Below is a guest post from regular MB reader, Phil Best:
In October last year, Leith Van Onselen drew attention to the extraordinary position of a Professor Michael Buxton of RIMT University on urban form policy in an article entitled Attack of the Housing BANANAs.
Now Buxton has authored another piece, on The Conversation, making similar arguments.
His position is one of opposition to both low density fringe development and to high-rise. Leith used him, therefore, as an example of a “BANANA” advocate: “Build Absolutely Nothing Anywhere Near Anything”
In Buxton’s latest piece, he says:
…….Higher average housing densities greatly augment outer urban land supply. A yield of 20 dwellings per hectare saves up to 46% of land and increases lot yield by up to 60%. Denser, better planned and serviced new outer suburbs, and attached row housing and medium rise apartments as infill, are achievable alternatives to unregulated low density outer urban. We should be emulating the best in European cities not the worst in Asian and American cities…..
……Planning regulation is a small component of housing costs in urban areas. Land banking by development companies and home purchase subsidies led to much greater reductions in affordability. The failure of governments to require smaller outer urban lot sizes has prevented more diverse housing options forcing the most financially stressed consumers to buy large houses and bear high running costs. In established city areas, high labour and construction costs, land speculation and financial policies such as negative gearing have led to a combination of high apartment prices and declining unit sizes……..
His gripe about excessive amounts of high-rise is fair enough, even though he probably does not realise that it is the presence of an urban growth boundary (UGB) that makes “building up” create more site rent. The land rentier class loves UGB’s with upzoning more than it loves UGBs themselves – in growth contained cities like Boston that still have density prohibitions, the selling price per hectare able to be captured by greenfields land owners is far lower. Vancouver has rammed through numerous upzoning ordinances without lowering the cost of apartments one iota.
But just how aware is Buxton of the density of new fringe developments right now? Are they not typically already around 20 units or more, to the hectare? And this is not because of any regulatory mandate, but because land costs developers so much that it would be impossible to sell housing with any more land incorporated in each home package. The houses are comparatively large because the site would be under-capitalised otherwise.
As has been pointed out repeatedly on MacroBusiness, most of the inflation in the price of land is due to developers being forced to bid against each other for the available supply just to stay in business. Blaming them for “land banking”, when their margins have been squeezed to the point that they need to sell at “top of the cycle” to make a profit at all, is just demagoguery. “Land banking” upstream from them, on the part of original rural landed gentry and some long-term speculators, is a problem also created entirely by handing these people a monopoly in the form of a UGB. Even moving the UGB merely enlarges the oligopoly; it is not free market liberalism at all.
Curiously, what Buxton sounds like he is advocating in the whole article, is something exactly like the UK’s system of growth containment combined with zoning against height in most locations, and in preservation of existing “heritage”. All this is inimical to achieving affordability, as experience in the UK shows. It is remarkable how little attention is paid to the UK by people advocating similar policies in other parts of the world.
Actually there are few features about typical American cities that can be said to be “worst” practice. There are a lot of myths prevalent on the subject. People like Buxton need to look at a bit of real life data for a change, on house prices, traffic congestion, average commute times, economic productivity, and local tax burdens. Australia’s problem is that it does NOT emulate the best in American cities (or in European cities).
Here is an interesting analysis of correlations that turn out the opposite way to what is commonly believed by compact city, pro-public-transport planning advocates.
From Phil McDermott, Congestion, Density and Transit, which quite logically concludes: (there are) “no grounds for suggesting that density is a prerequisite either to better commuting conditions or that congestion reflects the quality of transit systems…..”
I believe we can take McDermott’s exercise a step further if we focus on the UK’s cities which have all pursued an explicit policy of growth containment for decades with far more rigour than even the Europeans generally have; and make the most relevant possible comparisons with non-constrained US cities. The inconclusive European data contains too many variables in the background regulatory evolution of its cities.
The following are not cherry-picked; they are “representative”. Space prevents the provision of a full data set – for now.
- Manchester UK: Population 2.2 million; population density 4,000 per sq km; INRIX congestion score 25.3; House price Median Multiple (M.M.) 5.2
- Nottingham UK: Pop. 660,000; density 4,200 per sq. km; INRIX congestion score 21.9; House price M.M. 5.4
- Liverpool UK: Pop. 820,000; density 4,400 per sq. km; INRIX congestion score 21.5; House price M.M. 5.2
- Birmingham UK: Pop. 2.3 million; density 3,800 per sq. km; INRIX congestion score 20.5 ; House price M.M. 5.3
Versus:
- Indianapolis: Pop. 1.5 million; density 800 per sq. km; INRIX congestion score 5.0; House price M.M. 2.3
- Salt Lake City: Pop. 1 million; density 1,500 per sq. km; INRIX congestion score 3.1; House price M.M. 3.8
- Oklahoma City: Pop. 900,000; density 800 per sq. km; INRIX congestion score 3.1; House price M.M. 2.9
- Kansas City: Pop. 1.5 million; density 900 per sq. km; INRIX congestion score 2.8; House price M.M. 2.7
- Omaha: Pop. 730,000; density 1,000 per sq. km; INRIX congestion score 1.0; House price M.M. 2.5
House price median multiples in my data set are from pre-crash (2007) data to avoid accusations that US house prices in the data are unrepresentative post-crash lows.
The INRIX congestion index score methodology is explained here.
It is a very useful complex calculation that goes beyond the simple statement of “delay at peak time” of other indexes such as the TomTom one. It incorporates analysis of the state of the whole road network at all times. It clearly reflects the worse state of day-long congestion at all locations in denser cities, hence the massive disparities. The more blunt TomTom scores still show up the denser cities, but not as severely.
Ed Glaeser and Matthew Kahn in the excellent Sprawl and Urban Growth (2003) say in the conclusion:
“…..But while it is sensible to tax driving on crowded streets, it does not seem sensible to respond to congestion externalities by fighting sprawl. Before employment decentralization, when sprawl meant suburbanites driving downtown, the link between sprawl and added street traffic on congested roads was tight. However, the decentralization of employment actually reduces the pressure on crowded downtown streets. By moving to lower densities, the traffic problem is actually reduced. Indeed, one of the major appeals of sprawl cities is that they have shorter commutes than dense downtowns……”
Lest anyone think that because the urban area is smaller in the UK’s cities, congestion delays might not result in longer trips: BBC: UK commute ‘longest in Europe’
Or from Catherine Rampell: World of Commuters: “Americans have some of the shortest commuting times in the developed world, according to a new report from the Organization for Economic Cooperation and Development……”
As has been discussed before on MacroBusiness, decentralisation of employment has tended to keep trip distances short regardless of the level of density. The failure of the denser UK cities and now of more and more cities in Europe, Canada and Australasia is related partly to congestion delays due to sheer lack of road lane-miles, and partly to inefficient spatial sorting of workforces in relation to their jobs and indeed in relation to all desirable household trip destinations. Of course if 90% of the housing in an urban area is affordable to 90% of the population, it will be far more likely that people in that urban area will be able to choose housing conveniently located relative to any employment they might take; compared to an urban area where all of the housing is unaffordable to 50% of the population and desperate forced choices of housing with poor attributes including location as well as size and quality and neighbourhood quality, are the norm.
Alain Bertaud of course points out in his papers that study “the spatial distribution of density”, that high density fringes = longer average commutes as a rule, while an urban density curve that gradually slopes up from near-zero in rural land beyond the fringe, towards the city centre = shorter average commutes as a rule. This is because although employment is now always dispersed, its spatial distribution is still weighted towards the centre. Ironically, Australian cities are already trending towards the distorted “dense fringe” effect as explained above, even without Buxton’s already-redundant proposed mandates.
“Splatter” development on land beyond the urban fringe with later infill is not disputed to be inherently efficient in urban economics literature. This is because of the effect of maximum efficiency of co-location decisions and private sector land use decisions as the land is gradually utilised, rather than “take it or leave it” cramming on what planners deem to be “the right amount of land” for this decade’s growth. Even “clusters” that gain agglomeration economies are enabled under splatter and inhibited under planned incremental growth. Silicon Valley is a famous example of an uninhibited-growth exurban cluster.
The belief that productivity is increased by density per se is false. Any discernible correlations in national or international data between density and productivity are due to density and productivity both being endogenous outcomes in urban economic evolution under relatively free market conditions. It is absurd to assume that the higher productivity associated with “Manhattan” local economies can be replicated anywhere simply by containing spatial growth. The UK is the text-book example that proves the folly of such an assumption, suffering a productivity gap that is hard to explain by anything other than the planning-induced inefficiencies inherent in its cities. See, for example, Alan W. Evans and Oliver Hartwich: “The Best Laid Plans”; or Max Nathan and Henry Overman: “What We Know (And Don’t Know) About the Links Between Planning and Economic Performance”.
