Tax and transfer

With the welfare system being in the news I thought I share a book chapter I wrote on the topic a couple of years ago. It appeared in a volume edited by Gary Johns and published by Connor Court.


Milton Friedman has described the introduction of the welfare state in 1880s Germany as a combination of paternalism and shrewd politics.[1] Over the last 130 years the welfare state has grown dramatically across western Europe and its offshoots in the new world. That growth in government is now widely recognised has having imposed huge fiscal costs and social problems in many of those societies – as Margaret Thatcher has long recognised.[2]

Welfare benefits, distributed with little or no consideration of their effects on behaviour, encouraged illegitimacy, facilitated the breakdown of families, and replaced incentives favouring work and self-reliance with perverse encouragement for idleness and cheating.

While the shrewd politics has remained, paternalism as the underlying rationale for welfare has changed. Now there is an argument for ‘egalitarianism’ or ‘equity’ underpinning welfare. The apparently adverse consequences of inequality have most recently been set out in the 2009 book The Spirit Level by Richard Wilkinson and Kate Pickett.[3] Their conclusion is that society would be better off with higher levels of equality rather than lower levels of equality. To be sure, at some extreme levels of inequality this must be true – yet their argument related to the levels of inequality that are experienced in a sub-set of OECD countries.[4]

The solutions usually offered to ‘problems of inequality’ are for high levels of progressive taxation – a research report prepared for the Australian Council of Trade Unions argues, ‘Progressive taxation is a key policy mechanism for addressing wealth inequality’ – high levels government spending, and labour market intervention.[5] As Terry Carney and Peter Hanks indicate the fiscal costs of these ‘solutions’ are substantial so government has also undertaken ‘more preventive and more targeted measures’.[6] Means-tested welfare has problems of its own and the preventative measures have resulted in an expansion of the state.

Carney and Hanks trace out differing attitudes towards egalitarianism.[7] Classical liberals, they argue, view poverty as a product of personal choice and poor government policy. The solutions to poverty involve personal charity and increased economic growth. Another group, described as ‘true conservatives’, believe that inequality arises due to innate cultural diversity or rewards ‘elite’ performance, and responses – not solutions – involve reinforcing societal institutions such as the family, church, and other voluntary community groups. These views, it must be said, are now largely in the minority. The dominant view, at least, amongst elite opinion is that ‘inequality is a product of market failure, discrimination, [and] historical imbalances in the distribution of property’.[8] Peter Saunders, however, points to the puzzle that welfare spending has increased faster than has economic growth.[9] By definition, one might expect that the need for welfare would decline as the economy grew.

In order to provide insight into the growth of welfare spending, I collected data from the Australian Bureau of Statistics and annual Budget Papers on government spending. Four areas of government spending can be classified as being ‘welfare’ – Education, Health, Social Security and Welfare, and Housing and Community Development. These four items made up some 58.3 percent of government spending in 2000-01 and increased to 62.3 percent in 2008-09 before declining to 59.1 percent this financial year. That represents an increase from $104 billion in 2000-01 to $216 billion in 2011-12.[10] At the same time the participation rate in the labour force has increased, while unemployment has fallen, and average weekly earnings increased. Even a stricter definition of welfare to just include the budget item ‘Social Security and Welfare’ shows massive growth from $66.9 billion to $121.9 billion over the past 11 years – that is an increase of 82 percent.

The tax and spend aspects of the welfare state are highly wasteful and cannot be said to have been successful. As Friedman wrote over 30 years ago, ‘The repeated failure of well-intentioned programs is not an accident. It is not simply the result of mistakes of execution. The failure is deeply rooted in the use of bad means to achieve good objectives.’[11] Friedrich von Hayek is less tolerant; he describes any policy that aims to promote equality as ‘a wholly illusory ideal, and any attempt concretely to realize it apt to produce a nightmare.’[12] He makes the very important point that it is not intentions that matter but rather in ‘doing what in fact most benefits others’.[13]

Australian Egalitarianism and progressive taxation

Fred Argy defines economic egalitarianism as creating a welfare safety net, widely sharing the gains from productivity improvements, addressing the structural causes of inequality and proving a ‘voice’ for workers. He also talks about sacrifices being distributed by the ‘ability of bear’.[14] This is a somewhat pragmatic definition of egalitarianism and cannot be described as being in the Rawlsian tradition.[15] Argy argues, for example, complete equality is ‘neither achievable nor desirable’, not because the economic costs are prohibitive, but because people are inherently different. If people do have so very different skills and interests and risk tolerance and the like, it is not clear why equality is a desirable outcome at all. In other words, policies that promote equality are in search of a justification.

Despite this lack of principle Australia maintains a highly progressive tax system and redistributes a substantial amount of income from high income earners to middle and low-income earners.[16] The Australian Council of Trade Unions (ACTU) would like to increase the level of progressivity in the tax system.

Sixty per cent of Australians support raising taxes for higher income earners and the same proportion want lower taxes for middle income earners, according to new research commissioned by trade unions. [17]

The difficulty with this particular argument is that the Australian tax system is already very progressive. In August 2007, then Treasurer Peter Costello made the claim that ‘something like 60 per cent of Australian families are paying no net tax. That is, their family tax benefits are outweighing their tax liabilities and that’s something that’s actually really helped Australian families.’[18] In a careful analysis Peter Whiteford compares cash transfers and the direct tax burden and reports that the poorest quintile received ‘more than 30 times as much in cash transfers as they paid in direct taxes’ in 2005 – that figure was down from over 55 times in 2000.[19]

What is clear is that the personal income tax burden is very highly progressive. Each year the Australian Taxation Office releases statistical information relating to the tax system. The data is released with a lag and the 2008-09 financial year is the latest available. Using those statistics it is possible to determine the individual tax burden by income level.


Figure One: Individual Income Tax Burden

Davidson Fig 1

Source: ATO Tax Statistics, Author calculations

By collecting data from the Australian Taxation Office (ATO) and arbitrarily segmenting it into three segments – the bottom twenty-five percent, the middle fifty percent and the top twenty-five percent – and calculating how much net tax they pay, it is very clear that the tax burden has increased for the top end of the income distribution. The top twenty-five percent of taxpayers paid 67 percent of net personal income tax in 2008-09. The bottom twenty-five percent paid just 2.4 percent. It is possible the further refine the data; the top 5 percent of taxpayers paid 33.7 percent of net income tax in 2008-09, with an (average) effective tax rate of 42.24 percent.

Of course, that is how a progressive income tax system is meant to operate. High income earners are meant to pay more income tax than do low income earners, and in that regard the Australian tax system works as it was designed to work.

It could be argued that the income tax system is highly progressive in order to compensate for regressive indirect taxation so that the overall net effect of the tax system is roughly proportional. That proposition, however, is purely empirical. Once sin taxes (taxes of alcohol and tobacco) are removed from the equation it is clear that the Australian tax system is highly progressive even after accounting for indirect taxation.[20] In a 2008 analysis into inequality the OECD found that Australia and the United States have the most progressive tax system in the OECD collecting ‘the most tax from people in the top decile relative to the share of market income that they earn’.[21]

How does progressive taxation benefit the poor?

Peter Whiteford makes the point that the inter-relationship between tax and welfare has long been recognised in Australia. In particular, taxation is how the welfare state is financed and welfare spending constitutes the bulk of government spending.[22] To the extent that a progressive tax system raised more revenue than, say, a proportional tax system then it would be clear how progressive taxation benefited the poor. This, of course, would come at the expense of high-income earners, so we would also have to take the costs of raising revenue via a progressive tax into account but, in principle, the case could be made.

Figure two shows the distribution of the share of taxable income and the distribution of the share of net income tax paid by income quantiles for 2008-09. As can be seen taxpayers below the 80th percentile pay have a lower share of net income tax than their share of taxable income while taxpayers above the 80th percentile pay more in net income tax than their share of taxable income. The 80th percentile, of course, is well above average earnings. In other words, a progressive income tax does not only favour the poor, it also favours middle income earners who pay less tax than they would under a proportional tax system.

Figure Two: Distribution of Taxable Income and Net Tax (2008-09)

Davidson Fig 2

Source: ATO Tax Statistics

The great American economist Deirdre McCloskey specifically argues that the policy of taxing the rich so that the poor might prosper has failed.[23] She gives two reasons why that is so. First, following Robert Nozick, taxation is a form of slavery and is associated with various economic costs.[24] She also argues that most social spending is not designed to benefit the poor, but rather the middle class. She illustrates this point by means of a back of the envelope calculation. The US government collects about 25 percent of US GDP in taxation and if a third of that amount were distributed to the 34 million Americans described as being ‘poor’ then each one would receive about US$30,000. A family of four – two parent and two children – would have a combined income of US$120,000 and could hardly be described as being ‘poor’ with that level of income. She concludes by observing, ‘So it must not be true that the government’s taxes go mainly, or even much at all, to the poor’.

It is possible to replicate that thought experiment in Australia. According to the Budget Papers the federal government intends to spend $121.9 billion in the 2011-12 financial year on Social Security and Welfare. That sum of money – 33.3 percent of the federal budget – does not include health, education or housing – items that also constitute welfare spending. Making the assumption that there could be as many as two and a half million poor Australians that level of expenditure would be enough to transfer over $48,000 to each of them. Even if we considered that the government should distribute money to the bottom twenty percent of Australians that would still amount to over $29,000 each.

This is a practical illustration of something economists call Director’s law. This is the notion that government spending is primarily designed to benefit the middle class or middle income earners. The consequence of government spending money on middle income earners is ‘tax-welfare churn’ or middle-class welfare. Peter Saunders describes this situation as ‘many households find themselves paying money into the system only to get it straight back again in the form of government payments and services’.[25]

One particular problem associated with churn over and above the pure waste and inefficiency is the existence of high so-called effective marginal tax rates. Much of the Australian welfare system is means tested. As individuals income rises so they lose welfare benefits. The interaction between higher rates of taxation and lower welfare benefits means that work incentives are undermined.

Many recipients may not view middle class welfare as welfare per se. Former Prime Minister John Howard told the American Enterprise Institute in 2008 that, ‘The taxation system should generously recognise the cost of raising children. This is not middle class welfare. It is merely a taxation system with some semblance of social vision’.[26] The difficulty with this particular argument is that any social vision somehow exempts government spending from being welfare.

Saunders makes the argument that everyone is paying for everyone else’s welfare ‘but few of us seem to be aware of it’.[27] That could be true; voters could be under a fiscal illusion that allows them to think that they are getting a bargain from the government. [28] Of course to the extent that middle income earners pay less of the tax share and receive relatively more of the government spending share they are getting a bargain. On the other hand, government policy in a democracy is determined by the median voter and, as McCloskey tells us, the median voter ‘is not enthusiastic about helping the poor’.[29] Following McCloskey, it is important to realise that this churn is a feature of our democracy, not a bug in the welfare system.

What can be done?

It is only through an understanding that the expansion of the welfare system benefits the median voter and not the poor per se that we can appreciate why the welfare system continues to expand. The massive welfare system and the progressive income tax exist to redistribute income to the middle class. As Hayek as argued, ‘Much of what is today done in the name of ‘social justice’ is thus not only unjust but also unsocial in the true sense of the word: it amounts simply to the protection of entrenched interests’.[30] Unfortunately, that ‘entrenched interest’ is the median voter in a democracy, in other words, the status quo has majoritarian support. This makes it much harder to unwind the welfare state.

Proposals to unwind the welfare state usually involve tightening up the conditions for redistribution, while Saunders has proposed an opt-out feature.[31] Given the democratic demand for welfare, it is not clear that either of these approaches would gain wide-spread support. Rather the costs and failures of the welfare system should be emphasised. It is important to remember that fiscal illusion plays an important part in the demand for welfare. For as long as voters have the perception that welfare is a value proposition, it is unsurprising that welfare expands beyond the provision of a safety net for those individuals in actual need.

It is quite clear that the welfare system does not promote egalitarianism and that Australians are not egalitarian per se. Saunders has shown that Australians value a ‘fair go’ and are meritocratic but that is a long way from supporting massive income redistribution.[32] That suggests that greed and self-interest can be harnessed to drive welfare reform. If the costs associated with churn were emphasised it is quite possible that meaningful reform could occur. Rather than tighten welfare eligibility and so increase effective marginal tax rates, tax rates themselves could be lowered. This would increase the returns to work and lift many individuals out of the tax churn system. Other reforms such as providing tax deductions for school fees rather than funding schools themselves would provide an incentive for parents to invest in their children’s education and make the education system responsive to parents rather than bureaucrats.

Making the costs of the welfare system highly transparent would go a long way to reducing those costs over time. On the other hand, however, if the electorate chooses to bear the costs of the welfare system even after those costs become transparent then, at least, an honest choice will have been made.


The provision of ‘social justice’ via the welfare system remains shrewd politics. It provides for a massive redistribution while doing little for the actual poor. Of course, the poor do get some benefit from the welfare but not nearly as much as we might imagine. At a time of economic prosperity and continued economic growth the amount of welfare should be declining, yet we see massive growth.

An insatiable growth in the welfare system via demands for social justice is inconsistent with prosperity in the long-run. While nobody should begrudge a genuine safety net for people in actual need, it should come as no surprise that the last ten years have seen a decline in productivity just as the welfare state has expanded faster than economic growth. Ultimately the best guarantor of prosperity for all is a prosperous economy.


[1] Milton and Rose Friedman, 1980, Free to choose: A personal statement, Penguin Books, pg. 125.

[2] Margaret Thatcher, 1993, The Downing Street years, HarperCollins, pg. 8.

[3] Richard Wilkinson and Kate Pickett, 2009, The spirit level: Why more equal societies almost always do better, Allen Lane.

[4] For a comprehensive critique of Wilkinson and Pickett see Peter Saunders, 2011, When prophecy fails, The Centre for Independent Studies.

[5] David, Neal, Cassandra Govan, Mike Norton, and Dan Ariely, 2011, Australian attitudes towards wealth inequality and progressive taxation, A report prepared for the ACTU.

[6] Terry Carney and Peter Hanks, 1994, Social Security in Australia, Oxford University Press, pg. 2.

[7] Carny and Hanks, as above, pg. 1 – 2.

[8] Carny and Hanks, as above, pg. 2.

[9] Peter Saunders, 2004, Australia’s welfare habit and how to kick it, Duffy and Snellgrove, pg. 4 – 6.

[10] The $104 billion would be $134 billion in 2010 dollars.

[11] Milton and Rose Friedman, as above, pg. 124.

[12] Friedrich von Hayek, 1976, The Mirage of Social Justice: Law, legislation and liberty volume 2, The University of Chicago Press, pg. 85.

[13] Hayek, as above, pg. 72.

[14] Fred Argy, 2003, Where to from here? Australian egalitarianism under threat, Allen & Unwin, pg. 166.

[15] Rawls argues that inequality can only be tolerated to the extent that it benefits the least advantaged in society.

[16] Peter Whiteford suggests that almost 50 percent of Australian government spending can be described as social spending. See Peter Whiteford, 2010, The Australian tax-transfer system: Architecture and outcomes, The Economic Record, 86: 528 – 544, pg. 528. As shown above, however, that figure is under-estimated.

[17] Sabra Lane, 2011, ACTU pushes for tax reform, ABC News,

[18] The 7.30 Report, 2007, Peter Costello talks about the latest interest rate rise, August 8, 2007,

[19] Peter Whiteford, as above, pg. 532.

[20] Sinclair Davidson, 2004, Taxation with misrepresentation: Australia’s revenue lobby in denial, Policy 20(4): 31 – 37.

[21] OECD, 2008, Growing unequal: Income distribution and poverty in OECD countries, Paris, pg. 106.

[22] Whiteford, as above, pg. 528.

[23] Deirdre McCloskey, 2006, The bourgeois virtues: Ethics for an age of commerce, The University of Chicago Press, pg. 43 – 46.

[24] These costs are described in Alex Robson, 2006, How high taxation makes us poorer, in Peter Saunders (ed), Taxploitation: The Case for income tax reform, Centre for Independent Studies.

[25] Peter Saunders, 2007, The government giveth and the government taketh away: Tax-welfare churning and the case for welfare state opt-outs, The Centre for Independent Studies, pg. 23.

[26] John Howard, 2008, Keeping faith with our common values, The 2008 Irving Kristol Lecture, American Enterprise Institute, available at

[27] Saunders, 2007, as above, pg. 5.

[28] Fiscal illusion is a set of strategies that the state adopts in the hope of altering voters’ fiscal consciousness. In particular the state hopes to convince voters that the tax burden is not nearly as onerous as it is, while the benefits of public spending are greater than they are. See Sinclair Davidson, 2011, Fiscal illusion: How big government makes tax look small, in Robert Carling (ed), Taxploitation II: Tax reform for incentive, productivity and economic Growth, The Centre for Independent Studies.

[29] McCloskey, as above, pg. 46.

[30] Hayek, as above, pg. 96.

[31] Saunders, 2007, as above, chapter 5.

[32] Saunders, 2004, as above, pg. 75 – 78.

This entry was posted in Economics and economy, Hypocrisy of progressives. Bookmark the permalink.

13 Responses to Tax and transfer

  1. craig

    Oh no, there’s even global warming in the taxable income graph! Damn hockey sticks!

  2. Tel

    I recommend never using the word “welfare” to mean government wealth transfer. The word “welfare” already has a meaning, unrelated to any government. Remember Orwell’s warning about allowing other people to define your words for you.

    Given that we are talking about wealth transfer, the only “net tax” graph that makes sense is one where all government payments (both direct and indirect) are included, thus the lower end of the graph should be negative and the higher end should be positive, such that they almost balance to zero (of course, some money is lost in the transfer process). We can’t get an exact measure of how much benefit each income group is getting from medicine, education, etc. but we could get a reasonable estimate.

    I think you would find the people around the median end up paying tax and receiving benefits in roughly equal amounts. Whether they would have spent the money on exactly those things that government spends it on for them is another matter… maybe we could estimate some inefficiency factor to allow for that, but that’s quite arbitrary, and unknowable.

  3. Abraham

    The ‘progressives’ peddling the ‘wealth distribution’ nonsense seem incapable of comprehending the fact wealth is created by those who sensibly manage their income to maximise its growth and returns on its investment.

    If some individuals decide to seek welfare benefits instead of employment, then based on what moral principle are those who are employed expected to subsidise their existence?

    Is it even possible to prevent welfarism from killing Australia?

  4. Ant

    People on welfare, particularly those on long term welfare, should automatically relinquish their right to vote.

    They can hardly complain about removing this conflict of interest.

    It is in the interest of those who are employed to have ultimate control of who gets elected to office, because it will then be in the interest of those who seek elected office to enhance employment and productive opportunities.

    A recipe for stagnation and decline is creating a system that rewards most those whose ambition it is to exploit unproductive and lazy behaviour.

  5. Rabz

    People on welfare, particularly those on long term welfare, should automatically relinquish their right to vote

    The first step towards achieving this outcome?

    The abolition of compulsory voting.

  6. Rabz

    of course, some money is lost in the transfer process

    I’d argue that ‘churn’ is a significant and unjustifiable cost, given the intrinsic inefficiency of bureaucracies.

    It goes back to the old truism that people will spend their own money far more efficiently than any extortionist government claiming to act on their behalf.

  7. .

    Rabz #1164868, posted on January 24, 2014 at 8:13 am
    People on welfare, particularly those on long term welfare, should automatically relinquish their right to vote

    The first step towards achieving this outcome?

    The abolition of compulsory voting.

    Not necessary – many would choose not to vote.

  8. Rabz

    many would choose not to vote

    And even more would choose not to if they didn’t have to front up at a polling booth.

  9. Andrew of Randwick

    ADAM CREIGHTON wrote last March about ending-the-tax-money-go-round, and provides data and graphics.
    If the highest quintile of income earners ever get tired of carrying the rest of us we will truly feel the pain – but thankfully they are only 20%, and we are 80% – gotta love democracy!!!
    Seriously, your thought experiment on government ineffectively distributing to poor, could be done another way. $121.9 billion on Social Security and Welfare suffers friction going through the Government of 10-20% – so $18 billion a year is lost by the churn, even before you consider efficacy.

  10. .


    I’m actually making the same point as you, it probably wasn’t clear the way I wrote it.

  11. Johno

    Unfortunately, that ‘entrenched interest’ is the median voter in a democracy, in other words, the status quo has majoritarian support. This makes it much harder to unwind the welfare state.

    Kevin Andrew’s announcement that his review of the government’s direct wealth transfer system will NOT include the largest transfer item – aged pensions – is a clear demonstration of how hard it is to unwind the government’s direct wealth transfer system.

    Given that this is a very tame government that lacks the resolve to tackle hard problems, Andrew’s review will probably end up making the usual tame recommendations of tightening elegantly conditions.

    Your main point, that government direct wealth transfers are driven by the needs of political processes, not the poor, is spot on and any reform proposals need to keep that point front and centre.

  12. Andrew

    I think the term “spending” should be banned when referring to tax rebates (to net taxpayers, at least).

    If I’m asked “did you have private health cover? If so, pay less tax” then the result is a reduction in tax – because I deserve to pay less tax. I’ve taken myself out of the public hospital system. Calling it “spending” just plays into the hands of the likes of Blabbersac who calls increasing tax a “savings measure.” As did Gillard, every time her “structural saves” resulted in more tax.

    A rebate is a reduction in revenue measures. Abolition is a revenue measure.

    If the accounts were properly presented, Howard666 did not run anywhere near as “big” a govt as he’s accused of, or a high taxing one.

  13. Squirrel

    A politician who genuinely wanted to do something about this (as opposed to playing half-smart winner/loser politics with it) could begin by talking about the billions per year which we spend on the administration of the churn system through federal, state, territory and local governments, NGOs etc. etc. and then perhaps sketch out a simpler (although not so much as to frighten the horses) system and the administrative savings which it could produce. Some of the ingredients for this seem to be coming together with the Commission of Audit and the broader review process promised for Commonwealth-State roles and responsibilities, but it will still take dexterous handling for anything worthwhile to happen and endure.

Comments are closed.