City Futures Blog

News and research in housing and urban policy, from Australia’s leading urban policy research centre.

City Futures Blog random header image

Where is housing “affordable” in Sydney?

Posted by on March 24th, 2016 · Affordability

By Ryan van den Nouwelant, Hal Pawson and Laurence Troy

The affordability of housing in a particular neighbourhood is a tricky thing to reduce to a single metric. One option is to try and account for the complexity of dwelling composition and quality over space and time through data-intensive indices. But, more often, there is a tendency to reduce all housing market activity to the prices at the median. This is a limiting approach.

The median is only one point on the housing spectrum (the midpoint, to be precise!). It can be a reasonable indicator of activity elsewhere on the spectrum. But it is possible and preferable to directly examine these other points instead. Rather than asking simply whether the median dwelling is ‘affordable’ in a given place, it is possible to look at what proportion of dwellings is affordable. This is the approach we explore here.

Our starting point is a gross household income of $100,000pa in 2015. This is taken as a reasonable approximation for the midpoint of incomes for aspiring home owners in Sydney, and so is higher than the midpoint of all households by income.[1] For this income level, we can calculate the purchase price that could be afforded, based on certain assumptions.[2] This gives us an ‘affordable’ threshold purchase price of $513,180.

So far, this is nothing new. But instead of asking where this threshold value exceeds the median house price, we can ask (and show in the map below): in relation to this threshold price, what proportion of 2015 sales (using LPI notice of sales data) is affordable to our typical aspiring home owner in each postcode area?

This is a useful measure because we can model the geography of affordability for any household income, as well as chart affordability changes over time for households at that income. For instance, the map below shows Sydney’s affordability geography for a comparable household income (adjusted to $71,700 with NSW wage inflation) for 2005 sales. This map shows something closer to what you would expect to see for a mid-range purchaser: while there are many postcodes where the household could only afford a small proportion of the sales, there are also a number of areas where quite a high proportion would be affordable.

One of the most striking changes between 2005 and 2015 is in the heartland of Western Sydney. While swathes of the more exclusive suburbs have been largely unaffordable for people on this income for over a decade, affordable properties in postcodes beyond Parramatta and Liverpool have also become increasingly rare.

Around Blacktown (postcode 2148), for example, between 60-80% (actually 70-75%) of 2005 sales were ‘affordable’ to a household on $71,700. By 2015, however, between 20-40% (actually 25-30%) of sales in Blacktown were affordable to the equivalent income household. This 45 percentage point decrease, shown in the map below, is typical of Western Sydney. It also aligns with the findings of our recent and more in-depth AHURI analysis which identified similar trends in Sydney, Melbourne and Brisbane: in all three, house prices in low value ‘disadvantaged suburbs’ tended to rise at rates above city-wide norms during the 2001-2011 period.

Another story the maps tell is the loss of the last pockets of affordability in the inner suburbs. In 2005, there were still seven (of 81) postcodes within 10km of Sydney CBD in which at least 40% of sales were ‘affordable’ for this income. In 2015 there were none. Around Summer Hill (postcode 2130), for example, 40-60% (actually 40-45%) of 2005 sales were affordable. By 2015, under 20% (actually 5-10%) of sales were within reach for the equivalent household. This cheapest end of the house price spectrum is much more likely to be small apartments, rather than family houses. This example also shows a deterioration in affordability that would be lost when only considering the median, since even in 2005 under 50% of sales were affordable.

Combined, these two observations mean very few areas remain in Sydney where a proportionate number of sales are affordable to households at this income. A final observation from the maps is that the last sliver of affordability within 20km of Sydney is along the Bankstown train line, through Lakemba and surrounding neighbourhoods. This is notable because the area has been flagged for significant infrastructure investment in the near future. The commuter rail is being converted to metro rail, and there is the presumption of transit-oriented development following it.

This new development will inevitably come with the premium of new construction attached. And the associated urban renewal, improved services and better amenity will likely push up the relative value of existing homes in the area too. Unchecked, this trend could lead to the displacement of existing households, and a classic model of gentrification: whether directly because of insecure tenure, or indirectly as incumbents are marginalised by a changing community. Efforts to moderate or, at least, plan for this change seems lacking from the current planning emphasis on building heights, rail capacity and development viability.

All of this means that, for this income, home ownership (certainly for anything more than a small apartment) within 20km of Sydney’s CBD is likely to almost entirely vanish. Those on this income will be either compromising on what they purchase – the residual cheap properties might still be below the threshold, but aren’t likely to be more than a small apartment – or be left in the rental market or living with their parents. So while this income might have enabled access to home ownership a decade ago, it doesn’t now.

There are limits to the housing affordability analysis presented here, but it is a simple metric that offers a greater degree of insight than measures of median affordability. Plus, we can adjust the boundaries – show suburbs instead of postcodes, or vary the ranges – show a finer grain than the 20 percentage points used here. By drawing on other data sources, it is also possible to cut the sales by dwelling type and size. Or examine advertised rents, or lodged rental bonds, in a similar vein. We will continue to explore its potential as a metric of affordability.



[1] In 2013/14, the NSW median gross household income was $81,800. It’s outside the scope of this article, but the ‘typical’ or median household income of home buyers is even more problematic than the typical house price. Most measures will not account for all those not looking to buy a home because they either don’t need to, don’t want to, or don’t have realistic potential to. The figure of $100,000 translates to two earners making around $1,000 per week, which is above the 2015 minimum weekly wage of $640 but below 2015 average weekly earnings of $1,500.

[2] The assumptions are (a) no more than 30% of income goes towards mortgage repayments, (b) the interest rate is 5.0%pa and the loan term is 25 years, and (c) the household has sufficient savings to cover stamp duty and conveyancing, as well as a 20% deposit. These assumptions try to simulate ‘realistic’ conditions, but the savings needed for deposit and stamp duty stand out as atypically high. This is one of the symptoms of low interest rates: a given wage can service a higher loan principal, and so push house prices up. But the period of dropping interest rates in Australia was also accompanied by looser controls on high LVR mortgages (the so-called ‘no doc’ and ‘low doc’ loans), so the higher dwelling price wasn’t translating to a higher deposit in many cases. However, as regulators and lenders have reined in this risky end of the mortgage market, the deposit is now often mentioned as the limiting factor for many potential buyers, not servicing the loan. It is also why questions of stagnating inter-generational social mobility are being asked: is having home-owning parents the only route into home ownership for the next generation?

One Comment so far ↓

  • Janice Hall

    It is great to see it put so clearly. There is great need for concern for future citizens. Are Mr Turnbull and Mr Baird happy for them to be living in ghettos????

Leave a Comment