Walkable Neighborhoods have Higher Housing Vacancy Rates per Census 2010 Data


I pulled Census 2010 data for 20 walkable neighborhoods in the United States and compared the vacancy rates for those neighborhoods to the vacancy rates for their surrounding cities (and in 1 case, the county). One pattern emerged.

Walkable neighborhoods have higher vacancy rates than their surrounding cities.

The Details

After my Twitter conversation about walkability and diversity with @GvilleJen, I wasn’t too surprized by the pattern that “Blacks & Hispanics are Less Likely to Live in Walkable Neighborhoods per Census 2010 data” (This link also has an explanation of how the 20 neighborhoods were selected & what data were used).

But I hypothesized that a socially redeeming characteristic of walkable neighborhoods would be a lower than average vacancy rate. More people buying property in walkable neighborhoods means more real estate taxes–which fund our public services. And haven’t we all been reading articles about how Gen Y loves living in dense urban areas, and how 58 percent of American would prefer to live in a neighborhood where stores are within an easy walk?

Well, the Census 2010 data indicate that walkable neighborhoods have higher vacancy rates than their surrounding cities. Only Dupont Circle in DC and City Center East in Philadelphia have slightly lower vacancy rates than their surrounding cities. Every other walkable neighborhood had a higher vacancy rate than its surrounding city. Topping the chart is Core in San Diego with a vacancy rate of 24.6% while San Diego city has a vacancy rate of 6.4%.

Why is This?

Hypothesis 1

Perhaps homes for sales in walkable neighborhoods are more expensive than homes in the rest of the city. And more expensive homes were harder hit in the real estate downturn. But I haven’t loaded the new American Community Survey home values into Cubit yet, so I can’t test this hypothesis easily.

Hypothesis 2

My selection of 20 walkable neighborhoods is flawed. As Ben Shardlow comments, “The top 20 neighborhoods are going to have extremely high densities of amenities like grocery stores, restaurants, coffeeshops, etc., so it’s not surprising to see the most chichi neighborhoods in the country on this list. Conclusions drawn about this upper echelon can’t safely be extended to the rest of country.”

Why do you think that walkable neighborhoods have higher vacancy rates than their surrounding cities? I’m really interested to see what other folks think about this surprising pattern.

No TweetBacks yet. (Be the first to Tweet this post)

47 Responses to Walkable Neighborhoods have Higher Housing Vacancy Rates per Census 2010 Data

  1. Lynn Stevens June 20, 2011 at 10:19 am #

    For Chicago’s Loop (downtown) neighborhood I hypothesize that it’s because of the overbuilding of condos before the real estate crash which led to quite an over supply and huge price drops during the past few years. I wonder what the numbers would look like if you took a more neighborhood-ish but still walkable neighborhood like my own, Logan Square.

  2. Kristen Carney June 20, 2011 at 11:15 am #

    Interesting! Anyone else know if any of the other of the 20 walkable neighborhoods also overbuilt condos before the real estate downturn?

  3. DG June 20, 2011 at 12:22 pm #

    interesting…from what I’ve noticed, hypothesis 1 sounds about right. Here in Cleveland, Downtown is where many young professionals live…it’s also highly transient. Folks come down here to live for a few years then move to the suburbs when they have kids. Also, here in Cleveland, they’re constantly building but the city isn’t growing fast enough to fill these units. I do almost want to say that they have definitely overbuilt condos-too much of the ‘if we build it, they will come’ attitude, I think.

  4. Urbanchords June 20, 2011 at 12:25 pm #

    I don’t think Westport in Kansas City is a good example. The new entertainment district in downtown killed the life in Westport, causing more vacancy. The Plaza or Brookside would be a better candidate.

  5. Kristen Carney June 20, 2011 at 2:54 pm #

    Very good. So it seems that we have another +1 on overbuilding expensive condos in walkable neighborhoods.

  6. Kristen Carney June 20, 2011 at 2:59 pm #

    Hi Urbancords,
    Walkscore rated Westport Neighborhood as having a 99 walk score (out of 100). You can check out the ratings here–http://www.walkscore.com/2008/walkers-paradises.php

    Interesting to note that the Country Club Plaza has a 95 & the Plaza Westport has a 93. Brookside didn’t make the list, unless I just missed ’em.

  7. The Overhead Wire June 20, 2011 at 4:23 pm #

    Or you have under counting in walkable neighborhoods that are dense and harder to get all all the units for the census counter.


  8. Darin June 21, 2011 at 3:47 pm #

    In some of these cases, it might be that the units are owned as second homes rather than primary residences. I believe downtown San Diego had many units sold for that purpose, and other attractive/expensive places may have similar dynamics. For example, Aspen CO has a vacancy rate of nearly 38%. Can you account for that in the Census data?

  9. Ben June 21, 2011 at 3:54 pm #

    I feel that LoDo in Denver isn’t a good example either. I didn’t live here before the downturn, but I feel like it was also overbuilt (or at the least, overpriced).

    I also don’t feel that it’s the most walkable. It’s good for walking to entertainment/night-life and downtown offices, but I feel that Capitol Hill (the most dense neighborhood, from what I hear) is much more walkable, or perhaps even Uptown. Then again, it seems that Walkscore.com considers Uptown part of Capitol Hill.

    I do think that a walk-score is a powerful tool, but I feel it is still flawed when determining how walkable a neighborhood truly is. For instance, in LoDo (the Walkscore LoDo, not the actual LoDo), there are many bars, restaurants, shops, offices and convenience stores (which often get listed as grocery) within walking distance. However, I feel like my personal walkscore of 88 allows me to walk to more of the every day bars, restaurants, shops, offices, actual grocery, parks, etc. I guess what I’m trying to say is you should start with walk-score as a guide, but to truly pick the most walkable neighborhoods, you need to talk to actual residents.

    If the true Capitol Hill were compared with LoDo, it might make for an interesting comparison. Capitol Hill is more residential with commercial mixed in. LoDo is more commercial with residential mixed in. The true Capitol Hill (a portion of what Walkscore.com boundaries show) seems to have much lower rents. The residents also seem to have lower income on average. Then again, another neighborhood may not change the overall results.

  10. Ch.G, Ch.G June 21, 2011 at 4:48 pm #

    The Loop might be “walkable” but it only became “livable” within the last decade– after Daly poured millions into the development of Millennium Park and the Museum Campus, which encouraged high-rise residential growth where office towers had only existed. Like Lynn said, this led to huge speculation and a glut in the condo market when the economy tanked. (Many of the new buildings have since gone rental, proving popular enough in doing so that Chicago’s downtown is on the cusp of a “boomlet” in rental tower construction.) Neighborhoods that better approximate what I think you mean by “walkable” would be Lincoln Park, Lakeview, Andersonville, Wicker Park, Bucktown.

  11. research247 June 21, 2011 at 5:38 pm #

    I think cities like Portland and SF has to do with how much money the average working stiff can make. Rich executives can choose exclusive places to live. They have the money, time and vehicles to live far away in the best neighborhoods. The housing in the better part of walkable cities is expensive for those of us who like and want to walk to work, etc. I think these results indicate limited accessibility for entry level workers ability to afford high rents in walkable parts of cities. Rich folks may want to live a nicer place and worry less about getting to work exactly at 8 or 9 AM every day.

  12. Joe Cortright June 21, 2011 at 5:39 pm #


    I think you need to dis-aggregate vacancy statistics by rental v. owner-occupied housing.

    Owner occupied housing has far lower vacancy rates than rental–if I recall correctly its around 3% for owner-occupied and 10% for rental, nationally. There’s a lot more turnover in rental housing, and consequently higher vacancy rates.

    This matters because walkable neighborhoods usually have a different mix of rental/owner housing than the metro areas they are located in (typically, higher shares are renters).

    The interesting question would be whether rental vacancy rates in walkable neighborhoods were higher or lower than rental vacancy rates in the rest of the same metro.

    Can you split out owner v. renter in your vacancy rate tabulations?

    Hope this is helpful,


  13. Kevin Dickson June 21, 2011 at 10:53 pm #

    Somebody’s (the Census Bureau’s) definition of vacancy rate is messed up. Apartment investors (like me) live and die by the vacancy rate, but the vacancy rate we use to evaluate a neighborhood certainly doesn’t include “vacant for sale condos”. The vacancy rate by definition is the number of for rent units (condos, houses, and apartments) divided by the total number of for rent units. Capitol Hill in Denver, for example has never been over 13% since the 80’s.

    In a bad recession like we are recovering from, the sales of condos are always worse than the sales of single family homes, because they have always been perceived as worse investments (which they are*). So, when sales are bad, the number of vacant for sale units goes up. Guess what kind of neighborhoods have lots of condos and multifamily apartment buildings? Right, walkable neighborhoods. What kind of neighborhoods have lots of single family homes? Right, less walkable.

    *Condos are generally a worse investment than a single family home because prices are more volatile and drop quicker and farther in a downturn. This is due to the type of residents they have. Condo owners and residents are more transient than SFR homeowners, who typically have kids in school and a lot more money and time invested in their house, so they don’t walk away from their home or move away as quickly.

    Seriously, check your vacancy rate numbers against any of the real estate companies’ vacancy rates, like Marcus and Millichap.

  14. Rod Stevens June 21, 2011 at 11:15 pm #

    Joe Cortright is right. We used to live a block apart, in a turn-of-the-century neighborhood with bungalows and cottages. There was almost no rental housing, and it was blocks and blocks, then, to a grocery store. There are still fewer “amenities” than close-in redeveloped locations, but those are the places that large developers have gone into and put up both new condos and large rental projects. These are also the places that typically have mixed use projects, i.e. mid- and high-rise buildings with supermarkets in their base. They are also likely to be closer to bus lines. The census came along after the music had stopped on the housing bubble, and yet condo and higher density housing is decidedly a late-stage phenomena for most markets. Any increase lifts the single-family houses first, and as these go up in price it increases the residual value enough to make higher-density housing less risky to develop. Think of it this way: the residual land value on a high-density unit can swing 50-100 percent in the course of a cycle, that on a single family house less so. There’s simply more volatility in newly redeveloped places, particularly if those are heavy on density.

  15. Jamie June 22, 2011 at 12:25 am #

    I strongly disagree with the San Diego core data. First of all, where is “core”? I have lived here for years and work for the Community Planning Dept and have never heard of “core” before. If you go on SANDAG’s website, the planning agency for the San Diego region, then you can see that the Downtown/ Centre City vacancy rate is only 7.7% compared to the city vacancy of 6.8%.
    Other walkable areas include:
    Normal heights-7.1%
    Uptown 9.0%
    Greater North Park 7.3%
    Greater Golden Hill 10%
    Kensington-Talmadge 6.7%
    City Heights 7.5%

    Try controlling for factors such as median home price, crime rates, and avg rental rate, like of a 2 bedroom apt. Also, including a definition of “walkable” would be useful.

  16. Jamie June 22, 2011 at 12:30 am #

    I see you got the San Diego core from the walkable community website. Sample size looks small…Another note: are you getting your data from ACS? If so, please include statistical margins of error! <3

  17. S de la Haya June 22, 2011 at 11:10 am #

    Do take note that the Census Bureau’s definition of “vacant” includes not just units that are for rent or for sale but also those units that have already been rented or sold but not yet occupied, units that are for recreational, seasonal or occasional use, and the catchall “other” which could include units vacant because of probate, kept off the market by the owner, or are foreclosed. If you check for this, you may find that “vacancy rates,” as most understand them — also known as the real estate agents’ vacancy rates — may be at or lower friction level. Also, note that cities are made up of neighborhoods and not all city neighborhoods in the same city are ranked walkable; vacancy rates also differ by neighborhoods. Generalizing, as you did, is not advisable.

  18. Michael Lewyn June 22, 2011 at 10:46 pm #

    I am from Atlanta, and using Five Points as a sample “walkable” neighborhood is just stupid. Five Points, Atlanta’s traditional downtown, has been a questionable area for as long as I have been alive. So naturally it is a bit, um, wobbly.

    A more comprehensive examination would include neighborhoods that are nice and stable but still have above-average walkscores, like Virginia-Highlands, Inman Park and Morningside.

  19. kevin June 23, 2011 at 12:39 pm #

    LODO is both over built and over priced. Two downtown buildings came on the market within months of each other last year that increased downtown residences by 45%. While they are not both in LoDo, you can bet that had an impact.

    Furthermore the run up during the boom had both investors and owner/occupiers paying too much. Some have got to be sitting empty.

  20. Michelle M June 23, 2011 at 12:44 pm #

    I have always been skeptical of WalkScore as a valid measurement of walkability. You get density of destinations within walking distance, but nothing about the actual infrastructure, safety, quality and type of housing, site design, etc. As an Atlanta resident, I would never refer to the Five Points area as a representative walkable neighborhood – it is in the CBD with a few condos and lofts of mixed quality. Although the sidewalks and pedestrian amenities are pretty good, there are a lot of vacant storefronts, perception of crime and panhandling, and the whole area is nearly vacant after 5 pm. Not really the place you want to live unless you are a hard core urbanist (I’ve thought about it…). I agree that Inman Park, Virginia-Highland, Poncey-Highland, or Midtown are probably closer to most people’s interpretation of a walkable neighborhood.

  21. Urban_LJ June 23, 2011 at 12:46 pm #

    In Louisville you chose the Central Business District?? I am not sure what the physical boundaries of that are, but that is a fairly poor choice to compare to the rest of metro Louisville-Jefferson County as it is just that a CBD. There are many other places that are walkable that could have been chosen.

    One potential Hypothesis is that most walkable neighborhoods in cities are older historic neighborhoods were not much revitalization has happened and while they may be the “up-and-coming” places to live many are blighted.

    Also looking at one neighborhood and comparing it to an entire city seems a bit flawed from the start, sample is too small. Maybe if you compared a walkable neighborhood (and you should include your definition of walkable because that would also change a lot of the metrics of your study) to a sprawling neighborhood and then contrasted those results with a city to see the difference, maybe you will see there is no significant difference between walkable and non-walkable neighborhoods.

  22. Lou June 23, 2011 at 1:10 pm #

    I’ve lived in Baltimore, Brooklyn, Portland (OR) and now DC, and I think much of the discrepency between your hypothosis and your results has to do with the neighborhoods you picked. The Pearl District in PDX, while made to be walkable, was an industrial area until recently and consists predominantly of new condos. As noted above in other comments, it likely suffers greatly from the housing bubble busting. Federal Hill in Baltimore, on the other hand, with a lower vacancy rate, is one of the oldest neighborhoods in the city, and one that never fell on quite the same hard times as much of the rest of the city. Like the Pearl, it was made to be walkabale, but it was made 300 years ago, and from the get-go designed to be residential. Tribeca in NYC was built to be industrial and only relatively recently converted to mainly residential. DuPont in DC, another place with a lower vacancy rate, has been a desirable location for a long time and was built in the late 1800s. Perhaps your results say more about the design of density or the amount of time it takes a neighborhood to form.

  23. justin s June 23, 2011 at 2:00 pm #

    As someone who has always tried to live in walkable neighborhoods, yet often failed to have the means to, I would suspect that any vacancies have more to do with pricing than interest. Here in DC, there are many great neighborhoods filled with vacant homes miraculously existing within a supposed “housing shortage.” In my 6 years here, I’ve always wanted to live in those homes, however i’m not rich and many of those homes are for sale for over $450k or for rent for more than $1600 for a one bedroom unit. If the pricing lined up with demand, there would be no vacancies, however many owners have found ways to keep their properties vacant, hoping to use/sell them when (if) the market will let them turn a huge profit.

  24. Pat June 23, 2011 at 7:23 pm #

    For San Diego the difference is due to a perfect storm of events:

    “Core” San Diego, I assume, means the 92101 zip code. It was ground zero of the housing bust. At the time of the census, April 2010, hundreds of condos remained unsold, empty in default, or empty in foreclosure, for sale, or used as a second home. The flawed methodology of counting empty condos as “vacancies” skewed the data greatly.

    Just two condo buildings in that area (right across the street from each other, coincidently), Vantage Pointe (679 units) and 1050 B (229 units) account for nearly 10% of the the vacancies recorded. At the time of the census both were in financial limbo and, as I remember, completely empty: 908 empty units. Vantage Pointe has, subsequently, been changed to an apartment complex and has started to fill up. 1050 B has been converted to some of the nicest, high-rise low-income housing on the west coast.

    Other than the anomalous (and methodologically flawed) data for San Diego, I think there is a very simple explanation for why walkable neighborhood have higher vacancy rates than other areas:

    Walkable Neighborhoods command higher rent. (Especially in the “Core” San Diego area where most of the units are new, high-amenity, concrete highrises in comparison to the low-rise, aging, stick-and-stucco rental stock elsewhere around San Diego.)

    Landlords trade higher vacancy for higher rents:

    Cheap places fill up faster, have lower vacancy rates, but also have thinner profit margins, so their landlords can’t bear vacancies and need volume to turn a profit.

    Landlords in Walkable Neighborhoods have the luxury of higher margins and can bear higher vacancy rates waiting for a better paying customer.

    Simple as that.

  25. Kristen Carney June 24, 2011 at 3:13 pm #

    Not yet. Right now, the Census Bureau has released 2010 Redistricting data which doesn’t include recreation home break outs. As soon as they release SF1 Census 2010 data (by the end of this fall), I can pull more data on this point.

  26. Kristen Carney June 24, 2011 at 3:14 pm #

    Great idea! I’ll try to find a tool that let’s me survey folks online and let them tell me what neighborhoods are walkable in their city.

  27. Kristen Carney June 24, 2011 at 3:16 pm #

    Hi Joe,
    I’m using Census 2010 Redistricting data which doesn’t include owner versus renter break outs. As soon as the US Census Bureau releases SF1 Census 2010 data (by the end of this fall), we can get more detailed information about the housing units.

  28. Kristen Carney June 24, 2011 at 3:16 pm #

    That’s a great suggestion. Thanks!

  29. Kristen Carney June 24, 2011 at 3:20 pm #

    You can find a map of Core on Walkscore.com here: http://www.walkscore.com/2008/CA/San_Diego/Core
    You can find definations of walkable neighborhoods on that site as well.

    Because I’m using Census 2010 redistricting data, the only info that’s available right now is units and vacancy rate. Once the Census Bureau releases the SF1 data for 2010, we can look at those variables.

  30. Kristen Carney June 24, 2011 at 3:20 pm #

    It’s Census 2010 Redistricting data, not ACS data. So there’s no margins of error.

  31. Kristen Carney June 24, 2011 at 3:21 pm #

    Good point. We’ll have additional Vacancy data for analysis once the Census Bureau releases SF1 data later on this fall.

  32. Kristen Carney June 24, 2011 at 3:22 pm #

    Walkscore.com ranks Five Points Atlanta as the most walkable neighborhood in the city. Check out their ratings here: http://www.walkscore.com/2008/walkers-paradises.php

  33. Kristen Carney June 24, 2011 at 3:23 pm #

    I’ll have to start a survey to get better data on good examples of walkable neighborhoods by city.

  34. Kristen Carney June 24, 2011 at 3:28 pm #

    Walkscore.com ranks Central Business District as the most walkable neighborhood in Louisville. Check out their ratings here: http://www.walkscore.com/2008/walkers-paradises.php

    Great point about having a small sample size. If you check out the original blog post on this topic here: http://www.cubitplanning.com/blog/2011/05/walkable-communities-diversity/
    you’ll note that the whole conversation about pulling walkability data came from a Twitter conversation. We were looking for some quick data on walkable neighborhoods & race.

    That said, if you know of a recent study that uses something more scientific than my “pull-data-really-fast-for-a-blog-post” method, I’d love to hear about it!

  35. Kristen Carney June 24, 2011 at 3:29 pm #

    Great comment. Thanks!

  36. Kristen Carney June 24, 2011 at 3:31 pm #

    You can find a map of Core on Walkscore.com here: http://www.walkscore.com/2008/CA/San_Diego/Core

    I think your point about walkable neighborhoods have higher prices and rents is right on!

  37. Pat June 25, 2011 at 9:42 am #


    The census definition of “Core” San Diego and the Walkscore definition are completely different.

    Walkscore has the population of Core at 2,929 while the census counts 9,249 units.

    The census is counting ALL of 92101.
    Walkscore is counting just one section called “Core Columbia”.

    BTW: The Walkscore map shows a green area north of the San Diego airport in the Sports Arena / Midway area. While it’s true that there is housing there and a lot of businesses as well; it also one the least pleasant places to walk around in the entire city of San Diego. It’s a soulless warren of strip malls and big box parking lots, with long, pedestrian free blocks, driveway cuts, and fast moving or gridlocked traffic.

    The point?

    Walkability data is suspect until confirmed by shoes on the ground.

  38. justin s June 29, 2011 at 12:45 pm #

    I wouldn’t trust Walkscore to judge Atlanta. It’s brutally inaccurate, as demonstrated here: http://www.humantransit.org/atlanta/

  39. Mark Hickey June 29, 2011 at 2:00 pm #

    One of the earlier postings by Joe Cortright hit the nail on the head, if you are including rentals in your data it is going to cause a higher vacancy rate for walkable neighborhoods. I work for a compay that tracks both residential and commercial property and I can prove numerically that submarkets with a higher proportion of owner-occupied properties have lower vacancy rates. As for the new condo construction issue I don’t think this impacts the data much. In comparison to overall inventory there just aren’t enough to move the needle. However, my suggestion would be to check metros such as Phoenix, Las Vegas, and Miami (I didn’t see these on the chart). They had the most condo construction in the boom and are also the metros that have the most vacant condo stock. One would expect a disproportionately higher vacancy rate, though it might be difficult to seperate this from the rental stock as they also have higher vacancy rates in these metros.

  40. Dallas Condos For Sale Man July 12, 2012 at 12:24 pm #

    Education and, most importantly, experience is always the best way to learn the real estate business no matter if you’re an agent, broker, loan officer, home lender, inspector or contractor. Do not ever stop learning…that’s the key to success and the secret to treating your clients the right way.


  1. Walkable Communities & Diversity: A Look Using Census 2010 Data | Plannovation - June 19, 2011

    […] 6/19/2011 Update: If you like this post, you might check out: Walkable Neighborhoods have Higher Housing Vacancy Rates per Census 2010 Data […]

  2. Friday Links – 1 July 2011 | RealCentralVA.com - July 1, 2011

    […] – Walkable Neighborhoods have Higher Housing Vacancy Rates per Census 2010 Data […]

Powered by WordPress. Designed by WooThemes