houseEarlier this year, UW researcher Theo Eicher dropped a bombshell study purporting to show that regulation was responsible for adding $200,000 to the price of an average Seattle house. It was shocking. It landed on the front page of the Seattle Times.

At the time, I believed the study couldn’t possible be right. And I wrote three blog posts about it: here, here, and here. Among other things, I worried that it’s too easy to conflate “regulation” with “the growth management act.” In fact, that sort of conflation happened in the press coverage, and it’s been happening since in political circles.

But now, riding to the rescue, the Washington chapter of the American Planning Association has a paper out refuting the study — “Observations On the Costs of Land Use Regulations and Growth Management” (pdf). For anyone interested in housing affordability or growth management, the paper is definitely worth a look.

Washington’s Growth Management Act (GMA) was passed in 1990, but it didn’t become operational until the mid-1990s. And it’s fascinating to take a hard look at the empirical evidence about what’s happened—and what hasn’t happened — since the GMAs been with us. Or if you don’t want to read the paper, I’ve got a full summary below the jump…

  • Our work is made possible by the generosity of people like you!

    Thanks to So-Hum Foundation for supporting a sustainable Northwest.

  • 1.) Cost escalation actually slowed after the GMA went into effect. This may sound counterintuitive, but it’s importantly true:

    King County housing costs rose much more rapidly during the mid-1970s and early 1980s, prior to the passage of GMA and many other regulations now in place, than they have in the last two decades.

    The key here is housing “costs, “not “price,” and the data in this table make it clear what’s going on. See the last column in particular:housing costs

    The upshot? Some of the worst excesses in housing unaffordability actually happened before the GMA went into effect. That shouldn’t be surprising since one of the primary goals of the GMA is to ensure a supply of affordable housing.

    2.) Vacant land prices don’t suggest excessive restrictions. If the GMA were restricting land supply then we would expect to see the value of land escalating relative to the value of the house on the land. But that hasn’t happened. In fact, the share of home price that is attributable to the price of land has remained roughly constant, as this table shows:

    land cost

    All told, the study authors believe land prices indicate that “growth management regulations may be responsible for 4%-5% of housing price increases.”

    3.) Developer investment reports don’t indicate excessive regulation. By comparing local pro formas with national data, the APA authors were able to estimate the total price-increase due to all regulations (not just GMA). They concluded:

    The bottom line is that regulations are unlikely to contribute more than 17% of the final price of a typical home, and the impact in many communities may be much less. To use Seattle as a point of comparison, 17% would represent about $68,000 (in current dollars) of a $400,000 home. Presumably, since many regulations were already in place in 1989, the base year for the UW study, the increase in regulatory costs since that time would be even lower… This analysis suggests that the [Eicher] study has grossly overestimated the impact of regulations by an order of magnitude of 300% or more.

    4.) Income stratification contributes to rising prices. Rising income inequality — a national trend—has been pronounced in the Seattle area. To wit:

    …in King County, households earning above 150% of area median income have risen from a quarter to a third of the total.

    This accumulation of high-end wealth would have further contributed to a price run-up.

    5.) The Eicher study ignored the housing bubble of 2002 to 2006, incredible as that sounds. As I pointed out in my earlier posts, the bubble whiplashed prices everywhere, making it exceedingly difficult to discern underlying trends.

    6.) Larger economic trends bear on housing prices. The dot-com cycle pushed local money into the housing market:

    Manyinvestors shifted their assets from the stock market to real estate before and after the dot-com crash of 2001. Research has shown that people who have accumulated wealth through financial investments have reinvested a significant portion in real estate, driving up demand. This has been particularly evident in the Puget Sound region where beneficiaries of stock options from Microsoft and other thriving technology companies transferred wealth to real estate during the late 1980s and 1990s. Relatively poor stock performance in the early 2000s may have increased this effect, leading to the much-storied rise in speculation and purchase of second homes.

    7.) Inflation and local costs matter. Construction costs have soared faster than inflation. And labor costs are higher in Seattle than in many other areas. Plus, there’s been a pronounced trend toward ever-bigger homes with pricier finishes. All these factors would tend to push up Seattle-area house prices.

    8.) Natural constraints matter. The GMA does less to constrain land than pesky things like Puget Sound, Lake Washington, and the Cascade Mountains. This happens in other cities too:

    Urban areas such as the San Francisco peninsula, Manhattan, Honolulu, Boston, Vancouver, BC and Seattle, are severely constrained by water bodies, steep slopes, mountains, and other geological challenges. These factors make land more scarce, development more costly, and hinder the lateral spread of the urban footprint. As relatively scarce amenities, mountain views and waterfront can also increase the relative value of housing. Housing prices are notably high in all of these cities.

    9.) The Eicher study relied on suspect data from the Wharton Residential Land Use Regulatory Index. In fact, much of the Wharton data is based on subjective self-reporting from local planners around the country. There are at least anecdotal examples of incosistency among respondents.

    10.) The study doesn’t distinguish between types of regulation. To my mind, this is one of the most key points — both from an academic and a practical policy perspective. Eicher is broadly right that supply-restricting regulations will, all else being equal, tend to drive up prices. (Heck,
    that’s just Econ 101.) But not all regulations are supply restricting—some are actually force supply increases. From the APA analysis:

    A weakness of the study and the Wharton database it relies upon is a failure to distinguish between state policies and local regulations that accommodate housing growth and promote affordable housing vs. policies and regulations that restrict housing development and/or are exclusionary of more affordable housing types.

    Research shows fairly conclusively that growth controls increase housing costs. Growth management, on the other hand, may have little impact on housing cost, according to the research, and may indeed provide a wider variety of affordable housing choices than would exist under conventional land use regulations.

    11.) Recent local data suggests sufficient supply. The most recent Buildable Lands report, from 2007, shows that permitting and house production has been strong. Residential density is up, indicating a more efficient use of land. And according to the report, “each county and most cities were found to contain more than sufficient land to accommodate new households and new jobs through 2025.”