According to the 10th Annual Demographia Housing Affordability Survey of 360 housing markets worldwide, “San Francisco” is the third least affordable major metropolitan area, behind only Vancouver and Hong Kong, the least affordable market by a wide margin.

The Demographia rankings are based on “Median Multiples,” median house prices divided by gross annual median household incomes. The major markets are those with populations over a million. And for the purposes of Demographia’s survey, San Francisco includes Oakland as part of the greater metropolitan area.

Demographia ranked Pittsburgh as the most affordable major housing market in the word, followed by Detroit. In fact, the ten most affordable major markets were all in the United States.

Including areas with under a million in population, Honolulu, which should soon break through the one million mark, ranked as less affordable than San Francisco, as did Santa Barbara, and was the absolute least affordable metropolitan area in the United States, the third worldwide.

Demographia International Housing Affordability Survey 2014 []

57 thoughts on “San Francisco Named Third Least Affordable Major Housing Market, Worldwide”
  1. Yes, SF and many other cities are painfully expensive, but consider the source of this data. Wendell Cox is a well known anti-density / pro-car advocate. A city like San Francisco is his vision of hell.

  2. What do all the most unaffordable cities have in common? They’re all places where corruption money from China is laundered into real estate.

  3. nothing really surprising.
    Each of the top 10 has its own reason for being there. Yes foreign investment has a big influence in London, Vancouver. But Australian cities have the privilege to be in a country with a small population/high resources where only a small fraction is livable.
    But for SF, it’s as much a sign of a tech wealth boom as a sign of huge market imbalances.
    One thing to notice in SF is that the majority is renting, and many renters are subsidized, which allows then to stay here without paying the high market cost of living that newcomers have to pay.
    Therefore the median income is probably a tenant who will probably never think about buying. And the median buyer probably makes 2 to 3 times the median income.
    It’s a crazy situation. No wonder City Hall and the minions are all running around looking for solutions to an issue that has one main solution: abolishing rent control.

  4. NY is not limited to Manhattan. The common misconception is to think NY has to be more expensive than SF when actually they mean only Manhattan.

  5. I would assume that since NYC includes the other boroughs, it doesn’t rank as high as SF. I’d be surprised if Manhattan is less affordable than SF.

  6. NYC does not equal Manhattan (and even the western end of Brooklyn). There is more than enough affordable housing in Queens, Bronx, eastern Brooklyn to make the city as a whole affordable. If we just used Manhattan, they would be right at or near the top with these other cities.
    Now, on to SF.
    It is time for politicians to stop pandering to people who have no idea what they are talking about. The rest of the world sees right through our issues. We need to start building, and we need to do it now, not 7-8 years from now after “studies” and a painfully convoluted permitting process, all to just add a handful of four story boxes on the little remaining land we have.
    We can start by upzoning Western SOMA. That plan is a disgrace.

  7. With respect to “New York” not making the list, keep in mind that the New York Metropolitan Area not only includes New York City, but Long Island, parts of the Hudson Valley, the six largest cities in New Jersey (Newark, Jersey City, Paterson, Elizabeth, Trenton, and Edison), six of the seven largest cities in Connecticut (Bridgeport, New Haven, Stamford, Waterbury, Norwalk, and Danbury), and five counties in Northeastern Pennsylvania.

  8. I would still be curious to see where Manhattan alone would rank. Keep in mind that this is a price/income ratio — not just a price tracker.
    I wonder how the median income there compares to in SF. I have it in my mind that it would be higher, but that’s a pure hunch based on finance industry presence (which obviously employs only a fraction of Manhattan residents).
    Honolulu is not surprising, given that so many of the residents presumably have very low-paying tourism jobs.

  9. That’s what makes these sorts of lists meaningless.
    Comparing NY & SF is just irrelevant.
    NY is a city of 8 million people SF, covering hundreds of square miles, SF is 800 thousand people covering 49 square miles. The only sane comparison would be to break out each borough.

  10. OK, so fair points about NYC – I would guess that whoever came up with this regards “Paris” and “Tokyo” as also including a bunch of cheap areas well outside the city proper. Why the hell not include places like Antioch and Pittsburg under “San Francisco” then?

  11. JWS is on point about the western SOMA plan. It’s a travesty. Mayor Lee admits there’s a housing problem, especially for lower and middle income residents, but what is he doing about creating substantial infill in this particular area…an area which is walking distance to major transit, not to mention jobs and other amenities?
    Not shocking that many of the Rust Belt towns are listed here. Having grown up in Buffalo during the 70s and 80s I watched the city spiral down. Not much left except some housing stock and an extremely gray population.

  12. This index measures affordability to local residents, not the real price of housing. As such, it high values indicate (1) non-resident investment in housing (vacation properties, etc.), (2) political interventions in the housing market (e.g., rent control), and/or (3) a bubble in the local housing market.
    It’s not surprising that SF proper would rank high on this list (b/c of rent control), but it’s pretty interesting that the whole metro area ranks highly. Is the Bay Area witnessing an influx of non-resident money . . . or a localized housing bubble?

  13. First, the navel-gazing argument on here’s a bit silly; hard facts are readily available on this thing called the “internet”. From a quick search on the web, median Manhattan income is $66,700, and median home price between $840K and $880K. That’d place Manhattan, by itself, at a multiple of 12.6 to 13.2, above all locations except Hong Kong.
    For NYC as a whole, I found (somewhat more questionable) figures of median income of $57.6K, and a median home price of $515K, for an index multiple of 8.9 – putting it just below S.F. and Sydney.
    And some of the arguments make no sense when looking at Sydney (which geographically is actually quite large and diverse), or San Jose. San Jose?! How can San Jose, with all its tract homes spilling southward, be #4? The arguments about density and geographic constraints can’t support the rankings of Sydney, or San Jose, or even Los Angeles (which after all includes the Valley and a lot of fairly low-value housing stretching south from downtown to Long Beach).

  14. Look stop trying to use logic on this. San Francisco is just “special.” “Unique”! \
    #1 in everything!
    #1 in unaffordability.
    #1 in subsidized housing affordability.
    #1 in rent control.
    #1 in rent increases.
    #1 in fighting evictions.
    #1 in tenant buy-outs.
    #1 in employment.
    #1 in poor and unemployed people.
    #1 in everything!
    Don’t use logic. It does not apply here. San Francisco is a blissfully logic-free zone and always will be. It’s what makes us…

  15. outtahere has the right analysis. This report is intended to influence land use policy to encourage sprawl. Check out the chart on page 12 (22) which points indicates that land use regulation is correlated with the high cost of housing.
    This paper doesn’t factor in any of the externalities of sprawl. It doesn’t even account for the increased costs (in either time and $) of long commutes and “drive everywhere” land use patterns.
    SF and the bay area in general has plenty of room to grow without sprawl. Its just a matter of accepting increased density and adapting more efficient ways to move around.

  16. For this study SF also included Oakland and it still made the list…interesting. Also please note that Vancouver (where they are pro-high rise) is less affordable. Also interesting…

  17. for those asking about Paris, the Demographia survey looked at 360 cities in nine countries: Australia, Canada, Hong Kong, Ireland, Japan, New Zealand, Singapore, the U.K. and the U.S. I should have added this when I posted the link earlier today.
    Another interesting link is Expatistan cost of living index.
    SF’s high cost of housing is the main reason SF is very high on that list.

  18. What Sierrajeff said.
    From Slate’s Moneybox about two weeks ago, Brooklyn’s Median Household Income Is Less Than $45,000:

    Measured by median income, Manhattan and (especially) Brooklyn are much poorer than you think. Manhattan’s median annual household income is $66,739, while Brooklyn’s is a mere $44,850. Its less fashionable neighbor, Queens, outearns Brooklyn at $54,373 per year. New York City’s most suburban borough, Staten Island, is also its richest, with a median household income of $70,295, while the suburban counties surrounding New York are all richer than any of the boroughs.

    Meanwhile, the cost of living is astronomical in Manhattan, where the median monthly rent is $3,100; it’s $2,800 in the gentrifying northwestern quadrant of Brooklyn.

    I leave the calculation of what household income is required to cover the median monthly rent of $3k to the reader.

  19. I am sort of mystified as to who these people (or even families!?) are who live on $45,000 a year in a major city like New York or San Francisco. I don’t know any of them, none of them are my employees… and, so, where do they live? How do they eat? Pay rent? I just can’t conceptualize how you’d even survive on so little money.

  20. Home prices in Vancouver are just as high as in SF, despite Vancouver’s lower median income. However, one can rent an apartment in Vancouver for about half what it might rent for in SF.

  21. I’m also surprised not to see NYC. Yes, the other boroughs are included which brings down the average cost. But the SF-Oakland-Fremont MSA includes such high rent places as Oakland, Hayward, San Leandro, Castro Valley, San Pablo, Richmond, etc., etc. I can’t believe that housing in one of the outer boroughs of NYC can be any cheaper on a relative basis. I’m always suspicious of the methodologies of these best-of, worst-of surveys.

  22. “who these people (or even families!?) are who live on $45,000 a year in a major city like New York or San Francisco.”
    Two words Geeemie: rent control.
    And they still have bread to buy all those over priced 49er T-shirts. Before the game ($20+ each.) After game loss, tshirts hawked at only $5, poor.ass.millionaire pricing (only I’m not a sports fan; matter of fact I dislike all organized sports.)

  23. Stop. This chart is not based on rent. This is based on median home PRICE!
    Just because rent’s in the news here doesn’t mean the rest of the world’s obsessed about it too.

  24. As p.a.m said, yes rent control allows the lower tranches to afford to stay. Some situations I have seen in my RE hunting days were 2 or 3 grown-up generations living in one rent control mega apartment, like the one a few mentioned on SVN back in June. sub-800 in rent, at least 3 working adults plus a retiree. Paying almost nothing in rent gives you plenty of flexibility. Since it’s multi-generation with a retiree, daycare is not an issue. Public schools for kids of age, cheap enough food thanks to home cooking and proximity of local affordable produce and groceries.
    There’s nothing wrong with allowing this to happen in a rich city like SF. I just think it should NOT be financed through extortion aka rent control. The loser is a sorry-ass landlord, probably too elderly to know what to do. It’s the government job to fund its own social policy. Not private parties.

  25. Sierrajeff,
    I beg to disagree, these charts seem to be about sale prices but they are also based on rent in an indirect but essential way.
    Question: how come this median dweller can be able to live in SF on 1/2 of what is required to qualify for a median home mortgage?
    Answer: this median dweller is statistically very likely to be a tenant who pays much less than the current market rates.
    Therefore this “affordability” rate is not a true reflection of the pure resale market, but a window on the huge imbalances in the City between two HAVES: the rent controlled Haves who are given a quasi-lifetime annuity by their landlord, and the cash flow rich Haves who can afford the current resale market rates.
    Two segments are under-represented here: the old time owners who wouldn’t be able to afford their current dwellings, and the tenants paying market-or-close-to-market-rate.

  26. ^^^
    since rent control applies to about 75% of rental units in SF, it seems to be working great for about 75% of renters in SF, and not so well for the rest.

  27. ^No, sorry, not working for me. I rented a rent controlled unit last April, and I’m paying literally 15 times the rent of my next door neighbor, and still FAR above what I’d be paying in the absence of rent control.
    Now, if I decide to stay here for 10 years it might be in my interest at that point.

  28. Yeah, that is just a silly argument. It is true that a small number, maybe 20%, of leaseholders (> 10 years) are paying below the rents that would be prevailing in a but-for world without rent control. But the majority of renters are paying higher rents than they otherwise would, due to the restricted supply. Again, landlords who think it through should LOVE rent control because they rake in the bucks because of it. And it gives them a nice excuse not to keep up their properties. Tenants who think it through (except for that 20%) should HATE rent control because it leaves them with sub-par units and artificially-inflated rents.

  29. 20%?
    Way more than 1/2 of all rental buildings are pre-1979. They are therefore all rent-controlled.
    Due to lack of turn-over, probably 80%+ of all tenants have been there for more than 3 years, and someone who moved in more than 3 years ago is probably paying 20% under current market rate.
    Therefore I think that more than 50% of all tenants are actually benefiting from rent control.

  30. In SF, roughly half of all residences are rent controlled and roughly two-thirds of all residences are rentals. One half is three-fourths of two-thirds, roughly.
    According to the US Census Bureau 2012 ACS, in SF about 35% had lived in their current residence more than 10 years. In the 2000 Census for SF 33% had lived in the same place for more than 10 years and roughly half had lived in same place for more than 6 years.

  31. Yes, the ACS data is the source for my post. I was about to go reference it directly when I saw your comment, Jake 🙂

  32. Don’t confuse “households” and “residents.” Not the same thing. Renter households are smaller, and rent-controlled renter households are smaller yet (about half are a single person).
    Rents in the hypothetical non-rent-control world would be a lot lower than actual rents from three years ago. I’m sticking with my estimate that it’s only those living in their rent-controlled place for 10+ years that are now coming out ahead from rent-control (remember that for many of those years, they were paying artificially high rents caused by rent control, so you have to factor that in). So fine, it’s about 30% of renters rather than 20%. Point is the same. Rent control favors a minority of renters, and disfavors a majority. Landlords come out ahead — not only those who own post-1978 buildings but even those who own rent-controlled buildings. Rent-control is the landlord’s best friend. Don’t shoot the golden goose all you landlords! Fortunately for you, the tenant lobby is too stupid to realize that there efforts put loads of money into your pocket.

  33. aotheranon:
    According to the 2010 US Census, in SF
    Average household size for owner-occupied housing was 2.65 and for renter-occupied 2.04
    64.2% of housing units were renter-occupied
    58.0% of the population lived in renter-occupied

  34. “And for the purposes of Demographia’s survey, San Francisco includes Oakland as part of the greater metropolitan area.”
    This data is based on MSA data, not just SF proper so I think rent control is an even smaller factor than you two are calculating.
    Plus, do any of the other MSA’s on the list have significant amounts of rent control?
    What I see in common with the least affordable markets is that they are wealthy areas. High income people are more likely to have significant savings which can be tapped for housing purchases. Plus, there are some baseline minimum expenses that don’t scale with income so higher income people can afford to spend a larger multiple of income on housing than low income people. (i.e. buying a $100k house on $30k income isn’t the same as buying a $1M house on $300k income)
    It doesn’t really surprise me to see that wealthier areas have higher price to income ratios than low income areas.

  35. Thanks to Jake for posting some FACTS. Also , I am really getting tired of the false idea that building more housing will make it less expensive. Cities that build a lot more housing than San Francisco have seen similar price increases. I am at our other home in Kapalua right now, and this community has doubled its housing supply in the last 20 years but prices have gone up 3 to 4 times what they were 20 years ago.

  36. aotheranon:
    Massachusetts eliminated rent control 20 years ago and studies show that in Boston and Cambridge rents for non-rent controlled units went up not down, and that they went up by more than they would have if rent control had remained.
    A fair one-page explanation for why is at namelink.
    Basically, there was a surge in investment in rental property. A lot of poorly maintained buildings got better, making neighborhoods more valuable, which allowed all owners to capture more rent. Also, there was also a spike in evictions.
    Imagine how a spike in evictions and property refurb in the Tenderloin would increase rental value in Lower Nob Hill, etc.

  37. “don’t know who/what you are quoting but all the US Census data I provided in this thread was for San Francisco County, California.”
    Yes, but the Demographia study that is being discussed in this thread is based on MSA/Metro Area data. So it’s hard to see how you can draw any conclusions about rent control vs affordability with that data.

  38. anon2:
    sorry, I missed your context.
    Yes, I agree with your earlier post.
    For rent control vs affordability it would probably be more useful to look within the MSA at the different outcomes among neighboring jurisdictions.

  39. Plus, do any of the other MSA’s on the list have significant amounts of rent control?

    Number 9 from the top of the horizontal bar chart in the post above, The City of Los Angeles applies a Rent Stabilization Ordinance to all units pre 1978, with a lot of exemptions.
    The caveat about MSA in the survey results above vs. City proper applies.

  40. That just means that the community likely needed to quadruple or quintuple its housing supply. Austin has tripled its housing supply in the last 20 years and seen extremely small price increases.

  41. I couldn’t easily find same year figures or status from the year in question (1994), but consider this:
    • population density of Austin, TX: 2,828.44/mi² (2012 figure)
    • population density of Cambridge, MA: 16,422.08/mi² (2010 figure)
    Assuming anon knows what he’s talking about with Austin increasing it’s housing supply over the last two decades, it’s still less dense now than Cambridge was two years ago, and by a still-substantial amount, no?
    Perhaps when you’re talking about an already fairly densely-populated city, it’s not so easy for “the community” to “quadruple or quintuple its housing supply”.

  42. Indeed, Brahma. We have already exported our deceased to make room for the living. Density can be increased, but it will inevitably hit a wall at some point. We could become a 1M people city, but probably never 2M. That would mean tearing down many of those nice vics that sell today well north of $1M. You can tear down unpopular blight but not nicer housing.

  43. @Brahma, then we should increase zoning and see what happens. Of course it’s easier to increase supply when building on greenfields, but the claim I responded to was that increasing housing supply in SF would lead to increased prices.
    You can’t claim that the “market” can’t meet the supply desired, and then restrict the market from meeting the supply desired via zoning restrictions. That is a pretty easy to ensure that it happens, but it’s not particularly useful towards the goal of “proving” the “failures” of the market.

  44. I have lived in SF for 40 years. Rent Control is NOT the problem. The number of rent control units is minimal, this is a specious argument put out by property owners. It gained traction during the first dot-com boom when they wanted to evict long term residents and cash in by converting the units to market rate. Thousands of ‘lower’ class folks were evicted, now it’s also middle class, families, professionals etc. Apart from the limited land mass the problem is, and has always been, the cities failure to build truly affordable housing. The speculators and developers control whoever the current city officials may be, they decide what will be built.
    The money is in high rise commercial space and housing. Every time we go through one of these building booms the same thing happens, the bubble bursts and is followed by years of vacancy. The speculators have pocketed their money and left town, the complicit officials have moved on up the corrupt ladder of cooption that passes for politics here.

  45. SZ– that chart isn’t very accurate. Condo prices are higher in Vancouver than Toronto. And only one city per nation is listed.

Leave a Reply

Your email address will not be published. Required fields are marked *