Americans pay a premium for housing in walkable neighborhoods — $850 per point on a 100-point Walkability Score scale.

The scholars over at New Geography just won’t give up trying to make the case that most Americans prefer to live in single-family detached houses in the suburbs. Citing data from the 2010 American Community Survey, Wendell Cox wrote that 79.2% of the new households in 51 major metro areas moved into precisely such housing over the past decade. He also cited data that the occupancy rate for detached housing is marginally higher than that for attached, multi-unit housing. He concluded: “The trend of the last decade is evidence of a continued preference of American households for detached housing. The results are remarkable.”

Back in September I made two key points to a similar argument advanced by Cox’s buddy, Joel Kotkin: (1) the concept of housing “preference” is meaningless in the absence of price; and (2) the movement of people into detached dwellings is as much a function of supply (what builders are allowed to build) as of demand (what people actually want at a given price point).

Now comes Emily Washington at Market Urbanism, making the same points and tying them to walkability. All other things being equal, she says, people place less value on neighborhoods with low walkability scores (typically with detached, single family dwellings), and greater value on neighborhoods with high scores (which are more likely to include multi-family dwellings). I can do no better than quote her blog post, ‘The Value of Walkability.”

While people may not cite walkabilty as an important consideration in choosing a house, choosing a home involves weighing many factors, from size, price, distance to work and other amenities, aesthetic, and countless others factors. Consumers rely on tacit knowledge to weigh many of these factors because they can’t consciously enumerate all of them in making a decision of where to live.

For this reason, revealed preference theory is a more reliable tool than survey data for observing how consumers value one attribute of a complex good like housing. Building on a past project, my colleague Eli Dourado and I are studying whether or not consumers do pay a premium for greater neighborhood walkability. Using a fixed-effects model, across all metropolitan and micropolitan statistical areas in the United States, our preliminary results indicate that, on average, Americans are willing to pay a premium of about $850 for a house with one additional point in Walk Score. Because of the many restrictions that limit walkable development, consumers have to pay this premium for the scarce supply of houses in walkable neighborhoods.

This finding also indicates that, in a world with fewer regulations limiting the supply of walkable development, the free market would provide a greater supply of walkable neighborhoods because developers have opportunities to profit from doing so that are currently prevented by regulations. In a freer market, more people would have the opportunity to live in neighborhoods where completing daily errands on foot is feasible. …

“Market suburbanists” often cite survey data finding that most people prefer detached, single family homes to living in multifamily housing. They also often say that revealed preferences back up these surveys because most Americans live in single family homes. Indeed, this is true, even in the largest cities. However, looking at the housing choices that Americans make while ignoring both regulations that limit the potential choice set and without considering the prices consumers pay is misleading, like saying Americans prefer Fords to BMWs because there are more of them on the road.

An understanding of consumers’ complex decision process in selecting a home cannot be accurately gleaned from either survey or Census data; rather, this information should be observed based on the price that emerges between buyers and sellers in the market. While, all else equal, most people might prefer a large detached house with a big yard, in weighing the many factors like proximity to amenities, price, and house size, we find that people are willing to pay a premium for walkability.


Share this article


(comments below)


(comments below)


18 responses to “The Walkability Premium”

  1. would like to see survey’s of people who have chosen to move to the exurbs.

    they don’t generally move here for “walkability” as far as I can tell.

    they prefer subdivisions that are “stubbed” – have cul-de-sacs and do not connect to other subdivisions or other public roads.. i.e. just one entrance.

    they tend to be in favor of sidewalks and trails ONLY for their subdivision – in other words – not public-use amenities.

    It’s important to note that people who chose the exurbs are not the majority of people who live and work in places like NoVa.

    For instance 2 million people live in NoVa whereas only about 100-150K choose to live in the Fredericksburg Area and commute to NoVa jobs.

    but the vast majority who do choose to commute to the Fredericksburg Area – do so – to live in a “stubbed” subdivision that is not connected to anything else… via road, sidewalk or trail.

    People who choose to live in NoVa accept without question that roads, trails and sidewalks DO “connect” to other places and most all are public use.

    this is a fundamental difference between those who choose to live in an areas with public access to road/trail/sidewalk – as well as transit and those who choose to live 50 miles away in what are in essence private enclaves.

    Exurban commuters also utilize public transportation infrastructure at higher rates than those who choose to live in a more “walkable” NoVa.

    A 100 mile commute on public infrastructure is much more consumptive than someone who lives in NoVa 10, 20 miles from their job.

  2. DJRippert Avatar

    “Back in September I made two key points to a similar argument advanced by Cox’s buddy, Joel Kotkin: (1) the concept of housing “preference” is meaningless in the absence of price; and (2) the movement of people into detached dwellings is as much a function of supply (what builders are allowed to build) as of demand (what people actually want at a given price point).”.

    Why would builders supply a product (detached homes) that was in less demand (and, therefore commanded a lower price) than multi-unit walkable residences.

    There is plenty of available land in Richmond, Washington, DC etc to continue building multi-unit dwellings. If more are being built in the ‘burbs it is probably because that what people want and that what builders are producing.

    Your better argument might be the change in demand for urban, walkable, multi-unit dwellings. After decades of decay cities like DC are really humming with new construction.

    1. When you look on a map, it looks like there’s plenty of land. When you look at what the land is zoned for, that’s a different matter. Getting the land re-zoned is an expensive, time-consuming and risky proposition. The market for multi-family units, townhouses, etc. is supply-side constrained.

      1. DJRippert Avatar

        Getting substantial tracts of land rezoned for development within the suburbs is expensive, time-consuming and risky too. How many times do we have to hear about the sanctity of the so-called urban crescent in Prince William County. Good luck building anything in Fauquier County. Anything west of Rt 15 in Loudoun County is essentially “off limits” for development. East of Rt 15 proffers are the name of the game. The proposed Kincora development is expected to commit to pay the $40M it will cost to extend Glouchester Parkway.

        I am not at all sure that the zoning process in DC is any harder than the zoning process in NoVa.

        1. DJRippert Avatar

          Sorry …. meant rural crescent in Prince William County.

          1. Most folks don’t realize it but denser development is somewhat dependent on terrain and the volume flow of the river that receives the effluent.

            If you took a detailed watershed/tributary map of an area and overlaid it with water/sewer-served development – you’d see a pattern.

            Most water/sewer authorities do not pump-stations that pump out of one watershed into another. They’re expensive to maintain and operate and can fail during storms dumping sewage.

            so most water/sewer authorities attempt to keep as much of their system in gravity flow as possible.

            and it’s not just one size pipe. pipes have to be sized big enough to accept all the upstream flows so when you put in the bigger pipes down low in the watershed – they are sized for what engineers THINK will be the maximum development uphill from there.

            we take all of this for granted but it does affect what gets developed in what density and where.

            and developers have to pay pro-rata for water/sewer expansions.
            In other words, someone has to pay real money to contractors to put in water/sewer lines in advance of houses being hooked up and fees collected to pay for it.

            so if you have undeveloped land – and the downstream pipes were not sized or were sized some time ago – you simply cannot hookup new development until the receiving pipes are replaced.

            this can be pretty expensive and require a substantial up-front share from the developer.

  3. Builders construct what they think they can sell or rent at a good profit. If multifamily was number one in demand, we’d see more of it.

    People with kids are generally motivated by the quality of schools. And they often drive their kids to school. Walkability to an elementary school can be a positive factor, but that generally means crossing no busy streets. People without kids are not that concerned with the quality of schools.

    There are stories that builders on the south side of Route 7 in Fairfax County made lucrative campaign contributions to get the County to draw a finger around their subdivision to include in the Langley High School district instead of the South Lakes or Herndon High School districts. Click the following http://www.fcps.edu/images/boundarymaps/langleyhs.pdf and scroll to the upper left. See the gerrymandering.

    1. DJRippert Avatar

      I’d like to see evidence of the contributions. The boundary between Langley and McLean high schools is never much of a debate because they are both good. So, Dolley Madison and Leesburg Pike tend to work fine on that border. South Lakes / Herndon vs Langley is a whole different matter. People in that border zone are forever agitating to fill ever possible slot at Langley with students – including their kids. The core Langley district shrinks and grows with regard to the number of kids in Langley. As son as there is room there are people screaming about why they should be the gerrymandered area.

      If the builders got that “Langley finger” then the buyers may be disappointed. What the demographics giveth, the demographics also taketh away.

      Since Langley is widely seen as the best non-magnet public high school in the state I see no problem with spilling over the natural roadway boundaries if there is space available in Langley. Should it be based on political contributions? Of course not. Was it? I don’t know. The bulge across Rt 7 may prove that Langley is a desirable school but it doesn’t prove hanky panky.

  4. re: “plenty of land”, “right housing, right price”, etc.

    it’s not just raw land – the KIND of subdivisions that people want have to have water/sewer as well as cable/internet, local schools, libraries, fire/EMS.

    If you look at the exurban growth patterns – where there is density – the two most common components are that they are within 5-10 miles of an I-95 interchange – AND they are served by water/sewer. Land that is not served by water and sewer tends to be larger lots 2+ acres and up – and generally not the kind sought after by developers and major homebuilding companies.

    We do build apartments, townhouses, condos and mixed-use down this way and most of it is for local workers…. not commuters to NoVa jobs.

    I erred before by using 150K for commuters from our area. It probably half of that. We have about 300,000 for the two counties and the city of Fburg – about the size of Henrico.

    Significantly, Henrico’s tax rate is LOWER than all 3!

  5. accurate Avatar

    I guess you can make almost whatever you want out of a set of data, of course if you have bias to begin with … well, in the vast majority of time I know what you are going to find in the data. Let’s see, here are some of the things that living in a downtown, multifamily unit (think apartments) bring. Ummm, I get to carry my groceries (or whatever else I’m toting home) up those three flights of stairs – and if I’m relying on public transportation I can only tote what I’m able to carry. In my non-multifamily home, I drive into my garage and carry my groceries from my trunk to my counter top and I don’t get wet if it rains, I don’t HAVE to walk with my belongings from where public transportation left me off to my dwelling. In a multi-unit dwelling, I get to listen to my neighbors, I get to hear family fights, I get to hear when they share the music, I get to hear them come home drunk staggering down the hall. About the worse that happens in my SFH is the very occasional party at a neighbors house that goes on too long; easily managed by either talking to them or calling the cops. Sorry Jim, you’ll never convince me that multi-unit housing is better than SFH.

  6. I think younger folks, especially those used to dorm living find living in apartments not that different and are more used to the issues that Accurate dislikes.

    As people get older and especially so have families – they value more privacy and protection.

    but we know that families DO live in dense urban communities also.

    we also know that young people tend to want to be nearer to where activities external to home life are available… they’re generally not stay-at-home types as much as families are.

    It might be interesting to survey families only, half in dense urban locations living in multi-family dwellings and half in exurban subdivisions and learn how their values differ.

    I don’t know about commuting in Houston but my two times there, the traffic looked bad…. but commuting whether in Houston or from the exurbs to NoVa is a serious price to pay for living in a less-dense subdivision. It takes a hunk out of your day, your car and your finances
    and there are clearly families that live closer-in , in smaller homes than if they commuted.

    1. accurate Avatar

      Larry –
      Houston is basically 45 miles north to south and 45 miles east to west (including minor suburbs) with larger suburbs merely miles from where I’ve sent my boundaries. Depending on the time of day, it can take you from 45 minutes to 3 hours plus to make one of those trips. For a period of time, my work had changed me to where I was driving over 45 miles one way to get to work, I would leave at 5:30, travel mostly at 65 miles an hour and get there at about 6:45 (barring any accidents). I’ve changed jobs and now drive about 20 miles, am carpooling with my wife and get to the job in about 30 minutes. Yup, the commute, the traffic, the wear and tear on the car and the (roughly) hour each way trip was one of the (many) reasons I looked for and found this other job. Still, looking back to where two years ago I was living in an apartment – ugh. Like everything in life, I survived, I made it through, but ideal condtions? I think not. Yup, all the CRAP that I listed above, I had to deal with, and since I’m NOT in my 20’s or even 30’s any more, it was anything BUT fun.

      Could I live in a cramped small apartment in downtown – yup. But would I like it – nope. I went from a 1100 square foot apartment to a 2400 square foot home, wanna guess which one I like better? The cost? The apartment was roughly $800 a month, the house (PI&T) is roughly $1100 a month. I went from ‘shared’ BBQs to my own deck and private BBQs (with friends). I went from looking at a few trees to having several of my own. I went from having a ‘shared garden’ space to having my own, that I merely have to walk out my door to get to. I went from spending money (movies, shopping, cafes, etc) to be entertained on the weekends to weekend projects that never end and the only reason I go to the movies or whatever is because I’m finally tired and am ready to look at/do something else with the rest of my day. I know there are folks who just adore living downtown, Jim seems like one of them, he should buy down there. As for me, just can’t see it.

  7. Andrew Moore Avatar
    Andrew Moore

    Jeff Speck, author of Walkable City, gave a great lecture last night to a 300+ audience at VCU on the value of walkability, emphasizing the advantages to health, the environment and the economy. You can find a concise version as a TED talk here: http://www.ted.com/talks/jeff_speck_the_walkable_city.html

    Compelling stuff.

    As a follow up, Jeff met with a small group this morning to address the “hows” of walkability. The group included a cross section of business, non-profit and government leaders, including representatives from Chesterfield, Henrico and Richmond.

    1. reed fawell III Avatar
      reed fawell III

      glad things are getting traction on the ground.

    2. reed fawell III Avatar
      reed fawell III

      Have now sat through Jeff Speck’s Ted Talk.

      I agree with a fair amount of his specific comments but find the lecture to be all sails and no anchor – therefore, overall, I find the talk quite scary.

      What scares me the most, however, are facial expressions on many in the audience – caught in the thrall of rapture. Those face remind me of how one might imagine the last meeting at the People’s Temple at Jonestown. I regret to say this, but when these sorts of expression link up with awesome power government, federal or otherwise, we are in big trouble.

      Now, as for substance, for example, Portland is not Valhalla despite its description in Mr. Specks Ted Talk. It suffers from a growing problem of unaffordable housing combined with falling municipal revenues. Both these problems derive from Smart Growth too driven by ideology and top down mandates at the expense of smart economics and personal choice.

      I suspect that these Porland problems are now systemic and so may well be very difficult to unwind politically as well as structurally. If so, Portland will suffer major debilitating problems long term. I say this with regret as such problems can be avoided with the right kind of Smart Growth per the Fiscal Fix. See: https://www.baconsrebellion.com/2012/06/the-fiscal-fix.html

      Translation – the views and philosophies of the Sage of Houston deserve much respect. They are Accurate for much of citizenry, and rightfully so.

  8. Groveton Avatar

    If walkability is so desired, why isn’t there more of an effort to make existing places more walkable? Minneapolis has its Habitrail connections. Toronto has the underground tunnels connecting buildings. So, it’s possible. But why isn’t it being done?

    There are various gas pipelines criss-crossing Northern Virginia. As far as I know, none of them have had trails built alongside. There is a Farifax County Connector trail system in place. It is a patchwork of muddy trails with multiple river crossings that flood during the year. Many areas in Fairfax County have no sidewalks or even road shoulders. There is plenty of land beside the road but no interest in building the sidewalks.

    Even densely populated NoVa locations like Reston don’t have pedestrian overpasses.

    Everyone talks walkability but I don’t see much action.

  9. Tough issue. Guess I’m skeptical that consumer preferences can be measured based on “walk-score”, linking this to incremental home value seems like a stretch to me. And does the definition of “home” preclude renters?

  10. reed fawell III Avatar
    reed fawell III

    I agree with Groveton’s comment above. Walkablity is over emphasized.

    Recall the earlier post on this website describing why Ozzie, Harriet, David and Ricky, and their love of the brand new Chevy created the suburban subdivision. I suspect that most future nuclear families with children in US will prefer the suburban house on a lot if its available. But now it appears there will also be many Ozzies and Harriets without each other, much less children. These folks will not have each other so they will want a different lifestyle for many reasons, if only not to be lonely. On a different post I listed many reasons why. Not once did I mention or refer to walkability. Below is the list of reasons:

    “Here’s a partial list if buildings are part of a efficient mixed use commercial center served by mass transit. Like Connecticut Ave in DC, or within Arlington County Virginia’s new downtown, the Rosslyn – Ballson Corridor.

    1/ Compared to types of housing, multi-unit residential buildings generate by far the lowest auto trips per day. The reasons are obvious. Tenants rely on walking or mass transit, for work, shopping, and entertainment. Even the few auto trips taken to or from such buildings are typically far shorter, given the abundance and variety of nearby services. Thus typically high density tenants own fewer cars, and require less parking per unit.

    2/ The interior living space, and exterior walls, per occupant is typically far smaller and more efficient than single family housing. Electric, oil, and water consumption is far less. Less roof, less outside hard-scape plus underground garage = far less storm water runoff per unit. The advent of LEED building codes compound these natural advantage exponentially. As do fire and safely cost given more compact populations at risk. And all these advantages apply to all nearby stores, shops and restaurants.

    3/ Such residential tenants also typically put far less demand on schools. The reasons are obvious. They tend to be young, middle age, or elderly singles (whether alone or in groups), or couples whose children are grown.

    4/ Such tenants also use far more mass transit, and use it far more efficiently. So, to a great degree, they subsidize the mass transit system for outlying low density areas whose far fewer riders heavily burden the system, given the far higher capital / operating costs incurred to serve them.

    5/ Many such tenants lead far more efficient lives that most suburbanites. They have far shorter commutes, far more local conveniences, far fewer kids, so enjoy far more “free time” to work, play, and spend money locally. Plus more are single. So they often tend to go out more, to restaurants, bars, movies, galleries, plays, etc. than suburban cousins. Hence they pump vast sums into down-down local economies that would otherwise be dark and dangerous. These monies radiate throughout the city, whether in tips to waiters or sellers of expensive art, and everything imaginable in between. Much of all this would be lost without these multi-unit buildings.

    6/ Also many tenants in these buildings are not able to afford to rent or buy a home on a lot, or a townhouse in high priced down-towns. Thus these tenants would likely be living outside Wash. Dc. or Arlington Country but for these down-down rental units and condos. So instead to living and paying their taxes “in town,” they’d be paying sales, income, and property taxes to an outlying county while commuting daily into Washington or Arlington, often by car. This would clog the streets and poison the air of DC and Arlington after they’d lost the tenants’ tax revenues to maintain them, and their consumer buying power for the local economy to boot.

    7/ In addition, without such rental and condo units in-town, many such otherwise tenants would refuse to waste their time on long commutes into town. They’d find jobs elsewhere, outside DC and Arlington. DC and Arlington then would lose not only all the benefits that go along with these vibrant and highly productive citizens, they’d likely lose their employers as well. Without nearby affordable employee housing, many companies would move out of Arlington and DC or refuse to move in town in the first place. More and more employers today need to locate where the best workers for them are, if they’re to compete, given ever increasing commutes into single use commercial areas. Indeed bosses demand it for themselves as well.

    8/ Imagine if these multi-family units weren’t in Arlington’s downtown. Or on Connecticut Ave. I guess that roughly 18000 low mid to high rise occupants front Upper Connecticut Avenue alone. Without the buildings, where would their occupants go. Many elderly depend on their convenience for the lives. Many others can’t afford living any where else in Downtown DC. If DC lost these thousand of units, the already ungodly high housing prices in all the rest of NW Washington would sky-rocket yet again. It’d cause a forced migration by the thousands, including many of our most vulnerable citizens. Only the richest among us would be left. Connecticut Avenue, because of its high density apartment stock, is our last barrier to the exclusion of all but the rich in NW DC. Its been doing this invaluable service for generations …

    And this is only half the story. It doesn’t explain how such housing works in tandem with nearby office use to exponentially compound the benefits they spread throughout a city. Which is precisely what’s happening in Arlington County Virginia’s new Rosslyn-Ballston Corridor. More about that later.”
    In writing out that list walkability did not occur to me despite fact I live in an urban neighborhood less than 3 blocks from a subway stop. The Article and all its comments are found at:

Leave a Reply