Lessons from tax reform past
Address by Dr Ken Henry
Chair - Australia's Future Tax System Review Panel
and Secretary to the Treasury
Address to the Committee for Economic Development of Australia (CEDA)
15 October 2009
I would like to start by thanking David Byers for the invitation to speak here today.
This will be my last speech on tax reform before the Review Panel on Australia's Future Tax System hands its final report to the Government in December.
It has been a huge process for the Panel, that has engaged not only the dedicated officers of the Treasury and other departments, but many thousands of economists, businesspeople, community leaders and interested individuals. Obviously, the Review coincided with a major economic crisis. The fact that we have been able to address that crisis without neglecting the needs of the future illustrates a fundamental truth about Australia's economic success story over the past quarter of a century. It is, as Paul Kelly has argued in his most recent book, The March of Patriots, a success that is due to the determination of Australians of all political parties and backgrounds to fashion a modern, reformed economy that would become the envy of much of the world.
Tax reform has played a part in this economic transformation. I wouldn't call the tax changes that have occurred 'revolutionary' but they have been very substantial — improving the lives of Australians in many ways. And — as we look forward, hopefully, to the beginning of a new and sustained period of tax reform — I thought it would be useful today to reflect on past tax reform efforts and outline what I think are the lessons we should learn from them.
I don't offer these as scientific or economic proofs, although I think they are consistent with good economic analysis and recent research. What they represent are the conclusions of an experienced tax reformer:
- one who has spent his professional life advising governments about tax reform
- and one who has a big stake in seeing its findings make their way into the statute books for the benefit of the Australian people.
So I offer five lessons and a conclusion.
Lesson 1: The case for reform should be compelling
The first of my lessons is that the case we put for tax reform should be compelling.
We can't expect citizens to be enthusiastic about tax reform just because it has the word 'reform' in it. Reform may be regarded as self-evidently good in the think tanks and university departments in which many of this audience moves. But for everyday people, reform needs and deserves powerful justification.
Tax reform means changing prices in the economy, affecting existing incomes and wealth. If it is meaningful reform, it means that prices will change so that people have more incentive to do things that improve their own lives and those of others in the community. More incentive to work, invest, save; but also to consider how their actions impact on others, including others not yet born, such as through the impact they have on the environment.
Advocates of tax reform tend to focus on these benefits. But tax reform also has costs.
Many people can easily adjust to the new prices tax reform will bring. Some will adjust their sources of income, others will get new jobs or renegotiate existing ones, and some will adjust to the new pattern of investment. For many, there might be no perceptible impact at all. However, some people — those with stores of value based on previous prices and incomes — can find themselves worse off, especially in the short-term.
Whether people should be compensated for such 'windfall losses' is largely a question of fairness. Perhaps some – maybe most — should reasonably be expected to have anticipated that reform was inevitable. But we should not have this expectation for all. For example, some people may have only recently bought a tax preferred asset — long after the original concession was granted — at a price that reflects an expectation of the effect of the tax concession lasting indefinitely. This means they would suffer a loss of wealth in a reform that moved to a more neutral tax system, possibly making what most analysts would consider to be a fair and coherent tax reform seem, to them, both unfair and arbitrary.1 Others, on the other hand, may have accumulated their wealth from socially harmful activities and might not be considered worthy of compensation. But what about those who have earned their wealth from activities indispensable to prosperity, such as innovation or saving? Some will find change easy; others, such as the elderly, less so. Yet even if compensation is considered desirable in principle, it may be difficult to target those who need it.
Accordingly, a threshold issue is whether the size of potential compensation is so large, or its delivery so complex, that the tax reform may not be worth doing at all.
One of the best papers I have seen on this topic highlights how the benefits of moving away from the existing income tax system to a comprehensive consumption tax base quickly evaporate when it is concluded that compensation is required for certain groups.2 For example, such a change might mean governments want to provide additional assistance to low-income workers, who are worse off from the loss of progressivity, and the elderly, whose assets can buy less. The costs of providing this compensation can potentially outweigh the benefits of the proposed tax reform.
This highlights how the case for tax reform must be so strong, so compelling, that it outweighs probable compensation costs. It also means that the manner in which compensation is provided can determine whether a reform is a success or failure.
Can we find such a case in the current economy?
One possibility is road pricing reform. There would be few areas in economics where such a clear and rational set of policy directions have so consistently lagged in practice.
Most of the time, most cars impose minimal costs on other road users. However, when vehicles drive on a congested road they impose costs on other drivers. Each driver thinks of their own need to get to their destination, not considering how, by taking up space on the road, they impinge on the ability of other drivers to do so. There is no means for one driver to coordinate with others, to bargain about who should have priority, so that they can all be better off. This results in a predictable 'tragedy of the commons' which is estimated to waste around $9 billion a year in avoidable congestion costs, increasing to around $20 billion by 2020.3 Such costs will only increase with faster population and economic growth.
In the face of these costs, why have we stuck to the traditional 'fuel tax and rego' model for roads, when sensible pricing seems to offer such large benefits?
Personally, I think it is the inherent difficulty in deriving a feasible compensation framework.
Many people will have bought houses, cars and even taken jobs based on the existing set of road prices — explicit and implicit. I mention implicit prices because while many roads appear free, they are actually expensive in terms of time and taxes. Still, many people think that a change in road prices, or the burden of compliance, will necessarily make them worse off. And, with the history of urban transport services in some parts of Australia, this might not be an unreasonable assessment.
The case for change needs to be made. It is fast becoming one of the biggest public policy issues of the age.
We need innovative ways of dealing with the community's distributional concerns. For example, some truck operators might support road pricing as long as the costs they pay are reflected in better roads — their 'compensation' is a better functioning road network. They would also like assurance that the compliance costs of road pricing will be low and remain low. At least one automobile association supports road pricing as long as road users are 'compensated' by the abolition of fuel excise.4
There are some instructive examples overseas. In London, commuters were 'compensated' through additional funding for public transport. An innovative study in Seattle gave some drivers credits to pay to drive on congested roads and let them cash in the savings they made by driving at off-peak times or choosing other modes of transport.
Reforms like road pricing will be difficult, but these overseas examples show that it is possible.
Compensation is inherently a political problem, as it is the political process that identifies those affected and those in need. Our political history reveals a capacity to manage massive change, given sufficient time.
If the tax structure from early last century prevailed today, we would have to raise $40 billion from excise and $230 billion from tariffs to meet today's revenue demand. At that rate the excise on a schooner of beer would be around 7 times what it is today. And I shudder to think how much a television set would cost.5
Past politicians had the vision and foresight to avoid this outcome and the Australian people the good sense to agree with them. A similar vision is needed to make future tax reforms viable too — and it's our job to make the case for it.
Lesson 2: Uncertainty and risk impose costs too
The second lesson is not to frighten people by portraying tax reform as the equivalent of a short sharp jab.
The parents in the audience will understand that no one ever convinced a child to go quietly to the dentist by telling them it will hurt like hell but not for long.
There's an economically rational reason why people reject sudden upheavals in public policy; the mere prospect of change can cause uncertainty for taxpayers. Those of us who were around before the announcement of the introduction of Capital Gains Tax in September 1985 would well remember the almost hysterical atmosphere of the time. Scare campaigns about 'taxing the family home' were just as diabolical — and plain wrong — then, as they are now. Many people, including some of the most vulnerable in society, were made to feel less secure in their homes because of such misinformation. That is an extreme example. But there is a serious truth here; even by merely raising the prospect of change, tax reform imposes risk; and of course, risk can harm wellbeing.
Notwithstanding such risk, there are always calls to change the tax system. Our Review has received over 1000 submissions, none of which suggested leaving the tax system unchanged. This is somewhat refreshing to economic policy advisors like me who are keen to explore new ideas. But while the prospect of change makes refreshing reading for tax policy enthusiasts, it also has a very real impact on the economy and people's lives.
Change cannot be avoided. But it can, and should, be made coherent.
One way the Review can help reduce risk is by promoting a feasible long-term path for tax reform — in the way that the Asprey Review from 1975 created a reform framework that influenced the nation's tax reform direction for more than two and a half decades.
Like Asprey, the Panel understands that tax systems should evolve coherently towards some 'ideal system'. The tax system should evolve and its evolution should be coherent. Coherent evolution will be assisted by a clear articulation of an ideal system. The Asprey Review6 provided a reform agenda largely consistent with a comprehensive income tax benchmark. But the logic underpinning this benchmark has now largely been destroyed amongst tax policy specialists, and there is some uncertainty about the way forward.7
We have to be 'up-front' about this. But, at the same time, it is our responsibility to chart a clear course for future reform efforts.
Lesson 3: Simplicity often gets left behind
My third lesson is that we mustn't forget the importance of simplifying the tax system.
The history of Australian tax reform reveals a number of themes vying for pre-eminence. The very first taxes imposed by the colonies were customs and excise duties on rum and some other 'necessities'. Raising revenue was the overriding priority. It was the need for revenue to finance the First World War that led to the introduction of income tax in 1915, and the Second World War that led to its expansion in the 1940s. Not until the 1970s, and the Asprey Review, did equity gain prominence. This equity theme was highly influential in the 1980s, driving the development of general anti-avoidance provisions, capital gains tax, fringe benefits tax and the substantiation of claimed deductions. The 1980s also evidenced an emerging interest in efficiency issues, with a cut in the top personal marginal tax rate from 60 per cent to 49 per cent and the introduction of dividend imputation. In the 1990s and the current decade efficiency considerations became even more influential, as evidenced by company and personal income tax rate cuts and the Goods and Services Tax (GST).
What is striking about this potted history is how little attention simplicity has received in past reform efforts. While the relative importance of different principles seems to wax and wane, ensuring the system is simple to understand, and simple to comply with, has not. Our history of tax reform seems to reveal the following judgements: Is simplicity desirable? Yes. Essential? No. Achievable? Absolutely not.
But I would question these judgements, for two reasons. First, if simplicity is not the focus of a review, it will never be achieved. A complex system is the result of trade-offs along various constraints that are only relieved — and then only briefly — during a reform process. Second, ironically and perversely, a simpler system is more likely to be equitable, efficient and better able sustainably to raise revenue.
Consider how we tax individuals on their personal labour income. For tax to be fair it should reflect all of the resources a person has available to consume. For income tax to be efficient, people should not face incentives to derive their income from certain sources simply for a tax advantage. And, obviously, the more sources of income that are taxed, the more sustainable is tax revenue. Yet, currently, there are more than 40 offsets which can reduce taxable labour income that make the system very complex. And I am not aware of any study that has explored comprehensively the equity implications of this entire set of offsets. The task remains in front of us to explain to people how replacing complexity with simplicity will benefit them directly and benefit the nation as well.
Lesson 4: Perceptions of equity matter a lot
My fourth lesson is that equity is a powerful and under-utilised argument for tax reform.
I have been in the tax reform business for a long time, and have had the opportunity to observe how people respond to different arguments put to them about tax reform. People tend to take a keen interest in information that relates directly to their own interests. For example, the distributional tables outlining the impact of the GST were the most 'thumbed' part of the documentation, certainly by those Treasury officers answering phone queries. Of course, it helped that every individual and family represented across all income levels appeared better off.
Appeals to self interest are a seductive means of prosecuting reform. Several submissions to the Review emphasise 'how much larger GDP will be', perhaps in the hope that many people will confuse GDP gains with increases in their wage rates. In this way, the argument seeks to buy support for a particular reform proposal.
Perhaps it does.
But there are also costs emanating from this line of argument. By following it, we feed the mistaken belief that policy analysts are simply crude materialists. Reform, of course, does not just mean increasing GDP or increasing real wages. I will revisit this topic in a moment.
But this line of argument also fails to appeal — it might even be distasteful - to those people who are concerned about their community and the impact of tax reform on others. Equity has a strong base in all religions, most philosophies, good economic textbooks, and in the pubs and cafes of every Australian suburb and country town. On the Left and the Right; among Catholics, Protestants, Muslims, Buddhists, agnostics and atheists; even among economists — Australians care about social equity. Many deeply so. Self interest is crucially important to people's attitude to tax reform — but I believe strongly that we underestimate the power of equity to promote worthwhile tax change.
This means equity, or perhaps more correctly, distribution, matters. While many people find the concept of efficiency amorphous or even alienating, re-framing the same issue on equity grounds can sometimes make a powerful case for reform.
One example is the fringe benefits tax (FBT) debate in the 1980s. To tax policy specialists, a clear efficiency case existed for FBT reform. Resources were flowing to businesses and industries better able to renumerate their staff in fringe benefits. But this argument was not the one that claimed the most column inches. The argument was won by highlighting how unfair it was for 'fat cats' to enjoy tax-free lunches. Such images are more persuasive than graphs illustrating national 'deadweight' or welfare losses.
Framing some modern day issues in distributional language might also make the case for tax reform more engaging. Removing barriers to labour market participation can increase GDP. But it can also be presented as alleviating poverty, by providing people with the means of social inclusion. For example, lone parents can face barriers to employment that can work against their longer term earning potential and their full social participation. A worrying trend is that, in the last year, joblessness among families with children increased at a much faster rate than joblessness generally. As Professor Peter Whiteford noted last week, joblessness among families with children is a major contributor to overall income inequality, and leads to poorer health, lower educational attainment, elevated financial stress and increased risks of violence for lone parents. His assessment is one that I agree with — that government policy needs carefully to balance shorter term needs for an adequate income with longer term needs for lone parents to re-engage with work during their child's preschool years.
Consider also the case for reform to housing assistance. There is strong demand for public housing, as it provides greater support than does rent assistance. But this contributes to access being rationed by queuing. Queuing for public housing can harm a person's wellbeing. While queuing, people have an incentive not to take on jobs that make their income exceed the eligibility threshold. Also, eligibility is generally tied to a house. This may be one of the reasons why people on public housing waiting lists have lower employment outcomes than those actually in public housing.8 Assistance based on an individual's need, that supports the development, rather than depreciation, of capabilities would more effectively target poverty — a more efficient social housing market means a more equitable one.
While I have referred to the importance of 'framing', I am not saying that equity should be used simply as a 'sell job' or to 'sell' reforms that are otherwise unpleasant. I am saying that equity is usually a more intuitive and, indeed, a more meaningful way for many in the community to understand and become interested in tax reform.
Tax reform doesn't just mean increasing people's incomes. It means improving resource allocation to address social problems. This can be done directly by using taxes to target social costs, such as discouraging congestion or pollution. Or, it can be done indirectly by using the tax dividend from higher GDP to fund spending needs. A larger tax base means an increased ability to fund better health and education services, and improved capabilities for the disadvantaged, and better roads. Couched in these more approachable terms, tax reform avoids being dismissed as an exercise in arcane statistical composition.
Lesson 5: Governments need effective tools to improve peoples' lives
My fifth, and last, lesson is that we need tax and transfer reform because we need to give our governments the right tools to cope with the big changes going on in the world.
Not so long ago it was respectable to argue that the best counsel for governments was that they 'do nothing'. The Global Financial Crisis may have changed this forever — I don't know. But when designing a tax and transfer system for the future, it would be prudent to anticipate that our governments of the future will feel compelled to do many things. This provides a case for developing effective and targeted policy instruments, ready to be used, when needed.
How can I be sure? Well, it has always been the case. Look at the amazing period of reform in the mid 1940s when the Commonwealth took over and expanded some State programs that continue today. This includes child endowment in 1941 and unemployment and sickness benefits in 1945. Such instruments are more effective at providing assistance to those in need than Victorian-era soup kitchens or ad hoc state government programs. Another example is the introduction of the superannuation guarantee in 1992 to respond to the pressures of an ageing population. Governments want policy instruments that can finance a dignified retirement. Today, the Carbon Pollution Reduction Scheme (CPRS) is another example of a targeted policy tool, designed to address a problem perceived to be of major importance. Whether you are a climate change sceptic or not, there is no denying that market-based mechanisms, such as the CPRS, are better than existing programs for reducing carbon emissions.
Without these instruments, one wonders what costly, inequitable and, most likely, ineffective alternatives governments would have adopted to meet their policy objectives.
Some of the existing State taxes provide a clue. Many of these taxes – such as those on insurance policies and the transfer of property and motor vehicles – are generally recognised as being highly inefficient, sometimes even by the States themselves.
Yet in the absence of less distorting, more broadly-based taxes, the States have continued to use them to finance spending on critical services such as hospitals, schools and police. These taxes may not be noticed by people as much as other taxes, perhaps because they are only paid irregularly, possibly at times when people are thinking about the benefits they might receive from buying the house, or car or insurance. But what should we make of the fact that the burden of these taxes actually falls quite heavily on people who are more disenfranchised — such as those needing to adjust their housing circumstances due to relationship breakdown; or people in the lowest income quintile who spend a higher proportion of their income on motor vehicle purchases than the highest quintile; or those who bear risk where they would prefer to be insured but aren't, because of tax.
Taxes levied on broader bases would be more efficient policy tools, probably more equitable and certainly more transparent ways of raising revenue. Without such tools, governments will continue to rely on bad taxes to achieve their spending objectives.
In the course of this review, the Panel has identified a number of future opportunities and challenges governments will want to address:
- the ageing of the population, posing challenges for the financing of retirement incomes and of increasing health and aged care needs;
- environmental degradation;
- globalisation, particularly the ability to attract and retain investment capital in a diversified industrial economy;
- technological change, including such things as the increasing opportunities to improve the productivity of road infrastructure;
- improving the federation, policy formulation and administration; and
- the continuing effectiveness of government, including how well governments deliver affordable housing.
Governments will need modern tools to meet these modern challenges that lie ahead.
A conclusion: The goal is to improve wellbeing
Successful tax reform is not just about increasing GDP or revenue, or making the system easier to understand, or more sustainable, or fairer, or better able to assist governments to address various social problems. It is concerned with all of these things. Successful tax reform means improving the wellbeing of the Australian people.
This means evaluating reform proposals against the many dimensions that go into ensuring Australians have the capability to participate fully in society; that they are able to enjoy the freedom that comes with being able to make choices they have reason to value. The Review Panel considers that there is a set of tax and transfer reforms — compelling, clear, simple, fair and effective — that can, over time, deliver that goal. The tax and transfer system is not the only means available to governments to increase the wellbeing of our people, but it is surely the most important.
It is in that way I would like to think history might reflect on this particular tax reform effort.
1 Feldstein, M 1976 'On the Theory of Tax Reform' Journal of Public Economics, vol. 6 (July- August 1976), p77-104.
2 Altig, D; Auerbach, A; Kotlikoff, L; Smetters, K, and Walliser, J 2001 'Simulating Fundamental Tax Reform in the United States', The American Economic Review, Vol. 91, No. 3 (Jun., 2001), pp. 574-595
3 Bureau of Transport and Regional Economics 2007, Working Paper 71, Estimating urban traffic congestion cost trends for Australian Cities, 2007. Data relates to all capital cities in 2005.
4 Australian Automobile Association, Reform of Fuel Tax and other Motoring Taxes and Charges: submission to Australia's Future Tax System Review of Taxation, October 2008 http://taxreview.treasury.gov.au/content/submissions/pre_14_november_2008/Australian_Automotive_Association.pdf
5 In 1902-03, customs revenue raised 85 per cent of revenue and excise raised the remainder. Beer excise represented 34 per cent of excise collections. If we assume that the Federation tax mix prevailed, then beer excise in 2009-10 would represent $13.8 billion – absent, of course, any behavioural effects. This would be a 675 per cent increase on expected 2009-10 beer excise collections of $2,050 million.
7 Auerbach, AJ 2009, 'Directions in Tax and Transfer Theory'. Paper presented at Australia's Future Tax System Review Tax Policy Conference, 18 -19 June 2009. http://taxreview.treasury.gov.au/content/html/conference/downloads/attachment_03_Draft_Alan_Auerbach_paper.pdf
8 Dockery, M, Ong, R, Whelan, S & Wood, G, 2008, Australian Housing and Urban Research Institute Research Report 9 'The relationship between public housing wait lists, public housing tenure and labour market outcomes' http://ahuri.edu.au/publications/download/nrv1_research_paper_9.