Wednesday, June 10, 2015

Three reasons apartments may not deliver on density


A key thrust of higher density development plans for our various inner city areas was that this would lead to a significant increase in population densities. But what if the current wave of inner city apartment development isn’t likely to deliver as many people as the plans presumed? Does this mean we need even more apartments to meet infill population targets? Or do we need to revise the population numbers altogether? Here are three factors I doubt were taken into account when these targets were first devised

Many of our prevailing metro wide planning schemes are revisions of similar schemes devised in the late 1990s and early 2000s. A key plank of these schemes was to prevent ongoing ‘sprawl’ (being outward suburban development) and encourage higher densities in established areas, particularly inner city areas. The plans came in for their share of criticism at the time. Tony Powell – an icon of Australian planning – described them as a “sad parade of failing capital city strategic plans” that in some cases were “superficial to the point of ridiculousness.” Professor Brendan Gleeson disdainfully described the wishful thinking of much plan making at the time as “faith-based planning.” 

A key focus of the ‘faith’ and also of the criticism was the belief that population growth could be contained in higher density housing, as opposed to continued expansion of suburban markets. 

Tony Powell put it bluntly in a 2007 lecture to planning students: “The proposition in the latest crop of metropolitan strategy plans that 50 percent or more of future housing development can be accommodated in existing suburban areas of the major cities is patently ridiculous.  These are simply unexamined and unreliable hypotheses, not strategies.” 

Fast forward to today and the sheer volume of apartments now being developed in inner city markets might suggest Powell was wrong: indeed, there is widespread talk of oversupply and bubbles in particular markets. But what if, despite this unprecedented pace of apartment development in inner city markets, the population numbers still won’t materialise? Could we actually have an apartment boom without the people to match?

There are three factors at work which I suspect no one at the time predicted:

  1. There are very few owner occupiers in the current wave of apartment development. If it were true as widely claimed that we are witnessing a permanent change in Australia’s housing preference in terms of units over houses, you would expect to see more owner occupiers active as purchasers. But the numbers in some cases are very small. Michael Matusik in Brisbane has looked at a number of newer projects and in many of the larger scaled developments, the percentage of owner occupiers is only 3%. That’s not a typo. Smaller developments show a higher proportion of owner occupiers but across the board, the figure is still only 10% of sales to owner occupiers. Two thirds of investors of inner city apartments are from interstate or overseas. 
  2. The apartments are typically small and getting smaller. Again based on Matusik’s research, ten years ago the proportion of one bedroom apartments being developed was just 10%. It is now more than three times that at 35%. And while three or more bedroom stock accounted for some 30% of stock ten years ago, that figure is now just 5%. Plus they’re shrinking: one bedders have fallen from 70 square metres to 50 square metres in that space of time. 
  3. Vacancy rates are increasingly unreliable. The most widely quoted vacancy rates are provided by various real estate institutes. These are not audits of vacant stock – they only reflect a survey of member agents: not all agents are members and not all members bother completing the survey. The problem here is that with new apartment projects, the available rental stock is unlikely to be included in real estate institute numbers if being marketed through various investment or project selling channels, or strata managers or others. No one knows what proportion of properties available for rent are not being captured in ‘official’ real estate institute data because of this, but you can simply peruse available rental listings online to see how many are being offered outside of ‘traditional’ real estate agency channels to get some idea. It’s significant. Plus of course, the ‘official’ figures can’t possibly capture the number of apartments that are kept unoccupied for whatever reason. So while ‘official’ vacancy rates for inner city areas might point to a 4% vacancy, the actual vacancy (including those not recorded in ‘official’ data) could easily be double that or higher in the apartment sector. 


This poses a dilemma for planners who equated apartment development with population growth. Back when there were typically more (and larger) three bedroom apartments being developed, and more of these being sold to owner occupiers, the residency ratios were higher. I used to use an average of 1.6 persons per apartment as a rough guide. 

But today, with so many more (and smaller) one and two bedroom apartments being designed for absentee investor appetites, the residency ratios must be smaller. A one bedroom apartment of 50 square metres, for example, will on average not have more than one person living in it. And that’s only if it’s occupied. Take into account vacancy figures, factor in what’s not included in those figures and factor in the number of apartments that are simply unoccupied on a permanent basis and not available for rent, and residency ratios must fall further again.  

If you take a hypothetical 250 apartment project, it may once have housed 300 to 400 people - if those apartments were mainly two and three or more bedrooms, and a high proportion of owner occupiers. But now, with changes in the type of stock and the nature of the market, that project may only house half as many people. No one really knows, which is my point: markets and planners seem to be relying on untested assumptions about the relationship of apartment construction with population growth. 

This low residency ratio could also explain a couple of things. Many shops and businesses associated with new developments, who anticipated high expenditure from residents, have been disappointed. Perhaps the expectation of how many people are actually living in the new projects were based on wrong assumptions?

And the profile of those actually in residence is another factor. One bedroom rental units will equate to a high proportion of students or lower income rental groups. They certainly won’t be consuming like a family of four. Hence the actual discretionary income available from these residents needs to be rethought.

Most interesting though is the proportion of urban population growth that will actually find itself housed in the current wave of development. If the residency numbers are in reality that much lower than planners first presumed, this could mean we need more apartments per 1000 of population to achieve identified targets, or we need to rethink the targets.

Either way, it’s a fascinating question with significant implications.