Image by Scott Cresswell CC 2.0 vial Flickr http://bit.ly/2feyxGi

This is the final part in a four-part series on the topic.

The media and political debate about “zero net taxpayers” relies on these static measures which provide a static “snapshot” of individuals or households in a single year. Yet each individual (and household) moves through a lifecourse in which most (although not all) will be earning income and paying taxes, or receiving benefits, at different levels over their lifecourse. Individuals and households also face different risks over their lifecourse.

A lifetime of taxes and benefits

Households with a head aged 65 years or more receive by far the greatest net benefits from government. In Australia, as elsewhere, a high proportion of them receive at least part of the age pension; they are heavy users of the healthcare system; and they pay very low levels of direct tax. Their low tax payments reflect their low overall incomes, together with special tax offsets and the concessional tax treatment of superannuation.

It’s important to remember that most people aged over sixty-five were net taxpayers when they were of working age, so classifying them as “zero net taxpayers” fails to take account of any contributions they made earlier in life.

Many European countries have a contributory public social security system. This makes clear that in these systems that identifying a group in the population as “zero net taxpayers” makes little sense, since by definition you have to have contributed in the past to receive benefits currently. Of course, this does not necessarily mean that your past contributions were sufficient to cover your lifetime payments. Contributory social insurance systems generate widespread recognition of redistribution across the lifecourse as one of the primary aims of the welfare state.

Even in Australia, the point in time data in Table 4 showed that more than 70% of working age households are positive net taxpayers: they are paying tax while working, and receiving more in benefits once they reach age 65.

Similarly, many low-income families with children may receive more in family benefits and education support than they pay in taxes but they paid more in taxes before they had children and they will pay more after their children have left home; in future years their children will pay taxes too. As in most other rich countries, public spending in Australia is an important instrument for redistributing resources across the lifecourse.

Taking account of the vicissitudes of life

There are other reasons why the distinction between low- and high-income earners (and taxpayers) that is implied in the concept of “zero net taxpayer” is not hard and fast. Australia’s longitudinal survey of households, the Household Income and Labour Dynamics in Australia (HILDA) Survey, shows that many people ascend and descend the income ladder over time.

Between 2001 and 2010, only 2.2% of the Australian population stayed at the same percentile of the income distribution. Around 21% went up more than two deciles, close to 30% went up by less than two deciles, 27% went down by less than two deciles and 21% went down by more than two deciles.

People rise up the income distribution because they leave study and get jobs or because they are promoted at work or because they marry or because their children leave home. People fall down the income distribution because they retire or become unemployed, become sick or disabled or separate from partners, or because they have children (in particular, this affects female income levels).

In fact, HILDA shows that only around 47% of those in the poorest 20% of the population in 2001 were still there in 2010. Similarly only around 47% of those in the richest 20% of the population in 2001 were still there in 2010. So if we go back to the figures cited by Adam Creighton that only the richest 20% of the population pay net tax after taking account of non-cash benefits, it should be recognised  that nearly half the net taxpayers at the beginning of one decade will not be in that position at the end of the period.  The longer the time period over which income is measured, the greater the proportion of the population will change their economic status. Households that stay permanently either at the very top or the very bottom of the income distribution are relatively uncommon.

Mobility is greatest in middle income groups because there it is possible to experience both rises or falls in income, whereas if you start at the top you can only go down, and if you start at the bottom you can only go up. Only a quarter of those in the middle 20% of the population in 2001 were still in the same income group in 2010, with a third being in a lower income group and close to 40% being in a higher income group.

A US study, The Lifecycle of the 47%, analysed longitudinal data for households with observations from the US Panel Survey of Income Dynamics. They found that over a period of 10 to 40 years, nearly 68% of households owed no federal tax in at least one year, approximately 78% receive some type of transfer in at least one year, and more than 58% receive transfers other than Social Security in at least one year. Of those who do not owe federal tax in any given year, 18% pay tax the following year, and 39% contribute within five years. Of those who receive transfers other than Social Security within a given year, nearly 44% stop receiving such transfers the next year, and more than 90% stop within ten years. They also pointed out that the bulk of transfer spending over time (63%) goes to people over 65 in the form of social security payments.

A recent Productivity Commission report also modelled the distribution of taxes and benefits across the lifecourse. At a point in time, the lowest income Australians, with private incomes less than $25,000 per year[1], on average receive social security benefits of more than $18,000 per year, while the high income group between $175,000 and $200,000 per year receive benefits of about $120 per year. The low income group effectively pay no income taxes, while the high income group pay nearly $48,000 per year in taxes, with even higher income families above $200,000 per year (including GST) of close to $100,000 per year.

The lifetime distribution of benefits and taxes differs quite significantly, however, particularly for social security benefits. The middle income groups – those with annualised lifetime incomes between $25,000 and $100, 000 per year receive significantly more over time than these groups do in a single year – averaging between $4,000 and $7,000 per year. This reflects the fact that many in this group will end up receiving a part age pension after the age of 65, that many have children at some stage of their lives and that they can also experience periods of unemployment or disability.

However, these same groups pay more income taxes over their lives while higher lifetime income groups pay lower lifetime average taxes than annual taxes. Nevertheless, the distribution of net taxes (taxes paid minus social security benefits received) remains progressive across income ranges. Rather than increasing from minus $16,000 per year to $55,000 per year as they do on a current basis, net taxes rise from minus $14,000 per year to around $41,000 per year for annualised lifetime income.

A final issue that arises from this analysis relates to the question of whether people can be characterised as “lifters” or “leaners” and relates to the idea that it is only the rich that effectively pay (net) taxes.  A lifecycle perspective shows that people whose lifetime annualised income is less than $25,000 actually pay more than 10% of their lifetime income in taxes (rather than near to zero), and this doesn’t include indirect taxes.

In contrast, middle income people over their lifetime receive far more in social security benefits than do people in these income brackets at a point in time. The implication is that a much wider range of people benefit from the welfare state and pay taxes to support it than is often acknowledged.

A different approach to estimating lifetime social security benefits received (but not taxes paid) is contained in the Department of Social Services’ recent Baseline Valuation Report  as part of the Government’s new “investment approach” to social welfare. This report takes the population of Australia in 2015, and on the basis of past patterns of receipt of payments, it projects the amount of money the population will be paid over the rest of their lives (and converts this into the present value of this lifetime spending, with a discount rate of 6%).

The population modelled in the report includes: around 5.7 million people currently receiving various income support payments (of whom 2.5 million are age pensioners); 2.3 million people not receiving income support payments but who receive other payments (mainly families receiving the Family Tax Benefit); around 3.9 million who were previously receiving payments but are not currently; and just under 12 million people who are not receiving any payments currently and have not in the past.

The lifetime valuation is about 44 times the total amount of payments in 2014-15 (A$109 billion). But it also includes people’s future age pension entitlements. In fact, more than half the total estimated lifetime spending ($4.8 trillion in discounted terms) will be on age pensions. The average lifetime cost per current client is made up of $150,000 in age pensions and $115,000 in all other benefits. For previous clients, the corresponding figures are $114,000 in age pensions and $60,000 in other payments. For the balance of the Australian population, it is $88,000 in age pensions and $77,000 in all other benefits.

The implication of this is that the majority of the Australian population would be expected to receive a social security payment at some point in their lives, with the payment most likely to be received being the Age Pension.  But currently most Australians are not receiving payments and those of working age more likely than not will be paying taxes.

Should we be worried that nearly half the population are “zero net taxpayers”?

Underlying these figures is a more fundamental question about what the concept of “zero net taxpayers” actually means. As pointed out by John Quiggin, generally speaking, total government spending is roughly the same amount as government raises in taxes and other revenue. That is, on average, the whole population pays zero net tax.[2]

Once we consider the distribution of taxes and of spending – a distribution with an average around zero – it is not surprising that a large share of the population receive more in spending than they pay in taxes. Indeed, it is difficult to think of a distribution of net taxes where less than 50 % of the population pay less than the average.

Alternatively, imagine a country where the only tax was a poll or head tax, in which everyone pays the same amount of money, and the only spending was on classic public goods of defence and public order. Assume that the benefits of this spending were the same for the whole population. In this completely non-redistributive tax-benefit system, everyone is a zero net taxpayer.

Alternatively, imagine either halving or doubling both taxes and spending. That is, imagine a government that is half as big as our current government; and a government that is twice as big. As long as the distribution of taxes and spending remained the same, the proportion of the population who are zero net taxpayers would also remain the same. So, whether or not some proportion of the Australian population receive more in terms of benefits than they pay in taxes in any year does not tell us whether taxes or benefits are too high, and it also does not necessarily tell us anything useful about the distribution of taxes and benefits.

This point is reinforced by the finding that the proportion of zero net taxpayers is rather similar across countries with very different tax and social security systems. The fact that Korea – a country with very low taxation and very low transfer spending – has roughly the same proportion of the population who are zero net taxpayers as Denmark, which has the highest level of taxes in the OECD, suggests that the concept is not useful in policy debates.

The share of “zero net taxpayers” in the population is in a sense an “accidental statistic”.  It is the by-product of other decisions – where we want to set the income tax threshold, what is the mix of indirect and direct taxes, what is an adequate level of benefits, and  what withdrawal rate do we want to set on payments – presumably with the aim that we don’t wish to discourage work effort or savings.

This raises the issue of what policies are implied by a concern with “zero net taxpayers”. If we actually believed that there are too many zero net taxpayers, what would we do about it? To go back to the example of a single person eligible for Newstart, if we wanted to reduce the number of people in this situation who are zero net taxpayers, there are just four ways of doing this:

1. Increase the share of people earning over $25,000 per year;

2. Reduce the level of Newstart – already widely acknowledged to be inadequate – so that the cut-out point is lower;

3. Increase the withdrawal rate on benefits so that the cut-out point is lower; or

4. Increase taxes on lower income groups, for example by lowering the tax-free threshold to zero or increasing the first tax rate.

It can be noted that (3) and (4) may be counter-productive, as they involve increasing effective marginal tax rates on social security recipients and the low paid, which may reduce their work effort. In this context, the examples of Greece and Italy, which have very low effective tax thresholds and a high level of undeclared activity suggest that we should be careful what we wish for.

In addition, increasing taxes on lower income groups by either cutting the tax threshold or increasing the first rate also increases taxes on higher income groups unless there are some compensatory cuts further up the income scale.  This would simply be a way of reducing the progressivity of the tax scale – if this is what some people concerned with the number of “zero net taxpayers” want to achieve they should say this directly.

If we think that family payments are the cause of the problem, then we should recognise that if we reversed the policy direction of the last 40 years and we moved the payments as they are back into the tax system as rebates or refundable tax credits then we would have no impact on the number of zero net taxpayers.  If we wanted to reduce the number of people who receive more in family payments than they pay in tax then we would need to either reduce the base level of payments, increase the income test withdrawal rate or increase the income taxes they pay.  As noted earlier, there is an “iron triangle” in policies related to income-tested transfers – if we want to reduce the payment cut-out point then we must either reduce the payment or increase the withdrawal rate.

Spending on family cash benefits in Australia as a per cent of GDP peaked in Australia in 2003 (apart from the temporary stimulus payments at the time of the GFC) and has since fallen by more than any other OECD country (by around 0.7% of GDP). The proportion of families with children receiving Family Tax Benefits has been shrinking, due to changes in the indexation of payments introduced by the Rudd Government and the freezing of the higher income test threshold.  So contra the arguments put in The Australian cited earlier, any increase in the share of “zero net taxpayers” since 2008 is unlikely to be related to more generous family payments.[3]

Framing the “zero net taxpayer” debate

Media reports of the share of “zero net taxpayers” in the Australian population are broadly accurate. However, neither the numbers nor the concept are a useful guide to developing public policy.

First, recent trends are largely the result of tax cuts both for those of working age, but particularly for people aged 65 and over, and also related to changes in the taxation of superannuation withdrawals. Those concerned by the number of zero net taxpayers are usually not proponents of higher taxes. It would be possible to reduce the progressivity of the tax system but this would require either increases at lower income levels or cuts only for high income earners, or a combination of both. Nobody has come out to advocate this, probably in recognition of its likely political unpopularity.

Instead, the problem has been framed as a result of over-generous welfare payments or too much “middle class welfare”. But ABS statistics show that the share of Australian population predominantly reliant on welfare payments has been falling for most of the last two decades and correspondingly, the share  who are completely “independent” of welfare  has risen from 40% to nearly 50% of households. In addition, by multiple definitions, Australia has continued to have the lowest level of “middle class welfare” spending on social security in the OECD for the past 35 years.

There does appear to have been an increase in the share of “zero net taxpayers” in the working-age population between the early 1980s and the mid-1990s, but not strongly after that. It is also true that income tests were liberalised for unemployment payments in the 1980s and 1990s with the aim of encouraging recipients into part-time work as a “stepping stone” to full-time work. But since then payment levels have been falling relative to wages, and cut-out points along with them.

To reduce benefit cut-out points, it is necessary to either cut the already inadequate level of working age payments or to increase benefit withdrawal rates with likely work disincentive effects for low paid and part-time workers. The 2014 Report of the Commission of Audit  did propose increases in withdrawal rates to 75 cents in the dollar for Newstart and for pensions, but they did not discuss the implications for incentives to work, particularly in an economy now much more marked by part-time and casual work and underemployment than in the past.

The repeated discussion of “zero net taxpayers” and the sharing of the underlying idea across a range of English-speaking countries – but apparently starting with a conservative think-tank in the United States – suggests that the use of this language is about framing concepts for political impact.

It may be the case that it is desirable to emphasise the link between paying taxes and receiving the benefits of a government spending to emphasise the reciprocity necessary to maintain a welfare state. In a sense, this is what the contributory social insurance principle emphasises and it is closer to the principles said to be underlying the universal, citizenship basis for entitlements in the Nordic social democratic welfare states.

But the recent trends to seeing a change in the tax mix as desirable policy directions tends to work in the other way. One of the factors behind cuts in income tax after 2000 was the introduction of the Goods and Services Tax which also led to increases in benefits and reductions in withdrawal rates. If we measure “zero net taxpayers” without accounting for the effects of a shift in the tax mix then we are simply taking an internally inconsistent approach to thinking about tax and welfare reform.

In discussing the 9 May article about Duncan Storrar, Greg Jericho in The Guardian pointed out thatThe headline in Wednesday’s Australian –– reflects a longstanding view of the conservative side of politics and the media that we should worry about the rise of those paying no “net tax” – where you balance the amount of tax paid with benefits received.”  The proliferation of articles and opinion pieces using the concept of “zero net taxpayers” suggests that it strikes a chord – presumably most amongst those who believe they do not fall into this group.

Other terms that have been used contemporaneously or sometimes slightly earlier – “lifters and leaners” in Australia, “makers and takers” in the United States, “strivers and skivers”  in the United Kingdom are much more emotive. They are also more likely to be recognised as pejorative labelling of the sort Herbert Gans identified in Labelling the Poor (1995).

The concept of “zero net taxpayer”, in contrast, appears to be based on objective statistics and an apparently common-sense perspective about financial unaffordability. But like those other phrases, it is also a label that does not help us to understand the causes of the issue and whether it is something that politicians and the public should be concerned with. Most importantly, it does not help us assess the alternative policies that could be pursued to actually improve both public finances and outcomes for individual and families.

[1] The Productivity Commission report did not use income groups of the same size, so the low income group comprise about 35% of family units, while those with incomes above $175,000 per year accounted for less than 10 % of family units.

[2] If across all levels of government, there is a deficit, then net taxes are below zero and, perhaps, future generations may need to pay more tax because current generations are paying negative net taxes.

[3] In addition, the argument in The Australian that child care subsidies contribute to the problem is incorrect as child care assistance is not counted in the income surveys as a cash payment.

The full article on which this post is based can be found at the TTPI website.

Leave a comment

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

*