In the last five years house prices in Australian cities have in aggregate risen by 30 per cent in real terms, but this rise has been very much accounted for by the Sydney market, which has risen 64 per cent, with the Perth market in actual decline. Property markets are prone to cyclicality, with booms and busts featuring through history and in Australia’s case in the 19th century central to the great recession of the 1890s.1 In the period since the 1970s, there have been five major cycles in the Sydney market and a lesser number in the other markets (see Table 1).
For the Sydney market, this latest cycle comes in second, just beating the rise in 1987–89 of 59 per cent and behind the 1996–2004 rise of 85 per cent. But, it is well behind the rises recorded in other markets in the preceding boom, with the Perth market rising 173 per cent in the period 1998–2007. But while the rise in Sydney might not be the biggest, the longer-term story is that prices are rising off a high base, with these cycles part of a sustained long-term rise in all markets in the period since 1960. This has seen prices in the Sydney market rise by an average 3.8 per cent per annum in real terms (see Figure 1) and a lesser but still significant upward trend in all markets.
Rises part of a longer-term trend
This rise in the period since the 1950s contrasts with the period 1880–1950s when house prices in the Sydney and Melbourne markets did not show any growth (see Figure 2). This longer-term story is also observed internationally (see Figure 3). In the period 1900–1955, the international series has house prices declining by about 0.5 per cent per annum in real terms, led by declining land prices. Then from 1955–1996, prices rose by two per cent per annum and in the latest period 1996–2012 by 3.5 per cent per annum.
The 20th and now 21st centuries have seen the rise of cities globally but the growth in the first half of the 20th century was faster so, while urban theory tells us that land prices are higher in larger cities, simple reference to the size of cities is not the answer. In 1901, Australia was one of the most highly urbanised countries in the world with 36 per cent of its population in the six capital cities, but urbanisation accelerated and in 1961 these cities accounted for 59 per cent of the population, and their share has since risen to 67 per cent in 2016.(2)
The rise of cities is directly related to the transport revolution brought by the trains/ trams (in the 19th century) and most significantly the motor car, which lowered the cost of commuting and transporting goods around a city and allowed cities to grow. This transport revolution allowed people to move out of crowded inner cities, which went into decline, and it created a supply of land that drove down the price of land. In fact, some urban economists writing in the 1950s expected the price of land to continue to fall.(3)
Several factors have contributed to the change in direction observed since the 1950s. One was that in time as cities grew and infrastructure struggled to keep up, the volume of traffic led to increased congestion, which lifted commuting costs and started to shift the balance back in favour of inner areas. In 1960, when manufacturing’s share of economic activity reached its peak, a sizeable proportion of it was located in the old, inner areas of Australian cities near the main port and rail facilities. However, two trends changed that. One was the globalisation of manufacturing, which was related to the sharp decline in shipping costs and increased economies of scale, coupled with the rise of Japan, then South Korea and more recently China.
At the same time, within cities, in response to those same factors, manufacturing had been (since earlier) moving out to the outer areas where more space could accommodate the larger scale factories and the manufacturing workforce followed the factories out.(4) Meantime, agglomeration economies were attracting the fast growing financial and information service sector into the CBDs, drawing with it high income jobs and households. From being run-down in the 1980s, the period since has seen the regeneration of inner city industrial areas, a phenomena (gentrification) observed in many cities in the world.(5) This gentrification process generates some angst as the character and socio-economic mix of these areas changes, with the high rents signalling the exit of lower income households.
The other significant change has been a rising propensity for policy to impose supply constraints, a policy direction in which Australia has followed the UK and US. While financial and products markets have been deregulated in the postWorld War II period, the urban land market has become more highly regulated.
One element is the zoning of cities in terms of allowed use (urban vs rural at the edge, industrial vs residential, and in terms of density) which happened at a point in history, in Sydney’s case in 1951,(6) when cities were smaller. The objective of zoning and other planning policies was to limit the outward expansion of cities to curtail “urban sprawl”, but the potential impact of this supply constraint on prices was recognised early by economists.(7) As cities have grown, the resistance to re-zoning has increasingly distorted urban markets and it has become a HOUSING AUSTRALIA 40 more binding constraint. While restraining the outward growth of cities has been one objective, in the inner region where the high and rising value of land would increasingly favour higher density, it has also become a more binding constraint on an adequate supply response.(8)
In the Melbourne market, courtesy of good historical data, we can observe some of the intra-city shifts in response to these changes since 1971. In the period 1971–91 (see Table 2), price appreciation was higher in the inner but not significantly so. In this period land prices were pushing up the cost of houses on the outer. The position changed in the post-1991 period. While land prices and housing costs at the urban fringe continued to rise significantly in real terms, the lower interest rates in the later period underpinned a sharper rise in the value of the location premium in inner areas, so that in this period the inner urban area prices rose more sharply.
These trends are more accentuated in the Sydney market in the period from 1991 (see Table 3). All segments of the market have risen faster than Melbourne. In the outer, fringe land prices have risen significantly faster than in Melbourne, pushing up the prices on the outer. In the inner areas, the evidence of gentrification is clearly seen, with the prices in the old industrial (inner low) areas rising significantly faster than the established expensive (inner high) areas.
Inland vs coastal – the great divide
From an international perspective, we observe a significant difference between coastal and inland cities. This is best illustrated by the US where house prices in coastal cities, e.g. most cities in California have risen much more sharply than inland cities and have also exhibited higher volatility.(9) This also explains why European cities – mostly inland cities – tend to have lower cost housing. Coastal cities are naturally more geographically supply-constrained and have amenities (climate, harbours, beaches) which contribute to liveability and attract a premium. As incomes have risen, particularly for highly skilled workers in information and financial sectors, they are attracted to cities with the combination of deep labour markets and, critically, good amenities.(10) In addition, a relationship has been observed between coastal cities with high amenities (high prices) and the tendency for policies to restrict growth, further accentuating pressure on prices.(11)
Pertinently for Australia, its cities share similar climate and coastal settings with those of California, while Sydney shares a similar harbour and other geographic constraints to San Francisco, which tops the list in terms of prices in California. If we contrast the six capital cities, we see that for the five state capital cities other than Sydney, relative size explains the relative differences in prices (see Figure 4). Prices in Sydney are higher and have been persistently higher than the other capital cities over the period 1971–2016. Sydney has a high level of natural amenity (the so-called Sydney premium) and geographic constraints, and these factors may explain the higher prices in part but, Sydney also appears to be a good example of the connection between high amenities and policies to restrict growth. The Productivity Commission has indicated that in terms of policies that restrict growth, it is a problem in all cities but the planning system in Sydney appears to be the most restrictive.(12)
The late 1990s boom
The global recessions around 1990, ironically primarily the outcome of major property booms and busts, provides a neat divide in the housing story globally.(13) The 1970s and 1980s, culminating in those recessions, had been a period of poor economic performance (high inflation, high unemployment). The high nominal and real interest rates the high inflation era produced suppressed house prices for much of this period but saw a significant rise in rents (see Figure 5). The high inflation had also eroded the real value of debt, with a wealth transfer to house borrowers, so when we benchmark debt to income ratios against the 1980s, it should be remembered that the low ratio at that time reflects this devaluation by inflation. The high inflation had also imbued in all investors (including owneroccupiers) the value of property as a hedge against inflation.
The response globally to these adverse economic outcomes was a period of significant deregulation in product markets (lower protection), tax reform and financial deregulation. Tax reform in the 1980s saw consumption taxes introduced/broadened (albeit with a lag in Australia’s case), while top marginal income tax rates were cut sharply and capital gains taxes introduced in most countries.
Lower marginal tax rates reduced the tax advantages to owner-occupation, but otherwise the tax bias to owner-occupation was generally left untouched, while (in most countries) capital gains taxes and (in some countries) capping of negative gearing benefits worked to reduce the tax benefits to investors. Financial deregulation saw interest rate controls lifted and the new competitive environment shifted the balance of power to borrowers. The deep recession in the late 1980s then purged the economies of high inflation and, certainly in the case of Australia, set economies up for a significantly better economic performance in the period post-1991.
The decisions to deregulate the housing finance market were made in the mid- 1980s but the full impact of that did not come through until the 1990s. The combination of stronger economic conditions, but more importantly a significant decline in real and nominal interest rates (see Figure 6), and the deregulation of financial markets, were extremely favourable to the housing market. If we look at bank housing interest rates, the decline in the published interest rates does not tell the full story. Actual interest rates, particularly for marginal borrowers (low income households), were significantly higher in the period of regulated rates.14 In addition, also working against low income households, the constraints on supply meant that power was with lenders who imposed very conservative lending criteria (high loan to valuation ratios).
In a sense, while the supply side was being steadily more tightly regulated, the demand side was deregulated. The decline in interest rates happened in the early 1990s but prices did not start rising till the second half of the 1990s. While rates fell, borrowers had bad memories of the high interest rates in the 1980s and it took a number of years for them to be convinced that the low inflation/low interest regime was here to stay. The location premium in land values are, like any asset, a function of interest rates. In time, the lower interest rates started to be factored into land (house) prices, starting with Sydney, the market with the largest location premiums. While lower interest rates saw prices bid up, the lower rates also translated to a lower cost of capital and a period of slow growth in rents. Reconciling the two, the rent-price ratio for Sydney declined from over six per cent circa 1991 to under four per cent (see Figure 7). The rent-price ratio is the favoured metric for assessing housing valuation15 – while it has fallen roughly consistent with the decline in real interest rates, the experience of 2004 suggests the market was over-valued at that time.
In the Australian market, Sydney led the way with prices rising 85 per cent (in real terms) in the period 1996–2004. That boom ended due to a combination of the market overshooting and the lagged supply response that was putting downward pressure on rents in 2004, with prices declining nine per cent 2004–06. Other markets in Australia lagged the Sydney market and their rising prices caught the stimulus provided by the resources boom and a surge in immigration, which underpinned a period of sharp rises in rents. The upshot was that these markets showed much bigger price gains in this period, prices more than doubling in each, with even Hobart prices rising 115 per cent. At the epicentre of the resources boom, the Perth market topped the list with a rise of 173 per cent (see Table 1), which briefly had the Perth median price touching that of Sydney.
The previous time Perth prices had matched those of Sydney had been in the early 1970s, following the 1960s resources boom after which Perth prices fell 22 per cent and back to their normal relationship to Sydney prices. In each of those markets, the end of those booms circa 2008 roughly coincided with the GFC but without a serious fall in prices. The end of the Sydney boom had been cushioned by the resources boom, which meant that the overhang in the market was absorbed more quickly than otherwise. Rather than prices falling, rent-price ratios largely adjusted (see Figure 7) via a rise in rents, whereas typically it is prices that do most of the adjustment.
When the Sydney boom ended in 2004, the expectation was that this boom had been a one-off lift in prices in response to a one-off fall in interest rates. Things turned out differently for two reasons. One was that the parallel housing booms in the US and Europe (see Figure 8), which had started later, had ended badly for the US and European economies with the GFC. This then led to another ratcheting down in interest rates, which provided a significant boost to all asset markets, including housing markets. And, secondly, while the US and European markets suffered over-supply, that was not the case in Australia where the rise in immigration meant the markets were generally tight and facing upward pressure on rents. During this second phase of the resources boom (2009–12), interest rates in Australia were substantially higher than in other developed economies. The RBA was actually tightening and this was acting as a constraint on housing activity and prices. Then, in response to the end of the resources boom, the RBA cut interest rates aggressively (2012–16), looking to housing to help fill the growth gap created by the collapse in mining investment. This triggered the 2012–17 boom.
Sydney’s place as the strongest in this period is explainable; the previous cycle in the Sydney market had finished earlier, and the market was being more tightly constrained. Again, whether valuations (the rent-price ratio) have fallen too low, as they had in 2004, is the question that will be answered in time.
Home-ownership – beware?
The Menzies Liberal Government (1949–1966) has been widely credited by some commentators for the rise in home-ownership that occurred in the 1950s and 1960s. Indeed, there had been a sharp rise in the home-ownership rate from 52 per cent in 1947 to a peak of 73 per cent in 1966 (see Table 4), with a similar rise in the UK and US.16 There were concessional war service home loans given under the War Service Homes Act 1918, but this was not of scale to explain the rise.(17) In addition, with the imputed rental income of housing exempt from tax, the significant rise in marginal tax rates in this period would have also encouraged ownership.(18)
However, the largest factor in the rise in home-ownership in the period 1947–1972 was a policy failure, namely rent controls that were imposed in all three countries as a War-time measure in 1939 and were then continued into the 1950s. Coming during a period of high inflation, the sharp fall in real rents caused the supply of private rental housing to decline, a shortfall not made up by an increase in public housing. As with all controls, it created a group of winners – those established renters enjoying the protected low rents – and losers, the landlords and new households who could find nothing to rent.
The contraction in rental supply at a time of strong growth in demand led to, in effect, a forced rise in home-ownership, led by a generation of young households unable to enter the rental market. This was a period when land was relatively inexpensive so households did not need much capital to acquire a block of land. There were also few building regulations and Dingle (2000) estimated that over one-third of all new houses in the 1950s were owner-built.(19)
From this artificial high in 1966, the subsequent decline no doubt in part reflects the significant rise in prices. But to a large extent it could simply reflect a return to a more normal pattern as subsequent generations chose to be more mobile. Young households will choose to rent because they value their future earnings and that can be optimised if they are more mobile.(20) The high transaction costs of buying/selling are a major constraint on mobility so that renters are more mobile. So, while young households have less wealth and are credit constrained, household choice is as much about “settling down” (when investment in education is complete and career and family are established) as about “saving up”. Overlaying that, in the post-1991 period, the initial impact of lower interest rates would have assisted entry into the market but as those interest rates became factored into higher prices, the difficulty posed in building equity increased the hurdle to get into the market.
If home-ownership is the objective, the US provides a salutary lesson on the risks of over-promoting it. Between 1996 and 2006, a mix of policies encouraged young households into the market lifting the ownership rate from 62 per cent to 68 per cent before the GFC saw the housing market crash and the defaults and shattered dreams saw the ownership rate crash back to 62 per cent.(22)
In the period since the mid-1990s the most significant influence on the housing market has been the very significant decline in interest rates. A structural decline in interest rates delivers a structural (or one-off) rise in prices. It would be a mistake for buyers to assume that this period is any guide to the future. It should also be noted that the resources boom has played a significant role with its impact on demand. When the Sydney market turned down in 2004, it highlighted that markets can overshoot but, the downside was very much cushioned by the positive demand shock from the resources boom.
Looking beyond the cycle, the history of the past 50 years tells us globally that supply constrained and amenity rich cities, to which Australian cities very much fit the bill, have experienced much more significant rises and this is generating debate and angst in most countries. In terms of the housing affordability debate, in 2016 a US White House report was issued with a quote from then-President Barack Obama capturing the “yes we can” flavour of that report:
“We can work together to break down the rules that stand in the way of building new housing and that keep families from moving to growing, dynamic cities.”
About the author:
Dr Nigel Stapledon
Nigel has a PhD in Economics from UNSW and a Bachelor of Economics with Honours from the University of Adelaide. He started his career in Canberra, where he worked in the Commonwealth Treasury, following this he worked at Westpac where he was Chief Economist. Nigel has been at UNSW Business School since 2003 where he completed a PhD on the long-run history of house prices in Australia. Nigel is a regular commentator in the media on macro-economics and housing., contact Nigel on 02 9221 5211 or firstname.lastname@example.org