Note: This post was original published on 6 July 2015; I’ve updated it several times because both parties keep revisiting a decentralisation agenda.
Once again we’re hearing the argument that Australia would be a much better place if only we could actively “decentralise” population. The argument is we should encourage people out of our big cities – notably Sydney and Melbourne – and into smaller cities, like Wollongong and Ballarat. One recent claim comes from the Liberal Party’s Tim Smith, the member for Kew and Victorian Shadow Parliamentary Secretary to the Leader (Population Policy and Housing Affordability). In an article in The Australian, he argues:
Opposition Leader Matthew Guy’s vision is to decentralise Victoria and develop its regional cities, to take the pressure off Melbourne and grow country Victoria.
The state desperately needs a government that is committed to decreasing the percentage of newcomers who make their home in Melbourne. Our state needs a government that will engage in a mature debate about how to incentivise newcomers to move to country Victoria, or give them the confidence that if they move to a regional centre they can commute to Melbourne with reliability and ease …
… An effective decentralisation agenda is key to improving capital city liveability and the economic wellbeing of the regions.
In pursuit of this, various governments over the years have tried to move departments out to regional cities. Smith implies that Labor doesn’t want decentralisation, but the evidence suggests Labor is just as keen on the idea as Smith is. The Victorian government under John Brumby even ran an advertising campaign in Melbourne encouraging people to move out and resettle in regional Victoria.
This sort of argument has often been based on the idea that these regional areas have lots of existing infrastructure that we can exploit at little cost. It has been encouraged by talk of the “Death of Distance” and “The Flat World” – the idea that globalisation and modern telecommunications are making location obsolete, so you might as well live in the countryside. It’s particularly popular wherever there are plenty of marginal regional electorates.
And this argument seem to be spreading. So here’s the case against spending government resources to actively encourage decentralisation.
The first argument against encouraging decentralisation is that it is essentially a policy to encourage people to live where they don’t want to live. People are capable of figuring out where they want to live, and of moving when things don’t suit them. At least for most people, the reasons for living in the city and the country aren’t hidden; they’re well-known. Let people vote with their feet, and don’t rig the poll.
A second argument, advanced in earlier eras is that we need to fill the countryside with people in order to defend it from invasion. This view is now largely obsolete, although I suspect it is still in the backs of the minds of some older Australians.
A third argument is that regional infrastructure is not necessarily up to the job of coping with strong growth after all. A couple of years ago I read the CEO of the Committee for Gippsland arguing in the Herald-Sun that a coming influx of residents to rural Victoria meant we need to spend more on regional infrastructure. This is in part a bait-and-switch: first we need decentralisation to exploit unused infrastructure, then we need more infrastructure to cope with the new population.
And the fourth and strongest argument against active decentralisation policies is that big, dense cities are good for innovation and economic growth and personal growth too, and that those benefits should be embraced rather than resisted. Crowding people into cities has literally built our civilisation. We shouldn’t start trying to fight the process without good reason. We should be finding ways to support it.
Australia’s city-dwellers have a long tradition of not criticising country life. It seems almost rude to point out that living in big cities is, in very many ways and for the vast majority of people, better than living in towns or the bush. Big cities benefit from enormous economies of scale and scope. Small towns struggle to get doctors; big cities have specialists who will be able to recognise your specific type of epileptic attack. A gay Chinese 17-year-old in a small town is likely horribly alone; put them in a city and they’ll find their subculture. An industrial designer in Roma has to rely on social media for conversation about widget design; in Prahan they can go down the pub with half-a-dozen colleagues and talk more and more wildly about widgets as the evening grows late.
The growth of cities is no accident. They represent a lot of people voting with their feet.
Recent years have brought a flood of new US books in praise of cities. Ed Glaeser is probably the world’s leading urban economist, and his 2011 book Triumph of the City made the argument for cities. “There is a near-perfect correlation between urbanization and prosperity across nations,” he wrote. “On average, as the share of a country’s population that is urban rises by 10 percent, the country’s per capita output increases by 30 percent … Across countries, reported life satisfaction rises with the share of the population that lives in cities, even when controlling for the countries’ income and education.”
Also in 2011, The Economist’s Ryan Avent published The Gated City. And in 2012 Berkeley economics professor Enrico Moretti, a human capital specialist, published The New Geography of Jobs (available at Amazon), buttressing the case for cities. Moretti’s work includes an NBER working paper called Human Capital Externalities in Cities – download it from the NBER or read a summary at CNN.
Enrico Moretti’s argument is that by jamming smart people together, cities produce ideas and boost productivity. Think Florence in the fifteenth century, London almost anytime in the past 600 years, but also Silicon Valley and San Francisco in California. The California example is the strangest: a huge slice of the people who are supposed to making distance obsolete live and work within 100 kilometres of each other. And that’s despite the fact that this clustering has made homes wildly expensive.
The argument is less common in Australian. The Grattan Institute has done some work on cities as engines of prosperity. And in 2006, the Australian economist Glenn Withers published a CEDA paper, Can Distance Be Defeated?, in which he made essentially the same argument as Moretti. Withers referred in particular to the work of Frank Levy and Richard Murnane, who were among the pioneers of the idea that the high-level knowledge that fuelled innovation was tacit and serendipitous and hence was “conveyed in the direct personal interaction that is a function of co-location”.
Or to put it another way, some of us get our best ideas down the pub with our mates.
Since I wrote the first version of this post in 2015, an empirical study has also popped up – The Optimal Distribution of Population across Cities, an NBER paper by David Albouy, Kristian Behrens, Frédéric Robert-Nicoud and Nathan Seegert. The authors are blessedly modest about their ability to calculate the right size for cities. But they do say that their modelling suggests cities “may well be too numerous and underpopulated for a wide range of plausible parameter values”. (The paper is also useful for the history and analysis it provides regarding economists’ claims over previous decades about city sizes.) None of Australia’s largest cities ranks in the world’s top 50 by population. On the NBER authors’ numbers, they may be smaller than is optimal for us.
No-one quite knows what a pro-cities policy would look like. Moretti favours policies that increase labour mobility and education levels. The Grattan Institute argues for better urban transport. Avent gives a more important role to regulatory policy. Withers, one of the original authors of Australia’s successful skilled migration program, emphasises the need to keep managing skilled migration successfully. But there’s more work to do teasing out any positive policy implications of this thinking. And I’m not arguing for any of these policies here.
Nevertheless, if these analysts are right, one policy implication is clear. If you want people to expand their minds and grow the economy, don’t actively discourage the growth in our cities or spend tax dollars to encourage “decentralisation”. Yes, avoid growth that destroys what has made our big cities attractive in the first place. But don’t pretend that getting people out of cities is a sensible policy goal in its own right. It may help economic growth in Wollongong and Ballarat, but the evidence suggests it will do so at the expense of national productivity, innovation and creativity.
Update: More recent writing on the economic benefits of cities:
- Why Do Cities Matter? Local Growth and Aggregate Growth, by Chang-Tai Hsieh and Enrico Moretti – Argues that increased constraints to housing supply in high-productivity cities like New York, San Francisco and San Jose lowered US GDP.
- Urban Growth and its Aggregate Implications, by Gilles Duranton and Diego Puga – Another urban growth model where human capital spillovers foster entrepreneurship and learning in heterogenous cities.
- Order without Design: How Markets Shape Cities, by Alain Bertaud – A much-admired book which spends time on applying economics to cities. “Everybody should read it” – Paul Romer.