Matt Kahn and housing in liberal cities

Well, while I was brushing my teeth, Matt Kahn went out and published another very interesting paper:

Kahn, M E. 2010. Do liberal cities limit new housing development? Evidence from California. Journal of Urban Economicsdoi:10.1016/j.jue.2010.10.001.

The finding? Cities with a higher score on liberal voting also permit less housing.

It’s less clear what this finding reflects. Kahn brings up Berkeley’s attempts at growth control–which in some ways make no sense from an anti-sprawl perspective because already developed areas should be taking more housing, not less, to direct growth away from fringe areas. As my former colleague Jesse Richardson says, Smart Growth is not birth control, and if urban areas are to grow not on the fringe than existing areas have to take their lumps.

Kahn’s manuscript may be an indicator that California cities are not taking their lumps.

So why is this a problem? It is more evidence that cities can cherrypick Smart Growth and New Urbanist principles to take what they like and leave what they do not: the line from the New Urbanism as that we could, with more density, get plenty of housing–some of it affordable–if we simply stopped allowing suburbs and we deconstructed single-use, residential zoning. Unfortunately, the application appears to be to shifting housing to other jurisdictions or restricting housing permits overall–a far miss from the sort of densification that the New Urbanist and Smart Growth advocates envisioned.

This is, of course, only if the connection to liberal voting and housing permits occurs via this increased willingness to take on growth controls. It could be something else, as Kahn doesn’t have regulatory structure.

9 thoughts on “Matt Kahn and housing in liberal cities

  1. Thanks for quoting me! The results of this study do not surprise me. It has always puzzled me that “conservatives” favor affordable housing while “liberals” often favor “No Growth” policies that do great harm to low to moderate income families. A recent article in the Washington Post showed that most of the residents in the D.C. area are spending far more than the recommended percentage of their salary on housing than they should be. I don’t have a study to back me up, but my guess is that a lot of that comes from inadequate supply, particularly of multi-family residential.

  2. The idea of selectively using frames to justify strategic action rather than following a coherent ideological model to its logical limit is what we sociologists usually call “the cultural toolkit.” That is, we don’t view culture, ideologies, etc, as constraining action so much as providing them opportunities that they can more or less strategically exercise. In this sense it makes sense that people have wide varieties of often contradictory ideas. The Ur-cite for this is Swidler ASR 1986. (

  3. Maybe Matt’s lit review reflects this, but this connection has been part of the lit on growth control (with varying empirical results) since at least the 1970s. (David Dowell wrote an early piece on it, I think, like in the late 1970s or early 1980s. Mark Baldassare, Mark Gottdiener, Betty Deakin, Madeline Glickfeld, Mai Nguyen, and many others have also written about the determinants of growth control.) What hasn’t really penetrated the discussion is that the most effective growth control, whether we’re talking about all growth or about affordable housing or about apartments, isn’t a permit cap, a growth boundary, or even a moratorium. It’s low-density zoning. (Self-promotion: Pendall 2000, JAPA, Chain of Exclusion.) Republicans can’t do anything about it because a big chunk of their base live in the communities that use it to keep “the wrong kind of people” out. Look in the Boston area, and you’ll see that the most exclusionary (i.e., growth-controlled) towns are the ones with the highest Republican voter registration levels. Funny thing: I was in Loudoun County VA a couple weeks ago for a presentation at the County Planning Department, and someone from city government (who shall not be named but is in fact a Republican) said that the people there who want to limit high-density, mixed-use development don’t want it because they know places like that just breed Democrats. I.e., “those people,” yes, this is literally what she said, though she was saying it as if she actually disagreed with this attitude.

    WRT Berkeley, there are a couple left-of-center factions at war there over growth policy, with the recently ascendant pro-transit/ecocity progressives battering down exclusionary rules and building up five-story mixed-income apartments along University and Shattuck Aves. in spite of dogged opposition by neighborhood-conservation preservationists. Now there’s a downtown plan that will also add quite a bit of new housing near BART. Lots of “progressive” communities do have strong conservative tendencies; they are like other Tiebout consumers, right? just wanting to keep things like they were when they finally found a place with their own values. OTOH if I walk around my current city, Takoma Park, which is the Berkeley of the DC area, I see a tremendous mix of housing from single-family houses on modest lots, to 2-plexes, to 20-story towers. Of course the zoning here is done by Montgomery County, but in the less liberal-radical parts of the county there’s nowhere near this fine-grained a mix. There isn’t much new growth here, of course, but that’s because more growth would require serious urban renewal. (On the third hand, a lot of the land use policy in Takoma Park may also have been set out when it was still the center of the Adventist religion as a concession to religious leaders’ desires for enough housing, of varied enough price, to accommodate the faithful. Better do some history, right?) Manuel Castells has also written extensively (in The City and the Grassroots) on the paradoxical conservatism of many urban social movements. Parallels here?

    On affordability in DC: I suspect more and more (with back-up from the op-eds I read by Krugman) that the intense affordability crisis is driven in part by a combination of income inequality and the desire of some really high-income people, many of them only recently out of college and some still supplemented by $ from mom and dad, to live really close to centrally located federal offices. Plus all the college students at GW and Georgetown (whose parents now must earn 5-10X as much in constant dollars than the parents of the students in these schools 25 years ago). Too much money at the top, untaxed and unredistributed, means more money to throw at excessive housing consumption. Further, the article in the Post was about how many people earning middle incomes spent over 35% of their incomes on housing. I’d like to see how Stone’s measure of Housing Poverty would look instead of that standard. For the lowest-income households, even 30% is too much. But when a household earns $50K per year, they can easily pay 40%, especially if they can live closer to work and spend more time either making money or having fun and less money on transportation. The problem comes lower down the wage scale. Building more housing at the regional scale probably won’t solve this “problem” of middle-income affordability (because –hypothesis here — the middle-income people who “overpay” are doing so precisely because they want a place close in); building more in the locations where they really want to live will also not help, because the construction costs make the housing they want completely impossible for them to afford even at 60% of their incomes without subsidy. Not to mention the Congressionally mandated height limit … which quite a few DC progressives have argued ought to be eliminated, at least in the parts of the District that are twice as far from the federal core as Rosslyn. And then there’s the whole east-west divide, which means you can pick up houses in Prince George’s for a song, but you can’t touch anything west of Rock Creek and east of Leesburg for less than $600,000.

    I happen not to think that the appropriate response to such excessive demand is always “build more housing,” considering the durability of the stock and the negative impact on innocent bystanders after the crash. (Vacancy rates here are not especially low, I think.) Texas took the better part of a decade to recover from its speculative surplus in the 1980s (hair-trigger supply conditions), with the heaviest collateral damage to black and Latino home owners and renters, often elderly.

    California is a special problem, and one that is more intractable than any other I can think of. It is definitely threatened long-term by inadequate supply, unlike most other parts of the USA (since the most effectively growth-controlled regions, those with extensive exclusionary zoning, are now going to stop growing altogether if Dowell Myers is right), and that it ought to be a national priority to figure out how to deal with it instead of being treated as a curiosity. Oh, those crazy Californians! Academic and government-commission hand-wringing about local government growth control doesn’t seem to make a difference, however; it’s been a litany since the 1970s. There are many, many other reasons for the strangeness of its housing conditions: influxes of Asian investment capital in all kinds of real estate; some of the highest levels of income inequality in the USA; with those high earners, some of the strongest tax-related inducements to overconsume housing in the US; with Prop 13, some of the lowest inducements to move (including for commercial property owners, creating massive stickiness in the land market); tragic levels of underinvestment in infrastructure (which of course only makes people want to limit growth in their neighborhoods and cities); seismic conditions; endangered species; and highly empowered local residents. Probably all of these, and more, would need to be considered in any effort to increase housing supply in California. Would it be a better idea to stimulate economic development elsewhere in the United States? To reduce restrictions on immigration in hostile territories farther east and north? To start planning to market the boomers’ surplus suburban units (Myers & Ryu 2008) in New Jersey, Long Island, and Connecticut to cramped and fed-up Californians?

    Can California possibly build its way out of this huge backlog of housing demand now? Where would the housing go? Who would build the infrastructure? Not in the Central Valley, by any means? With its nation-high rates of poverty and unemployment, its Tule fog, its dreadful weather summer and winter…? More housing in Riverside and San Bernardino, where it’s 100 degrees and smoggy and transit is mostly unviable? Or, should builders be allowed to pave the vernal pools, the artichoke fields, the coastal scrub?

    And it still drives me completely batty that whereas environmental economists who study pollution control have been working in a cost-benefit framework for two generations (testing the costs and benefits of the best available technology), those who study planning and land-use regulation still stop at the cost side, as if, QED, regulation is bad because it associates with higher prices. And therefore the policy is self-evident, just take the regulation away and the problem will go away. Ed Glaeser says New York should change its regulatory environment to be more like Houston’s. Really? How many people actually believe this would fix the affordability problems caused by the people who can live in New York and nowhere else, including those for whom money is no object: bond traders, students paying full tuition at private universities, the list goes on. No doubt that supply would help some of it, and New York has in fact built loads more housing than its suburbs has in the last 15 years and will probably build loads more. But part of the problem, again, still, is that there are just so many really, really rich people in America that urban real estate markets have become extremely tough places for housing affordability. And noplace more so than in the Northeast Corridor and California.

    • I DO believe you, of course, but…why the voting share result? You’d think that the vagaries of the California market would be pretty constant in urban areas across the state due to the factors you mention.

  4. Rolf, by “most effective growth control” do you mean most effective in excluding low- to moderate-income families? In practice, low density zoning is an extremely popular exclusionary tactic.

    I’m from Virginia, and Republicans have come over to the exclusionary side, at least in some instances. Our prior Governor, Tim Kaine (a Democrat), perfected what I call “NIMBY politics”. He ran on a promise to “stop runaway development”, with commercials showing traffic jams in Northern Virginia. I wondered how he was going to do that- ship the jobs to West Virginia?

    After he was elected, he set a goal of placing perpetual conservation easements on 400,000 acres during his administration. This number was apparently arbitrarily chosen and there was no targeting of the easements- the first 400,000 acres that come in the door. Note that Virginia also offers very generous (and transferable) state income tax credits, on top of federal income tax benefits, for easement donations. Governor Kaine’s Republic successor, Governor McDonnell, has adopted the same goal for his administration. Forget about a chicken in every pot, it’s a conservation easement in every backyard. Planners should be concerned, but are mostly cheerleaders for indiscriminate conservation easements- more is better. By the way, lots of conservation easements form a pretty good exclusionary strategy also (obviously lowering densities).

  5. I studied the article just now. The best regression models still had R2s in the mid-0.50 range (leaving a lot of unexplained variation). Liberals buy hybrids, yes; they also have different preferences for housing when they live in metro areas. If Berekely, Takoma Park, Ithaca, and Santa Monica are any guide, they like places with sizeable shares of historic housing units; lively and fine-grained downtowns; an interesting mix of housing types in their neighborhoods; access to mass transit; and maybe even narrow frontages (in the 50-60 foot range). So, instead of liberal ==> low levels of permitting, the real causality may run historic + built-out + highly fragmented property ownership structure ==> low levels of permitting, with “liberals” being the proxy for those conditions in the urban form that depress new construction. (Several JUE articles also address how land fragmentation depresses construction and can promote leapfrog development.) What would happen if, instead of liberal share, something more like “median age of the housing stock,” or “percent of the city land area with no housing units on it and fewer than XX jobs per acre in 2000” were the controls?

  6. PS Jesse, by “most effective,” I mean “most effective at slowing growth,” at least in the 1980s, not just most effective at reducing low-income population. It would be fun to re-test my models with my 2 land-use regulations databases (1994, N= 1100, 2003, N=1550, more or less, 25 biggest and 50 biggest metros, respectively) now that I have better access to modeling expertise.

  7. Rolf, I maintain that you cannot “slow” growth in a region with land use tools (mandatory birth control, immigration reform and sending jobs elsewhere may do it). Growth “controls” merely deflect inevitable growth elsewhere, often with negative consequences. Any type of growth “control” ultimately hurts low- to moderate-income families- they are the ones that have to go elsewhere to find housing. I hope you aren’t advocating for low-density zoning as good growth “management”. Loudoun and the surrounding areas are good examples of what a disaster low-density zoning (often in the guise of “farmland protection”) creates on the ground.

  8. Jesse: Agreed. Low-density-only zoning is evil, precisely because it *is* the most effective policy for reducing the amount and density of development (and therefore at excluding low-income people and minorities, but mainly, I contend, by limiting rental housing construction rather than by raising single-family house prices). Very low density zoning in some places coupled with medium- to high-density zoning in others (enough for a 20- to 25-year projection of growth), marginal-cost pricing for infrastructure, congestion pricing, open-space amenity taxes, and pro-growth investment in established areas is good. California cities (those in Matt’s study) don’t practice low-density-only zoning, though many cities — and more counties — do impose much larger minimum lot sizes in parts of their territory than most places in the Northeast could ever get away with.

Comments are closed.