Update: Ryan posted a reply to some of the points I raise here. It’s definitely worth reading.
The Gated City is a mini-ebook by Ryan Avent that makes the case for removing restrictions on densification in cities. In addition to being a left-leaning economist, Avent is also a journalist who is an editor at the Economist magazine and a principal contributor to its Free Exchange blog.
Avent’s journalism skills make him one of the more articulate and easy to read economists out there. This book brings Avent’s signature readability to the table and also has the virtue of being brief. This makes the book very much a recommended read. To me though that is as much for his introductory treatment of the economic value of cities as it is about the virtues of density. The primacy of urban regions in our new economic order is something that is hardly a matter of left or right. But I continue to be amazed at how few policy makers actually seem to know or care about this, particularly at the state level. Our governments continue to implement policies that are not just indifferent to the success of our cities, but in many cases outright hostile to them. More reminders of the real state of affairs are clearly needed.
Having read several other reviews of the book, I was expecting big and controversial claims to be made for density. In fact, Avent’s claims for the benefits of increased density are fairly modest. He suggests that removing artificial barriers to urban density might raise the growth rate in GDP by 25 to 50 basis points. That’s not a radical amount, though of course compounded over time adds up. He also doesn’t claim that more density is always better, that we should all live like Manhattanites, or that he professes to know what the optimum level of density is. In fact, he explicitly disavows all of those. He also says that he doesn’t care where people personally choose to live, and, while I would have preferred to see a more forthright statement on policies like urban growth boundaries and such that actively restrict suburban development, Avent does include New Urbanist critics of suburbia as among those who want to more strictly regulate land use.
In short, Avent more or less wants to reduce barriers to densification such as zoning, FAR limits, historic districts, etc. in order to allow the market to determine the optimum level of density. He hypothesizes that this would lead to an increase in density overall, and given that both the goal and practical result of much planning practice is to restrict density, I agree. (See: Planning and Free Market Density). While clearly Avent is sympathetic to urbanism and densification, in this book at least he makes a clear effort to avoid statements that might suggest he wants to force density.
On the whole, I’m personally very supportive of Avent’s goals here, though of course the best course for overcoming the huge structural biases against density is still speculative.
There are a few areas that weren’t clear and that I weren’t quite as sold on. One of them is that in some cases Avent talked about employment density and in other cases residential density. Those might be related in many cases, but I don’t think they are necessarily tightly linked. Most of the economic benefits would appear to occur from employment densification not necessarily residential. For example, until fairly recently, Chicago’s Loop business district was almost exclusively office oriented and had little residential density at all. While adjacent urbanized areas are higher density in terms of people, I don’t believe that’s the only factor supporting the Loop economy, or even the most important. (See: The Big City CBD Advantage and Employment Challenges Facing Small City Downtowns). Fully distinguishing between employment and residential density in terms of impact and policy would have been helpful.
Another is that the book heavily focuses on the supply side of the equation (i.e., the ability to build) as opposed to the demand side. But arguably demand had a bigger role to play in driving up housing costs in places like NYC. It came as no surprise to me that the metros Avent cities most in his dense cities list were New York and San Francisco. Those places are home to the two most successful macro-industries of the last 20 years in terms of wealth generation for their practitioners – finance and high tech. Avent notes increasing income inequality and stagnant wages for the majority. This by itself explains why the middle class is increasingly priced out of these cities. When the Fed pumps $14 trillion into propping up the financial sector in the last handful of years, it’s not unrealistic to expect that to show up in New York area housing prices. No amount of new supply would have been able to keep up with the significant deregulation of finance that occurred in the last two decades (repealing Glass-Steagall, raising the ceilings on the share of national deposits held by a single institution, etc) and Helicopter Ben’s printing presses.
Also, assume we were able to moderate the price of housing in these places somewhat. What would that do for us? Well, NYC and SF would still remain among the most expensive places in America to live. This means that they are going to continue to draw principally those who are able to tap into the particular wealth generating functions of those regions, along with those who particularly value the amenity set of those cities or who are following network path migration. However, the industry groups that are now predominant in those cities are ones that employ almost exclusively high end labor. So what this would mean in my view is that instead of say the top 1% being able to live and work in these places, we extend that down to the top 2% or 3%. That might add more new net economic growth, and even benefit the people who fall into those newly eligible brackets, but a few more near luxe jobs aren’t going to solve the fundamental problem of employing middle class Americans and restoring the engine of upward social mobility. The factories that used to be a big part of the economic life of Silicon Valley are now gone, and they aren’t coming back no matter what building restrictions we remove. Even if we could grow employment there, it would be almost certainly be more jobs for those who, as Joel Kotin put it, “already reside at the apex of society or expect to soon” thanks to their education, skills, or family connections.
With the current economic development paradigm for tier one cities emphasizing talent, higher end activities, and “global city” type affairs, I also question whether local leaders would actually want to make it easier for what are basically lower end activities to take place there. Everyone seems to want to keep climbing the value chain, not descending it. If you lower prices, you open the market to more economically marginal people and activities, and thus lower items like per capita income and GDP, unless somehow on a lifecycle basis this somehow changes the overall economic structure of a region (which would probably raise housing prices again….)
Another part of the challenge is that the business models of high tech and finance have evolved to thrive in an environment in which they a) now find themselves in a globally competitive market and b) have to pay for sky high real estate and expensive people in SF and NYC. I’m most familiar with tech. I’ve written before how the onshore tech market has reinvented its business model to focus on lean methodologies, design, being close to the customer, and rapid iterations. When you combine this with innovations in open source frameworks and cloud hosting, the average web company simply doesn’t employ that many people these days, and likely will never scale its labor force by much because the economic model of something like SaaS is almost 1:N – virtually all costs are fixed. (See: More Thoughts on the Programmer Shortage).
So to really boost employment, we’d need more than just cheaper land making labor more available at a reasonable cost. That might provide some boost. But we would also need business model innovation.
One nitpick: the choice of Phoenix as the primary example of a low density city wasn’t one that resonated with me. When I think of Phoenix I think of a retirement or leisure community, not a business center. Texas cities like Houston would have been a better choice. (Avent does mention Houston at times, however).
Avent doesn’t exactly slam sprawlvilles like Houston, but he has a clear affinity for the high end city. I can’t really begin to address this, but a few things come to my mind. Firstly, the amenity gap between cities can be seen partially in terms of life-cycle. While certain things like the collections at the Met can never be reproduced in a place like Houston, that city will only grow more sophisticated and amenity rich over time. Cities almost invariably get rich, then decide they want to use some of that money to buy class. It was true for Chicago, and it’s increasingly true now in places like Houston and Dallas, as well as emerging market cities around the world.
Second, how much of the perceived productivity of places like NYC and SF result from them having excluded all but high end activities in the first place? Houston not only has its energy cluster that has created fantastic wealth there, it has also managed to build a city where workaday businesses can do their thing. If a similar range existed in the Bay Area, it would look a lot less impressive in its stats. While it might make sense that transplanting tech works to Silicon Valley would make them more productive, it’s not clear that would be the case for workers in most other sectors. Whereas in Houston not only can the energy worker be more productive, other types of businesses and workers aren’t chased away. I’m sure he’ll correct me shortly, it strikes me that Houston could in fact be very much what Avent wants to see. It’s home to a powerful macro-industry cluster – energy – but has low enough housing costs that it doesn’t deter people who want to work in that sector from coming to Houston.
Third, Avent does believe increased density pays dividends across the board, so even places like Houston would benefit from raising their densities. And suburban type densities are very easy to raise without fundamentally changing the built form or quality of experience that people have there. So long as somehow this didn’t end up compromising Houston’s low housing costs, presumably even Houston itself would benefit from more density in Avent’s view. I mention this lest I leave the false impression that this book is all about a handful of already very dense places.
One last comment: I did not see rent control mentioned anywhere in the book. This is clearly a factor in the housing markets in both NYC and the city of San Francisco.
My main takeaway was that if we increased housing supply, we could boost the labor force in proximity to some high value economic clusters that are critical to the US economic future. Avent’s belief in the benefits of density are broader, but this is what it stressed. While I agree with the policy goals, I find it difficult to believe that what amounts to some marginal benefit is going to be sufficient to overcome the vast anti-density inertia and structural forces out there. Nor do I believe that this fundamentally solves the economic problems facing the United States, especially boosting employment over the long term, raising middle class wages, and addressing the broken engine of upward social mobility. What Avent’s prescription amounts to is tweaking the dials on the engine to squeeze out a few more horsepower. That’s a worthy goal and a potentially useful action, but only a small part of the solution to the challenges we face.
While he doesn’t specifically address the housing cost issue, Kevin Drum has recently demonstrated how supply of housing affects demand:
http://motherjones.com/kevin-drum/2011/10/future-of-innovation
“Wait a second. Video games? Am I joking? No indeed. Give some thought to just what innovation and productivity gains are for.
Initially, of course, they help provide a better basic standard of living. But what happens after that? Once you have a certain level of food, shelter, sanitation, and so forth, you start adding nonessentials. Basically, luxuries, whether you call them that or not. Entertainment. Vacations. Restaurant meals. Fancier clothes, faster cars, and bigger houses.”
In other words, lower the cost of living of essentials like housing, and you immediately increase demand of non-essentials.
Even if it didn’t raise demand, and subsequently employment, a higher housing stock increases wealth because it lowers costs. It is a virtue in that respect alone.
Well except if lowered housing prices causes Keynesian economists to panic about the coming deflationary spiral and collectively decide to destroy everybody’s wealth. Oh wait…that already happened.
This is a great and thought-provoking post. I have two thoughts I’d like to elaborate on:
(1) I don’t think SF is a great example of a dense city, especially with respect to the Tech industry (which I also know intimately). SF’d employment density is much more finance-related than tech-related – the tech industry clusters in San Jose/Sunnyvale/Mountain View, a decidedly suburban area, albeit a very expensive suburban area. I see SF as a mix of (a) a medium-sized, finance-dominated dense employment district, and (b) a dense “bedroom community” for techies who want to live in “the cool city” before they settle down and have kids. As an example, Google has shuttles to bring workers who live in SF to their corporate complex in Silicon Valley ~20 miles south of SF. No Wall Street investment bank does anything analogous in NYC.
This is not to criticize SF, which is a fine and interesting city. But the city of SF is not really driven that much by the tech industry (as opposed to the metro area).
(2)You raise a great point about city life-cycles. Having lived extensively in both Boston and Dallas, and liked both, I know exactly what you are talking about. Dallas is simply a younger city than Boston in both physical and mental senses. The museums in DFW are younger and can’t rival the collections of old masterworks. The transit system grew up in an auto-centric world. There is a very different attitude towards development and preservation. Both cities are quite successful, but I would summarize them as follows: Boston already is what it wants to be (a refined, historic, university-centric city with many luxuries, but on a smaller scale than its old rival NYC) and is more fearful of screwing up what it has than of missing the next opportunity; Dallas is still growing fast and hasn’t (yet) become what it really wants to be (an internationally-respected major commercial city on a tier with LA and Chicago) and is determined not to miss any chance to grow.
Aaron wrote: When I think of Phoenix I think of a retirement or leisure community, not a business center.
You should read Jon Talton’s Rogue Columnist blog. He’s excellent at showing the dark side of the Valley of the Sun.
Talton, a former Phoenician, is an author and now a Seattle Times business columnist. He used to have the same duties at the Arizona Republic.
He doesn’t ease on the venom for his former hometown. And he’s a very engaging writer.
From reading Rogue Columnist, you could best describe the Phoenix region as “Fresno with light rail.”
Phoenix is a state capital, a county seat and the principal city, yet with the public sector stretched thin and held in wide contempt, you wouldn’t know downtown was the government center. Its downtown only seems to exist when there’s a home baseball or basketball game.
Otherwise, the Valley’s economy suffered acutely from Sunbelt Syndrome. Massive population growth just fed off itself, and its economy oriented around that (house-building, real estate and mortgage financing). The rest of Phoenix’s economy is being a branch office metro, with a bunch of nonspecialized work functions that are found in other cities, as well as low-wage retail and tourism.
More bad news: Migration slowdown is Sun Belt’s loss and North’s gain. Then again, the story is by USA Today, a newspaper nice to look at but thin to read.
Does Ryan Avent address how to get over density being its own worst enemy?
High density always calls attention to itself, and the solution is to stop any more of it. New York and San Francisco are marvelously dense, but that’s not because they embrace it. Communities there try to frustrate further density and tend to be anti-development.
One problem is that future residents are an invisible constituency and cannot influence the politics of a place where they don’t live. Present residents are indifferent or hostile to the interests of future residents. The political process then splits the difference. Density is not put where it isn’t wanted, and if there is development, it flows to the path of least resistance.
Much urban development has been concentrated either in deactivated commercial or industrial zones, or in poor areas with transient populations. (Poor areas with stable populations fight development fiercely.)
People forget that tax law as much as urban planning seems to determine how we live and work. When the bean counters realized that they could write off the cubicle partitions on a faster depreciation schedule than the office building itself, office work departed offices and went into cubicles.
In 1986, voters in San Francisco were asked to decide whether or not to strictly restrict the amount of new office construction in downtown San Francisco in Measure M. Proponents of Measure M argued and the consults reports seems to support the argument that because of Prop 13, what cities received in property tax revenue from building large office spaces in California did not cover the cost of new services created by new office spaces. See SPUR report below for more details. Since prop 13 has not been repealed and seems to have no chance of being repealed in California, the arguments of the proponents of Measure M still seem equally germane today.
http://www.spur.org/publications/library/article/propositionM07011999
http://calcoastnews.com/2011/09/we-still-love-prop-13-field-poll/
Ed:
Thanks for the reminder about depreciation schedules. Have you read Malcolm Gladwell’s article about the suburban mall? It makes a similar point – malls were unprofitable for investors until accelerated depreciation created phantom losses they could write off for tax purposes.
(One wonders if the unsustainability of American suburban development comes from this: accelerated depreciation means that for the first few years you get to pretend you’re losing money and not pay taxes on your profits, but afterward it balances out, and so the only way to maintain profit margins is to build newer things.)
Again the reasons for suburban development are more prosaic. I basically think regions are designed indirectly as a consequences of local governments trying to capture the most sales tax dollars.
In California (and I suspect elsewhere in this country) local governments compete fiercely for sales tax proceeds. In order for an area to incorporate (form a local city government) the region needs to convince the local agency formation committee (lafco) that there is sufficient tax revenue to sustain the proposed city. Generally if an area has a mall, an auto-mall or enough undeveloped space adjacent to a freeway to attract either the mall or the auto-mall then generally the lafco will approve the proposal to bring to voters the plan to form a new local government.
This has several consequences. First existing cities and local government over zone retail and over zone parking for that retail in an effort to keep sales taxes in there jurisdiction and to prevent it from leaking into other jurisdictions.
Second, new cities have a lot of discretion over what type of housing to go permit and they use this discretion to try to maximize sale tax receipts while minimizing direct government expenditures. If the new city mandates lots of large lot development, the developers will respond by building large McMansion style homes to recover the higher cost of land. So in effect a new edge city can zone most of the poor out of the region by mandating large lot sizes. More importantly the large retailers make location decisions based upon average incomes of residents in the immediate area. If your edge city has lots of wealthy commuters then your edge city mall will attract more retailers while expanding the local sales tax base.
Who is getting hammered right now are the older inner suburbs built post WW2 to about 1980. These areas are mostly built out. The housing stock is old enough where its starting to show its age. The local public schools in these areas tend to attract lots of immigrants who don’t speak English at home so the schools aren’t performing well. Because the demographics are in decline, the sale tax dollars are fleeing as the national retailers relocate out of the area for greener pastures.
Jane Jacobs explained to activists a theory of how to gentrify pre-WW2 neighborhoods but no one has a really good model of how to bring back the older inner suburbs. These neighborhoods are housing the poor and lower middle class people that can’t find housing in the edge cities and poor and lower middle class residents displaced by yuppies moving back to the city centers for shorter commutes. These are also the areas losing there sales tax bases right when the demand for government services seems to be going up. These are the neighborhoods with the dead malls. Other than Walmart the big national retailers generally aren’t interested in this demographic. But the capital cost of leasing a vacant big box retailer location is still too large for an entrepreneurial immigrant to take over these locations. Even if they mortgage there house and borrow from all of there friends and family, they still can’t cover the cost of the lease and the cost of inventory to fill the space so the space stays vacant.
My hunch is that the solution to this problem is some sort of regional sales tax sharing. But making that happen is problematic. Alternatively the local redevelopment agencies may need to buy up the vacant big box locations and subdivide the locations or subsidize the process to get developers to subdivide these locations.
Interesting points, ed. It seems to me that regional arrangments will work in young growing metros as it is easier to share a growing pie, but development corporations will work better in stable or declining ones because it is harder to share one that isn’t. The Cincinnati area has been more aggresively using various development corporations to successfully develop several central areas in recent years.
I’d be careful about generalizing based on sales taxes. Some states (Indiana, for example) have no or limited local-option sales taxes and primarily finance local government through property and income taxes, while state government relies on income and sales taxes.
In the Northeast, municipalities fund themselves through property taxes. The incentives are different from in California and people who know where to look will notice the difference, but broadly the results are the same: rich suburbs use zoning and other regulations to turn into gated communities, and the inner and outer limits of the poverty donut are both moving outward steadily.