Housing Supply, Demand and Affordability

It could be worse -- you could live in Hong Kong.
It could be worse — you could live in Hong Kong.

by James A. Bacon

The Hampton Roads and Richmond housing markets are “moderately unaffordable,” according to the 10th Annual Demographia International Housing Affordability Survey: 2014. While not exactly a kudo, that classification puts the two of Virginia’s three largest metros in the top one third of housing affordability for major markets (one million people and over) in the English-speaking world (the United States, United Kingdom, Canada, Australia, Ireland, New Zealand) as well as Japan, Singapore and Hong Kong.

Demographia classifies the Washington metro region as “seriously unaffordable” based on 2013 data, yet the capital region rates better than Denver, Portland, Boston, New York and San Francisco, not to mention such expensive outliers as London, Sydney and Hong Kong. (The study uses a multiple of 3.0 as the threshold of affordability — if median housing prices exceed median household incomes by more than three times, a region is deemed unaffordable.)

The authors contend that housing affordability is largely a function of regions’ “urban containment policies,” or land use controls that restrict the supply of developable land on the grounds of livability, sustainability or smart growth.

Housing affordability has deteriorated sharply in the past decade in Australia, Ireland, New Zealand, the United Kingdom and in some markets of Canada and the United States (evidenced by sharply higher Median Multiples). In every market where there has been a sustained and significant increase in the Median Multiple, more restrictive land use policies have been implemented.

I have extracted Demographia’s data for Virginia’s three major metropolitan regions below and bracketed them with the most affordable and least affordable major markets in North America, Pittsburgh and Vancouver.


Demographia maintains that its methodology actually understates the differences in housing affordability. The average new house size in the United States, which tends to have more affordable housing markets) is significantly larger than in other countries: 250 square meters in the U.S. compared to less than 1oo square meters in Ireland, Singapore, the U.K. and Hong Kong. Not only is the median house size more affordable in terms of median income, it is bigger, which translates into a higher standard of living.

Bacon’s bottom line: The Demographia study drives home a partial truth: Land use controls restrict the supply of developable land and, all other things being equal, a restricted supply of land drives up housing prices and diminishes the standard of living. I don’t see how any serious person can dispute this conclusion.

However, all other things are never equal. The Demographia analysis is shorn of critical context.

Transportation. It is highly deceptive to focus on the cost of housing without also considering the cost of transportation. It is axiomatic that the price of housing is largely influenced by its accessibility to jobs and amenities. Many Americans (and undoubtedly citizens of other countries) are willing to trade higher transportation costs and longer commutes for lower housing costs, and vice versa. Housing cannot be considered in isolation from transportation. If higher housing costs are offset by lower transportation costs made possible by restrictive land use policies, livability is not necessarily sacrificed. Therefore, the Demographia analysis is incomplete. 

Infrastructure. The pattern and density of land use impacts the cost of utilities and public services, including water and sewer; fire, police and rescue; and school busing, just to name the most visible. In the United States, the cost per dwelling unit of providing these urban amenities rises when new development occurs in more scattered, low-density locations. Inefficient settlement patterns lead to higher costs — and higher taxes — than a municipality would incur otherwise. Demographia’s analysis does not consider these costs.

Demand. The price of housing reflects the equilibrium between supply and demand. One reason that housing prices are so high in San Jose (Silicon Valley) and neighboring San Francisco is that there is so much wealth creation occurring there that far more people would like to live there than can afford to. San Francisco has added 1,500 housing units a year over the past two decades, according to Gabriel Metcalf, hardly a no-growth policy. The problem is that the slowly expanding housing stock could not keep up with demand. Contrast that to super-affordable cities like Pittsburgh and Detroit. The population of the Pittsburgh MSA is lower than it was in 1960. Need I even mention the situation in Detroit? Sad to say but many of Demographia’s most “affordable” cities have been economic laggards over several decades. That path to housing affordability is one that few others would want to emulate.

That’s not to say that Demographia is wrong about the link between urban containment, housing affordability and standard of living. But the issues are far more complex than Demographia lets on. If study authors Wendell Cox and Hugh Pavletich want to persuade the unconverted, they need to up their game.

Share this article


(comments below)


(comments below)


8 responses to “Housing Supply, Demand and Affordability”

  1. re: “partial truth: Land use controls restrict the supply of developable land and, all other things being equal, a restricted supply of land drives up housing prices and diminishes the standard of living. I don’t see how any serious person can dispute this conclusion.”

    it’s a canard with no real proof in my view and looking at it in a world context more or less bears this out as countries and cities vary widely in hos they control land use – yet there is no real rhyme or reason as to affordability with regard to density… which ought to be a proxy for “restrictions”.

    you’d have to show that density or FAR has some kind of relationship with affordability… something more than just making the assertion… and using really weak correlations when there are available much stronger and precise metrics available that could make a more convincing case because so far – there seem to be major inconsistencies between the premise and the reality when you look at the cities … look at the top ranked cities.. and confirm that those top ranked have less restricted land-use policies…

  2. NewVirginia Avatar

    Of course Virginia has a significant supply of land and natural resources supporting a relatively low population – at least compared to most of the developed world. So space (which is what the affordability of low-density housing is about) won’t carry as much of a premium as it does elsewhere.

    Indulge me in a little conspiracy theory for a minute…

    The future may not be as bright. We currently have no metro areas outside of NoVA that have actually reached the maximum physical size that is easily supportable by our automobile-centric lifestyle (Hampton Roads is close). People will generally sprawl out in a 30-45ish minute commute radius from an urban core. If highways are abundant, land is inexpensive and available for development, and driving is cheap – and it is very cheap here – that means 30 miles or more in every direction. Diffuse employment centers can push it out even more. Only land development controls, a culture of urban living, or carefully constrained transportation options can stop it. The land is there… so we use it. Housing costs have a lot more to do with the construction market, the age of the housing stock, what areas are popular, etc.

    But when that limit starts to be reached, either artificially – in the case of an urban growth boundary, or naturally – because we’ve built as much sprawl as we can as far away in every direction as people are willing to drive, then space becomes a much scarcer good. Land becomes more valuable and people live more densely. No biggie – it happens in every city.

    Northen Virginia’s and Hampton Roads’ problem (and that of many U.S. suburbs) is not restrictive growth rules on the fringe. Rather, there is one development rule in the United States that is a far more powerful market distortion than almost any other, but has the support of mostly conservative voting blocs. That rule is the density limit. It’s the single hardest thing to get people to change. Approving greater densities where it’s demanded can spark the ire of entire suburban towns, which carefully cultivate single-family homes of moderately wealthy taxpayers with minimum lot sizes, maximum DU’s per acre, and bans on having more than a certain number of unrelated adults living on a single property.

    It’s incredibly hard to change these rules or (even more critically) to re-configure suburban infrastructure to handle higher densities. By this I mean converting winding roads into interconnected webs and alternative transportation options, as well as reconfiguring utilities and public services.

    So the suburb goes as far as it can in a place like NoVA, then prices start to rise. But rather than this resulting in similar prices for gradually higher and higher density housing, the density of the area is fixed by sunk costs and density limits. Unsurprisingly, the price of housing soars and incredibly high-density development happens in the few places it is allowed. It’s sheer scale scares other neighborhoods into locking down their support for density limits and drives more people to just give up and drive further rather than compete on this market.

    If Hampton Roads or Richmond did experience dramatic growth on the scale of NoVA, my guess is they would quickly find their populations and infrastructure unable and unwilling to allow a natural transition to higher-density living as the price of space increased. We could see the same kind of dramatic drop in affordability that we have in NoVA.

    That’s my theory anyway. It’s obviously a simplification.

    1. DJRippert Avatar

      Great comment. Well articulated.

      So, what’s the answer? You almost got there with, “Unsurprisingly, the price of housing soars and incredibly high-density development happens in the few places it is allowed.”.

      The answer is to buy out the homeowners and landowners who cling to the density rule thereby creating more places where high density living is permitted.

      Which, I submit, is exactly what the Rail To Dulles program is doing.

  3. Yes, the issues are a lot more complex than Demographia lets on.

    Nowhere do they admit that the more affluent and desirable land markets where growth is constrained are that way for a reason: new housing growth isn’t politically desirable. In this sense it isn’t “urban containment” by any means, rather preservation of the status quo. Hard to blame urban planners for that when they’re employed by politicians.

    The fringes of Boston, New York, and San Diego are pretty nice places to live, if you have the dough. Doubtful they’s be as nice if they attempted to cram in as many subdivisions as possible, all in the name of housing affordability..

    1. DJRippert Avatar

      Points well taken. The other issue is the unholy alignment of politicians with developers. Developers are major campaign contributors to local politicians. Therefore, the local politicians are beholden to the developers. When a developer wants to build, the politicians work hard to let them build. However, this has a “knock on” effect. Each new dwelling typically costs more in local services (especially schools) than it generates in additional real estate taxes. Citizens complain partly to preserve the status quo and partly for the very valid reason that additional development of dwellings will ultimately force and increase in property taxes.

      Politicians could demand substantial proffers from developers and some jurisdictions do exactly that. However, in many cases, the politicians are reluctant to saddle their developer-benefactors with substantial proffers.

      I have seen informal estimates of how much a new dwelling has to cost before it becomes a net economic positive for the county given existing property tax rates. In Prince William County one such informal analysis came up with $400,000 as the break-even point. Is it any wonder that the residents of Prince William County balk at additional development of new dwellings below that threshold?

      1. Proffers or impact fees are critical. Profits should be made by constructing and then selling or renting dwelling units or other buildings and not from getting rezoning when the rezoning results in overstressed infrastructure and higher taxes to construct the necessary supporting public facilities for such denser land use.

        A developer needs to spend money to clear and grade the land, etc. No one thinks this is unfair and proposes, instead, taxpayers fund these activities. Yet, many feel taxpayers should pay for the roads, parks, schools, libraries, etc. that are also needed to support the rezoned uses. That is absurd. The costs incurred to build or enhance infrastructure necessary for the rezoning needs to be paid by the rezoned development in the forms of proffers, impact fees, special taxes and the like. We may find out building is a profitable business, but development is not so much. Also, we won’t see uneconomic development proposals.

        Now I can hear the smart growth folks arguing that this will push out dense development in close in areas in favor of sprawl. But if the hinterlands development pays for the additional public facilities needs it causes, so what?

  4. I concur. some well articulated thinking…

    I need to see an urban area with density by census block or equivalent…
    because I will accept that restrictive land use policies are in play.

    second – let’s forget single family dwellings and look at the price per square foot for residential apartments, condos and perhaps townshouses – on a region-wide basis that crosses multiple jurisdictions like the DC Area AND INCLUDE that data out from the core by 40-50 miles.

    let’s look at the data – first… then – at the least, acknowledge supply and demand… before we get to WHY the supply/demand varies.

    My view is that you’re probably going to see supply/demand variances within the same jurisdiction – that has the same policies across that jurisdiction.

    Further – I think when you jump boundaries between adjacent jurisdictions – with different land-use policies – that you’re not go to see dramatic differences in densities and per square foot costs between the adjacent jurisdictions.

    The problem with the “govt is doing bad stuff” argument is that it distracts us from some observable realities .. that we should seek out .. and .. we should look for consistencies and inconsistencies with regard to the land-use restriction claims.

    this kind of thing – happens – around the world – and it’s just hard to believe that – around the world – no matter the city – that the explanation for higher residential prices – is – govt restrictions.

    we spend so much time these days on the real and imagined sins of government that I fear we just ignore other realities right in front of us.

    You want super cheap residential housing in an urban area??? go to Detroit.

    no restrictive land-use policies have changed… but big changes in “affordability”..

  5. […] Housing Supply, Demand and Affordability The authors contend that housing affordability is largely a function of regions' “urban containment policies,” or land use controls that restrict the supply of developable land on the grounds of livability, sustainability or smart growth. Housing … Read more on Bacon’s Rebellion […]

Leave a Reply