The recent article “Defining Poverty” (On Line Opinion) by my namesake Peter Saunders (hereafter PS) from the Centre for Independent Studies (CIS) contains a mix of sensible reflections, mis-reading of the evidence and muddle-headed thinking, neatly wrapped in the usual CIS package of pro-market ideology.
The guts of his argument can be reduced to two propositions: first, claims that poverty in Australia “is a huge and growing problem” are based on dubious evidence and greatly exaggerated; second, past attempts to reduce poverty by increasing welfare spending have failed and have probably been counter-productive.
In relation to the first issue, it is important to note the claim “poverty is a huge and growing problem” is attributed to unnamed “activists and academics” without further attribution - although it is clear from the article that I am among “the guilty”, even though I have never made such a claim. In fact, in his article there is no discussion of whether poverty has grown at all or not. Most of the discussion focuses on the claim poverty has been greatly exaggerated in studies that have attempted to measure it. (A task that PS or his CIS colleagues do not appear to have attempted.)
The only time that the change in “poverty” (PS’s quote marks, not mine) is referred to is in the discussion about changes in income inequality, where he notes correctly that those with the lowest income brackets experienced slightly below-average real income gains between 1995 and 2003. Yet this is an example of confusing poverty with inequality - a mistake that the author regularly condemns others for committing.
In fact, most of his argument focusses on demonstrating that the level of poverty is much lower than others have suggested. This includes by implication, Minister for Family and Community Services Senator, Kay Patterson, who earlier this year released findings from the third wave of the Household, Income and Labour Dynamics in Australia Survey (HILDA), which shows that in 2002-03, about 12 per cent of the population had incomes less than half of the median - a common measure of economic poverty. Those same findings, summarised in a recent paper by Wooden and Headey (2005), also indicated the numbers of those living in poverty had fallen from over 14 per cent in 2000-01 to about 12 per cent in 2002-03 putting to rest the claim that relative poverty is always increasing (another PS furphy).
PS supports his claim that poverty is less than half of what HILDA and other research indicate, by saying there are problems with the income data used to estimate poverty (though not with the HILDA data used in the studies just referred to). These problems are a cause of considerable concern among poverty researchers, and have been the subject of an extensive research collaboration between myself and colleagues at the Social Policy Research Centre (SPRC) and the Australian Bureau of Statistics (ABS), to identify the problems and seek solutions to them.
Incidentally, while PS bemoans the lack of attention given to this issue in my recent book The Poverty Wars, the main motivation for the book and its principal message, concerns the need to move beyond measures that conceive of poverty simply as a lack of income, partly as a way to avoid the limitations of the income statistics.
In any case, he is quite happy to use these “flawed” (my quotes this time!) income statistics to show those on a low income have benefited from economic growth over the last decade (as noted above). Yet one has to wonder how much of this “trickle down effect” is itself a statistical illusion, reflecting the increased effort made by ABS to correct the under-reporting of income in the most recent years but not (yet) in the earlier ones. We at SPRC have been urging ABS to make the corrections in all years so we can get an accurate picture of changes over time, but until this is done we need to be cautious when describing what has happened to the Australian income distribution over recent decades.
When PS discusses the extent of poverty, he suggests the overall rate is “no more than 5 per cent of the population, probably less”. This figure appears to be based on research conducted by the Department of Family and Community Services (FaCS), which shows that only 3.1 per cent of the population experienced hardship in 1998-99 (Bray, 2001). Further, data from the first three waves of HILDA indicate that only about 4 per cent of the population had incomes below the half-median cut-off in all three years between 2000-01 and 2002-03. I know of no one other than PS who has suggested either of these measures should replace those currently used to estimate poverty (and the only merit of the latter, three-year measure appears to be that it produces a figure close to that based on hardship). That aside, the hardship estimates do warrant a closer look.
They were derived from a series of 13 financial stress questions included for the first time in the 1998-99 Household Expenditure Survey (conducted by ABS). The hardship measure, however, covers only four of these indicators and the cited rate of 3.1 per cent refers to those who experienced two or more of the following four conditions: had gone without a meal; unable to heat the home; sought help from a welfare organisation; had to pawn or sell something (all because of a lack of money). This is a measure of chronic poverty that captures only those in the direst situation.
If we look instead at those who experience one or more of the four hardships, the rate increases from 3.1 per cent to 8.2 per cent. If we go further and define poverty as involving a high level of financial stress associated with experiencing at least five out of the original list of 13 indicators, then the ABS has shown the rate rises even further, to 12.6 per cent - close to the half-median income measure discussed earlier (McColl, Pietsch and Gatenby, 2001).
We have no basis for choosing between these different measures, since this involves making a judgment about how much hardship we are prepared to tolerate as a society. The judgment made by PS produces the very lowest estimate, so that his 5 per cent poverty rate figure is an absolute lower-bound estimate, not an upper-bound as he implies.