The conventional wisdom is that Australia has an especially high rate of poverty among the elderly. Advocates of increasing the Age Pension point to OECD research which shows that in 2014, 26 per cent of Australians aged 65 and over suffered income poverty, compared to 13 per cent across all OECD countries.
But the conventional wisdom is wrong. Traditional estimates of old-age poverty are deeply flawed. We need a better way to measure poverty among older Australians.
In this post, we show why the OECD’s approach to poverty rates for older Australians misses the mark, and present our alternative that accounts for high-rates of home ownership among the elderly.
The OECD’s measure is deeply flawed
As the OECD explains, its definition of poverty is relative – the proportion of households with disposable incomes (meaning after income tax is paid and income support payments are received) that are less than half of the median disposable income of all Australians.
But there are a number of issues with the OECD measure. Small changes in reality produce apparently very different outcomes, calculation of the benchmark rests on arbitrary definitions of equivalisation, it does not take into account drawdowns on savings outside superannuation, and it does not adequately account for housing costs.
The OECD poverty rate measure for older Australians is volatile from year to year, even with very small changes, because the full Age Pension is close to the OECD benchmark, and so a large number of Australian retirees cluster close to the benchmark. For example, a recent report by the Australian Council of Social Service (ACOSS) found that old-age poverty in Australia apparently fell sharply from 22 per cent in 2011 to 13 per cent in 2016. But the big apparent shift merely reflected the maximum rate of the Age Pension (including related supplements) oscillating around the benchmark of 50 per cent of median incomes.
This clustering close to the benchmark also means that outcomes on the measure depend a lot on somewhat arbitrary definitions. For example, the poverty rate among over-65s varies from 12 per cent to 23 per cent depending on how households with different family sizes are compared. According to the ABS-preferred definition of equivalisation, previously used by the OECD, households are ‘equivalent’ if they expend 0.5 times more for every extra adult and 0.3 times more for every child under 15 than a single household. According to the new OECD definition, households are ‘equivalent’ if a household of n members expends the square root of n times as much as a single household. The choice of benchmark relative to median incomes is also arbitrary, but the most commonly used benchmark is 50 per cent of equivalised median disposable income.
Traditional old-age poverty metrics ignore housing costs
Most importantly, a relative poverty measure based on disposable incomes tends to overstate poverty in old age because it ignores the differences in housing costs due to high levels of home ownership among older Australians. Four in five Australian households over the age of 65 own their own homes. Even among the lowest income quintile of seniors, home-ownership rates are above 70 per cent.
Home ownership provides them with big benefits: they have somewhere to live without paying rent, and they are insulated from rising housing costs. The benefits that a house provides to its owner-occupier – which economists call imputed rents – are worth more than $23,000 a year to the average household aged 65 or over, roughly the same value as the maximum-rate Age Pension.
Grattan Institute’s new estimates
Our latest estimates of poverty rates confirm that fewer retirees are in poverty than previously thought. After taking account of the benefits of homeownership, by treating imputed rents as disposable income, the percentage of Australians over 65 in poverty under the OECD’s square root scale more than halves – from around 20 per cent to 9 per cent.
Under this new metric, Australian retiree poverty rates are more in line with the other comparable OECD countries, albeit still on the higher end.
And even though changing the definition of poverty has a material impact on old-age poverty, adjusting for housing costs still has the same effect under different definitions. Using the ABS poverty definition, adjusting for housing costs also significantly reduces Australian old-age poverty – from around 13 per cent to 8 per cent.
Poverty among over-65s in Australia is still higher than in many other OECD countries after accounting for housing costs, but only marginally so. The 8 per cent of older Australians in poverty after housing are likely to be maximum-rate pensioners in the private rental market.
Other evidence suggests most pensioners aren’t struggling unless they rent
Our results accord with other evidence that most Australian retirees are keeping their heads above water. Retirees today are more likely than working-age households to say they feel financially comfortable. Across the income distribution, people typically have enough money to sustain the same, or a higher, living standard in retirement as when working. Many retirees are net savers, and current retirees often leave a legacy almost as large as their nest egg on the day they retired.
Of course, there are older Australians living in poverty. The evidence suggests pensioners who are really struggling predominately don’t own their own homes.
About 14 per cent of pensioners who don’t own their home say they have suffered financial hardship – they skipped meals, did not heat their home, failed to pay utility bills on time, or failed to pay their car registration on time. The comparable figure for home-owning pensioners is only 4 per cent.
Rental stress among pensioners in the private rental market has worsened for a number of reasons.
First, Commonwealth Rent Assistance, which provides financial support to low-income renters, is indexed to CPI. But rents have been growing faster than CPI for a long time. Between June 2003 and June 2016, CPI increased by about 38 per cent, while average rents increased by about 62 per cent.
Second, rents paid by low-income earners grew faster than average rents and low-income households are spending more of their incomes on rent.
Third, the stock of lower-rent social housing has not kept pace with population growth. In the past, more than half of retirees who rented did so from housing authorities. In recent years that proportion has fallen to less than 40 per cent. Public housing provides a much greater average level of assistance than Rent Assistance.
Many working-age Australians are struggling more
Pensioners are not the only group likely to face hardship in Australia today. In fact, they are on the whole less likely to suffer financial stress than working-age Australians.
About 8 per cent of working-age homeowners with no welfare payments report experiencing financial stress, a rate about two times higher than pensioners who own their own homes. And 20 per cent of working-age renters with no welfare say they suffer financial stress, higher than the rate reported by renting pensioners.
Neither of these groups is struggling, however, compared to the levels of financial stress reported by the most disadvantaged people in Australia today – working-age households surviving on income support payments such as Newstart. More than 40 per cent of working-age Australians who receive income support payments and don’t own their own home – and 15 per cent of homeowners receiving income support – report experiencing financial stress.
It is not surprising to find higher rates of financial stress among working-age Australians than among pensioners. Pension payments are indexed to grow with full-time wages, whereas allowance payments only grow with CPI. Over the past two decades the maximum rate of the Age Pension has consistently risen faster than both minimum and average wages, and much faster than income support payments to working-age Australians, such as Newstart. For singles the maximum rate of Newstart has fallen from 89 per cent of the base Age Pension in 2000 to just 66 per cent today.
Any moves to alleviate poverty in Australia must start by boosting payments to working-age income-support recipients, not pensioners.
The policy priority for retirees should be to boost the maximum rate of Commonwealth Rent Assistance by 40 per cent, or roughly $1,400 a year for singles or $1300 for couples. Such an increase would cost $1.2 billion a year, but it would do much more to reduce financial hardship in old age than lifting the Age Pension.
And the next Australian Government should take a close look at lifting Newstart too.