Wednesday, December 2, 2009

What price a roof over their heads?

If you’re a Baby Boomer or Gen X, talk of the problems of housing affordability might be of limited personal relevance. But for the following generations of Australians, the cost of housing - both rented and ownership - has escalated to a point that our society in the future may be significantly altered. For Families like the Kerrigans of the future, things will be very different.

Is there a problem?

No matter which measure you use (there are many published) it’s clear that housing costs relative to incomes have soared, especially since the late 1990s. As a multiple of average incomes, the median house price has escalated from around 4 to 5 times incomes for almost all of the post war period (or since records began) climbing rapidly in the late 1990s and continuing to climb now, reaching seven or even eight times average incomes for many capital cities.

Historic low interest rates have had only marginal impact on the affordability issue, because the size of mortgages are now so much larger, especially for first home buyers entering the market. In other words, housing was on average more affordable when interest rates were above 10 per cent or 15 per cent for the 1970s, 80s and early 90s, than when interest rates fell below 7 per cent - because houses were cheaper relative to incomes and mortgages easier to service.

Home ownership rates, especially in the generation of under 35s, are falling to historic lows, from 44 per cent a decade ago to just a third now (see here for example). This leaves many in the rental market, where rental costs relative to incomes have also climbed and eat up much higher proportions of the household budget, making saving in turn harder and the idea of building a deposit that much more elusive.

Causes now undisputed

The causes of the decline in housing affordability are now better understood. Many economists, not trained in the mechanics of supply-side housing delivery, wrongly assumed Australians’ “love affair with housing” and growing demand fed by falling interest rates was the primary cause of rising prices. But now the price pressures on the supply side created by state and local government planning policies are better understood for their role in increasing house prices. Limited new land supply (the result of artificially imposed growth boundaries), compounded by new and exorbitant levies on new housing (the combination of which can readily exceed $100,000 per dwelling), plus the high compliance costs created by labyrinthine and uncertain planning regulations, have combined to create a generational price pressure on supply that simply did not exist prior to the late 1990s (after which point, unsurprisingly, prices began to escalate relative to incomes).

As much has been acknowledged by the Reserve Bank Governor,27753,25847312-462,00.html himself, who has publicly complained that our inability to create new housing supply at such a time in our economy, while house prices continue to rise, is a cause of concern that policy makers need to address.

The consequences

If the extent of the problem and the causes are now widely understood (though there remain some in denial), it’s the consequences of this generational change that are worthwhile thinking about. Home ownership has been a cornerstone of Australian economic, social and family life for generations. What happens if an entire generation finds ownership so much more costly, or entirely elusive?

The first consequence is already apparent. Our parents, and for many of us aged 40 or more, probably coped with a mortgage on a single income. Today, young couples or families entering the housing market need to rely on two incomes to service the debt. Even then, the combination of two incomes allows little comfort room - a loss of work by one member of the family can lead to immediate financial distress. This places additional pressures on young families. It’s also hardly a coincidence that the rapid escalation of child care in Australia can be roughly traced to the point where house prices started to escalate out of proportion to incomes (roughly a decade ago). What the long term social impacts of dual income families with children in child care from an early age - and pre- and post-school care for their school years - will be, only time will tell.

Deferrment of children by an entire generation is already being observed (PDF 1.88MB). To some this is a sign of a selfish generation, splurging on the here and now. To others, it’s a sign that the prospects of starting a family, and buying a home, have become financially too remote for people in their 20s, so they defer these plans until their 30s. The health risks rise in proportion to the age at which women have their first child: what consequences will this create for health care costs and prenatal care, given the supposed ideal age for having a first child is in the 20s?

At the other end of the scale, project this current generation forwards in time. The family home and real estate generally has proven the single biggest form of savings for Australians for generations. It’s been relied on to help fund retirement, and even to help fund people into aged care. If rates of home ownership decline substantially for a generation, this form of retirement savings is no longer there. Superannuation has been a revolution but for average wage and salary earners, remains largely insufficient to fund a generation of workers into retirement. Will this mean greater dependence on social welfare for the aged as a higher proportion of the current generation moves to retirement age?

The economic consequences of high housing costs are also apparent. With greater proportions of total household incomes being devoted to mortgage payments than ever before, this also means less is available for other forms of consumption, or saving. The paradox of thrift shows that reducing consumption in favour of saving has economy wide effects by decreasing demand and putting a brake on economic activity. The same surely happens if thrift occurs because housing mortgages are consuming so much economic energy that consumption elsewhere suffers.

Australians, the reports tell us, are working harder, longer and taking fewer holidays. Is that because of necessity, triggered mainly by the single biggest impact on household finances - the mortgage? The economy-wide consequences of less consumption generally, mean reduced leisure travel, less non-housing or business investment, and so on, simply because for a generation the cost of servicing the debt on the family home has become financially all consuming.

Consider this: if the combination of supply shortages, up front levies and red tape have conservatively added $100,000 to the cost of a new home, that additional cost by way of a larger mortgage is worth an extra $675 per month alone! ($100,000 at 6.5 per cent variable over 25 years). The total interest bill is over $100,000, plus the principal, over the life of the loan. If that money was directed instead into domestic consumption, would the economy be stronger and Australians more prosperous?

Limited economic growth is a further consequence that could derive from needlessly high house prices. For states like Queensland, for example, population growth has been driven not just by lifestyle seekers, but also by the lure of relatively lower cost housing and costs of living. That relative advantage has now eroded, interstate migration numbers have slowed (topped up for the present by international migrants) and - due to the slowdown in the building industry as public policy chokes new supply - new job creation has also slowed. Ironically, Queensland now trails the (more affordable) states of Tasmania and South Australia in terms of economic performance, according to this report,23739,26198157-952,00.html by Commsec. (West Australia is also concerned - see here

You could speculate forever on other possible consequences of maintaining this high price regime for housing via current public policy settings, but among the most worrying is that we are witnessing the creation of a new landed class structure in Australia. For Gen X and Baby Boomers, rising house prices have added equity to family balance sheets - which have been leveraged to acquire additional investment housing, often more than one. While this generation is building wealth through house price growth, the other generation is being denied that opportunity. They are becoming the rental generation, whose economic efforts will go in rents to the landed generation, whose wealth will rise further.

Australia has long held dear the ideal of a classless society, of equality of opportunity and ‘a fair go’. But are we now witnessing a new class structure, defined by those who own property (and quite a bit of it) and those who don’t? That this is occurring largely at the hands of Labor state governments who claim a charter of social equality, is just as worrying. And if we fail to remedy the problem now, what are the possible consequences of that new social division as we move forwards in time?

The demise of evidence

Has evidence-based planning fallen from grace in favour of catchy slogans and untested assumptions? Has ‘must – debating’ replaced the search for factual examination leading to workable strategies for our urban future? In the case of urban planning, arguably that is just what’s happened. The evidence, in Australia at least, is worrying.

“We must get people out of cars and onto public transport.” “We must stop urban sprawl and the consumption of valuable land.” “We must build higher density communities to achieve sustainable environmental outcomes.” Phrases like this are now de rigueur across many discussions about urban planning in the media, in politics and in regulatory circles in Australia. They have become defacto international statements of fact, rarely challenged on the basis of what the actual social, economic or scientific evidence is really saying. So chronic has the march of the ‘must-debaters’ become that attempts to question the assumptions on factual grounds can produce Animal Farm like dogma in response: ‘Four legs good, two legs bad.’ Or ‘Napoleon is always right.’ Denial, followed by ‘pass the buck’ and ultimately ‘shoot the messenger’ are responses to legitimate questions.

But given the far reaching social and economic changes which will invariably flow from some of the regulatory planning schemes now legislated, it does seem valid to ask whether the various policies will actually achieve what they say they will. After all, these regulatory planning schemes are intended to govern our urban growth over the next 20 years. It would be a shame to get it badly wrong, simply because assumptions weren’t tested.

The rise of the big plan

In Australia, there have since the late 1990s been a raft of regional planning schemes dealing with urban growth in our major centres. The common theme has been the creation of urban growth boundaries and increased density in established urban areas, with an emphasis on public transport as opposed to the private vehicle. These schemes have generally passed without considered public scrutiny or challenge, although Tony Powell – a highly respected urban planner – described them as a “sad parade of failing capital city strategic plans [which are] superficial to the point of ridiculousness.” Hardly big raps there but his was largely a voice in the wilderness.

One of the most recent of these schemes was ‘The South East Queensland Regional Plan 2009-2031’. It is the State Government’s ‘plan to manage growth and protect the region’s lifestyle and environment. The plan responds to issues such as continued high population growth, traffic congestion, koala protection, climate change and employment generation. The plan balances population growth with the need to protect the lifestyle residents of South East Queensland value and enjoy.’

Because it bears much in common with similar schemes around the country, the SEQRP serves as a reasonable model with which we can examine some of the underlying assumptions, and test them against the evidence.

First, some context.

Australia’s total population is currently around 24 million people, in a land mass roughly the size of continental USA. This puts us below Nepal and Uzbekistan but ahead of Madagascar in population rankings. Reports that Australia’s population may reach 35 million in another 40 years (the current population of Canada) have raised domestic fears that we might become over populated. (See my blog post ‘Australia Explodes’ for more on this).

The State of Queensland is the second largest state by area, but contains only 4.4 million people in total. Its population growth rates have in the past been amongst the highest of any region in Australia, growing at up to 1500 people per week (close to 80,000 per annum). Much of this growth has occurred in the south east corner of the state, surrounding the capital city – Brisbane. While modest by global standards, this rate of growth has thrown governments and some sections of the community into apoplexy. How will we ever cope? South east Queensland (population 3 million) has been compared to California (population 38 million) in terms of its growth rates and population pressures.

Against this context, the SEQRP identifies the need to provide a further 750,000 dwellings in the period to 2031, with roughly 50% to be developed in established urban areas via infill, and the balance through new detached housing development on land within an urban growth boundary. The challenge for infill is greater in Brisbane, where 138,000 new dwellings are expected to be developed in established urban areas, especially around transit centres (typically rail).

Against this context, it’s time to examine some of the many assumptions that underpin the core strategy of the SEQRP.

Assumption: We’re at risk of sprawl.

This is the ‘Mouse that roared’ assumption, somehow suggesting that modest and manageable growth rates of 1500 people per week are somehow tipping the big end of the global scale. The region’s current population of 3 million shows obvious signs of urban expansion as a result of growth to date, but if sprawl is defined as the unplanned and unserviced expansion of land for housing, there is no evidence of that. Growth to date has been orderly and regulated. With some notable exceptions in recent years, infrastructure has generally kept pace with the growth. Even at the urban fringe, new housing development has been at higher rates of dwelling density than in years past (lot sizes are shrinking). The region is largely auto-dependent, but there are reasons for that (we’ll discuss later).

Assumption: We are running out of land.

South east Queensland has vast tracts of land suitable for urban expansion. Established regional centres at the edges of the urban fringe (from Kilcoy in the north to Beaudesert in the south) are readily capable of servicing new housing development to varying degrees as the infrastructure and town centres are largely in place, and capable of upscaling. Any quick examination of the region via Google Earth will reveal swathes of land suitable for urban development. The urban growth boundary imposed by the SEQRP is approximately 300 kilometres in length as it curtains the urban area. An expansion of this boundary by as little as a kilometer (under a mile) would create a notional land supply suitable for an additional 500,000 detached homes at 15 to the hectare (or six to the acre). No, we are not running out of land.

Assumption: We can no longer afford the dream of the quarter acre block.

Australians have a folklorish attachment to the notion of a house on a quatre acre block. The evidence, however, suggests this is now ancient history: lot sizes have not been anywhere near a quarter acre since the 1960s. The typical lot size now is 400 square metres, or around one tenth of an acre. Hardly an irresponsible over-consumption of land for housing. As for the dream of a quarter acre block, that’s been long dead.

Assumption: Broadacre growth will consume valuable farmland.

This assumption seems to be most favoured by latte lovers in inner city coffee shops, with only a vaguely remote understanding of farming practices. In the south east corner of Queensland, typically two types of land have been conserved for this reason. The first is land devoted to growing sugar cane. The farming of sugar cane in the south east is no longer economically efficient. Farmers are given government subsidies to continue, despite the local sugar mill having closed years ago. Sugar is efficiently farmed in the state’s tropical north but for some reason, the biodiversity desert that is a sugar cane farm needs preservation in the sub-tropical south east corner. The second type of land conserved under this rationale is land historically devoted to cattle grazing. This was always marginal grazing land in the main – dry, shallow soils that struggle to hold moisture or grow pasture. As technology improved and transport economics developed, more efficient grazing country has been opened up further from city markets. But as farmers are prevented from selling their land for housing, despite its logical location for that purpose, herds of bony cattle continue to roam the urban fringes of the metropolis.

This assumption also seems to hold dear the notion that, for sustainability reasons, regions should source their food needs from within a nearby catchment, minimizing transport costs (and hence saving the planet). Were that the case however, Queenslanders would not enjoy apples (grown in southern temperate zones) and neither would Tasmanians (our cool climate southern state residents) ever enjoy bananas (two thirds of Australia’s crop of which are grown in Queensland). It would also mean our agricultural industries, which rely heavily on export, would fail.

Assumption: infrastructure is more economically deployed in established urban areas.

The cost of infrastructure provision is a subject that preoccupies governments in growth regions. Perhaps for this reason, the suggestion that infrastructure is more economically deployed in established urban areas, as opposed to newly provided in outer growth areas, found much support in treasury corridors. However, the evidence suggests otherwise. In established urban areas, underground services (water, sewerage, stormwater) can be approaching 50 or 100 years since built. These were not initially designed for higher usage levels based on higher urban densities, and anyway are approaching the end of their shelf life. The replacement and upgrade cost of retrofitting these services is demonstrably higher than the cost of installing new services in new growth areas. The same applies to roads, rail systems, schools, hospitals. Ironically, as if to counter their own arguments about the alleged efficiency of exploiting existing infrastructure capacity in established areas, local councils and state authorities still charge very high per-dwelling infrastructure levies in infill areas (levies which can now total more than double the cost of the land for an infill unit housing project).

Related to this assumption has been a concern that migration away from the inner core would deplete population numbers, rendering existing infrastructure (from schools to hospitals to transit and so on) under-utilised. That may have been the experience in a number of US cities, but here the evidence in Australian cities is to the contrary: school class sizes in inner areas remain high, hospital beds in short supply, and infrastructure generally reported as over whelmed by demand.

Assumption: Outer suburban growth will mean worsening urban congestion.

If it was true that residents of new outer suburban growth areas were predominantly employed in inner city areas, this might follow. But according to the Census and other official government data, most jobs are in suburban locations – 90% of all jobs in fact. Teachers, nurses, shop workers, manufacturing and transport, professional and personal services – all are predominantly suburban by nature. The CBD (our downtown) is a high density focus for many headquarter operations, but at 2 million square metres of office office, it cannot by any stretch of the imagination provide sufficient space for the majority of the region’s workers. Congestion has worsened in the south east corner, but arguably this has been from under investment in transport infrastructure – public and private - in the past 20 years (including the absence of ring roads such that the historic hub and spoke arterial road pattern continues to direct traffic into the city, even though more than half of it is trying to get to the other side). The private vehicle is for most not a modal choice of transport but an essential requirement of daily life.

Assumption: infill and higher density will get more people using public transport.

Perhaps this is true to an extent. Current public transport usage represents under 15% of all trips. With higher density housing in established areas, especially in and around transit nodes (TODs), that figure could theoretically increase. But even the most heroic of assumptions would put the future rate at little more than 30%. Meaning that 70% of future trips by the new residents of inner city locations would continue to be by private vehicle. Based on the infill targets for the city of Brisbane, that could mean another 150,000 people using cars on already congested inner city roads. This is hardly likely to diminish congestion. There is also an unanswered question on the capacity of existing rail and bus services to cope with additional demand (frequent reports mention chronic overcrowding) combined with the high level of public transit subsidies per passenger, which will somehow have to be funded.

Assumption: Higher density in established areas is better for the environment.

Again, the evidence here tends to suggest the opposite to what is assumed. Unit and townhouse dwellings, based on several University studies, actually consume more energy than equivalent detached dwellings. Common area lighting, lifts, clothes driers and airconditioning are all more commonplace in high density dwellings than detached (where natural light, cross flow ventilation and solar power for drying clothes are the norm). Factor in the higher number of persons per dwelling in detached housing, and the per person energy consumption of inner city, high density housing looks ordinary.

No less an authority than the Australian Conservation Foundation actually proved this in their Consumption Atlas which revealed that inner city high density residents had much larger carbon footprints than their suburban cousins. Built form was part of the explanation – the rest was explained by wealth (inner city residents tend to be wealthier, buying the privilege of inner city real estate prices, and thereby able to consume more). Wealth also tended to mean that these people eschewed public transport, despite its ready availability.

Assumption: it works in Europe, it will work here too!

This is my favourite assumption, especially when trotted out by a new convert to the density matra, recently returned from an overseas ‘study tour’ which took in all the sights of various European downtowns. The reality of course is that European cities were largely designed (in their inner areas) during the middle ages. Walking was not a choice but an economic necessity, and density was a by-product of both this and the prevailing economic systems (such as they were) and also the climate. Studying the downtowns and inner urban areas of European cities and applying those observations to the Australian context is patently silly. If more study tours instead took in the middle and outer areas even of European cities, where the majority of the ‘workers’ (as opposed to elites) live, they might return with different conclusions. (Wendell Cox has produced a series of ‘rental car’ tours which are worth reviewing in this context).

And so the evidence says?

On balance, many of the assumptions that underpin the central strategic intent of regulatory planning schemes such as The South East Queensland Regional Plan, just don’t stand the test of evidence. Indeed in many cases, the evidence suggests the opposite of what is assumed. But evidence, it seems, is out of favour and slogans are in.

The consequence of this demise of evidence based rigour though is becoming immediately clear. Despite the global economic downturn, policy induced housing shortages due to land supply constraints, accentuated by high and extortionate infrastructure levies, layered on deficient and largely incomprehensible local planning laws, are seeing house prices rise even further. In Brisbane, the median house price is now around $450,000 – or close to 7 times average incomes. Dual income households are now the norm, and minor fluctuations in mortgage rates are enough to send shudders through the entire economy as household budgets are tightened. Even recent warnings by the Reserve Bank Governor (see this for a summary) that we are failing to produce new supply at a time when it is most needed, have so far fallen on deaf regulatory ears.

Four legs good, two legs bad. Napoleon is always right. Why consult the facts when mantra will do?