[Editor’s note: a version of this post originally appeared in Dissent]
Why is housing in booming U.S. cities increasingly unaffordable to everyone but the wealthiest? In early September The New York Times published a provocative op-ed that answered this question from a market-oriented perspective. Drawing on their widely cited 2015 paper, “Why Do Cities Matter? Local Growth and Aggregate Growth,” urban economists Chang-Tai Hsieh of the University of Chicago and Enrico Moretti of the University of California, Berkeley contended that “[s]ince the 1970s, a property-rights revolution—what critics call Nimbyism, from ‘not in my backyard’—has significantly reduced the development of new housing stock, especially in cities where the economy is strongest,” thereby driving prices up to their current astronomical levels. Moreover, by impeding worker mobility and recruitment, “too-stringent housing regulations in high-wage, high-productivity cities” have resulted in “slower economic growth, fewer jobs,” “lower wages across the nation,” and ultimately “forgone gross domestic product” of $1.4 trillion.
Hsieh and Moretti had reason to think that their op-ed would be well received. Since its publication two years ago, “Why Do Cities Matter?” has been routinely cited by influential purveyors of the market creed, including some with liberal credentials—among them the Obama White House, the California Legislative Analyst, Vox cofounder Matt Yglesias, and economist Paul Krugman. Two days before Hsieh and Moretti’s op-ed appeared, Krugman opined in his Times column that “Nimbyism is bad for working families and the U.S. economy as a whole, strangling growth precisely where workers are most productive.”
Although they mention Boston, Seattle, San Francisco, and New York, Hsieh and Moretti home in on California and specifically the Bay Area, where, thanks to challenges brought by “neighborhood groups,” the “main effect today” of the “well-intentioned” but ill-used California Environmental Quality Act (CEQA) is to “mak[e] urban housing more expensive.” The two economists applaud a controversial bill making its way through the California Legislature, SB 35, that “would significantly curtail municipalities’ ability to delay urban housing projects that meet certain planning and environmental standards.” With a closing nod to the “debate in Washington about the costs of regulations for economic growth,” they assert that the reformation of “exclusionary land-use regulations in our most dynamic labor markets” would primarily benefit America’s middle class and thus deserves bipartisan support.
Only Sort-Of Anti-Government
Their contrarian tone notwithstanding, Hsieh and Moretti only advance the neoliberal agenda that has dominated U.S. public discourse for forty years. That agenda is often construed as anti-government, a view that the op-ed’s attack on zoning and CEQA may seem to confirm. But as indicated by the co-authors’ endorsement of state intervention in local land use, “supply-side” pundits are not against government per se, only government that hinders market forces. Today, they’re using the urban housing crisis as a pretext to roll back environmental protections, curtail local democracy, and deregulate, or more precisely, re-regulate land use on behalf of property and finance capital. The Times op-ed, however, isn’t just another neoliberal diatribe. It significantly extends the case against regulation through its contention that zoning, a municipal function, has national effects.
Despite its currency among policy wonks, this argument fails on empirical grounds. Hsieh and Moretti conceded in their academic paper that their findings were highly conjectural. Nevertheless, they’ve presented those findings as justification for an aggressive, market-oriented, democracy-adverse approach to land use. Their disciples have followed suit, to considerable effect: in mid-September, the California Legislature passed SB 35 with bipartisan support. Two weeks later, an enthusiastic Jerry Brown signed the bill into law. In the interest of rational public policy, the blue-sky thinking that legitimates such politics needs to be brought down to earth, and down to California terrain in particular.
First up for a reality check: the idea that over the past fifty years “increasingly draconian zoning restrictions” have significantly stifled housing production in “expensive coastal U.S. cities.” As Hsieh and Moretti noted in “Why Do Cities Matter?,” their cited source of this charge, Harvard economist Edward Glaeser, provided only “anecdotal” evidence in its support.
A Crude Approach to Measuring Growth Control
A lack of empirical traction also vitiates what the two economists call “the best available measure of differences in land use restrictions,” the Wharton Residential Land Use Regulatory Index. For starters, the Wharton Index relies on data from only a short period at the turn of this century. Far more problematic: the sources of that data—a national survey of 2,649 municipal planning directors and chief administrative officers; 50 state profiles of residential land-use regulation based on judicial and legislative activity over 10 years; and “measures of community pressure” registered in environmental and open space-related ballot initiatives—do not register on-the-ground outcomes. As sociologists Kee Warner and Harvey Molotch have observed, the “crudest approach” to identifying “‘growth control’ as a variable” is
to simply lump together all places that have some new way of regulating growth or that have the words ‘growth control’ written into some legal measure or stated as part of a local policy by a staff person answering a questionnaire. . . . This approach blurs great differences in the content of various local policies, not to mention how well policies are carried through in daily administrative practice.
Implementation of public policy is always uncertain. Impervious to contingency, the Wharton Index is a dubious guide to the actual effects of local land use regulation.
Nor do Hsieh and Moretti’s allegations of CEQA abuse stand up to empirical scrutiny. The attack on California’s premier environmental law as a deterrent to growth, a stock-in-trade of the state’s growth elites, was refuted by the in-depth 2016 study commissioned by the Rose Foundation for Communities and the Environment. The researchers found “no evidence” to support the assertion that the law is “‘a major barrier to development.’”
Moreover, a survey of projects undergoing CEQA review statewide since 2002 revealed a “surprisingly low” rate of CEQA litigation,” with an average of only 195 lawsuits a year. Meanwhile, “the vast number of CEQA projects . . . go unchallenged.” The researchers acknowledged that meeting the law’s complex procedural demands takes time and money. That said, “the cost of CEQA compliance [and] its impact on development projects” have never been quantified. Nobody has shown that, as Hsieh and Moretti assert, the law’s “main effect” is to increase the cost of urban housing.
Regulation Is a Bogeyman
Instead, as planner and University of Southern California faculty member Murtaza Baxamusa has written, “regulatory hurdles are a bogeyman for the housing crunch.” Baxamusa backs up this claim with evidence from his own city of San Diego, where, downtown, “there is virtually no NIMBYism, and development permitting is mostly by right,” yet “private developers are building fewer units than the zoning allows, and avoiding building affordable housing altogether, despite a tower of regulatory incentives.” More affordable housing is “being demolished than [being] built.” Since 2015, the unsheltered homeless population downtown has spiked 60 percent.
To explain this seeming conundrum, Baxamusa spotlights a blatant factor in the supply of affordable housing that Hsieh, Moretti, and their fellow supply-siders ignore: private developers don’t take advantage of permissive zoning or incentives to build affordable housing, because doing so doesn’t yield the profits that they and their investors demand. In the supply-side narrative, developers are at the mercy of local authorities. “Cities and counties,” writes California Legislative Analyst Mac Taylor, “generally decide when, where, and to what extent housing development will occur.” That’s true, insofar as new construction requires entitlement and a building permit.
What’s not true: the notion that cities and counties build housing. Developers build housing, and what they decide to build—and when and whether they decide to build it at all—depend on factors that over which local governments have no control: the availability of credit, the cost of labor and materials, the cost of land, the current stage of the building cycle, perceived demand, and above all, the anticipated return on investment.
Because affordable housing doesn’t yield acceptable profits to real estate investors, the only way a substantial amount of it is going to get built is if it’s publicly funded. In California, as elsewhere in the United States, public funding is paltry. And California has an extra deterrent to housing production of any sort: Prop. 13, passed in 1978, severely limits property tax increases, impelling cities to favor commercial development, especially retail, with its sales-tax revenues, over new housing. These are the major constraints on the supply of affordable housing in California. None of them figure in Hsieh and Moretti’s analysis.
The co-authors’ treatment of demand and affordability is also deeply flawed. Blaming the Bay Area’s exorbitant housing prices on a regulation-based failure to meet demand, Hsieh and Moretti disregard the stunning wealth effect generated by the latest flood of highly compensated tech workers. Between 2010 and 2015, the Bay Area added 640,000 new jobs. More than a third of those jobs—roughly 230,000 positions—were in the so-called knowledge sector, which includes technology. According to the 2017 Silicon Valley Index, in 2016 median wages for “Tier 1 occupations,” a category encompassing managers, professional positions (lawyers, accountants, and physicians) and highly skilled technical occupations (scientists, computer programmers, and engineers) were $108,700.
Plan Bay Area 2040, the land-use and transportation “blueprint” approved by the region’s planning agencies in July, reported that between 1990 and 2015, the number of households in the region with an income greater than $150,000 constituted 80 percent of household growth in the region, jumping from 375,000 to 741,000. In 1990, such super-affluent households accounted for 17 percent of Bay Area households; in 2015, they made up 27 percent of the total. In the succinct appraisal of economist Steve Levy, who runs the Center for the Continuing Study of the California Economy: “This is a very, very hot area to live and work, and the wage growth is pushing up housing prices.” That boost is exacerbated by the fact that housing, especially housing in “very, very hot” areas, is now traded in a globalized speculative market.
The inflationary impacts of what Jane Jacobs called “cataclysmic money” escape the authors of the Times op-ed. That oversight evinces the blinkering effects of their methodology. Like other mainstream economists, Hsieh and Moretti assume that, absent state “interference,” the economy is propelled by businesses seeking the most efficient means of production and consumers seeking the greatest utility—that is, personal satisfaction based on a trade-off among wages, costs, and amenities. Equipped with perfect knowledge about their options and enjoying the complete freedom to exercise those options, these putatively rational actors move at will, achieving a “spatial equilibrium” that maximizes both efficiency and “welfare.” High wages indicate high productivity; their wealth effect is indiscernible.
In “Why Do Cities Matter?,” Hsieh and Moretti used the equilibrium model to come up with their estimated $1.4 trillion-plus loss in GDP. Marshaling data from 220 cities, they argued that between 1964 and 2009, local restrictions on housing supply in municipalities with high productivity and high demand for housing—most notably, New York, San Francisco, and San Jose—kept workers from low-productivity areas from moving to the high-productivity locales.
To portray the attendant losses in employment growth, they elaborated two counterfactual (what-if) scenarios. Counterfactuals are useful aids to evaluating proposed policies, but to merit serious consideration, they have to be based on credible assumptions and sound data. In their counterfactuals, Hsieh and Moretti hypothetically “reallocated” U.S. workers from low-productivity to high-productivity cities. In the more sweeping, “full adjustment” scenario, they relocated just over 50 percent of U.S. workers.
As a result, New York experienced “a staggering 787 percent increase” in jobs; in San Jose and San Francisco, employment jumped more than fivefold. Flint, Michigan, lost 98 percent of its workforce. In the “intermediate scenario,” 20 percent of workers moved. In New York, employment increased by “only 179%”; in San Jose, it grew by 149 percent; in San Francisco, by 148 percent. Flint’s workforce shrank by 77 percent.
The figure of $1.4 trillion cited in Hsieh and Moretti’s New York Times op-ed presumably represents the additional amount of GDP that would have been generated in 2009 if the “full adjustment” scenario had been a reality. I say “presumably,” because that figure, cited in the op-ed, appears nowhere in “Why Do Cities Matter?” Instead, the academic paper refers to a lost $1.95 trillion. In another discrepancy, the updated May 2017 version of the paper presents the additional $1.95 trillion as “the effect of changing the housing supply regulation only in New York, San Jose, and San Francisco to that in the median US City,” rather than, as the co-authors stated in 2015, the effect of changing the “spatial dispersion” of workers in “All Cities.” More precisely, or, as it turns out, imprecisely, Hsieh and Moretti write in the updated paper that the “net effect is that US GDP in 2009 would be 8.9% higher under this counterfactual,” adding in a footnote: “US GDP in 2009 was $14.5 trillion so a GDP increase of 8.9 percent implies an additional aggregate income of $1.95 trillion.” No, it doesn’t: 8.9 percent of $14.5 trillion is $1.29 trillion. The co-authors have gotten their own numbers mixed up.
And there’s another, more notable discrepancy. The Times op-ed presents the figure of $1.4 trillion without qualification; the word “trillion” is even italicized. But in “Why Do Cities Matter?” Hsieh and Moretti cautioned that “the number [“of output and welfare losses stemming from an inefficient geographical allocation of labor”] we present should not be taken as precise estimates of the losses but rather as guidance on the general order of the losses, as they are based on a number of untestable assumptions,” including “perfect mobility” for workers: “The assumption of inter-industry mobility,” they concede, “is clearly false in the short run. For example, it would be hard to relocate a Detroit car manufacturing worker to a San Francisco high tech firm overnight.” Quite.
Also admittedly problematic is the “extreme” nature of “the changes in the economic geography” in the “full adjustment” scenario. In Hsieh and Moretti’s own judgment, these imagined shifts “are massive and probably not realistic,” given that “less than 20 percent of workers change MSA every 10 years.” (MSA stands for Metropolitan Statistical Area, defined by the U.S. Census Bureau as a place with “at least one urbanized area of at least 50,000 or more population, plus adjacent territory that has a high degree of social and economic integration with the core as measured by commuting ties.”) In light of these caveats, the two economists designated the “partial adjustment” scenario as the “more plausible” eventuality and indeed as their “benchmark scenario.” But the flaunted figure of $1.4 trillion dollars in lost GDP seems to derive from the less plausible scheme, in which more than half of American workers relocate.
Not that their benchmark scenario is all that plausible. Volunteering that the employment numbers in their “partial adjustment” scenario “are not completely implausible,” the co-authors offer as evidence projections in the 2013 version of Plan Bay Area. That “formal economic development plan,” they write, “calls for the addition of enough housing units to increase the region’s population by 80 percent in 2030. This increase is smaller than the one [we estimate] for the San Francisco MSA, but not too far off.”
Hsieh and Moretti don’t seem to have read the report very closely. In fact, they don’t seem to have read their own report very closely. The increase that they estimated for the San Francisco MSA under the “partial adjustment scenario,” a jump of 149.2 percent, refers to growth in employment, not housing units. Corresponding population growth, given their assumed labor share of 0.65, would reach some 230 percent. As for their misreading of Plan Bay Area: its projections span 2010 to 2040, not 2030; those projections refer to either the whole region or to specific cities; nowhere does the plan forecast growth for any MSA; and none of those projections—a 30 percent increase in population, a 33 percent increase in jobs, and a 24 percent increase in housing units for the entire region—come anywhere near an 80 percent increase in population, much less a 149.2 percent increase in employment.
Last April, Glaeser, the superstar dean of neoclassical urban economics, commended Hsieh and Moretti’s calculations. “Whether these exact figures are correct,” he advised, “they provide a basis for the claim that America’s most important, and potentially costly, regulations are land use controls.”
Yes, they do: a basis in a free-market fantasia whose speculative prognostications should have no place in the formulation of public policy.