|Page tools: Print Page Print All RSS|
THE DISTRIBUTION OF HOUSEHOLD INCOME
There are many measures used to assess the distribution of income, of which the more commonly used are presented in the table on the following page. Most of the movements shown have been small and differences across the income distribution are not statistically significant,(SEE FOOTNOTE 1) showing little or no overall change in the level of income inequality among households during the period 1994-95 and 1997-98.
The average income measures given in the commentary Economic disadvantage and inequality (presented in the headline graph) are shown again in the first rows of the following table. The next measure shown in the table focuses on the relationship between the income levels of those at different points of the distribution, namely at the 20th, 50th and 80th percentiles from the bottom of the distribution. As can be seen, the trends over time of the two series are very similar - percentage increases between 1994-95 and 1997-98 in the income of lower income households have generally been very similar to those in the middle and towards the top of the distribution.
Changes in incomes of the respective (low, middle and high) income groups relative to each other are highlighted by the ratios of incomes of households at the respective percentiles. For instance, when comparing households towards the top of the distribution with those at the bottom (illustrated by the P90/P10 and P80/P20 ratios) it is apparent that the households at the bottom of the distribution have generally maintained their position compared to those at the top (the ratios have remained much the same). In addition, as seen by the P20/P50 ratio (i.e. the income of those at the twentieth percentile compared to those exactly in the middle of the distribution), the relative position of the more disadvantaged group compared to those in the middle of the distribution also remained virtually the same. Moreover, when seen in terms of the changes in the share of all income received by high and low income households, the changes have not been significant. In 1997-98 the share of all household income of the 20% of households with low incomes (i.e. those in the second and third deciles) was 10% while that for households at the top of the distribution, the highest 20% of households, was 39%. These proportions were the same as in 1994-95.
The Gini coefficient (a commonly used summary measure of income distribution) also shows minimal change. The Gini coefficient is a single number that summarises the distribution. It takes values between zero and 1 - higher values indicate greater inequality in the distribution of income; lower values indicate greater equality. While a little lower during the intervening years the coefficient in both 1994-95 and 1997-98 was the same (at 0.32).
Other commonly used indicators of economic disadvantage are the proportion of households with income less than half the average weekly equivalised disposable income of all households and half the median weekly equivalised disposable income of all households. These income cut-offs are sometimes used as poverty lines.(SEE FOOTNOTE 3)
The number of households below the half median indicator shows a small increase between 1994-95 and 1997-98 (2%), but this change is not statistically significant. In addition, this indicator needs to be used with caution for two reasons. First, there is a considerable clustering of households with income relatively close to the half median value of income, reflecting the fact that the half median value is close to the value of benefits provided by the age pension and other benefits under the social security system. A relatively small change in the benefits available from social security can result in relatively large changes in the percentages shown by the half median indicator, and therefore it has the potential to be relatively unstable. Second, many of the households in the first decile and reporting very low incomes are not believed to be the most economically disadvantaged (see footnote 1 of Economic disadvantage and inequality). Therefore changes in the half median measure may also be influenced by changes to the proportion of households reporting very low incomes which are not economically disadvantaged.
While the increase from 1994-95 to 1997-98 in the number of households receiving less than half the mean household weekly income is statistically significant (up 5%), as with the median, the mean is also affected by the incomes of households in the lowest income decile and interpretation of change in this indicator can be difficult.
Selected measures of household disposable income(a)
Source: Data available on request, Surveys of Income and Housing Costs.
LOW AND HIGH INCOME HOUSEHOLDS
The circumstances associated with economic disadvantage among households are well known. These include a lack of savings or other assets and the absence of a substantial income from sources other than pensions and benefits provided by the social security system. People whose major source of income is from pensions and benefits include those who are unemployed, those unable to work due to disability, illness or old age, or those with caring responsibilities for others (such as persons with disabilities or dependent children). Aside from factors associated with life-cycle stage (i.e. relative youth or old age), not having a paid job or having a low paid job is often associated with lower levels of educational attainment and limited job opportunities within the areas in which people live. It is also often associated with problems in other aspects of life, including those related to physical and mental health and the size and strength of people's social networks. Of course changes in life fortunes can also be factors (some people can benefit from windfall gains while others can suffer unexpected losses through crimes committed against them or their own misadventure).
Comparing households of various types in low and high income groups helps to characterise the types of households most likely to be economically disadvantaged. Households are classified in several ways in the table above, presenting data for 1997-98. The patterns shown could be expected to be broadly representative of the patterns seen throughout the 1990s.
There are some substantial differences in the representation of certain household composition types in low and high income groups. One parent families in one family households with dependent children, for example, were over-represented (11%) in the low income group (those in the second and third lowest deciles of the income distribution) and under-represented (1%) in the highest income group (those with income in the top two deciles of the income distribution). Possibly associated with their age and employment circumstances, couples without children were over-represented in the highest income group and under-represented in the middle of the income distribution. Lone person households were over-represented in the lowest income group and many would have been receiving the age pension.
In contrast, while significant proportions of households in the lowest and highest income groups were couples in one family households with dependent children only (18% in both groups), households of this type were over-represented among middle income households (34% of all middle income households).
In terms of the principal source of household income, households in the low income group (i.e. with income in the second and third lowest deciles of the income distribution) were, as might be expected, most likely to have government pensions and allowances as their major source of income (79% of households). In sharp contrast to those in the low income group, most households (93%) in the highest income group had employment related income (from wages or salaries or their own unincorporated business) as their principal source of income.
The number of earners (i.e. with employment related income) present in a household is an important determinant of household income. Clearly those with two or more earners will tend to have higher incomes than those with only one earner. In 1997-98, the average number of earners per household in the low income group was 0.3 persons, which contrasts with 1.2 and 1.9 earners per household in the middle and high income groups respectively.
Household characteristics of selected income groups(a), 1997-98
Source: Data available on request, Survey of Income and Housing Costs 1997-98.
ABORIGINAL AND TORRES STRAIT ISLANDER PEOPLES
Data from the 1996 Census of Population and Housing show the average household income of households with at least one Aboriginal or Torres Strait Islander person aged 15 years or over to be well below that of households in the wider community. The household income per capita in 1996 was $158 per person per week for households which included at least one adult Indigenous Australian, compared to $310 per person per week for all households.
In terms of those who were employed at the time of the Census, the median gross weekly income of Indigenous Australians ($365) was substantially less than the median of non-Indigenous Australians ($493).
1 Differences in the various estimates of income and income distribution, when compared over time, may or may not be significant as there are errors that can arise as a result of the data being obtained from sample surveys. Tests of significance show that there is a high level of confidence that there have been no real changes among most of the measures used in this commentary to compare income distributions from 1994-95 to 1997-98.
2 For details of data sources used and further information about the distribution of household amenities see ABS 2001, "Household Amenities", in Australian Bureau of Statistics 2001, Australian Social Trends 2001, Cat. no. 4102.0, ABS, Canberra.
3 Harding A., Lloyd R., and Greenwell H. 2001, Financial Disadvantage in Australia 1990 to 2000: The Persistence of Poverty in a Decade of Growth, NATSEM and The Smith Family, Camperdown, NSW.