Triumph of the City

How our Greatest Invention makes us Richer, Smarter, Greener, Healthier, and Happier

by Edward L. Glaeser 2011

Like Glaeser, I think cities are a good idea. Like Glaeser, I live in the suburbs outside of one. He wants soft grass and good schools for his kids. I want room for a garden, and distance from ground zero. Color us both hypocritical.

Cities are greener, cleaner, and a good place to meet people. Glaeser meets his people at Harvard in BosWash. I meet my people mostly on line, because they are all over the planet. Cities can also be a good place to meet people on line, but in the Portland, Oregon area, if you want fiber optic internet, you move to the suburbs west of the city.

Glaeser says we evolved to hang out in groups, not online, and he's right. But we did not evolve among groups of thousands, either; my father lived about 30 years of his life in Portland, but all my ancestors from grandparents back came from rural communities. "Right sized" groups are about 500 people, and people in cities tend to cluster in overlapping subcommunities of that size. Glaeser has probably met tens of thousands of people, as have I. His "group" is probably his family, his neighbors, his professional colleagues, and temporary alliances to produce books, papers, and conferences. My groups are similar, but are more globally distributed; I have many writing collaborators across the US, but quite a few in England and Japan as well. Economists are thicker on the ground than chip designing space enthusiasts.

In the past, the only way to meet people and exchange ideas was at face-to-face meetings, arranged or accidental. Today, I met a publisher from Carlton, Oregon, at a class in Portland, 10 miles east of where I live, and 40 miles east of where he lives. We both burned gasoline to get there. Glaeser is correct that such meetings are more likely in cities (with present technologies), and that carbon-based transportation to such meetings is unsustainable. But inertia is not the same as inevitability. Younger people do much of their "getting around" with smartphones, and with ten million times as many transistors 30 years from now, immersive haptic telepresence may be a much cleaner and more immersive way to get around.

Cities have the BIG advantage of sharing infrastructure; a big rectangular building subdivided into apartments is easier to heat, plumb, and maintain than an acre of suburban homes. But there is a third way, demonstrated by the small-but-multistory towns that housed most people in most of the world until the 20th century. 50 family units made of concrete and steel, clustered but within walking distance of countryside, and surrounded in a cloud of electronic commerce and telepresence maintenance, may be just as clean and resource frugal as a city - except for transportation.

Glaeser points out that apartment owners and condominium governance are irksome, driving autonomy-seeking individuals towards single-family dwellings. But imagine a high automation alternative, based on standardized framing and utility hookups, and modular "skins". Imagine I have a fight with the condominium president in the morning over how I use my 3000 square feet of space in the building. Peeved, I summon Robot-Moves-R-Us, which finds an empty 3000 square foot hole in another structure three blocks away. It negotiates a good lease for me there. A giant robot climbs the side of my old building, uncouples my 3000 square feet, and rolls them over to the new place and installs them in the new place while parents are at work and the kids are in school. R.M.R.U. subdivides and sells my current lease to two other families (whose psychological profiles are more compatable with my condominum president, 1200 sqft to one, 1700 sqft to the other, with a new shared entry) and their units are moved the next day. My family might move three times a year, but change living arrangements twice a decade, or vice versa, all mediated by very powerful software agents relentlessly working to improve my satisfaction. Those agents become 200 times more affordable, or 200 times more powerful, as transistors become 200x more plentiful every decade.

The automobile is not the last word in transportation, nor is the current crop of electric cars that emulate it. Current thinking is dominated by appeals to the past, when there was no automation and little complex organization. Imagine an all day meeting, 40 miles from my home, with a break for lunch. I go there, not by driving a car, but by climbing (in my bathrobe) into a rented "travelling meeting pod", where I put on my VR helmet, enter the meeting, shower and exercise while I listen to my colleagues speak, all while the pod glides towards the meeting venue at a sedate (and low friction!) 20 miles per hour. By lunchtime, all the local participants have gathered for the common meal, handshakes, one-on-ones, etc. At 2 pm, we go back to our pods, continuing the meeting while we travel home. Our colleagues overseas may also participate, and never leave home. If I lived in a conurb with the density of "sprawling" Houston Texas, 1000/square mile, my 40 mile "commute" radius would encompass 10 million people, but I would waste less time and energy than a Brooklyn-to-Manhattan commuter. This is by way of illustration - I'm sure that hypertechnology and the ingenuity of billions of people will do much better.

The fuel for this new world will be massive amounts of bandwidth and computation, powered by more transistors per person than exist in the entire world today. That won't change what we want or who we are, but it will create new easy and cheap ways to satisfy our needs.

Glaeser bemoans (perhaps rightly) that land use laws have moved growth from the most sustainable places on earth (coastal California at this writing) to brown places like Houston. However, given inevitable climate change, San Jose is the new Houston anyway, much dryer and hotter than it is now. The problem with cities is that they are very hard to pack up and move. Perhaps we should stop building them like deeply anchored towers, and instead build them from transportable modules. Since our first cities were built, we learned how to make standardized, interchangable parts; a 100,000 square foot building module, configurable as an apartment building or a semiconductor fab, might be an interchangable part, carried around the country and around the world like a jumbo shipping container. Again - billions of smart people will come up with even better alternatives.

Glaeser bemoans the regulated flatness of Paris and Mumbai, but partially excuses Paris because it is historic (I agree). But he doesn't mention the engineering/seismic reasons for such restrictions. Mumbai, like much of San Franscisco and other large west coast cities, is built on sediment and fill, which turns into jello in earthquakes. When a big earthquake hits these places, a lot of buildings come down. Mumbai has been spared so far, but one look at the geology of the region shows that this is lucky happenstance. Or perhaps very unlucky - it may be the first "million dead in 30 seconds" disaster the world has ever seen, and millions more may die until sufficient help can arrive. India needs tall cities, but not in that part of Maharashtra, or in similar places in California.

Stupidity emerges with every new generation, and is destroyed by disaster. Glaeser is correct that single function, education-poor cities like Detroit or New Orleans are dying, and that education rich cities like Boston, that move from strength to strength, have endured over time.

From his east coast university, Glaeser forecasts a Detroit-like fate for "single industry" "low density" Santa Clara County, entirely missing the essentials - the place is new, it has already moved from agriculture to aircraft to electronics to software, and has many world-class universities, from Stanford and Berkeley and UCSF, through many mid-tier centers of excellence like Mills and Santa Clara University and San Jose State, down to a lot of crazy little "woo-woo U"s that may specialize in eastern mysticism now, but may evolve, like religiously founded Harvard and Tufts, into diverse and world class secular powerhouses. Harvard was founded almost 400 years ago; what will Stanford accomplish by the time it is that old? Glaeser scolds Santa Clara County for its low density and nimbyism and low growth rate of housing. But how does that stack up against Boston's Suffolk County?



2010 population

Land area




Suffolk ( Boston MA )



152 km2




Santa Clara ( San Jose CA )



3,342.9 km2




With a 200 year head start, Suffolk county is 9.2 times more dense. At current growth rates, Santa Clara County will be that dense in 72 years. However, much of Santa Clara county is mountainous; it contains Mount Hamilton (home of Lick Observatory) at 4,370 feet, taller than any east coast mountain between New Hampshire and North Carolina. Given seismic, water, and steepness constraints, Santa Clara may never grow to 16 million people, but its stronger connection to the world's largest economies around the Pacific Rim may exceed all current expectations. If it sustained that growth for another 207 years, when it will be as old as Suffolk County is now, it would house nearly a billion people, and that is unimaginable.

Glaeser's ancestors came from Germany - whose cities were bombed to rubble in World War II. By some amazing miracle, we haven't launched the thousands of nuclear warheads we've built. If we are blessed by a far more amazing miracle, we may manage to get rid of all of them, and never ever build more. But over the coming centuries, the fate of Germany may be visited 1000 fold on the world's dense cities. As horrifying as this is to think about, cities may someday disappear forever. Perhaps we should learn to duplicate their virtues in our existing sprawl, rather than move all mankind into dense target zones.

Glaeser's book has many useful ideas. He is absolutely correct that education is the key to economic growth, though education may be a very different animal a century from now. Hopefully, that education will be the cure for the nimbyism and sprawl-promoting tax codes that favor suburbs. More hopefully, generations to come will learn how to live sustainably and convivially at all densities, ready for anything that nature or human cussedness can throw at us.

TriumphCity (last edited 2014-02-17 18:05:57 by KeithLofstrom)