The latest HILDA report is out! The HILDA survey is an extremely valuable resource – it asks a large sample of people a whole bunch of questions about income, family life, and other things, and tracks respondents over time. We learn things from HILDA that we can’t learn from any other Australian data source.

You might have seen coverage of the latest HILDA statistical report on the front of The Australian today. The article includes a version of this chart from the HILDA report, which shows two different measures of the poverty rate in Australia:

Screen Shot 2014-06-16 at 9.32.52 am

‘Relative poverty’ here means the share of the population that lives in a household that has an income of less than half the median income, adjusted for differences in household size. But what does ‘absolute poverty’ mean?

The Oz described  absolute poverty as follows:

…the ability to provide the minimum requirements of food, clothing and shelter…

But that’s not how it is defined in the HILDA report. The report (p.29) defines it as follows:

An absolute poverty line is an income poverty threshold which has its real value held constant over time rather than adjusted for changes in average living standards. It is ‘absolute’ in the sense that the purchasing power of the poverty line—the basket of goods and services that it can purchase—remains fixed over time. The level at which an absolute poverty line is set may nonetheless be based on the level of a relative poverty line obtained at a particular point in time, for example the beginning of the time period under study.

In the chart above, the ‘absolute poverty’ line is the 2001 relative poverty line, frozen in real terms. It’s not based on some assessment of the income required in each year to afford the basic items necessary for subsistence in Australia. It’s just half the median income in 2001, indexed to the CPI. So ‘absolute poverty’ in the chart above tells us that around 6% of the population in 2011 had an income that was lower, in real terms, than 50% of the 2001 median income. This is an interesting and useful metric, but it doesn’t necessarily reflect “the ability  to provide the minimum requirements of food, clothing and shelter.”

Some countries (notably the US) have domestic poverty lines that are based on an assessment in the past of the minimum income required to meet some basic standard – the US poverty line was fixed in 1963 and has been indexed to inflation since then. The ‘absolute poverty’ line in the HILDA report isn’t like this.

Critics of relative poverty measures sometimes like to point out that relatively poverty thresholds are arbitrary. This is true. Why set the poverty line at 50% of median income, instead of 51% or 49% or 60% or 40%? But the ‘absolute poverty’ measure used here does not remove this arbitrariness. Instead, it is arguably more arbitrary, as the choice of base year (2001) is added to the mix. 2001 is used solely because that was the first year of the HILDA survey; there’s nothing in particular about living standards at that time that makes that year any more or less suitable as the base for an absolute poverty measure.

Poverty, particularly in advanced economies, can be a nebulous and difficult-to-measure concept. Reasonable people differ on how to define it, and measuring it is inevitably arbitrary at the margin and requires judgement. While it is useful that the HILDA report provides these two measures of poverty alongside each other, you should not imagine that these contentious issues and arbitrariness are avoided by using the absolute poverty measure presented above.