You get the city you pay for. For anyone keen on a “quick and dirty” look at selected expensive/affordable locations across the English-speaking world, Demographia’s third annual International Housing Affordability Survey published last month is a treat.

Demographia’s fascinating dataset caught the attention of our major newspapers, with headlines highlighting the story that Sydney (#7), Perth (#11), Hobart (#20) and Melbourne (#23) are among the 26 most expensive cities in the English-speaking world.

But to get an understanding of why that’s the case, it is worth taking a look at the characteristics of the particularly expensive and affordable cities.

I’ve added some simple city markers in the categories below: coastal centres are italicised, “global cities” are bolded, cities located in North America’s infamous “Rust Belt” are labelled “RB”, and US locations whose “city” (not necessarily “metropolitan”) populations peaked several or many decades ago are earmarked with *.

What’s striking is that the bulk of Demographia’s top 26 expensive cities are coastal, mostly either “global” or relatively large, “s-xy” and coastal. Australia’s state capitals are a natural fit on this list.

Demographia’s Top 26 expensive markets: Sexy cities (US unless otherwise noted):

1. Los Angeles-Orange County, 2. San Diego , 3. Honolulu , 4. San Francisco , 5. Ventura County (LA region) , 6. Stockton (San Francisco region) , 7. Sydney (Oz) , 8. San Jose (“Capital of Silicon Valley”) , 9. London (UK) , 10. Bournemouth-Dorset (UK) , 11. Perth (Oz) , 12. Riverside-San Bernardino (suburbs of LA) , 13. Vancouver (Canada) , =14. Miami-West Palm Beach, and Modesto (San Francisco region) , 16. Cardiff (UK) , 17. Bristol (UK) , =18. Fresno (San Francisco region) and New York, etc. , 20. Hobart (Oz) , =21. Auckland (NZ), and London Exurbs (UK) , =23. Melbourne (Oz), Sacramento (San Francisco region), Sarasota, and Victoria (Canada)

Meanwhile, Demographia’s top 42 “affordable” cities all are inland, nearly half of them in America’s Rust Belt, and to some extent “shadows of their former selves”.

Demographia’s Top 42 affordable markets: Dull Cities (US unless otherwise noted):

=1.Youngstown* RB (a shout out to The Boss) and Fort Wayne RB; and Regina (Canada) , =4. Buffalo* RB , Dayton* RB, Indianapolis RB, and Rochester* RB , 8. Akron* RB; =9. Grand Rapids RB, Omaha, Toledo* RB, Wichita, and Quebec and Winnipeg (both Canada),  =15. Des Moines*, Huntsville, Northwest Indiana* RB, Pittsburgh* RB, Syracuse* RB, and Saskatoon (Canada); =21. Augusta, Cincinnati* RB, Dallas-Fort Worth, Detroit* RB, Harrisburg* RB, and Lansing* RB; =27.Cleveland* RB, Columbia, Kansas City*, and St. Louis* RB;  =31. Atlanta*, Columbus RB, Houston, Louisville*, Nashville, Oklahoma City, Scranton-Wilkes Barre* RB, and Ottawa (Canada);  =39. Little Rock, Tulsa, and London and Oshawa (both Canada).

By my count, the two categories above reveal the following distribution of characteristics:

  • “Global” cities: 10 of top 26 expensive markets, only one of top 26 affordable;
  • Coastal: 22 of top 26 expensive markets, whereas all affordable markets inland;
  • Rust Belt cities: No expensive markets, but about half top 42 affordable markets;
  • Long-term “city” population decline: No US expensive, half US affordable markets.

Make your own judgements, but to me it seems obvious that the “affordability gap” between Demographia’s expensive and “affordable” markets mainly reflects the fact that desirable (HIGH-demand) locations are expensive, while less-desirable (LOW-demand) locations are more affordable. No surprises there: rump steak and shiny BMWs always will cost more than sausage meat and beaten-up Fords.

Unfortunately, the cross-country evidence above suggests that Australia’s state capitals — all coastal and the dominant city in their region — are predisposed to be relatively expensive.

Extraordinarily, Demographia so far has chosen not to notice that its expensive cities tend to be “global”, coastal, “s-xy”, etc, or that half of its “affordable” cities have been on the wrong side of two of the biggest structural shifts in US history – the decline of Rust Belt economies and the US population shift from the Northeast and Midwest to the South and West.

In particular, St Louis, Pittsburgh, Buffalo, Detroit, Cleveland, Cincinnati, Rochester, Syracuse, Dayton, Akron, Toledo and Louisville are identified as “large cities that declined continuously”, in a thorough analysis of US city/metro/regional/national population changes from 1950 to 2000.

So, about half of Demographia’s top-42 “affordable” markets – including nine of its US top-11 – are in the Rust Belt. At the same time, half of Demographia’s top-35 “affordable” US markets – almost all of them in the Rust Belt – have “city” (not necessarily “metropolitan”) populations that peaked several or many decades ago. Stop Press: homes are cheap in places with centre-cities going backwards!

Another point that needs to be made is that housing demand in any market is a function of personal wealth (great or modest), as much as it is of incomes (high or low) and population (rising or falling). Demographia’s affordability rankings are based on the ratio of median home prices to median household incomes. That’s fine, as long as it is understood that cities particularly popular with the affluent – individuals with wealth (a stock) on top of an income (a flow) – naturally will tend to have higher home-price/income ratios, reflecting that extra spending power at work.

In summary, there are pretty strong indications that relative demand is critical in determining relative home prices. Faced with the choice between a flash house in one of Australia’s coastal capitals and a flash house in a less-interesting place hundreds of kilometres inland, most Australians would be prepared to pay more for the former than the latter. Thinking back, Prime Minister Howard’s decision a decade ago to break with tradition by living at Kirribilli House in Sydney rather than at The Lodge in Canberra met with little controversy because most Australians read the move as perfectly understandable.

Just as home prices vary according to location across suburbs within cities – say Bellevue Hill versus Penrith in Sydney – home prices vary according to location across cities within countries. Just as homes in Sydney are much more expensive than those in Dubbo and, in turn, Kickatinalong, homes in LA on average are much more expensive than those in Youngstown, Ohio. As observed earlier, you get the location you pay for.

From a policy perspective, if Demographia and the IPA understood the extent to which the “affordability gap” between expensive and affordable cities simply reflects relative demand – and to a lesser extent natural supply constraints (coastal versus inland) – they would be less inclined to blame “urban planning” and “artificial constraints” on land supply for Australia’s capital-city affordability problems.

Similarly, if Demographia and the IPA accepted that the structural drop in inflation and halving of mortgage interest rates in the wake of the early-1990s recession was, in fact, behind much of the massive increase in home/land prices right across Australia over the past decade, they would be less inclined to (inadvertently) exaggerate the potential for open-slather homebuilding on (say) Sydney’s periphery to reduce city-average home prices, and thus fix affordability.

It would be nice if fixing housing affordability were that simple, but it’s not.

This is an edited extract from a discussion paper “Getting at the facts on cross-country housing affordability” written by Macquarie Bank Interest-Rate Strategist Rory Robertson.



Institute of Public Affairs’ Alan Moran writes: Macquarie Bank’s Rory Robertson never lets his ignorance of facts and analytical inadequacies get in the way of a good spray against those who have been more systematic in addressing facts. He accuses the Institute of Public Affairs and Demographia of being unable to understand his own idiosyncratic and crazy logic about house price movements. Among the reasons that have been offered for the increase in prices is the effect of the market overheating through increased demand. Mr Robertson has maintained that the increased prices are caused by increased demand especially in urban areas he refers to as “s-xy”. He defines these as those cities he likes. Alongside those cities Robertson refers to as “s-xy” are those he considers to be too “dull” to attract the monied people. (He dismisses Houston’s attractiveness because he claims an ex-girlfriend lived there and found it boring!). Like Beauty, s-xy is in the eye of the beholder. What really distinguishes the cities that have affordable prices from other cities, sadly including all of Australian capitals, is the supply of land on which the authorities allow new houses to be built. Land that is readily released on demand enables cities like Houston, Dallas and Atlanta to be kept affordable (house/land packages roughly half the price of Melbourne) despite growing faster than any Australian cities. Australia is better placed than any country in the world in terms of land availability. Yet, planning restraints on land for housing purposes has meant that in spite of high house prices current levels new housing starts in NSW at under 30,000 per annum compare with 50,000 in the early 1990s. Victoria was suffering from the Kirner ALP blues in that period but its present level at 30,000 is fewer than the 36,000 of 2002. Queensland in the early 90sturn and again in 2002 was starting 40-50,000 new homes a year, well above last year’s level and South Australia’s 11,000 compares with 14,000 in the early 1980s. Even Perth is commencing fewer dwellings than in the late 1980s. It would be sad for all those who are priced out of housing by government land restrictions to have their misfortune maintained and compounded by policy makers being swayed by callow analyses of the sort promoted by Macquarie Bank.

Rory Robertson, an economist at Macquarie Bank, writes: The IPA’s Alan Moran is much stronger on abuse than on facts. Yet again he failed to acknowledge that about half of Demographia’s top-42 “affordable” cities are in America’s “Rust Belt”. And that no major coastal city in the English-speaking world is “affordable”, according to Demographia’s six-country dataset, whereas important coastal cities dominate its top-26 list of expensive places. Of course, it’s not news that homes are affordable in America’s rust belt, and expensive in high-demand coastal cities. Yet Dr Moran claims – apparently seriously – that the affordability gap here is solely a function of excessive urban planning in the expensive cities: “…there is no indication that demand factors separate the highly unaffordable cities from the highly affordable cities”. Dr Moran also refuses to accept that structurally lower interest rates since the early-1990s recession have been the principal driver of higher home prices in Australia. That’s despite (former) RBA Governor Macfarlane last August explaining again that it’s as plain as the noses on our faces: “…why have the prices of the eight million houses in Australia basically doubled in the last decade? The answer… I think, is almost entirely on the demand side. Basically, because we returned to low inflation, interest rates were halved. People could now borrow, if they wished, twice as much. …The incentives were mainly tax incentives, plus a history of high inflation. So they borrowed the money and drove up house prices, so the whole stock of eight million houses basically doubled in price”. Having studiously ignored key facts that don’t suit his story, Dr Moran is left with an exaggerated sense of the potential for “fixing the fringes” to improve housing affordability. Sure, we should develop new suburbs on our city fringes as rapidly as is feasible – it will help to absorb our elevated level of immigration – but don’t hold your breath waiting for a big drop in city-average home prices. Important coastal cities are expensive everywhere. We’ll find over time that it’s much easier to make our capital cities larger than it is to make them affordable. If Australia’s population, average incomes and wealth all continue to rise, so too will average home prices in our coastal capitals – despite Dr Moran’s insistence that homes in America’s rust belt are cheap only because of inspired hands-off planning regimes.

Peter Fray

Get your first 12 weeks of Crikey for $12.

Without subscribers, Crikey can’t do what it does. Fortunately, our support base is growing.

Every day, Crikey aims to bring new and challenging insights into politics, business, national affairs, media and society. We lift up the rocks that other news media largely ignore. Without your support, more of those rocks – and the secrets beneath them — will remain lodged in the dirt.

Join today and get your first 12 weeks of Crikey for just $12.


Peter Fray
Editor-in-chief of Crikey