Academic journal article
By Hill, Trish
The Economic and Labour Relations Review : ELRR , Vol. 20, No. 1
The conventional measure used to assess relative economic disadvantage and poverty status is income. Conventional income poverty measures use household or income unit income levels as a proxy for the standard of living for all members in that household. The total income is then equivalised to take into account the size of households and economies of scale in larger households. All members of the household are allocated the same level of equivalised income and are thus considered to have equivalent standards of living. This approach has been criticised for assuming that all resources within households are pooled. Income measures have also been criticised for focusing on income level whilst ignoring the sources, circumstances, costs or time taken to earn or use the income. Feminist research has argued for a more complete depiction of differences in living standards or relative disadvantage, including the value of time use and a better description of the intrahousehold distribution of resources (eg. Aslaksen and Koren 1996).
This article proposes an alternative concept of 'full income', which constructs a broader measure of individuals' resources, including the value of time use. It compares the distribution of conventional income and full income to gain further insights into the extent of gender inequalities in living standards and identify vulnerable population subgroups on both the conventional income and full income measures.
The concept of full income was inspired by the work of Peter Travers and Sue Richardson (1993), who proposed it as a broad measure of material wellbeing. Their model of full income aimed to provide a broader and more complex picture of wellbeing than that provided by the current income of individuals or households. Their full income model of wellbeing included some of the major sources of material wellbeing: income, the value of time in leisure, the value of benefits from goods and services provided by government--called the social wage--and the value of goods and services received from assets owned, such as a car or house. They then sought to find 'sensible money values' for the goods and services that did not have a price in the market in order to construct a single monetary index of relative wellbeing. One advantage of this approach is that it showed whether disadvantage in one sphere such as income was counteracted by advantage in another sphere, such as ownership of assets. The theoretical approach used here to analyse wellbeing or living standards is located in their work, although it significantly differs from it in some respects. Its key aim is to outline a method for providing a more comprehensive picture of the value of each individual's total resources than that provided by income alone.
The analysis departs both from conventional income measures and from the approach of Travers and Richardson (1993), whilst remaining within the latter's conceptual framework. First, it uses the individual as the unit of analysis and does not assume equal sharing of income within the household. Second, it uses the Australian Bureau of Statistics (ABS) Time Use Survey to add the value of unpaid work performed by partners into each individual's bundle of resources. Third, it identifies much more precisely the amount of 'uncontaminated' leisure each individual has, rather than using the residual model of leisure that was employed in Travers' and Richardson's original study. Fourth, its estimate of the individual's benefit from the social wage is derived from the ABS Household Expenditure Survey. Its estimate of the value of services provided from assets is limited by data availability to the degree of home ownership.
Methods of including time use in measures of living standards are not without controversy. This article outlines one such approach and identifies its effect on our picture of who experiences disadvantage in Australia. Both ABS surveys used in this analysis are sample surveys, and thus for reasons of statistical reliability in the estimates, they do not include indicators of Aboriginal and Torres Strait Islander status. …