses. You can have a very high income in NYC or SF and still not be able to buy your own home. In Atlanta or Houston, you can have a relatively modest income and still manage to do so. This is why the fastest growing areas of the country are auto oriented. People are moving to places where they can find a job that will let them buy a home.
To me the biggest enviromental puzzle of our time is to figure out how to make home ownership affordible in enviromentally efficiencent type cities like NYC and SF.
The tall dense cities with strong central business diestricts that are the most effective for transit in the US aren’t growing much. Until someone can figure out how to make homeownership affordible in those communities we are probably going to keep building more cities that look like Houston and Atlanta.
Excellent point. If we live in a next city that is very dense, as I think we must, and if cities are inherently more energy efficient and environmentally appropriate, and they are, then affordability becomes a huge issue. As I have said here previously, cheap fuel has induced suburban housing in southern Pennsylvania for folks who commute to DC. As crazy as that may be, they do so because housing costs are much lower. (Of course they do it for other, less rational reasons too, but I will leave that til later).
So assuring that there are places in the next city that are affordable is critical. What to do? Well, there are several things I can think of right away.
First, we need to change any public policy that offers incentives to live outside of cities. That would include policies related to energy, fuel, infrastructure, and taxes. In the US, our policies have unquestionably induced suburban housing – and obviously this is changeable.
Second, we need to reevaluate what constitutes a reasonable housing unit. The current size of an average home in the US is 2,459 square feet. The average house size in Britain is less than half that – right around 1,000 square feet. Building and living in smaller houses would affect cost.
Finally, in the very dense next city, without all the cars, think of all the building types we can repurpose. Parking garages, shopping malls, office buildings – a host of targets for smaller and more affordable units.
I will keep thinking this one over – it’s a key question. Thanks.
Look at the cost per sqft to live in various zip codes in the bay area. In the City of San Francisco housing prices per sqft range from 271 to 816 per sqft. That means a 1000 sqft home costs from 271,000 to 816,000 depending on which zipcode you live in the city of SF. In Solano County on the fringe of the Bay Area housing costs range from 84 to 158 per sqft. That means a 2459 sqft home is 206,559 to 388,522 depending on what zipcode you live out in Solano County.
http://www.dqnews.com/Charts/Monthly-Charts/SF-Chronicle-Charts/ZIPSFC.aspx
Shrinking home sizes helps but still really doesn’t overcome the site rent problem. More significantly increasing density of the urban core might make the site problem worse.
All of the really dense city urban cores are very expensive. The most expensive real estate in the world is in the high density urban cores – think Manhattan, Paris, Hong Kong, Tokyo. Especially in the walkable areas highly served by transit.
There are also several collective action problems, (ie prisoner dilemna) that you are ignoring.
The more aggressively the bay area has pursued smart growth measures, the more suburban Sacramento developers have benefited as the people priced out of the Bay Area moved to suburban Sacramento where they could more easily afford a home. A big chunk of the Sacramento economy is turning farmland in Sacramento into housing for people priced out of the Bay Area.
Additionally current landowners (and voters) benefited greatly from growth controls in the bay area and especially in SF. The more building restrictions that were passed in SF, the more valuable the land of the people who already owned land in SF. There is plenty of open space in Marin suitable for high density development. But if it was developed, housing prices and average incomes in Marin would not be among the highest in the US.
Various growth controls measures in the bay area are probably responsible for turning these communities from reliably Republican to reliably Democratic districts. In the 60’s Marin county had a bunch of wealthy republicans and now it has a bunch of wealthy liberals.
Similiarly, suburban growth has turn Placer County in suburban Sacramento a community that in the 60’s was reliably Democratic into the most Republican county in California.
I am not sure what will be more difficult, getting more growth to occur SF and the inner core of the bay area or slowing down growth in Placer County and along the suburban fringe of Sacramento. Too many incumbants in both areas are gaining too much from the current policies.
Interestingly, the high value you reference, $816/sf, is in zip code 94107. Now a bit of analysis shows that this zip code has something like 7,000 units per square mile. This works out to a density of just under 11 units per acre. So while the price per square foot is very high, the density is very low. Real urban density would be up near 100 units per acre, not 11.
So density alone is not driving the ’site rent problem.’ Something else is in play in 94107 to drive value to $816/sf. The highest real estate values in our city, Washington, do not correspond with highest density either. Our neighborhood, Capitol Hill, is thought to be desirable, and has mostly stable housing prices (make no mistake – not cheap), but a density of only about 25 dwelling units per acre (du/a), which is pretty typical for urban rowhousing. The real hook for value here in our city is access – to what are thought to be good schools, shopping, transit and high value cultural targets.
The folks with the real prisoner’s dilemma here are not so much the urban dwellers. Instead, they are suburbanites watching value disappear and wondering if or when they should defect. It’s getting pretty hard for them to play the self interest and greed game when the market has tanked and their values have sunk. They would love some closed-bag exchanges with city dwellers, to extend the prisoner’s dilemma lingo.
Interestingly we have friends who are in the process of relocating to Manhattan. They want to end up in the Murray Hill neighborhood just south of midtown. They have been aggressively looking of late, and report that rents have noticeably declined in NYC – they showed us places they are considering with rents that are very much like, say, Dupont Circle here in DC.
Anyway, the real dilemma is that high housing cost is affected by density, but not density alone. It’s quite a bit more complex, in my opinion.
As to growth controls, the real problem for most cities has been another prisoner’s dilemma. Can a city institute growth control at the regional, rather than the simply municipal level? Portland has seen solid escalation of home values in the wake of their growth boundaries, though not nearly at the $816/sf level. But in Portland, as in the Bay area, once you are outside the growth boundaries, the same old suburban crap appears, and in many places around but outside Portland, those values are nothing to sneeze at either.
But if growth controls could happen at the regional level, we would have a chance of making much more intelligent public policy. The prisoner’s dilemma here is that every region has many, many units of local government, and no one wants to give up their fiefdom. If you have a couple hundred units of local government, getting them to cooperate instead of defect is nearly impossible. Maybe some kind of real crisis will change this. Like running out of water, something like that….
You’re right its not just merely site rent. Yet site rent is still a big factor. Let me explain.
First you need to know a little about the history of building regulation in SF. First there is measure M. It puts a cap on new construction of 475,000 square feet per year. That is about the size of the Transamerica Building. There is an excellent article about the history of measure M here.
http://www.spur.org/documents/990701_article_04.shtm
In addition to the overall cap of 475k sqft, there are pretty stringent growth restrictions in most neighborhoods. The argument was in order preserve the traditional character of Chinatown and the historic Italian neighborhood of Northbeach, there needed to be severe restrictions on height in these and in most SF neighborhoods. Thus SF has not Vancouverized adding high rises throughout the city. There are only a few neighborhoods where new high rise developments are permitted.
Additionally there is the issue of rent controlled units. In fact this is probably the biggest obstacle to adding new development into SF.
When the rent control ordinance was passed in the 70’s rent control was limited to buildings built at the time the ordinance was passed. The argument was that if new construction was exempt from rent control, there would be incentives to invest in building new buildings.
The problem is that to build a new building in SF, because the city is built out, you generally need to tear down an existing building. 2/3 of the voters in SF are renters the majority of those living in rent controlled units. They fear developers will buy up their building and they will lose their cheap rent.
As a result there is a tension between people who feel that adding new units will make the city more affordible vs the existing residents who feel that tearing down rent controlled units will kick them out of the city. Almost always the advocates of strengthening rent control at the expense of new development win.
Chris Daley is the progressive city councilman who is locally famous because of the huge fees he extracts from developers as a condition of granting development.
See the $11 dollar a sqft for residential development here.
http://sf.curbed.com/archives/2007/12/10/rincon_hill_ripoff_outdoor_funds_funneled_elsewhere.php
or the impact fees of $30 million or $130,000 dollars a unit here.
http://triangle.bizjournals.com/triangle/othercities/sanfrancisco/stories/2009/05/04/daily77.html
Between the high impact fees, the higher cost of construction of high rises and the huge amount of time (and large cost of money tied up in very long development processes) It can be very expensive to build new developments in SF.
In the City of SF there are limits on the use of land sales as a proxy for site rent. A lot of the price of building has to do with how many rent controlled tenants are in a building and how long they have lived there and your ability to get them out of the building.
Here is a zipcode map of SF.
http://maps.huge.info/zip.htm
Right now on paper, this area is near the new ballpark. That area formerly was the site of a railyard. This means the area has few rent controlled units and a lot of brand new property. This area is one of the few areas growing in SF. But the high building impact fees ensure that prices per sqft will be really high. I doubt if land use regulations were loosen in this neighborhood this zip code would not have the highest prices per sqft in the city.
But overall what is driving costs up in SF is site rent. People want to pay a lot to live near where they work.
You can argue that rent control should be eliminated to loosen building restrictions. But rent control is one of the major strategies locals have employeed to protect themselves from really high site rents. It too is a symptom of the problem.
Rent control is not an issue in Sacramento or Solano Counties.
It may be helpful to have a little more background about the history of development in SF. First there is something called Prop M. Basically it put a cap of 475,000 sq ft a year of new high rise construction space. That is enough space to build one Transamerica size building a year. In practice that space is allocated among several buildings meaning that it will be difficult to build any large skyscrapers in SF. There is an excellent discussion of it here.
http://www.spur.org/documents/990701_article_04.shtm
Residents of SF were afraid that tall buildings would destroy the historic character of established neighborhoods. SF’s historic Chinatown and historic Italian neighborhood North Beach (where the Beats spent time during the sixties) are adjacent to the financial district. In addition to the cap of 475k sqft, in most neighborhoods there is a cap of I think 45 ft. Existing buildings taller than that are grandfathered in, but no new construction going forward is permitted about the height restrictions.
The net effect is that there are just a few neighborhoods where tall buildings are permitted and even then there are limits on how tall you can go.
SF has a fairly extensive system of rent controls. The rent control ordinance exempted new construction from rent control as inducement to build new housing stock. But because SF is mostly built out, to build a new building generally you need to tear down an existing structure.
2/3 of the people living in SF are renters. Rent control and preserving rent controled units are locally very politically popular.
Supervisor Chris Daly is locally famous for his ability to extract money from developers as the price for permitting new construction. See 130,000 per unit impact fees here
http://www.bizjournals.com/sanfrancisco/stories/2009/05/04/daily77.html
and 50 million from developers back in 2005.
http://www.sfweekly.com/2005-08-31/news/the-daly-deal/
Since 2005, he has used that Rincon Hill model to get impact fees from new development throughout his district. He has now extracted much more than 50 million. These impact fees price out all but the trustfarians from living in SF.
In SF two identical buildings on the same street may sell for dramatically different prices depending on how many tenants in the building are protected by rent control, how long these tenants have lived in the units and your prospects for getting them to leave the building.
There are a couple of other factors that also distort the market. Rent control and growth caps are city measures. They don’t apply to land owned or controlled by the Feds or the State of California. The transbay terminal is a joint powers authority ran by the State of California. It is exempt from the growth caps and from rent control. When you see pictures of the high rise rising over a planned high speed rail corridor. That is the reason why that is happening there. I believe the land owned by Caltrain out near the ballpark is in the same category.
Your intial reaction might be that rent control should be eliminated and land use restrictions eased. But remember rent control was how the locals decided to remedy high site rents.
Look at the history of rents in SF. During the 50’s and 60’s a lot of new freeways were built in the bay area and both levels of the bay bridge were opened up to traffic. The Irish and Italian working class were moving out of SF to the burbs and the professional class was moving even further out to Concord or down the Pennisula. Because the freeways were new and those areas weren’t built out, it was easy to commute from from the cheap land into SF. As a result rents got dirt cheap in SF. This was the time when the Beats were hanging out in NorthBeach and the hippies and homosexuals were moving into the Haight and the Castro.
As the areas of the Bay Area around the freeways built out, congestion got worse, some wanted to expand the freeways, but the nimbies didn’t want their homes demolished to expand freeways. During the 70’s Gov. Jerry Brown was called Governor Moonbeam for advocating carpooling lanes. Bart was built out.
As traffic got worse, it was no longer feasible to commute from Concord or other outlying areas into SF. People started moving clsoer to SF to live closer to work. Rents went up. Some of the hippies were moved back to the countryside. Others passed rent control so they could stay in SF.
As the bay area grew over the past 30 years and little new freeway capacity was added for nimby reasons, the time to commute to SF has gotten longer and longer. That puts increasing pressure for people to live closer and closer to SF. Without rent control, it is argued that landowners would capture most of the gains of higher local wages in the city.
Holland does provide a wonderful example not just because of its cities but because the entire land is man-made. If it looks natural it is because man can make natural things.
But we should be careful not to make a critical mistake. The cities of Holland were not sustainable because they were dense, they were dense because they were sustainable. It was the rich life and activity of the city that drew people to live there despite the crowding.
Forcing density has never created a sustainable city because it doesn’t bring the same living qualities. If you were to build a wall or no-growth-boundary (same thing) around today’s modern cities, it would only make the problems of sustainability worse.
What makes a city sustainable is it ability to grow within itself and diversify its tissues, and that implies a low-density starting point.
Well said, as usual, Mathieu. I especially enjoy your observation that everything in those wonderful paintings is essentially a giant land-fill project.
In the 17th century, and in the Dutch city of that age in particular, density and sustainability were simultaneous. It is possible in the 21st century to imagine a dense city that is not sustainable, and a sustainable city that is not dense, though the latter seems much less achievable than the former.
Most North American cities do in fact have a low density starting point. This is relative, of course, but I define low density as anything around 25 to 30 du/a and below. Real densities of 100 du/a would begin to resemble Amsterdam in 1752.
The real question, I think, is how increase the carrying capacity of existing North American cities, both to enhance quality of life (access to diversity in every sense), and to achieve much higher levels of sustainability. Clearly we can’t afford to start over.