Address to the Sydney Institute
14 November 2016
[Check against delivery]
Northern Australia makes up 40 per cent of Australia but there is no city there that can be called a “middle city”, or a city of more than 500,000 people. That is a sobering fact for me as the Minister for Northern Australia. Why is it that 228 years since British settlement we have failed to truly settle the north?
Then again, I should not get too down. Southern Australia has just one city that is a middle city - the Gold Coast. Southern Australia has the five mainland capital cities which are all larger, but the spread of Australia’s population is unusual relative to the rest of the world, and is skewed to a few very large cities and many very small ones.
Tonight I will argue that the abnormal distribution of our population costs us greatly. It contributes to Australia’s housing affordability issues, it costs us vast sums of money to build infrastructure in crowded cities and it inhibits policy and economic innovation by monopolising the cultural influence of Australia into a handful of major cities.
Australia has an unusual distribution of its population. Compared to other countries we have far more people in our top two cities. In other countries, the population of cities follows a ranking order that is remarkably predictable. In effect, the top city is usually around double the size of the 2nd biggest city, triple the 3rd biggest city and so on.
Take the United States: New York at 8.5 million people is around double the population of Los Angeles at 4 million and about triple the population of Chicago at 2.7 million people. It is not a perfect relationship but it is one that is repeated across enough countries to have earned its own name, albeit an unpronounceable one: Zipf’s Law. 1
Australia is an outlier when it comes to following Zipf’s Law. Our biggest city, Sydney, at 4.6 million people, is only just bigger than our second biggest city at 4.4 million people in Melbourne. If you graphed our cities on a logarithmic scale – with the rank of a city on the vertical axis and its population on the horizontal – Australia’s chart is flatter than those of other countries. Other countries, that follow Zipf’s Law, demonstrate a rough 45 degree line. A recent economics paper looked at the Australian situation and concluded that it had the flattest line of a range of comparable countries, including those with similar population levels or federal structures. 2
This means that Australia has a greater share of its population in large cities relative to other countries. To compare us to the United States again, our top five cities – Sydney, Melbourne, Brisbane, Perth and Adelaide – account for 60 per cent of Australia’s population. The top five cities in the United States account for not even 10 per cent of their population. This is a stark contrast that cannot be due only to the lack of a Mississippi River, or as big a population.
I find it remarkable that we are such an outlier on the distribution of population among our world peers but there is almost no commentary on this fact or what its consequences are. In my view the consequences are significant and negative.
They manifest most directly in competition for land. Because there are fewer cities of a sizeable population, Australians seeking a job do not have as many options for where they can live. This increases pressure on land values, leading to higher house prices and the difficulties we see young people having in affording their first home.
As the Reserve Bank stated in a submission to a Parliamentary inquiry into housing affordability recently:
Australia also faces particular structural constraints in its capacity to provide new housing at a reasonable cost because of its population structure … Unlike many other comparable countries, Australia lacks the medium-sized cities (500,000–1,000,000 inhabitants) that could provide alternatives to households seeking to avoid high housing costs in the largest cities, while still offering the range of job opportunities that cannot be supported in smaller towns. 3
There is a strong theoretical case therefore that our population structure does increase house prices. If we could create more middle sized towns how much would house prices be lower?
The median house price in Sydney is approaching $1 million, while in our larger non-capital cities, like the Gold Coast, Newcastle, Wollongong and Geelong, average house prices are around $500,000.
Fifteen years ago the Reserve Bank tried to estimate how much our distinct population spread impacted house prices. Their conclusion was that our urban structure could explain perhaps a third of the difference between Australia’s house price to gross income ratio compared to that in the United States. 4, 5
Australia’s housing stock is valued at around $6 trillion. If instead we only had roughly $4 trillion tied up in housing, that would free up $2 trillion of capital. That amount of capital represents more than our annual GDP and close to the amount of assets we have in superannuation.
There is nothing inefficient about investing in housing per se, but the capital that is used to save for and buy a house, could be used elsewhere. There is an opportunity cost. A family that could buy a house at a more affordable price could use it for the husband or wife to start a business rather than pay off the bank. They could inject capital into a family farm, or to invest in venture capital or innovative stocks. These higher housing costs are a real cost, albeit an unspoken one, to our economy.
As the Treasurer recently pointed out, not being able to access secure and affordable housing has flow-on effects to educational outcomes, health, work force participation, social cohesion, welfare dependency, family formation and household consumption. 6
The difference in our population spread is also significant in terms of people. If our 24 million people were spread across the country in accordance with Zipf’s Law, then we would have around five million more people living in regional areas, that is living outside the five mainland capital cities.
Imagine the stimulus that would create to regional areas and the greater dynamism that our country would have. Townsville would be a city of 350,000 not 200,000, Rockhampton would have more than 200,000 people and Tamworth 140,000 people. We would have 44 cities with populations of more than 100,000 rather than only 16. Perhaps half of our National Rugby League teams would not need to be based in Sydney! This may seem like a radical shift but population in other countries are spread in exactly such a fashion.
There is an unhelpful stigma in Australia about calls to develop our country. As a member of the National Party I get the odd rebuke that I am an agrarian socialist. To which I often retort, I am only an agrarian socialist because there are too many urban socialists in this country. Someone has to fight back!
When in fact, given the potential gains I have outlined, the development of the north, and other regions, should be seen as an economic reform on the same platform as tariff reduction, financial deregulation and the floating of the dollar. A potential $2 trillion benefit is nothing to be derided.
Indeed, it makes some other often-trumpeted reforms, such as pharmacy deregulation, Uber and trading hours reform – that we as Nationals are sometimes blamed for holding up – pale into insignificance.
As meritorious as some of these other changes might be, we are not going to get rich by letting people catch an Uber at midnight to Woolworths to fill their prescriptions. We may very well increase our wealth substantially, however, if we make Mackay, Lismore, Coffs Harbour, Mildura or Morwell a place where more people can live and more importantly find a job.
And this is just the start. Providing services in major cities is getting more expensive by the day. In the mid–1980s Joh Bjelke-Petersen built the Gateway Bridge over the Brisbane River. The project cost around $140 million. Twenty-five years later we built exactly the same bridge next to it and this project cost $2 billion. Even after accounting for inflation, to build another one in 25 years’ time would cost around $17 billion in today’s dollars if similar growth in costs occurs.
Our major cities are so developed that they need costly tunnels and public transport systems to ease their congestion. This is true of course for all major global cities but it does gall people from regional areas when the wealth of our mining industries goes to pay for all of this infrastructure.
Australia’s public transport fares cover some of the lowest share of the costs of providing public transport in the world. Where does the tax revenue come from to pay for the subsidies for Sydney’s public transport system? Well much of it comes from mining royalties. What is the public policy rationale that sees the wealth created from a miner in the Hunter Valley help pay the transport costs of a professional in Sydney?
In regional cities, congestion is generally not an issue. From an infrastructure planning perspective it would be a much cheaper option to have more Australians live in areas where there is little urban sprawl and underused roads.
Concentrating our population also reduces the diversity of our culture and innovation. I grew up in the outer suburbs of Brisbane but now live in Central Queensland. There is a different mindset and different priorities. If we had more people live in regional areas, we would encourage different ideas to emerge and have richer culture for it.
This all begs the question: why hasn’t Australia developed middle-sized cities like other countries have? That is a question I don’t pretend I can answer alone but I would like to highlight two relevant issues. First, our federal structure has encouraged the development of what is known as “primate” cities. Second, our tax and regulatory framework is set for the conditions of a developed country, which inhibits growth opportunities in developing parts of our country.
All of our state capitals are also the largest cities in their respective states, and it has always been that way. This has perhaps encouraged the development of large primate cities within Australian states, as the largest city monopolises both the political and economic influence of the region. A primate city is a city that is much larger than would be predicted by Zipf’s Law.
Zipf’s Law does not hold for Australian states either, even though it often does for regions of other countries. Remember if Zipf’s Law is to hold, the biggest city should be double the size of the second biggest city. But Sydney is 11 times the size of Newcastle, Melbourne is 24 times the size of Geelong, Perth is 27 times the size of Bunbury and even in Queensland (the most decentralised mainland state), Brisbane is still four times the size of the Gold Coast.
Canada has developed over a similar timeframe to Australia and does have a flatter rank-size city distribution although not as flat as Australia’s. Unlike Australia, the provincial capitals are usually not the largest cities and this helps provide alternative administrative jobs in smaller towns. In addition, Canada has 10 provinces to Australia’s six states.
This suggests that moving Australia’s state capitals, or creating more states, would help us balance our population distribution. Given political realities, however, I don’t think that Brisbane will move to Townsville anytime soon. That said, we can move individual departments, or parts of departments, to regional areas and that is why the Federal Government’s push to move agricultural research and development corporations should be supported.
Further, while I have expressed support for the creation of more states, primarily for the reasons I have outlined here, the political difficulties are insurmountable in the medium term. So we have to look for other solutions.
Investing in smaller towns is often more risky than larger ones, especially if the investments are large. Because there are fewer customers to spread costs, prices are often higher, or payback periods longer. Any risky investment is more exposed to higher tax rates than lower risk investments due to the asymmetric nature of taxation. If you win, the government takes its cut, if you lose the government is not there to help. The more likely you are to lose, that is the more risky an investment is, the bigger issue higher taxes become.
Most of Australia’s cities were founded before the 20th century. Three of the five youngest Australian cities with a population above 100,000 were founded in Northern Australia during the decade after the Burke and Wills expedition.
When Robert Towns financed the creation of Townsville in 1865 the tax to GDP ratio in Australia was less than 5 per cent.7 It is now around 30 per cent. This massive increase in taxes reduces the incentives to take a risk, go to the frontier and develop Australia. If Robert Towns was around today I am pretty sure he would invest in north shore property, not building a port on the Ross River in North Queensland.
Lower tax rates would help. Other countries have, in effect, done this through the creation of special economic zones. Providing special treatment to particular regions, such as lower corporate and other taxes and other incentives to business investment, might seem radical in the Australian context, but internationally it isn’t. Special economic zones have boosted regions in countries as disparate as Hong Kong, Singapore, Ireland, the US and the UK. China’s experience in the Shenzhen region is one of the more well-known examples of how effective they can be.
The UK has also had success with its Enterprise Zones which were first used by Margaret Thatcher in the 1980s and 1990s – and more recently under David Cameron.
One identified problem with special economic zones is that they displace activity; that is firms simply relocate from other areas, and do not create new investment or employment. But then, as I argue here, this movement would itself create benefits as it would make house prices more affordable and spread our population throughout Australia.
Unfortunately, creating special economic zones is difficult for the Australian Government to do because of constitutional limitations. Section 99 of the Constitution says that:
The Commonwealth shall not, by any law or regulation of trade, commerce, or revenue, give preference to one State or any part thereof over another State or any part thereof.
That generally has been interpreted as prohibiting special tax zones even though we have had a zone tax rebate for Northern Australia since 1945. It is now largely redundant because the most common payment is just $57 a year. No one is going to move to Northern Australia for a slab of beer a year.
Taxes are not the only issue for developing the north, however. Regulations too have impact. We now have laws that mean it takes years to get approval for a mine, a road or even a housing development. New environmental regulations that prohibit the dumping of spoil in the Great Barrier Reef Marine Park area are adding hundreds of millions of dollars to the cost of expanding ports in Cairns, Townsville and Gladstone.
We must protect the environment but if all of Australia is going to have a say about the environment of, say, the Great Barrier Reef, then all of Australia should help pay for that protection. Further, we should not let our environmental laws be abused by those seeking other political aims, like the shut down of our coal industry. That is why the Government has already taken steps to help facilitate approvals, through the creation of one-stop-shops, and plans to do more depending on support in the Senate.
If we can’t create more states, move state capitals, provide tax advantages or have more balanced environmental regulations what is left to do? Well, given the benefits of spreading our population we have to invest in regional areas directly to correct for a stubborn federal structure and punitive tax system.
That is why growing the north should be a national economic reform priority. It is why our Northern Development agenda is important for all Australians not just our north. It is why the terrific work that my colleague, Fiona Nash, is doing in developing regional Australia more generally is so important too.
Our Northern Development agenda is about creating bigger cities and more opportunities for all Australians. Over the past year we have announced almost $700 million in road funding through the north, more than $150 million for water infrastructure and established a $5 billion development bank in Cairns.
Our investments are focused on delivering the infrastructure that can create economic opportunity and lower costs for businesses willing to take a risk and invest in the north. So, for example, we are building the Rookwood Weir on the Fitzroy River. The Fitzroy is the second biggest catchment in our country after the Murray-Darling but it has only one major dam on it at the moment. The Rookwood Weir could double agricultural production and create 2100 jobs.
A lot of people in southern Australia are sceptical about our ability to grow food in the north but that is not the reality. Some recent CSIRO work shows that the record of irrigation projects in the north is not one of failure. For almost every failure there has been a success. For every Lakeland Downs that does not go to plan, there is a Burdekin. For every Ord, there is an Emerald, and for every Humpty Doo there is a Katherine-Mataranka.
Moreover, the failures of agriculture in the north are not because of disease or pest or the tropics. Almost all of the failures have been due to poor financial planning or bad luck with the weather in the crucial early years. There have been failures in the south too.
This scepticism about whether we can grow food in northern Australia is matched by the scepticism of why we would seek to develop the north in the first place. For me that answer is not because people in Northern Australia deserve money or welfare, it is because investing in Northern Australia will make our country stronger.
The Pilbara region produces more economic wealth than 119 countries yet just 60,000 Australians live there. That should be a national disgrace. We can do more to build regions like the Pilbara so that we do not just keep stacking people up in our major cities. We can relieve the pressure of congestion, high house prices and declining support for immigration.
To do so we need to invest in building new places for people to work and live. Australia is not short of beautiful places to live. I live near the Great Barrier Reef, the Keppels and Byfield forest but people need something to do between Monday and Friday too. That is why we need to invest in the regions, create new industries and jobs, and we need to do so for the benefit of the nation and its economy.
- This relationship was first noticed by Alexandre Le Maitre in 1682 when comparing the population of French cities to Paris. In the 20th century, Felix Auerbach and George Zipf formalised the relationship for cities, and it is also apparent in the frequency of words used in a language, the number of people that watch different TV channels and many other relationships, see Eeckhout, J. 2004, “Gibrat’s Law for (All) Cities”, The American Economic Review, December, pp. 1429–1451.
- “While most of the results obtained for other countries show a Zipf coefficient around one, regressions in this paper show Australia has a much lower one, around 0.7, which means a more uneven population distribution among the cities of the system … In order to compare results obtained for several different countries when applying a Zipf’s regression, we pick examples spread around the world with some apparent similarities, such as area (Canada), population or GDP (Netherlands). As we can observe, Australia’s coefficient scores as the lowest one.” (Arribas-Bel, D., Sanz-Gracia, F. and Ximenez-De-Embun D. 2012, “Kangaroos, Cities and Space: a first approach to the Australian urban system”, Region et Developpement, no. 36, p. 166.
- RBA 2014, Submission to the Inquiry into Affordable Housing, submission to the Senate Economics References Committee, February, pp. 8–9.
- Ellis, L. and Andrews, D. 2001, City Sizes, Housing Costs, and Wealth, RBA Research Discussion Paper, 2001–08, October.
- The Economist 2016, ‘Hot in the city’, 2 April, http://econ.st/1X47PRj.
- Morrison, S. 2016, ‘Keeping home ownership within reach’, Address to the Urban Development Institute of Australia, Sydney, 24 October, http://bit.ly/2esCnjy.
- Barnard, A. 1988, Australians: Historical Statistics, Vamplew, W. ed., p. 256.