Worlds Apart: Postcodes with the Highest and Lowest Poverty Rates in Today’s Australia

The 1996 sociodemographic profiles shown for each Census Collectors District in the 1996 Census have been uprated to estimated 2000 population levels using ABS data on dwelling commencements and estimates of demolitions for each CCD, along with labour force survey data on labour force characteristics by region. The incomes are as recorded in the HES unit record file.

Marketinfo is derived from the Census and the HES and hence provides data at the household level. Consequently, this paper uses households as the unit of analysis. This effectively assumes that there is complete income sharing within households. The HES person file was also used to derive some variables.

The results of this study are based on simulated data and the techniques are at the cutting edge of poverty research. However, the model has solid foundations in terms of the original data and the techniques. Its predecessors have been used for market research purposes since 1993 and have been benchmarked against other data sources. Policy makers frequently make us of simulated data from high quality model when actual data are not available. Nonetheless, it should be appreciated that this is among the first of NATSEM’s attempts to use the new synthetic regional income database for poverty analysis and that subsequent efforts will no doubt embody more sophisticated techniques and more recent data.

In a major report for the Smith Family written in 2000, NATSEM estimated the before-housing Henderson half-average income poverty rate in 1999 to be 13.3 per cent (Harding and Szukalska, 2000). These results were derived using the 1997-98 Survey of Income and Housing Costs (SIHC) with incomes uprated to 1999 (Harding and Szukalska, 1999). The Australian average poverty rate in this study, which uses the 1998-99 Household Expenditure Survey with Marketinfo weights, is estimated to be 10.3 per cent.

Part of the difference can be attributed to the fact that this study uses households as the unit of analysis while the Smith Family report was based on income unit analysis. An analysis of the poverty rates at income unit and household level in the 1997-98 SIHC showed that household level poverty rates were approximately 2.5 percentage points less than income unit level rates.

Another possible source of difference is the varying income distributions between the HES and the SIHC – with our initial explorations suggesting that the HES income distribution may be more equal than that shown in the SIHC. For these reasons, this paper should be seen as a work-in-progress and the results as indicative rather than definitive.

The financial circumstances of a household are dependent not only on the income of the household but also on its composition. For example, a single person with a disposable income of $19 000 is unlikely to suffer from the same degree of poverty as a couple with four children on the same income. Equivalence scales provide a way of defining poverty levels for families of different composition.

Results can vary greatly depending on the equivalence scale used. The detailed Henderson equivalence scale, which was used in this study, was derived from a survey of household budgets and costs in New York in the 1950s. Despite this, it has been widely used in Australia as a standard method for equivalising incomes. The detailed Henderson equivalence scale takes account of the gender, age and labour force status of the head, the age and labour force status of the spouse and other adults, and the ages of dependent children. The original Henderson approach assigned higher ‘working’ points to people who were either working full-time or unemployed and looking for full-time work. In this study, the ’working’ points have also been assigned to those who are working part-time and to those who are unemployed and looking for part-time work.

The Henderson equivalence scale has been applied at the household level. Because of this, we have given all non-dependent adults who live in the household the same points as a spouse. In the case of a household consisting of three unrelated single people we assign the reference points to the person deemed to be the household head and points equivalent to spouse points to the other two. Other studies using the family or the ABS income unit as the unit of analysis would, in contrast, have assigned head points to all three adults. This is one key reason why our results vary from other poverty estimates (for example, in Harding and Szukalska, 2000). The 1996 Census and consequently Marketinfo/2001 do not, however, allow easy analysis at any level other than the household.

Poverty lines are levels of income and are different for each type of income unit – in our study the income unit is the household. If a household’s income falls below the poverty line for that type of household then the household is considered to be in poverty.

The Henderson half-average poverty measure sets the standard poverty line at half of the average equivalent disposable household income for a standard household. The standard household consists of a couple both under 40 years old with the husband working and the wife not in the labour force with 2 children, a boy aged 6-14 and a girl aged under 6 years old. The Henderson half-average poverty line was $400 per week. (This is looking at all people in Australia, not just those living in the selected postcodes.) Poverty lines for other family types are derived using the Henderson detailed equivalence scales.

There is no consensus about whether the median or the average is to be preferred as the poverty benchmark, with available studies using both (for example, Layte et al., 2000; Strengmann-Kuhn, 2000). The median has the advantage of being less affected by extreme values than the average. For example, large increases in the highest incomes will cause the average to increase but alone will not have an effect on the median. On the other hand, during an era of rising income inequality, there is concern that the incomes of those at the top end might increase substantially, while still leaving median incomes — and thus the poverty rate — unaffected if the poverty line is set at half median income. For that reason we have chosen to use the half-average income poverty line in this study.

Poor households in Lightning Ridge are more likely than poor Australian households generally to have government cash benefits as their principal source of income. While 57 per cent of poor households in Australia rely on government benefits as their main income source, almost seven in every 10 households in Lightning Ridge do. Conversely, while 16 per cent of poor Australian households have wages and salaries as their main income source, less than one-tenth of households in Lightning Ridge do. This parallels the fact that the proportion of people in Lightning Ridge living in a household where the head is unemployed is significantly greater than the national average, while the proportion where the head is a full-time employee is considerably lower. Poor households in Lightning Ridge are more likely than Australian poor households generally to have a head born in Australia (80 per cent compared with 65 per cent) and almost none have a head born in Asia. Poor households in Lightning Ridge are less likely than the Australian average to live in public housing and more likely to be single person households.

The majority of poor residents in the postcode of Carlton South live in a household headed by a man. Seventy per cent live in a household where the head is not born in Australia, which compares with 35 per cent of all poor Australians. The head of the household is more likely to have never married and be under 25 than for poor households nationally. Figure 1 shows that almost one-quarter of the poor in Carlton South live in a household where the head is unemployed and almost six in 10 live a household with the head is not in the labour force. Only 18 per cent have a head that is working. These figure are quite different to the national average. About half of all poor Australians are not in the labour force and another 14 per cent are unemployed. Over three out of 10 poor Australians have some sort of employment. Thus the poor of Carlton South are much more likely to live in a household where the head is not working. In Carlton South, 37 per cent of poor households live in public housing – about 3.5 times the Australian average. Overall, therefore, poverty in Carlton South seems to be due to large numbers of students, unemployed, migrants and a concentration of public housing.

Figure 1

Download asset Open asset

The picture of poverty in Gin Gin is quite different from that of Carlton South. Less than one percent of poor residents live in a household headed by a person under 25. The head is more likely to be middle-aged – 7 in 10 poor households are headed by a person aged 35-65 compared with 58 per cent of all poor Australian households – and are much more likely to be married. Compared with the Australian average, there are more poor households that are couples, either with or without children. A significant proportion of the poor in this postcode live in a household where the head is employed – 4.7 per cent are employed full-time, 16.2 per cent part-time and 18.3 per cent are self-employed. Correspondingly, over one quarter of households rely on income from wages and salaries or self-employment as their principal source of income. Compared with the national average, a greater proportion of the poor in Gin Gin live in public housing or are home owners/purchasers, and they are more likely to be Australian born.

The poor of Ferryden Park tend to live in households where the head has a very high chance of being unemployed (28.7 per cent compared with the national average of 13.8 per cent), never married (30.1 per cent compared with 13.0 per cent generally) and not born in Australia (58.3 per cent compared with 34.6 per cent) (Figure 2). Most poor households in this postcode live in public housing (a striking 78.4 per cent compared with the national average of 10.3 per cent) and almost three-quarters have government cash benefits as their principal income source.

Figure 2

Download asset Open asset

The picture of poverty in the Perth City postcode is one of young, single people without dependents. Over 23 per cent of poor people in this postcode live in a household headed by a person under 25 years of age. This compares with 4.6 per cent of poor Australians. Almost half have never married (compared with 13 per cent nationally) and a similar proportion lives in single person households (compared with 15 per cent of Australia’s poor). Over 68 per cent live in a household without dependent children. The majority of the poor in Perth City live in households headed by a male (51 per cent, compared with 38 per cent for Australia). Over one-quarter is unemployed and 55 per cent are private renters.

The poor in St. Mary’s in Tasmania have a profile more like the Australian average than any of the other postcodes profiled here. Poor households in St. Mary’s are more likely than the poor Australian households generally to have government cash benefits as their main income source and less likely to rely mainly on wages and salaries. An Australian-born person heads almost 9 in 10 poor households in St. Mary’s and only 3.7 per cent of poor households there live in public housing, compared with one-tenth nationally. Poor households in this postcode are more likely to have a head that is married and more likely to be a couple with children.

As we are using an income-based measure of poverty, it is not surprising to find that postcodes with high poverty rates have relatively low average household incomes and postcodes with low poverty rates generally have high average household incomes. For example, the estimated average 1998-99 household disposable income in Ferryden Park is just $20 600 per annum, compared with an average household disposable income of $68 000 in Spit Junction.

However, it is worth noting that some of the postcodes that have the highest and lowest poverty rates do not have particularly low or high average incomes. Both Perth City and Carlton South have average incomes much closer to the Australian average than other poor postcodes, while the average disposable income in the low poverty suburb of Taroona is just over $40 000, less than $4 000 greater than the Australian average. Because poverty lines are based on equivalent income, household composition is also an important factor in determining poverty rates. In addition, the degree of income inequality within a postcode is also important in determining poverty rates. For example, two postcodes may have the same average income, but one might have all households with income close to the average, while the other might contain some households with very high incomes and some households with very low incomes. The latter would have a higher poverty rate. This seems to be one of the factors underlying the high poverty rates in Carlton South, where professionals on relatively high incomes co-exist with poor young students and unemployed.

Postcodes with high poverty rates tend to have one of two age profiles. Lightning Ridge, Gin Gin, Ferryden Park and Taroona have older age profiles than the Australian average, with a greater proportion of households with a head aged over 55. Perth City and Carlton South have a much younger age profile with a large proportion of households headed by a person aged less than 35. This suggests a large student population.

Moving to postcodes with low poverty rates, while Roxby Downs has a young age profile the other postcodes with low poverty rates have a greater than average proportion of households with a head aged 45-64 — and in Brighton, City Beach and Taroona, a greater proportion of households with a head aged 65 or over.

The high poverty postcodes are typified by a greater than average proportion of households headed by a female. Postcodes with low poverty rates are all above the Australian average of 63 per cent of households headed by a male. In Roxby Downs, 90 per cent of households have a male head.

Postcodes with high poverty rates all have a smaller than average proportion of households headed by a full-time employee and more headed by a part-time employee, self-employed person, unemployed person or someone not in the labour force. There is some variation between the different states, however. Ferryden Park has a strikingly high percentage (56 per cent) of people in households where the head is not working. Lightning Ridge, Gin Gin and St. Mary’s have a low proportion of households with the head working full-time, but with a higher than average proportion of part-time employees and self-employed people. Lightning Ridge has a large proportion with an unemployed head.

Postcodes with low poverty rates all have a significantly greater proportion of households headed by a full-time employee and a lower proportion headed by an unemployed person or someone not in the labour force.

Principal source of income is closely linked to the labour force status of the household head. High-poverty postcodes have a large proportion of households relying on government cash benefits. In Ferryden Park, 57 per cent of households have government benefits as their principal income source. Conversely, the proportion of households relying on government cash benefits in the low-poverty postcodes is significantly less than average and most households have wages and salaries as their main income source.

Three of the high-poverty postcodes, Ferryden Park, Carlton South and Gin Gin, have high levels of public housing; in Ferryden Park 68 per cent of households are government renters. Perth City has a very high proportion of private renters, as does Carlton South. However, Lightning Ridge, Gin Gin and St. Mary’s have about average numbers of households that are owners/purchasers. Similarly there is no clear trend among the low-poverty postcodes. There is a tendency for there to be an above average proportion of owners/purchasers, but in Spit Junction and Roxby Downs there are a large proportion of private renters and City Beach has an above average proportion of public housing tenants.

Again there is no clear pattern among high-poverty postcodes. In Perth City, Ferryden Park and Carlton South, a greater than average proportion have a head who has never been married and a greater than average number of single person and multiple family households. Lightning Ridge and Ferryden Park have more heads who are separated, divorced or widowed and Ferryden Park has almost one quarter sole parent households. However, a married person heads eight out of 10 households in Gin Gin.

Among low-poverty postcodes, generally a greater than average proportion are headed by married person and the proportion of households headed by someone who has never been married or is separated, divorced or widowed is less than the Australian average. In all of the low-poverty postcodes, the proportion of sole-parent households is less than the Australian average.