Skip to Content

Reconsidering the Sacred Cow of Housing: Policies for an Ageing Australia

Chairman's speech

Peter Harris gave an address at the Curtin University Public Policy Forum 2016 on 18 February 2016 in Perth.

Download the speech

Read the speech

I don't know about you but I find breakfast meetings to be the very worst idea civilised society has thrown at itself in the last couple of generations.

And now I'm going to add to the burden of this early start by talking about demography and age pensions and the intersection of them with that sacred cow of Australian political and social existence, the family home.

I can only say that the issue is worth your time as, unlike many other elements of public policy, it's going to affect almost every person in this room at quite a practical level: we're all going to get old, we all have stake in quality of life in retirement and most of us own a home.

So we can be both high-minded and self-interested here this morning.

And indeed that's the context nationally for such a discussion: the inevitable clash between the legitimate self-interest of individuals to make what are very personal decisions about their retirement; and the need for design of public policies affecting those decisions to offer a welfare improvement across society, in an era that will be quite different from the period in which many were introduced.

In the last three years, the Productivity Commission has published self-generated research work1 focusing on the implications for public policy of imminent demographic change for three key policy areas: workforce impacts from ageing, retirement incomes and housing policies.

Together they cover:

  • the broad sweep of the large demographic shift we face in the next forty years
  • how current policy envisages us maintaining incomes and assets, while absorbing a large addition to our non-working population.

In addition, in 2011 the Commission comprehensively reviewed the aged care sector in a report Caring for Older Australians.

Collectively, these reports provide the basic foundation for my remarks today.

Internationally, there has been intense interest in demographic effects interacting with growth and employment. Japan and Germany are good examples, entering or in this demographic shift already; China has significant social policy needs created by its one child policy and increasing longevity; and most recently the Bank of England has looked at the effect of ageing in the UK on interest rate policy.

Locally, other entities - Commonwealth Treasury, the Grattan Institute, academics and most recently the CSIRO - have also contributed.

All this analysis may make it seem like we have solved the consequent policy conundrums.

But we haven't. Consistent with the current preferred public policy stance of only responding when faced with a crisis, our awareness may be high but our preparedness is low.

Substantially, that's because the question of policy in relation to retirement is highly prone to populist thinking, making leaders very wary.

There should be a way to get past this wariness. Today, I would like to suggest a pathway, drawing on our three reports.

With the breadth of commentary and analysis available to them, leaders in Australia should be pretty well informed by now on the following inevitable set of policy shifts:

  • we will have a big rise, over an extended period — much longer than a single generation - in relatively healthy but retired Australians, and they will disproportionately outweigh the addition to the population of younger Australians, at a net cost to participation in the workforce
  • the Age Pension will continue to be accessed by most of these retired Australians, notwithstanding any of today's mooted alterations to retirement incomes policy
  • the consequent size of the fiscal increase for the Age Pension is quite notable, at roughly 1% of GDP; but bigger increases through that same period will occur in Aged Care (1.8% of GDP) and in Health (2.9% of GDP)
  • of these latter two, Aged Care is clearly driven by ageing, while health costs are driven by more than just demographic change
  • moreover, quality survey work and longitudinal analysis found in both our housing policy and superannuation research reports confirms a loose perception of policy makers: older Australians are strongly inclined to save rather than spend in retirement, almost regardless of circumstances, and despite substantial increasing housing wealth
  • and research also disproves a similar loose perception : there is no evidence for a strategy by retirees to deliberately blow the super, cushioned by access to the age pension
  • finally, even in very old age, most Australians remain net savers - as the wealth embedded in their housing asset persistently but invisibly increases, even as their personal health or aged care needs also grow.

At the personal level, there's no reason to be critical of an inherent tendency to save in old age.

But poorly-constructed policies should not exacerbate such tendencies. Nor should they pander to populist commentary.

There is good evidence that they do.

The powerful housing policy nudges that most of us have received throughout adult life — the status given at all levels of government to the acquisition of a primary residence; the clear tax advantage it is offered as an investment; and the substantial exemptions it receives when qualifying for the aged pension or aged care — are collectively skewing the judgments of individual older Australians to never consider the housing asset as a means of improving standards of living in old age.

This isn't perhaps a conscious aim of policy, but it is certainly where we have arrived at today. That is surely undesirable for individuals, and for the country.

The rest of this speech explains that view.

Whenever any aspect of housing policy is critiqued, the voice of populism quickly guides the media debate. Housing is that most sacred of cows.

This response does not serve the interests of older Australians.

It may simply reinforce fears of rampant reform, even though history would suggest no such thing has occurred in decades.

And fear is a serious political weapon.

But worse it prevents any thoughtful exchange on how to create, without putting individuals at risk, a mechanism by which older home-owners can retain their home and yet also draw on it for better living in their later years.

This should become the primary objective of considered commentary: how a retiree can learn to milk the sacred cow, rather than consume it; or worse still lose it.

Until this housing belief structure — created over generations - is altered, reforms in any part of retirement income and welfare swim against an almost impossible tide.

Shifting beliefs is hard work. More on that a little later.

Accompanying that private interest, we should have a clear view of the public interest.

A fiscal argument is often proposed.

While there's nothing undesirable with seeking to improve our fiscal circumstances as a consequence of also improving the welfare of individuals, it should be just that: a consequence, not an objective. The fiscal risk to future taxpayers — who, from this point on will be a diminishing proportion of the population - is well-demonstrated, including in our reports.

But the better course is to define the public interest as objectives that align with the private interest. These preferably would be:

  • that with a different mindset and recognition of the asset value available to them, the pressure to pay for new or better aged care or health services on future taxpayers will diminish
  • that relatively higher levels of consumption are available to retirees than would otherwise be the case.

On this latter point, relatively little attention has been paid to the potential for change in patterns of consumption in Australia over the medium term as the population ages.

Two facts stand out: older generations consume more publicly-provided services, where productivity is already known to be low and efficiency incentives are not strong; and their demand for privately-provided goods and services diminishes.

Forecasting the actual impact on consumption beyond those general direction type comments is likely to prove complicated. Many other factors will affect consumption in the economy at any single point in time in the future.

Nevertheless, the broad context is clear — there are going to be a lot more of us older, wealthier and (mostly) healthier than our predecessor generations:

  • already one-quarter of us are aged over 55; inexorably, over the next thirty years we will more than double the number of persons aged over 65, to around 7 million by the middle of the century
  • in 1909, the original age pension recipients generally had spent near enough to 75% of their life after the age of 15 in full-time work; for the baby boomers, that figure has fallen to about 60%; and for the generation today in high school, that figure will fall to around 50%, mostly based on significant health gains, including in retirement
  • in 2012, there was roughly 1 person aged over 100 for every 100 babies born; by 2060, assuming nothing more than current improvements to longevity, there will be 25 such centenarians.

Yet the special incentives we apply as housing policy in this country were perhaps most appropriate for a quite different Australia: the Australia of the 1950s and 60s. We sought to encourage home ownership for a society then dominated by young couples — often with only one at work — and a big jump in births; while at the same time supporting a generation that had endured world wars and the Great Depression and so entered retirement with few assets and often poor health.

These policies have been successful in creating a deep attachment to home ownership, and in ensuring the majority of Australians now enter retirement with a very valuable asset.

Today, housing is the largest asset by value for households aged over 24.

About 70% of families own or are in the process of buying a home, a level stable for some decades.

Older Australians unsurprisingly have the highest value holdings of all.

Home equity per household was worth $480,000 for Australians aged 65-74 in 2011-12.

By comparison, after twenty years of saving, the median superannuation balance at age 60 for the comparable year was worth $95,000.

Yet despite its size, the housing asset is the very last to be drawn upon, if at all, almost regardless of circumstance.

It's a very big — and growing — asset to remain untouched during a long retirement.

In the last decade, household wealth on entering retirement has lifted significantly, about 30% on average, driven mostly by the housing asset.

As such, the lock-in of this level of wealth by the fastest-growing population cohort in a form that is intended to remain untouched for twenty years or more will, if unaltered, shift the obligation for maintaining consumption strongly towards younger people in the coming decades. At the same time, their tax obligations too may rise, as dependency ratios move against them.

We can say what the intentions of older Australians are because we have recently tested them.

One of the three reports in this series - Housing Decisions of Older Australians report, published last December - undertook a professionally-designed survey of older Australians. Being consistent critics of poor quality survey work at the PC, we made significant efforts to ensure that the results were reliable.

The most common answer across all the older age groups questioned in relation to potentially drawing down on housing was 'not under any circumstances'.

Offering some hope for future reform, some respondents also indicated it was possible that they may draw down on the housing asset for health or aged care expenditures.

But across the questionnaire, the clearest preference was to leave the house to their heirs as a bequest, with 88% support.

Further validation of individuals' desire to preserve the housing asset can be found in studies drawing on both the University of Melbourne's HILDA data base and on Centrelink data.

A particularly interesting study2 covering Centrelink data from 1999 until 2007 showed that at death residual wealth is equal to 90% of the amount recorded at first observation.

And that the even lowest income groups — where consumption is traditionally dominant - were net savers from a very early point in retirement.

Compounding the 'preserve at all costs' approach to the principal residence is invisibility.

There is very little by way of a trigger or incentive for older Australians to make themselves aware of the value of the asset and its ability to re-shape their lives in retirement.

A decision to down-size might be a trigger, but it is far less common than is often imagined.

The means test for the age pension might also be such a trigger, but by conscious design — i.e. exempting the home from consideration regardless of value - it is not. And the fearful complexity of dealing with government regulation — e.g. simply trying to assess whether this is a viable option in any one person's particular circumstances — also nudges any reasonable person only in one direction.

The one new factor that may force reconsideration is the catalytic effect of what might for effect be characterised as the arrival of 7 million 70 year-olds.

In years to come, this sizable group may ultimately see means tests incorporate the housing asset in some way. Future taxpayers may have no choice.

The idea of addressing big, medium-term shifts of this nature is pretty uncommon today. Kicking the can down the road (as they say in the US) is the common response.

But here we are dealing with policies that people rely upon to make serious long-term life choices. They deserve better than to wait til the inevitable imposes (possibly) quite sudden change.

Since I've just mentioned it, I should briefly outline what our reports have to say about making an adjustment to the means test for the Age Pension.

The detail is in the documents, the bottom line is this: today, complete removal of the exemption of the primary residence from the means test for the Age Pension would see a reduction in the proportion of older Australians receiving the Age Pension of 11 percentage points; and nearly 50% would incur a cut in their benefit.

That's big.

But — even if it had no other policy implications — such a change obviously runs strongly counter to the encouragement towards retention and enhancement of the principal residence that governments and society has so deeply embedded in us.

Perhaps, you might ask, there is a less bold option?

We looked at that too. Different parties will have different interpretations of the result. I'll leave you to determine yours.

We tested the concept of including in an assets test just the value of the home above the median value of homes in Australia ($440, 000) would see the proportion of older Australians who qualified for an age pension reduced by 2.5 percentage points, and reduce benefits for another 8% of recipients.

The generosity of scope allowed for owning a range of other assets - including financial ones - means in effect that adding a solid dollop of cash (i.e. the excess above $440,000) to the total asset test doesn't put many people over the threshold.

Of course, even small changes can have big fiscal numbers attached, so this isn't the end of the story.

In fact, the story is only complete when a change like this is examined in the context of its overall impact on incomes in retirement.

For that, the implications for availability of superannuation and aged care would also need to be examined.

One final point will illuminate this need to look at all retirement assets and income sources together if policy is to have its greatest chance to work for both the public and the private interest. It concerns superannuation balances.

Superannuation balances in future will definitely be higher than the median value ($95,000) I referred to earlier.

Could these be a better substitute for using the housing asset in retirement to improve quality of living?

The answer of course is of course yes, they might.

But the trend to acquire homes later in life and have unpaid mortgages on them at or near retirement is also now well-established, and will act as a counter-balance.

In 2011, 32% of Australians aged 55 to 64 were still paying off mortgages, double the rate of a decade previously. Some housing policy commentators suggest the numbers are even higher today.

And as our second report in this series established, when Australians today access their super in lump sum form, the single largest reason they do that is to pay off the mortgage, with 25% of retirees using it this way.

In the notes published with the invitation to speak today, there was a challenge posed : how to put the sacred cow to work, voluntarily.

In order to expect people to change their attitudes voluntarily, one initial step must be to provide better knowledge for retirees on the options available and the value accumulated.

There are good examples both locally and overseas of how, with better information and the ability to make better judgments, individuals change behaviour when conventional wisdom says change is unlikely.

Perth itself has a good example. Facing significant persistent step-changes in rainfall starting in the seventies and repeated in the nineties, local water authorities launched an initiative to put on the face of a water bill the water use of a household versus the average of equivalent households. This change alone — no price shift, just better information - yielded significant water savings for no additional cost.

Use of these kind of comparisons is now quite common across other utilities such as gas and electricity, here and overseas.

Another example: in the US, re-designing a letter to veterans to inform them that they had earned - rather than simply qualified for - education and career counselling benefits resulted in a 9% increase in take-up of the scheme.

Just the alteration of language can help provide a statistically significant variation in human behaviour.

In the UK, tax collections were lifted by targeted personal information campaigns telling people that their neighbours have already paid their tax.

Personal text messages reminding people to pay fines on time cut numbers of bailiff interventions by 150,000.

And a failing program offering subsidies for home ceiling insulation was turned around by offering to remove rubbish in lofts at cost price, a nudge towards the real reason people had failed to insulate their houses.

Giving people better information at a time when they have good reason to consider it and an incentive to do so is the crucial common element to these successes.

If we adopted a similar approach to alter perceptions of the family home during retirement, we might require that at the time of the first application for aged care benefits (simply accessing programs to obtain in-home support) a free valuation of the family home was offered.

There would be no obligation to use the information, just to gather it.

This could be complemented by an offer of free (accredited) financial advice as one of the optional benefits that comes with accessing in-home care.

Superannuation funds might, for example, be the sort of trusted party that could assist here.

A second nudge, in the parlance.

As with many behavioural economics concepts, the sheer novelty of this kind of approach makes it easy to treat with derision.

But most of these examples that I mentioned a moment ago had critics who doubted their ability to alter behaviour. Sometimes even after they achieved their notable successes.

Unusual problems like this — how to unwind the effects of risk perceptions and public policy designed for a different era — do require unusual thinking.

To close, let me note that the work published across the three papers I've mentioned today could be used to support a much wider, comprehensive review of sustainability of income during the retired years of all Australians.

We suggest a comprehensive review because it is probably the only way to generate the necessary public understanding that change may be needed; and what that might look like.

It might even offset populism.

The demographic changes we are entering into now are going to last into the long term, and will have an array of effects. We may not have the worst outlook in the world for ageing — some countries' populations are likely to contract — but that is hardly an argument not to pursue better policies when they are reasonably evident.

I hope I have also made it obvious that simply tugging on any one of the pieces of string so oddly entwined in this area is not likely to make the knots untangle.

The purpose of a review should be to ensure that no Australian lacks decent quality of life in old age, in a context of significant demographic shift.

A review of this kind can take the long view, and that is deeply desirable when the problem itself is easily put off. The political cycle is three years and the Forward Estimates are four years. Policy makers generally tend to be dismissive of change that might apply outside those time frames.

But setting a review now with intent to make change that applied later than the normal cycle is likely to generate more acceptance, and offer more scope for behavioural change and preparation, than any other approach.


1. An Ageing Australia; Superannuation Policy for Post-Retirement; and Housing Decisions of Older Australians. Return to footnote 1 2. Wu, 2015: see Housing Decisions of Older Australians. Return to footnote 2