Category Archives: Inequality Research

Highlighting income and wealth inequality data, issues with inequality data, or research (mine or others’) on income and wealth inequality.

South Australia’s Small Gender Pay Gap: Is it Good News?

The gender pay gap in South Australia is around half the national figure – but is this really good news?

This is the sixth post in the series on inequality in South Australia, with the early posts dealing with inequality between households and the later posts looking at structural inequalities. This post continues the focus on structural inequality, and looks at gender inequality – or at least one aspect of gendered economic inequality.

Gender Pay Gaps – Official

The “official” gender pay gap figures used by the government’s Workplace Gender Equity Agency show that the gender pay gap is significantly narrower in South Australia than in the country as a whole. At 7.4% the gender pay gap in South Australia is around half the national average, and is the lowest in the country. However, there is devil in the detail of this official figure.

While the gender pay gap is lower in South Australia, this gap is on lower wages overall and does not necessarily reflect women doing relatively better. ABS Labour Force data shows that the $1541 per week full-time ordinary time earnings for women in South Australia was 95.8% of the equivalent figure at the national level. For men in SA, full-time ordinary time earnings were 89% of the national figure. In theory, this difference could mean that SA has proportionately more women in higher paying jobs, but more likely it reflects labour market segmentation where proportionately more women are in jobs with wages set nationally (e.g. minimum wages or modern awards), while men are overly represented in non-award industries where wages may differ more across the country.

This suggests that much of the difference in the national and SA gender pay gaps is not about women’s pay, but rather because men’s wages in SA are disproportionately lower than the national average. Put another way, rather than the labour market in SA bringing women’s wages up closer to men’s, it is an “equalling down” based on relatively lower men’s wages.

This creates particular challenges for the left and the union movement: how to increase wages in South Australia to closer to national averages without also increasing the gender wage gap? Obviously, a focus on gender segmentation and increasing wages in highly feminized industries like childcare is a start, but it should also be a factor in changes to enterprise bargaining currently under consideration.

Bigger Gender Pay Gaps Beyond the Headline Figure

The “official gender pay gap” based on full time ordinary time figures is only one measure of gendered pay inequality. The table below shows a range of key gender pay gap and wage share data for Australia and South Australia and is necessary because the full-time ordinary time data is limited. It ignores (and arguably institutionalises) gender differences in access to overtime and bonuses.

The first line in the table is the official full-time ordinary time data, but as can be seen in the second line of the table, when overtime, bonuses and other extras are taken into account in total full-time weekly earnings the gender pay gap is much larger. It is still less of a gap in South Australia, but the difference in the national and SA figures is narrower.

However, even these full-time figures ignore the over-representation of women in part-time and casual work. As evident in the third line in the table, the gender pay gap jumps markedly when the average weekly earnings of all employees are included. The gap between South Australian and national figures is much lower here, presumably because of a higher proportion of women in part-time jobs in South Australia. This is confirmed by the gender wage share data in the bottom line of the table, where the difference between the national and SA figures is lowest.

Gender Pay Gaps, May 2022

 AustraliaSA
F/T Ordinary Time Earnings14.1%7.4%
F/T Total Earnings16.5%9.6%
Average Earnings29.7%27.7%
Women’s Participation Rate62.2%58.4%
Gender Wage Share39.0%40.0%
Source: Calculations from ABS Average Weekly Earnings and Labour Force

Gender Wage Share

I have set out elsewhere a rationale for considering, alongside the traditional average pay gap data, the gender wage share – that is, the female share of the total wage pool. The importance of that wage share data can be seen here.

Both nationally and in South Australia, the female half of the population takes home around 40% of the total wage pool. In actual dollars, the May 2022 figures show that men as a whole in South Australia take home $221m more than women each week, for an annual return to gender of $11.5bn. To put that in the context of the local economy, the total sales revenue of South Australia’s biggest company, SANTOS was around $6.9bn in 2021, around 60% of the gender wage share differential.

As I argued in an earlier post about the corresponding national figure of a $200bn annual wage differential, the aggregate gender wage share difference is a significant economic flow that not just reflects but contributes to the reproduction of that inequality.

Changes Over Time

The 2022 data above shows, in short, that while the official gender pay gap is much lower in South Australia, the benefit of this is diluted by lesser participation in paid work by women (both relatively lower participation in the workforce, and relatively fewer hours by those engaged in paid work). As we will see below, these patterns are not new.

The following graph shows the full-time (all earnings) gender wage gap in Australia and South Australia over the last 25 years. While the data on South Australia is more volatile, the gender pay gap in South Australia has generally been below the national figure. In both data sets, the gender pay gap shrinks in the late 1990s, but grows again from 2000 nationally and from 2004 in South Australia through to 2014. The gap then shrinks over the next few years before flattening out in the last few years.

Time series of gender pay gap in full-time total earnings in South Australia and Australia.

However, the gender wage share data tells a slightly different story. There is little difference in the time-series data on South Australian and Australian women’s share of their respective total wage pools. In both data sets there is a small but sustained increase in women’s share (apart from a small dip after the Global Financial Crisis in 2008) (see the black SA line in the graph below). In South Australia women’s share of the total wage pool increased from 35.1% in November 1994 to 40% in May this year. The national figures showed an increase from 33.1% to 39% over the same period.

Crucially though, this story is set against the backdrop noted in a number of my previous posts of a general decline in South Australia’s economy relative to the national economy. This was evident in SA’s share of household income and the labour share of the economy. The graph below shows that this decline also impacts on the gender wage share. While the top line shows that women have been increasing their share of the South Australian wages pool, with that pool in relative decline, South Australian women’s share of the national wages pool has actually shrunk from 2.6% in 1994 to 2.41% this year.

Time series (1994-2022) of female share of total wages pool in South Australia, and South Australian female wages as proportion of Australian wage pool.

While this decline in South Australian women’s share of the national wage pool is alarming, it is made even more problematic because, as noted in the previous post, the wage pool itself is declining as a proportion of the economy.

Summing Up

The official gender pay gap data shows South Australia doing relatively well with the smallest gap in the country, but a closer look at the data tells a less rosy story. The smaller gender wage gap is based on lower women’s participation rates and relatively lower male wages in South Australia. It is not a story of greater female agency in the labour market. The national/SA differences almost disappear when the share of the total wage pool is considered.

And regardless of the comparative story, the gender wage inequalities in South Australia remain significant – with a 27.7% gap in average earnings leading to an aggregate $11.5bn annual wage gap. Further, while women’s share of the South Australian wage pool has increased over the last 25 years, these gains are undermined by the shrinking of that wage pool relative to the national wage pool – a national pool which, with historically low labour shares of GDP, is itself shrinking relative to the economy as a whole.

Beneath the headline gender pay gap data, for South Australian women workers there is a particularly problematic intersection of gender, geographic and class inequalities.

The Labour Share: No Paradise for Workers

This is the fifth post in the series on inequality in South Australia. While previous posts used ABS data to identify and track inequality between households, this post shifts the focus to structural inequality, and questions of class in particular – as evidenced by the changing labour share of the economy. Future posts will look at other structural inequalities based on gender and race.

Structural Inequality

The shift in focus is important because while the household data reveals much about patterns of inequality, it also has many limitations. Household income and wealth data is based on an assumption that households share resources. This assumption is probably unwarranted for many households, including the 13,000 “group households” in South Australia (272,000 nationally) where some expenses may be shared, but income is likely not shared. Further, viewing the household as the basic economic unit hides dynamics of potential inequality within households – most obviously gender and age dynamics, which then have ramifications in society more widely.

Further, the household data produced by the ABS, and used in many studies of inequality (for instance, the Productivity Commission report and the ACOSS/UNSW research) presents a mono-dimensional stratification of income distribution: a continuous spectrum from lowest to highest income/wealth on which all household/individuals are located. The measures of inequality are then based on the relationship of arbitrary points along this continuum such as income quintiles, top-bottom deciles. They do not tell us much about what is driving this stratification.

While it is possible to look at different demographic characteristics of households at various points on this spectrum, it is still largely only a scoring of an end point of distribution – not the mechanism of unequal distribution itself. By contrast, structural inequality is not simply a different location on a spectrum, it is a systemic and often conflictual relation, buttressed by a range of institutional arrangements where the economic inequality drives or at least contributes to the reproduction of that inequality.

Class – the Labour Share of GDP

This description of structural inequality and some of terms above are loose and contested, and there are mountains of academic writings on the subject. I have previously discussed Erik Olin Wright’s attempt to draw together classical Marxist, Weberian and Durkheimian approaches to class. All these approaches offer different insights, and all are added to and cross-cut by analysis of other structural inequalities. However, with no claim to being comprehensive or determinative, in this post I want to keep it fairly simple with just one measure of one structural inequality: class, as measured by the relative financial returns going to capital and labour.

The usual measure of these class-based economic flows is the share of national income going to labour (“compensation of employees” in ABS-speak), or alternatively, the labour share of the whole economic pie (that is, the share of Gross Domestic Product going to labour). I am sure there is a PhD somewhere critiquing the notion that “GDP = the economy”, but nonetheless, the labour share of GDP is a convenient measure as there is a robust ABS data set and changes in the labour share reflect changes in class power within the economy.

The Labour Share Data – National and South Australian

As is well-known, labour’s share of GDP in Australia has been falling in recent times. In 2018, the Journal of Australian Political Economy dedicated an excellent issue to highlighting the issue and examining the causes, and the labour share remains a cause of concern for the labour movement and many on the left.

Rather than add to the debate about causes or solutions, in this post, I simply want to look at the figures and compare the situation in South Australia with the rest of the country. As can be seen in the table below, despite lower average wages and lower workforce participation rates, the most recent ABS data (June 2021) shows that labour share of the South Australian economy was slightly higher than the labour share nationally.

AustraliaSouth Australia
Average Total Weekly Earnings$1,797.10$1,626.00
Participation Rate66.2%62.7%
Labour Share47.7%49.3%

While this suggests a slightly better distribution of income to workers in South Australia (i.e. a more equitable outcome in class terms), as the graph below shows, this is not historically consistent. The labour share in South Australia is more volatile and slumped below the national average in the decade from 1996 to 2006.

Time series showing labour share of economy in SA and nationally from 1990 to 2021. Long term decline in national series, while SA slumped but regained ground from 2010.

It should also be noted that the national data here does not show the current “historically low” labour share discussed in the media. That discussion is rightly based on seasonally-adjusted figures, but these are not published at the state level so I have used the ABS “original” data series to ensure a like-for-like comparison. That said, the seasonally adjusted figures show an even lower labour share nationally (46.1%), so either way, it would appear that SA labour’s share of the state economy is higher than labour’s share at the national level.

While this may appear to be good news for South Australian workers, this is a higher share of a decreasing part of the national economy. In 1990 the SA economy (Gross State Product) accounted for 7.7% of the national economy (GDP). In 2021, SA’s share was 5.7%. As the graph below shows, South Australian labour’s share of the national economy (in orange plotted on the right axis) has declined over the last 30 years, even while largely retaining its share of the state economy (black line plotted on the left axis).

Time series graph repeating SA labour share of GDP, but also showing long even decline in SA labour's share of national economy (GDP)

For South Australian labour there is then a double-challenge – the class challenge of retaining and increasing its share of income, and the development challenge of growing the economic pie overall. Of course this struggle is not unique, but it is a particular challenge given the arguments in previous posts about being at the economic periphery.

Caveats and Conclusion

It is worth emphasising again that the labour share is a class distribution, an economic process rather than a positioning of people or households. Many people and households (particularly higher-income households) have multiple sources of income including government transfers, bank interest, investment income, imputed rents and capital gains. These are different economic flows (class processes) and those households’ labour income does not necessarily determine their total income, lifestyle options or position in society.

That said, labour income is the major income source for the majority of Australian households. Accordingly, changes in the labour share of income are a key contributor to the household income spectrum, while also being important in their own right as a reflection of structural inequality.

In this context, my interim conclusion (to be added to in future posts) is simple: with a declining share of the national economy, it is no paradise for workers[1] – either in South Australia or nationally.

[1]  With apologies to Ken Buckley and Ted Wheelwright for stealing their classic title.

Household Wealth Inequality

This is the fourth post in the series on inequality in South Australia. The previous posts focused on different aspects of inequality in the distribution of income. This post looks at the distribution household wealth, again based primarily on the 2019-20 ABS Housing Income and Wealth data released this year.

Wealth Distribution Between Households

It is well-known that the distribution of wealth is far more unequal than the distribution of income. This is a function of both the cumulative impact of income differentials over years, as well as the ability of wealth to beget wealth (through capital gains and higher average returns on capital).

Nationally, the households in the highest wealth quintile hold 62.2% of all household wealth in Australia, while the bottom 40% of households account for on 6.1% of wealth. The P90/P10 wealth ratio sees the high-wealth households (90th percentile) owning some 52 times the wealth of the low-wealth households in the 10th percentile. (By contrast, households at the 90th percentile earn around 9 times the income of the low-income (10th percentile) households.

There is clearly a massive inequality in wealth distribution, but it is important to note that these high wealth households are not necessarily high-income households. Most obviously wealth accumulates over a life-time and at some point people retire which may leave them with some wealth but low incomes. The ABS data suggests that the wealth-income correlation is not straightforward. For instance, 10% of high-wealth households had low-incomes, while only one-third of low-income households also had low wealth. However, Piketty’s work highlights the correlations at the very top of the income/wealth spectrums where income from capital (wealth) becomes the dominant income stream.

Unfortunately, the ABS does not publish wealth ratios or wealth-by-quintile of data on a state basis, so there is limited data on wealth inequality within South Australia. However, there is some useful data on wealth distribution between states.

SA share of total household wealth

As with previous posts, I start with aggregate wealth rather than the ABS household averages because the aggregates adjust for differences in population. The graph below shows the various state and territory share of total wealth held by households in 2019-20, but plotted against the share of total income and population. It can be seen that NSW and Victoria have a greater share of wealth than of total current income, while Queensland and WA a higher share of income than wealth. South Australia’s share of the total wealth held by Australian households is quite close to its share of income (wealth 6.5% of total, income 6.3%), but both are below the state’s population share.

Column graph showing state/territory shares of national household income, population and net household wealth.
(Note: NT data does not include remote areas)

While this graph is a snapshot of 2019-20, the long-term trends are not easy to discern. My previous post showed that South Australia’s share of national household income has declined over the last two decades, driven in large part by a declining share of the population. However, the columns in the graph below, which show the state’s share of total household wealth, do not show a similar pattern of decline – or any pattern really. This could be a trick of data unreliability (wealth data is less reliable than income data), or of different structures of household wealth (see below) or different levels of indebtedness (which would impact on net wealth – although the data is not available over the long term).

Time series showing SA declining population and income share (as % of national aggregate), but no real pattern in share of wealth bouncing between 5.6% (2016) and 7.2% (2006), and overall approximately the same in 2020 as in 2004.
(Note: data not available for 2000-01 and 2007-08)

I have included this time series data for completeness, but I draw no conclusions from it. Nor can I draw any conclusions on the distribution of wealth between households in Adelaide and the rest of the state. Even though the ABS publishes it, the regional wealth data has such high margins of error I would not rely on it. I think the primary take-out of the data above is that South Australia’s share of national household wealth is lower than its share of population. Or put another way, wealth (like income) is disproportionately held elsewhere – and with only around 6% of both wealth and income, this makes it difficult for South Australia to have an impact on Australia’s political economy.

Structure of Household Wealth

Beyond these aggregate figures, the published data based on average household wealth shows some interesting differences between South Australia and other states in the structure of wealth holdings.

The graph below shows a state-by-state comparison of different components of average household net wealth (that is, the value of household assets minus liabilities). At the highest summary level, it is clear that South Australian households on average have less wealth than Australian households generally (approximately 88% of the national average), and significantly less than states like NSW, Victoria and the ACT. This is seen in the height of the columns on the graph, but it is interesting to note that South Australian households have the fourth highest average wealth. By contrast, SA households have the second lowest average household income. Both WA and Queensland have higher average household incomes, but lower average wealth.

South Australian households fare relatively better in wealth distribution than income distribution, but this is even clearer when we look at the components of this total net wealth. One of the key reasons for SA lagging behind the richer eastern states is the higher value of wealth tied up in owner-occupied dwellings there (the blue at the top of the columns). That is, households in other states where housing prices are significantly higher are likely (on average) to have more of their wealth tied up in their homes. Not counting owner-occupied dwellings, the wealth held by the average South Australian household is just $15,000 less than the national average. The SA average is 98% of the national average.

Column graph showing different components of wealth (real estate, and various financial assets) - net of liabilities. SA has 4th highest level of wealth, but below national average - but much closer if owner-occupied housing assets are excluded.

The amount of wealth held in owner-occupied housing is important because there is considerable debate as to whether owner-occupied housing should actually be considered as wealth. At one level it is clearly an asset which can be sold for money, and (as I have argued in an earlier post) provides housing services which is a form of in-kind income which can be imputed as non-cash household income. On the other hand, as Anwar Shaikh and others have argued, it is not real “wealth” because housing is a necessity and cashing in the asset is not simply transferring a physical asset to a cash asset (with an attached transfer of income from imputed rent to other investment income). Liquidating an owner-occupied house actually leaves the owner worse-off as they will now have to pay rent.

Obviously though, the same is not true for housing properties not occupied by the owner as these are held as investments and can be liquidated in a simple swap to another form of asset.

The wealth tied up in owner-occupied housing is also important because lower housing costs in South Australia (relative to other states) can also translate to SA households holding relatively more financial assets (because their “savings” are not tied up in housing costs to the same extent). This is evident in the graph in the relatively high shareholdings and business assets held by South Australian households (seen in orange), although the ABS warns that there is a high margin of error on these particular figures and some caution is required.

Further though, (despite lower incomes) lower relative housing costs also potentially translate into lower average household debt. In 2019-20, the principal outstanding on loans for owner-occupied dwellings in SA averaged out at $88,500 per household, which was 77% of the national figure, while total household debt as a proportion of household wealth was also lower in South Australia (14.4% of total household assets in SA, 16.4% nationally).

The debt figures are averages of households with and without debt, so those in debt will have much higher amounts owing than these average figures, and the averages will be impacted by the proportions of households with/without debts. However, on its face it does suggest interest recent and predicted interest rate rises will impact slightly less in South Australia than in states (and territories) where average debt is higher.

Conclusion

What all of the above suggests is that, despite the average income of South Australian households perennially lagging behind the national average, SA fares better in the distribution of household wealth. Indeed, when the wealth data is considered without reference to owner-occupied housing (which is not transferable or income-earning in the same way as other assets), average net household wealth in South Australia is pretty close to the national average.

However, at the aggregate level, South Australia’s small share of both income and wealth shows that there are still significant issues around the state being relatively marginal to income distribution and wealth accumulation in the national economy. This has not (yet) flowed down to impact on average household wealth, but the wealth data does little to relieve the concerns highlighted in previous posts about SA being at the economic periphery.

And finally, this focus on geographic inequality should not blind us to the massive inequality of distribution of wealth within the state. There is little reason to assume South Australia is immune from the national pattern where the majority of wealth is held by the richest households and the lowest quintiles hold almost no wealth. The dual challenges remain: to get a bigger share of national economic development for South Australia, while distributing that share more equitably within the state.

Regional South Australia – Inequality at the Peripheries

Context

This post on regional South Australia is the third of a series of posts on inequality.

The first post looked at inequality of household incomes between states. It found that South Australia’s share of income is below its share of the population, and that the state’s share of both national household income and population has decreased over the last two decades.

The second post looked at inequality within South Australia. It found that while the broad pattern reflected the national spread across income quintiles, the South Australian distribution was characterised by relatively lower incomes at the top end of the income spectrum.

This post looks at regional South Australia. Again, the data is drawn from the ABS Household Income and Wealth series, with “regional” being defined by the ABS “Rest of State” data beyond Greater Adelaide. This categorisation obviously conflates very different regional and remote areas, but it does still reveal some important economic dynamics beyond Adelaide.

Again, I use the share of total/aggregate income as a key analytic. The share of aggregate income captures changes in both income and population, and in a regional context it highlights changes in a key material base of community viability and development.

Regional South Australia’s Share of State and National Income

Incomes in regional areas vary greatly from area to area and from year to year depending on the local industry-base, weather patterns and commodity prices. The 2019-20 data collection also overlapped with bushfires and COVID, but the ABS methodology was designed to minimise the impacts of those events so the data reflects long-term trends.

South Australia has a particular regional demographic (shared with WA) of a particularly dominant capital city. Around 80% of the SA population live in the Greater Adelaide area, while the rest of the state is large and lacking other big cities. This in itself creates economic, social and service-delivery challenges in much of regional South Australia, but it is made worse by lower average incomes.

Average (mean) gross household income in regional South Australia in 2019-20 was $1679 per week, which was 81.6% of the Adelaide average – although household size was also smaller in regional South Australia (2.2 people per household in the regions, 2.4 in Greater Adelaide). However, given average household income in Adelaide was also below the national average, this meant that the average household income in regional South Australia was just 72% of the national average.

At the aggregate level, regional South Australia accounted for just 17.7% of the total household income in the state, while housing 19.9% of the population. By comparison, regional areas in all states accounted for 31% of the Australian population and 27% of all household income. So regional South Australia is struggling relative to both Adelaide and to other regional areas.

Inequality within Regional South Australia

The overall income spectrum in regional South Australia appears to mirror Greater Adelaide. Adjusted for household size, the income share for the bottom two quintiles, the P90/10 ratio and the Gini Coefficient are all fairly similar between Adelaide and the “rest of the state” (more details here). However, the census data mapping the proportion of the population in different income brackets tells a more nuanced story.

The Adelaide and regional SA numbers are shown in the graph below in the same way that my previous post mapped the national/SA figures. In both cases, for the relatively poorer jurisdiction, we see a greater proportion of households in the middle-income categories and significantly fewer among higher income households – in this case, households earning over $2,000 a week.

Line graph showing the proportion of households in Adelaide and Regional South Australia in each of the census income brackets.

Some of the differences evident in the graph above are an effect of household size, where bigger households in Adelaide more likely to contain multiple income-earners, but it is also just a simple question of geography. To extend the conclusion of my previous post: the rich don’t live in regional South Australia (at least not in the same proportions as in Adelaide, or the rest of the country).

Changes over Time

However, the biggest concern about regional inequality may be the changes over time. Just as the national figures show that South Australia has lost both population share and income share over the last twenty years, so too regional South Australia has lost both income and population share within the state.

The graph below is similar to one used in the previous post on South Australia’s share of the national household income, but this shows regional SA’s share of aggregate state household income. Regional SA’s share of income and population is plotted on the left axle, while average household incomes (as a percent of those in Greater Adelaide) are plotted on the right axle.

Line graph plotting regional South Australia's share of state income alongside its two components: changes in population share and average income.

In all years, regional South Australia’s share of aggregate household income was below its population share, but there has also been a clear decline of the regional income share over the last two decades (from 22.6% of SA’s total household income in 2000-01 to 17.7% in 2019-20). As the graph shows, the early years of this decline was associated with a fall in average regional household incomes relative to Adelaide. However, even when this relative fall in average income levelled off and improved, regional South Australia’s share of aggregate household income still fell – driven by significant declines in population share.

Given that this regional decline is relative to a state which is also declining in the national data, the end result is particularly worrying. Regional SA’s share of national household income has declined by nearly a third over the period, from 1.6% in 2000-01 to just 1.1% in 2019-20. This makes maintaining economic (and political) viability harder relative to the rest of the country, a fact which can drive further decline.

Conclusions and Speculations

I have suggested that the picture of regional inequality – the unequal incomes between regional SA and the capital city – mimics the differences between South Australia and the national income distribution. This obviously suggests that there are some similar dynamics at play.

David Peetz (and others) have argued that the majority of the gains of neoliberalist economic growth over the last 30 years have gone to finance capital. The data under consideration here is consistent with that analysis. Such financialisation processes tend to concentrate income in the geographic centres of finance, leading to a relative decline in income in the areas outside those centres. This in turn creates a periphery where households are living at a base rate (arguably propped up by nationally regulated incomes), but where it is more difficult to sustain higher incomes as population and income is attracted to core areas.

This type of effect is compounded for regional South Australia. The core-periphery process happens between South Australia and the financial centres of the Australian economy, and is then repeated between regional South Australia and Greater Adelaide.

I deliberately use the academically contentious (and somewhat dated) term “periphery” because it summons a long history of debate over colonial and post-colonial development. While I will not go over those debates here, some of that “development” should ring warning bells about policies we might adopt to address the under-development and subsequent inequality in our own periphery.

For instance, the Australian regional development default often appears to be to intensify the extraction of natural resources. This is intellectually lazy and driven more by private profit than community development, but crucially, as Joe Collins argues in a recent JAPE article, such “extractivism” does not guarantee development and comes at great costs to the environment, and often to local communities – Indigenous and settler.

Further, the low-wage strategies adopted in parts of the post-colonial world may also be counter-productive. Given the particular shape of the income graphs above, low-wage strategies may simply increase numbers in the low-mid income brackets, but do little to close the gaps identified at the higher end which are drivers of geographic inequality (albeit while making internal distributions more even). At best, it is a levelling-down, rather than levelling-up, approach.

I may return to these policy arguments in future posts, but my main purpose here was simply to examine the data on regional inequality in South Australia. However, the results, which posit regional South Australia as the periphery of the periphery, clearly bode badly for equality and for the economic sustainability of those regional areas. A significant migration and regional development policy response is required.

The super-rich don’t live here: Inequality in South Australia

Context

This is the second of a series of posts on inequality, with a particular focus on South Australia. The first post highlighted differences in the share of the national household income going to different states, and the long-term decline in South Australia’s share of household income.

This post shows that, while household income may be lower than other states (except Tasmania), the distribution of income between households in South Australia is more equal than the national average.

Again, unless otherwise stated, the data here is drawn from the ABS Household Income and Wealth series. However, the data on the “share of total income” that I flagged as crucial in my previous post is less important here because the proportion of households in each quintile is the same. That is, there is no dynamic from changing population numbers.

The ABS Data on Inequality in South Australia

While the overall pattern of income distribution and inequality within South Australia is similar to the national pattern, by almost all measures in the ABS data, income distribution within South Australia is slightly more equal than for the country as a whole.

The table below summarises key measures for both gross household income, and for equivalised disposable income (that is, after-tax income adjusted to equivalent household sizes).

Table showing ABS measures of inequality in South Australia by both gross and equivalised income quintiles.

The first line is the income share of the lowest two income quintiles, with those households in South Australia having a slightly higher share than the national average.

The second line is the P90/P10 ratio which refers to the ratio of income of high and low income households, specifically, high-income households at the top of the 90th percentile by comparison with those in the 10th income percentile. These high-income households in South Australia received 8.19 times the gross income of households at the low (10th percentile) end of the income spectrum. This represents significant inequality within South Australia, although it is less than the Australian average ratio of 8.98.

The third line is the Gini Coefficient, which is a complex statistical measure where lower numbers represent more equal income shares (0 would be perfect equality, and 1 would be absolute inequality with one person receiving all income). While the Gini coefficient is widely used, I don’t find it particularly useful. The number is meaningless of itself (I still have to look it up every time I see it) and the importance of a difference at the second or third decimal point is hard to grasp. Again though, the Gini Coefficient shows SA has a slightly more equal income distribution than the national average.

It is also worth noting that in each case, the equivalised disposable income figures are much more equal than the gross income figures. This demonstrates the importance of both household size and the redistribution impact of taxes. But in each metric in both gross and equivalised data sets, the South Australia figures are slightly more egalitarian than the national ones.

Top End Drivers

It is important to note that the differences between South Australia and the country as a whole are not the same across all income quintiles. The incomes of those on the lowest incomes in South Australia are much closer to those on the lowest incomes nationally than is the case for those in the highest income brackets. The average (equivalised disposable) income of the lowest income quintile households in South Australia was $403p.w. in 2019-20. This was 97% of the national figure for the lowest income quintile. By contrast, for the highest income quintile, the average income in South Australia ($2,015p.w.) was only 90% of the national figure.

The same pattern is evident in the graph below, which shows the income at the top of the selected percentiles. The gap between South Australia and the national average increases as we move up the income spectrum. (The equivalised data provides a more like for like comparison, but the gross data [not plotted below] shows proportionately higher incomes at the top of the range in South Australia. The difference in the two data sets suggests that that relatively higher incomes in gross data are driven by larger household sizes rather than higher individual incomes).

Line graph of equivalised income at the top of percentiles, 2019-20 for South Australia and Australia. Both lines upward sloping but with a gap widening as we move up the income scale.

This overall pattern is perhaps not surprising given that the incomes of many of those in the lowest income brackets are set nationally (e.g. social security and minimum wages) and are the same regardless of location. However, the patterns do highlight a key issue in the analysis of inequality – the importance of what is happening at the top end. In this case, it appears that the incomes at the high end of the South Australian spectrum are much lower than those evident in the national data. This reduces income inequality in South Australia, but also contributes to South Australia’s relatively low income share overall.

Census Data

The recent Census data provides a different perspective on the same phenomenon. The census only collects data in brackets of total household income, but the graph below shows the proportion of households in each bracket. South Australia has proportionately more households in the low-to-middle income brackets, but proportionately fewer in the brackets over $2,500 income per week. In particular, the top income bracket has the biggest gap of any bracket, with 12.4% of households nationally receiving more than $4,000 per week, by comparison with only 8% of South Australian households.

In short, the rich and super-rich don’t live in South Australia (or at least not in the same numbers as elsewhere in the country).

Line graph showing proportion of households in each ABS income bracket for Australia and SA.

Inequality in South Australia Over Time

While my previous post highlighted a long-term decline in South Australia’s share of national household income, the changes over time in inequality within South Australia have been less clear. The graph below shows both the P90/P10 ratio (on the left axis) and the income share of the lowest income quintile (right axis) since 1994/95. Some caution is needed as there were methodological and data series changes in 2003-04 and 2007-08, but both indicators have remained fairly steady over the last 25 years.

The blue line of the P90/P10 ratio does show an increase in inequality in South Australia between high and low income markers from the mid-1990s until the Global Financial Crisis in 2007-08. However, the income gap began closing from that point until 2013-14 when inequality began rising again – at the same time as the income share of the lowest quintile began to fall. It remains to be seen if this is simply a fluctuation or the beginning of a period of further rising inequality. But at a minimum, we can say that the level of inequality in South Australia is long-standing and not getting any better.

Line graph showing time series of two measures of inequality in South Australia: P90/P10 ratio, and the lowest income quintile share from 1995 to 2020.

Conclusions and Caveats

In summary, the data above shows that while South Australia is not getting an even share of national household income, inequality within SA is at least a little less than the national spread. Crucially though, both these phenomena appear to be driven at least in part by a proportionate lack of high-income households.

Again, all this data comes with caveats around the limitations of analysis by income quintile (and the lack of accounting for non-cash incomes). In particular, the crucial top 1% of income earners are invisible in the ABS data.

Perhaps though, the biggest caveat is that South Australia is not a singular entity and the analysis above is based on only one formulation of income inequality (between households). There are other important ways to consider income inequality beyond households, and there are also geographic differences within South Australia. These will be the subject of future posts in this series.