How fair is Australia’s welfare state?

Cross posted from  Australian Policy Online

IN ITS 28 May edition the Economist carried a long feature about Australia, praising our resilient economy, criticising the quality of our political discourse, and highlighting our social egalitarianism. “The Evolving Platypus: A Distinct Society, Perhaps Becoming Less So,” was the magazine’s summary of how we do things here.

The feature referred to an article in Policy, the journal of the Centre for Independent Studies, by David Alexander, a former senior adviser to Peter Costello. Under the title “Free and Fair: How Australia’s Low-Tax Egalitarianism Confounds the World,” Alexander argues that Australia offers a genuine alternative to both the low-spending but high-inequality United States and the high-taxing but egalitarian countries of Northern Europe. This “unique form of low-taxing egalitarianism,” he concludes, “is both more successful and more sustainable than other models.”

Is this characterisation of Australia’s social protection system accurate? Alexander presents a wide range of evidence to support these arguments, and it is not surprising that I agree with much of it, since one of his sources is a paper I wrote for a conference during the Henry Review of Australia’s tax system.

The most recent data on social spending in OECD countries shows that in 2007, the year before the global financial crisis, Australia spent 16 per cent of GDP on cash benefits (including pensions and unemployment payments, healthcare and community services) compared to an OECD average of just over 19 per cent. We actually spent a little less than the United States and Japan, and the only countries that spent substantially less than we did were lower-income countries like Mexico, Chile, Turkey and Korea.

In most rich countries, the welfare state is the largest single component of public spending and therefore the main determinant of how much tax income needs to be collected. About half of all the taxes collected in Australia are directed to social spending, but because we spend less than average we also have lower taxes than average. With taxes at about 27 per cent of GDP in 2008 compared to an OECD average of close to 35 per cent, Australia is the sixth lowest-taxing country in the OECD.

So it’s fair to say that we are a relatively low-taxing country compared to other rich nations, and to a significant extent this is because we have lower levels of welfare spending. But is this spending particularly egalitarian and are our taxes progressive?

To answer these questions, we need to look at how social spending and taxation is distributed across income groups. And to do that, the most up-to-date comparisons, using 2005 data, are in a 2008 OECD study, Growing Unequal? Income Distribution and Poverty in OECD Countries. (Although the data is six years old, the ways in which benefits and taxes are distributed tends not to change significantly over short periods of time, as was confirmed by an analysis prepared for the OECD Ministerial Meeting on Social Policy in May this year.)

It’s important to remember that the Australian social security system differs markedly from those in other OECD countries. In Europe, the United States and Japan, social security is financed by contributions from employers and employees, with benefits related to past earnings; this means that higher-income workers receive more generous benefits if they become unemployed or disabled or when they retire. By contrast, Australia’s flat-rate payments are financed from general taxation revenue, and there are no separate social security contributions; benefits are also income-tested or asset-tested, so payments reduce as other resources increase. The rationale for this approach is that it reduces poverty more efficiently by concentrating the available resources on the poor (“helping those most in need”) and minimises adverse incentives by limiting the overall level of spending and taxes.

Economist Nicholas Barr from the London School of Economics has pointed out that the main objective of social security systems in most countries is to provide insurance against risks like unemployment, disability and sickness, and to redistribute income across the life cycle, either to periods when individuals have greater needs (for example, when there are children in the household) or to periods when they would otherwise have lower incomes (such as in retirement). Barr describes this as the “piggy-bank objective.”

A second objective of the welfare state can be described as “taking from the rich to give to the poor” – or what Barr calls the “Robin Hood” motive – and Australia is the strongest example of a country emphasising this approach. Our system relies more heavily on income-testing and directs a higher share of benefits to lower-income groups than any other country in the OECD (and probably in the world). The poorest 20 per cent of the population receives nearly 42 per cent of all the money spent on social security; the richest 20 per cent receives only around 3 per cent. As a result, the poorest fifth receives twelve times as much in social benefits as the richest fifth, while in the United States the poorest get about one and a half times as much as the richest. At the furthest extreme are countries like Greece, where the rich are paid twice as much in benefits as the poorest 20 per cent, and Mexico and Turkey, where the rich receive five to ten times as much as the poor.

Because of these design features, Australia has the most “target efficient” system of social security benefits of any OECD country. For each dollar of spending on benefits our system reduces income inequality by about 50 per cent more than the United States, Denmark or Norway, twice as much as Korea, two and a half times as much as Japan or Italy, and three times as much as France.

Other countries that are similar to Australia in this regard include New Zealand, the United Kingdom and Ireland, and also Denmark and Finland. In fact, nearly all of the high-spending Scandinavian welfare states target to the poor more than does the United States.

Australia also has one of the most progressive systems of income taxes of any OECD country and, like our social benefit system, our income taxes are the most “efficient” at reducing inequality of any rich country. It is important to note that the progressivity of taxes in Australia is not a result of high taxes on the rich; rather, it’s due to the fact that lower-income groups in Australia pay much lower taxes than similar income groups in other countries (with the exception of the United States and Ireland).

The extent to which the Australian welfare state redistributes to the poor is determined by the interactions between the tax and social security systems, both in terms of the size of taxes collected and benefits paid and the distribution of these taxes and benefits. The chart shows an estimate of “net redistribution” to the poorest 20 per cent of the population in 2005. This is calculated by estimating the level of spending on social security benefits as a percentage of household disposable income and then taking account of how much of this goes to the poorest fifth. The same procedure is used to calculate how much tax is paid by people in that group, which is then subtracted from the benefits received to give “net redistribution to the poor.”

The chart shows that there are large differences in how countries redistribute income to low-income households, ranging from more than 5 per cent of household disposable income in Australia, Belgium, Denmark and Sweden, to around 2 per cent in Japan, Poland and the United States and less than 0.5 per cent in Switzerland and Korea. Nordic countries transfer large amounts of gross benefits to low-income people but also levy a significant amount in taxes from them; conversely, most English-speaking countries pay less generous benefits to the lowest-income households but partly offset this by levying lower taxes on them.

As a result, even though Australia spends below the OECD average on social security benefits, the distribution of benefits is so progressive, and the level of taxes paid by the poor is so low, that Australia redistributes more to the poorest 20 per cent of the population than any other OECD country except Denmark (which spends about 80 per cent more than Australia).

These figures suggest that in important respects the debate over Australia’s welfare state is misconceived. Following the federal budget earlier this year, for example, the issue of “middle-class welfare” attracted considerable media attention. But the OECD data shows that Australia actually has the lowest level of middle-class welfare of any OECD country, a position it has consistently held for at least the past thirty years. Organisations such as the Centre for Independent Studies have also argued that the Australian welfare state is marked by a high level of inefficient and wasteful “churning,” meaning that many people who use welfare state benefits and services finance most or all of what they receive through the taxes they pay themselves. But the OECD data shows that Australia has the lowest level of churning of any OECD country except Korea – and Korea only has lower churning because it has very little at all in the way of welfare payments. Claims by the Institute of Public Affairs that the main beneficiaries of the welfare state are the middle-class bureaucrats who administer the system are equally misleading.

While our social security system has a lot of strengths, this certainly does not mean that there are not real shortcomings to deal with, including the inadequacy of unemployment benefits and rental assistance. The fact that poor Australians get higher benefits than many poor people in European countries or the United States doesn’t actually help them pay their bills. It is always possible to be more efficient, and every year governments go through the laborious process of incrementally adjusting our benefit system to try to produce better outcomes. These changes often look like tedious fine-tuning, but the evidence suggests that it is changes of this sort that produce real improvements, rather than grandiose plans for completely replacing our welfare arrangements. •

This entry was posted in Economics and public policy. Bookmark the permalink.

16 Responses to How fair is Australia’s welfare state?

  1. wilful says:

    This post is only about the second or third one I’ve ever read (apart from ones by Ross Gittens and Tim Colebatch) that actually gives credit where credit’s due, and doesn’t bemoan the fact that we’re a hopeless useless high taxing and inefficient country.

    So thanks!

  2. Sinclair Davidson says:

    Peter – can you interpret the table please? It looks like less than 6 percent of household disposable income is welfare.

  3. Peter Whiteford says:


    These are the welfare payments received by the poorest 20% of the population expressed as a proportion of total household disposable income for all households (hence comparable across countries).

    If you expressed it as a proportion of the total household disposable income only of the poorest 20% then it would be much higher (over 100% in some countries) – but not necessarily as comparable because the poorest 20% get different shares of HDI in different countries.

  4. James Farrell says:

    That’s a very clear and useful piece, Peter, and a good achievement in distilling the complexities of welfare targetting to a single comparative statistic. (The Powerpoint presentation has lots of useful graphs too — thanks for linking that.)

    I gather that the numerator includes cash rebates for health services purchasesd in the market. Presumably a measure that somehow includes health (and education?) transfers in kind would give an more complete picture, for the purpose of comparing countries.

  5. conrad says:

    Thanks, that’s a wonderful post.

  6. Peter Whiteford says:


    On the health benefits, none of the countries include cash rebates for health services (either public or private) in their definition of cash benefits, or indeed in their definition of disposable income.

    Including health and education services would change the picture . The OECD Growing Unequal report does have a chapter on this but does not analyse it in the same way as I have done here.

    Norway, Sweden and Denmark spend much more than Australia or anywhere else for that matter on non-cash benefits, but their spending is a bit less pro-poor, so when you calculate the proportional effect of including services on income inequality, Australia , Sweden and Denmark reduce inequality by about a further 40% (measured by ratios of income between the richest and poorest quintile.) The USA is actually a bit more pro-poor in their health care spending, since at the time public spending on health was directed to the poor and the old, so they reduce inequality by 42%.

    However, indirect taxes which should also be included for comprehensiveness are far higher in the Nordic countries and more regressive than in Australia or the USA, and I am certain that if you included them Australia could very well be at the top of the OECD in terms of overall transfers to the poorest 20%.

    In a sense our problem is that we have higher inequality in market incomes which is why despite a very redistributive welfare state we only end up with average inequality.

  7. paul walter says:

    Yes, nice to finally find a thread on welfare without it being welfare-bashing hysteria.

  8. Fred Argy says:

    Well done, Peter.

  9. derrida derider says:

    “In a sense our problem is that we have higher inequality in market incomes …”

    Indeed, and this causes us all sorts of problems – not least in forcing us to have a very redistributive tax and welfare system (with its attendant costs) if we don’t want our society to look like Brazil or South Africa.

    I can think of several reasons for that market inequality. Industry structure is definitely one – Australia has often been described as a First World country with a Third World economy, and there is such a thing as a resource curse.

    But I wonder if in turn it’s not also because of a tax and welfare system that is obsessed with targeting. I think you can get some insight from the “optimal social insurance” literature about how protection against life risks actually pushes people towards middle market incomes, so what we would call “middle class welfare” may be functional. But I need to think it through.

    BTW, this is another great post from Peter.

  10. paul frijters says:

    another word for a highly targeted system is a system with a large poverty trap. Its the perennial catch 22: if you only give money if they are really poor then you implicitly impose a high level of taxation on efforts towards escaping that poverty.
    I miss that sense of trade-off in Peter’s otherwise great post, particularly when I think of the many posts I see on clubtroppo about groups that clearly are in poverty traps, such as the Aboriginal communities Ken often talks about (though there is obviously more at stake there than just the high EMTRs that come with a targeted system).

  11. Peter Whiteford says:

    Thanks everyone for kind comments.

    I’ll just address the issue of targeting raised by DD and Paul. This is something that I have intended to write about for a long while, but need to do more research before reaching firm conclusions.

    While we have by far the most targeted overall welfare system in the developed world, our EMTRs on moving from a position of zero market income to a substantial part-time or low-paid full-time job are actually amongst the lowest in the OECD. The only countries with substantially lower “participation tax rates” are places either with very little welfare for the poor – Greece and Italy – or much lower benefit levels like the USA.

    The reason for this is that we have reduced withdrawal rates for most categories of recipients. Uup until 1987 as I recall we effectively had a 100% withdrawal rate on unemployment payments from the first dollar for people renting privately (because we withdrew rent assistance at 100% from the first dollar of private income).

    With some exceptions, social assistance systems in other countries still have 100% withdrawal rates. Now in most other countries people may instead be entitled to social insurance benefits. While these are not asset-tested they are implictly income-tested – for example if you work more than a certain number of hours you no longer qualify as unemployed or to maintain eligibility for disability payments you also have to work fewer than a set number of hours.

    Where we do have higher EMTRs is for second earners, and we also tend to have higher EMTRs for people moving from part-time to full-time work – although even these are lower than in most other OECD countries.

    We also have generally high employment rates relative to other countries – even before the GFC. But while we have moderately low overall joblessness it is more concentrated in households where there is no one in paid work than nearly all other countries.

    In part this may reflect the higher EMTRs on second earners – except that a substantial proportion of our jobless are lone parents, where there is only one potential worker.

    So we appear to have significant poverty traps – but I don’t think that this is because of income-testing. Rather I think it is due in the past to the lack of activation of benefit recipients. For example, until 2006 lone parents were not required to look for work until their youngest child was 16. In Norway it was 10 years, and in other Scandinavian countries it was effectively around 3 years of age.

    So to me, the poverty trap was lack of activation and the lack of the supportive services that you need to make activation work, poor public transport and unsafe neighbourhoods for many, and the lack of job opportunities and relevant skills in the most disadvantaged groups.

    So while there are important reforms to be made to public housing income tests for example, I don’t think that the solution to poverty traps is lowering EMTRs, but in addressing these other barriers.

  12. paul frijters says:


    I agree that from zero to what I would call pocket-money is not the region of high EMTRs in Australia but my stylised understanding of the weird and wonderful world of the ‘Pet-caculations’ on the many different types of family arrangement (with or without young kids, old kids, rent assistance, day-care assistance, etc.) is that EMTRs are around 80% for a long range after pocket-money. That’s pretty high and would obviously be much lower if you adopted things like the US Earned Income Tax Credit Schemes that push the claw-back further up the income distribution.

    So whilst I agree that there are a lot of nuances in the basic targeting = poverty trap story, these nuances (i.e. EMTRs are not 100% at the start) cannot disguise that the quicker you take the welfare money back, the more targeted but also the more trapped.

    Its an empirical question as to whether EMTRs are really 100% at the start in the rest of the OECD. The world of housing subsidies itself is weird and wonderful (because they are administered by many different layers of government and subject to many caveats about family composition and area) so i would be interested to know where the latest figures on these are.

  13. derrida derider says:

    At least in recent decades the designers of our welfare system (I suppose that includes us, Peter) have tried to replace poverty traps with low-wage traps – that is, to put the high EMTRs inevitable in a targeted system into areas where they affect the “battlers” more than the non-employed.

    Now I’d like to think that was a clear-eyed response to the findings from the labour supply literature that elasticities (and therefore the welfare losses from adverse incentives) are far higher at the extensive margin of participation than at the intensive margin, so we prefer to encourage people to be trapped in a lousy job rather than be trapped in no job. I suspect instead that political economy rather than pure economics was what really enabled the change.

    Also, while this argument from supply elasticities is absolutely true in static partial equilibrium I worry that the dynamic (eg incentives to human capital investment) and general equilibrium (ie what types of jobs are offered) effects may have widened the dispersion of earnings. I still think a combination of changed industrial relations, changed industry structure (each of which fed on the other, BTW) plus increased rents from resources at the top end accounts for most of the rise in market inequality, but it may well not be the whole story.

    On the second earner issue, ya gotta be clear what you’re trying to achieve here. Fixing that problem would require Scandinavian institutions (free high quality childcare, individualised payment entitlements, etc) that will cost other taxpayers far more than the extra tax we’ll get from the new arrangements – and that’s before we attempt to value any loss of household production.

    You can argue for that on the grounds of social policy, horizontal equity or even non-utilitarian arguments (“labore est orare” or “arbeit macht frei” ). But you won’t be helping GDP much – quite the contrary.

  14. Pingback: More ladders, fewer snakes: Two proposals to reduce youth disadvantage « Department of Management and Marketing

  15. cameron says:

    It sounds really practical when I read about the welfare in Australia. It is obviously a good cause that the authorities are doing to ensure that the really “needy” people get as much as they need and the more fortunate ones get lesser as their assets are more. The authorities disregard the amount of taxes paid by the people and only take into consideration of their status at the moment of time and how much assets they own in their personal storage when they are in need of social security.

Leave a Reply

Your email address will not be published. Required fields are marked *

Notify me of followup comments via e-mail. You can also subscribe without commenting.