Peak City
I admit it. I’m a progressive urbanist.
That is to say, I like bike lanes, I like density — to a point (there’s a reason I left Manhattan after 10 years and having kids. A decade was enough). I believe mayors, planners and developers have done untold damage to cities in the name of private auotmobiles that we’ll still be grappling with a century from now.
That said, I don’t embrace every new idea about urbanism. I think food trucks are a fad. Tactical urbanism is appropriately named. It has no strategic value whatsoever. Placemaking is important, but face-to-face against the forces that global economic restructuring exerts on cities, its like putting a band-aid on a leak in the Hoover Dam.
Most importantly, I am open-minded. I’m also very much a market urbanist.
So when the evidence tells me things I don’t want to believe are true — like markets are the dominant force shaping urban form — I don’t dismiss it out of hand. I read Joel Kotkin, Ed Glaeser, City Journal, and so on. Because as controversial as these voices are, they often bring new evidence and insights to bear in support of their ideas.
All too often, I find progressive urbanists behaving like hypocrites — and ignoring basic science in the same idelogically-driven fashion they decry their enemies (e.g. climate change skeptics) of doing. They ignore facts about urbanization because they don’t align with how they want the city to be, or perhaps more importantly how they want it to be made. Cities are made by developers. With all the baggage that brings.
Sprawl Is Baked Into Urban Expansion
Now the biggest disconnect between my progressive self and my scientific self is around sprawl. Western urbanists have so thoroughly gorged themselves on anti-sprawl ideology (and perhaps rightly so) that it is the only thing they see, and it taints all suburban and exurban development as inherently wrong. I can’t tell you the number of times I’ve found myself across the table from some New York/San Francisco/DC/Boston-based planner bashing the rest of the country… in a way that goes far beyond just a sound urban design / land use critique. How can they live like that?! Surely they would choose better if they understood! — this is the arrogant, culturally-biased subtext, and sometimes it comes right out in the open. And I think its really ugly. People don’t choose those products — and we all acknowlege that much of Amercian land use is a product — because they are stupid or bad or evil. They choose it partly because its all that is offered, but believe it or not, they do seem to like it. Honestly, they can’t seem to get enough of it. On some reptilian level, it fulfills their needs and desires.
Furthermore, as Glaeser and Kotkin and others remind us, there are sound economic reasons that these families choose those awful places. And those reasons are far more thoroughly baked into the fundamental underlying dynamics of urban growth in our world than any marginal reforms the black-rimmed glass-wearing critics have managed to build a larger case for. Any way you build them from Denver to Dhaka, cities are becoming more land-hungry over time.
The evidence base here is so overwhelming, and despite an almost total lack of recognition among urban intellectuals, it’s not even worth debating. It is a fact of urban expansion in our world. A stunningly detailed body of research looking at decades of data from cities around the world gathered by Shlomo Angel and his colleagues at NYU’s Urbanization Project have found that urban density peaked just before the turn of the 20th century, and urban land area is currently expanding at about 1.5 times the rate of population growth.
When I hang out with urbanists who chatter on about their last visit to the skyscraper-studded bubble towns of Dubai or Shanghai, I simply recall Angel’s unambiguous and earth-shattering conclusion — the future, barring some fundamental and sudden global economic transformation, is one of declining urban density. Wasteful land use (sprawl) is now firmly built into urban expansion. To the point that it’s almost not even worth fighting anymore, because by the time the fight is won, the expansion will be over.
Some of you are no doubt looking at the chart above and saying — “but wait, the trendline is upwards! Manhattan is re-densifying!” No doubt, that is true, but if the slope of that line holds, and there is evidence everywhere that the increases from 1980 onwards were a one-time boost (find me a vacant tract of land in Manhattan and I’ll write you a fat check so you can afford a Uber instead of getting to work on our sardine-packed subways), it will be about 2100 by the time Manhattan surpasses its peak 1910 population density.
And c’mon. Manhattan is Manhattan. Can you really point to it as a model for anyplace else in the United States?
We can try all we want to reduce that population:land consumption ratio for new develoment and steer that density trendline back to something that contemporary urban planning and elites prefer. I won’t get in the way of that.
But if we are going to be honest with ourselves, its a pretty good starting point for forecasting. The future isn’t sprawl. It’s what geographers observing urbanization patterns in Southeast Asian nations like Indonesia called ‘desakota’, a weird of mashup of urban and rural conditions co-existing side-by-side in expansion zones. It is a place where megacities bump up and infest the rural hinterlands they draw strength from. Their inhabitants’ livelihoods and social lives span both urban and rural, formal and informal economies. (a good technical paper on desakota is here.)
My bet is that while this concept came from the rainforests of Indonesia, it will be useful model for fast-growing regions everywhere — including North America as it inevitably resumes its metropolitan expansion in the next decade.
Peak City
The problem with this is that there is also some powerful evidence mounting for a phenomenon, that for lack of a better term, I’ll call peak city. What I mean is that we are starting to see the potential for tremenous efficiencies in the urban economy from the widespread application of technologies of cooperation to match supply and demand for infrastructure, space and capital-intensive goods like vehicles.
Of course, I’m talking about the sharing economy, or more specifically the shared city, which includes all the technologies that allow us to cram into and use the same stuff more efficiently and effectively. There is a lot of excitement about this, understandably. Broadly adopted these innovations could, and already are, having a big impact:
- Car sharing schemes like ZipCar can replace a dozen private automobiles with one shared vehicle and allow zipper, 9 to 13 fewer cars per shared car.
- In New York City, the spreading of rush hour (persumably through more flexible scheduling, and a shift in the economy away from banking and other industries tightly tied to clock time) has allowed the transit system to support growth despite limited capacity growth through physical expansion and increase in service frequency.
- In Singapore, Urban Engines is trying to pay transit riders through their mobiles to shift their commute.
- AirBnb has added about 2–3% the number of existing rooms to the U.S. short-term / hotel housing stock. That’s an estimated $10 billion in deferred capital investment alone.
- Frank Duffy, the founder of workplace design shop DEGW, argued convincingly in his 2008 pamphlet Work and the City that the developed world basically has all the commercial buildings it will ever need. Any future demand for space can be accomodated simply through more effective utilization.
- I’m sure there are parallels in energy, water and food as well (though energy has its own peak tweakers like renewables that are very comlpicated and actually often make the supply peaks worse). Please add comments to share them.
Another way of putting this — there is so much spare capacity in existing urban systems that we can grow without physical expansion — making this moment in history, the moment of peak city. And thats before we even really get into demand management, which can range all the way from gentle nudges to draconian rationing. That’s a lot of wiggle room to keep the municpal bond underwriters idled.
Townsend’s Law
But how much growth or excess demand needs to be redirected through increased utilization to really make a dent in urban problems? Only time will tell, but there is one hint out there that points towards a remarkable answer to this question — not that much, and maybe as little as 10 percent.
That hint is MIT resaerchers Marta Gonzalez and Pu Wang’s 2013 study published in Nature that found “canceling or delaying the trips of 1 percent of all drivers across a road network would reduce delays caused by congestion by only about 3 percent. But canceling the trips of 1 percent of drivers from carefully selected neighborhoods would reduce the extra travel time for all other drivers in a metropolitan area by as much as 18 percent.” (Update: Via CityMinded, which notes that “A 2–5% reduction in peak-hour traffic volumes has been shown to lead to a 27%-35% reduction in total traffic delay” from a 2009 Japanese study)
Let’s just be conservative and say 10 percent instead of 1 percent. Super-conservative. A ten percent surgical cut in demand equals a 20 percent increase in infrastructure performance.Presumably, rebound effect / latent demand will come out of the woodwork to soak up that newly liberated road space, and we’ll be back to equilibrium but with a whole lot more throughput.
Since I’m the first to note this, I get to call it what I want. How about Townsend’s Law? Let’s summarize it as follows:
A ten-percent reduction in peak urban system load allows for twice as much growth to be accommodated at the same level of service and corresponding overall quality of life.
So, think about it. If such amarginal decrease in peak load can have such disproportionate systemic impacts on infrastructure requirements, it opens up new design possibilities that are more livable, more sustainable, more productive and probably more resilient too.
Now, apply this to something like the ride-share services offered by companies like Uber and Lyft, and it quickly becomes clear that a) it doesn’t matter if they catch on much beyond their current market penetration / mode share as they are already probably having a huge effect and b) any externalities of their business model and operating procedure (e.g. wage depression, dangerous drivers) are probably something tha can be managed. In other words, as soon as rideshare reaches a level at which it is producing big gains for the rest of us, we put a cap on it before it starts to cannibalize transit or other vital services.
We don’t need to make the hard choices that putting 50% of private car drivers into shared taxis would entail. Ten percent will change our society fundamentally. That’s all we need.
Reconciling the Two Views
The irony of the term peak city is that what it really implies is a city with fewer peaks, and an infrastructure that can shrink disproporationately to accomodate a lower, steadier level of demand. That makes peak city an inflection point on the way to a peak-less city. Or more accurately, a city of fewer and lower peaks. Funny, that.
All of these leads to some intriguing future scenarios for human settlements. What could the city look like when we shave 10 percent off the peak load requirements for EVERYTHING?
Light-Touch Sprawl
As a small town boy at heart, my personal favorite is imagining a kind of light-touch suburban form typified by more sustainable homes, cleverly scattered across the coutnryside amidst a much less intrusive backbone of connective infrastructure. Land consumption would be higher than the kind of transit-oriented urban density progressive planning favors today.
A garden city in the garden, I guess. Maybe it’s possible. All I know is that the world seems to be providing more possbilities for design innovation in human settlemetns going forward, and we need to expand our thinking to how these possibilities could play out over decades across regions. Not just the next infill redevelopment project by the rail station.
Serviced Hyper-Density
Another way it might play out is a substantial increase in density in urban centers as shared resource schemes allow the same infrastructure base to serve a much higher, more demanding population. You can essentially keep putting new high-rise housing and commerical space on the grid without any major construction simply by squeezing excess capacity out of what’s already in/on the ground. (e.g. see Manhattan, 2015).
Are there other possibilities here? All my trains of thought seem to lead to one extreme or another, but I don’t see how, short of other land use constraints, we get more Brooklyns anywhere anytime soon. Sorry, friends.
Conclusions and Next Steps
There are a lot of leaps of logic here. I’m not sure this is anything different than the dynamics that Luis Bettencourt and Geoffrey West‘s’ work on urban scaling has documented — which we should not, is largely agnostic of any technology. What the ‘Santa Fe school’ or urbanism has shown is that cities grow because they do exactly what I describe here — they find ways to scale ever more efficiently. But perhaps at some point those same dynamics could be applied to making a city work better without growth.
Maybe it’s nothing new, but I feel like I’m onto something here.
