Tax Reform in New Zealand
Submission to the 2001 Tax Review
by Keith Rankin
Dept. Accounting Law Finance
Private Bag 92025
31 July 2001
The fiscal history of a people is above all an essential part of its general history. ... Nothing shows so clearly the character of a society and its civilisation, as does the fiscal policy it adopts.
Joseph Schumpeter 
The 20 June 2001 Issues Paper on New Zealand's taxation system raised a number of questions about income tax. The most controversial of these was the proposal to tax the imputed interest income on home equity. Partly as a result of this controversy, the more general discussion of income tax - in particular the arguments favouring a flat or two-step scale - was minimal.
My principal concern is to take a more comprehensive view of income tax in general, and to suggest ways in which changes in our perception of income tax can open the door to future reform.
My second concern relates to the 'straw man' argument used by the Review Committee to dismiss universal basic income (UBI). It is unfortunate that loose thinking - by advocates and by sceptics - about this means of integrating the tax and benefit systems has made the concept too easy to dismiss.
Income, the National Accounts and Home Equity
In our system of national accounts, the gross domestic product (GDP) of a nation represents the final value of new goods and services produced for sale. It is equal, in essence, to the total income paid to the owners of the factors of production (classically defined as land, labour and capital) and to the government by way of indirect taxes.
The GDP can be thought of as the income tax base. Ideally the conceptual framework used in our system of national accounts (SNA) should be the same as that used by the Inland Revenue (IRD).
Problems arise however because not all interest and rents arise from new production. For example, we may rent or mortgage a second-hand car. Where this is done by say a rental car company there is some new production: the services provided by the company. But consider a purely private arrangement. My second hand car is sitting in the garage. So I rent it to my neighbour. I do not produce anything, yet I receive a rental income, which is tax liable.
Further, in order to raise a sum of capital, I might mortgage this same car to my mother. I pay her interest, which is tax liable. We now have created two income streams without producing anything. Should either or both of these income streams be taxed? At present both are.
Yet, my first car - the one I drive myself - can be said to have both an imputed rental income (I'm renting it to myself) and an imputed interest income (I'm paying interest to myself). Why should the same taxes not be liable re my first car as my second?
In our national accounts, we regard this kind of income as production when it applies to the renting and mortgaging of second-hand houses, but not when it applies to other second hand goods. This is because we regard the house as a capital good and hence as a productive resource; as a factor of production. The argument therefore follows that we should tax the imputed income arising from owner-equity. But if we tax both the rental income and the interest income - eg where a landlord has a 100% mortgage - should we not tax two streams of imputed income when there is neither a lease nor a mortgage?
The IRD at present takes a different approach to the SNA. While it ignores imputed income on used housing, it taxes both rental income and interest income on used cars and other property items that the national accounts regard as second-hand consumer goods.
Generating consistency between the definitions of income adopted by Statistics New Zealand and the definitions adopted by Inland Revenue will be a difficult, possibly futile, exercise.
One resolution might be to regard all second hand property the same, and to ensure that it is subject to zero net taxation. In other words, going back to my first example, my mother will pay tax on the interest she receives from the mortgaging of my second car, but I will receive an equal tax credit. And I will pay tax on the rental income I receive from my neighbour's use of my second car, while my neighbour receives an equal tax credit. (This would be a bit of an administrative nightmare if we have - as we have now - multiple rates of income tax.)
Following this process, the tax I would pay on the imputed income from my first car (the one I own and drive myself) will be exactly equal to the tax credit that I would be entitled to. In other words, there should be no tax on imputed interest and rental income.
There are some implications to note, however, if we start to treat houses as consumer goods (ie like cars) rather than as capital. Already a rising percentage of our nation's savings is being used to support consumer credit rather than business credit. If we redefine residential property acquisition as consumption rather than as investment expenditure, a huge chunk of our savings will be seen to be funding present consumption rather than future economic growth. This considerably weakens the argument for using tax concessions as a way of encouraging more personal savings.
The Equivalence of Income and Production
There is another way to resolve the problem that I have outlined. It turns out that it resolves other difficult problems as well; in particular, the problem that a market society has of giving adequate recognition to the productive resources that are in the public domain.
Income is a derivative of production. We learn that from our circular flow diagrams in Economics 101. So we could re-label income tax as production tax.
A production tax sounds like it might be inefficient; it sounds like a disincentive to produce. Of course it is no more of a disincentive to produce than an income tax, because an income tax is a production tax. A production/income tax is a revenue tax rather than a way of fixing a negative externality. What matters is that it applies equally to everyone.
Like all revenue taxes, a production tax must be designed to raise revenue without creating undesired disincentives. To this end, horizontal equity is an absolute must. Vertical equity (as is implicit in our present graduated scale of income tax) may or may not be an important issue. (I believe that vertical equity is not a desirable attribute of a revenue tax. Vertical equity is the domain of social welfare.)
Ideally, horizontal equity exists on a global scale (Simon 2000). The fact that it doesn't constitutes the main single reason for calls to have a low flat rate of income tax. The argument for low income tax rates generally boils down to the alleged need to maintain or create international competitive advantages through price-cutting (ie cost short-cuts).
I would like to leave aside that issue for now. My argument in the next section considers our nation to be a closed economy.
A Theoretical Basis for a Production Tax
When we think of income tax as being what it really is, a production tax, then we come to realise that income taxes are actually paid by firms rather than households. Indeed, a number of tax reforms in the 20th century reflected that reality.
The introduction of PAYE ('pay as you earn') meant that 'our' taxes were deducted without us ever receiving the payment in the first place. Likewise deductions of tax on interest earned from bank deposits and the like is deducted before we receive it, as withholding tax. And we now receive dividends as after-tax income; profits are taxed at source as company income tax.
Production/income tax is simply a fraction of the value-added by the activities of any firm or non-profit producer.
What we call 'company tax' today is a small fraction of the total of production tax that is paid. We could, if we wished, call the whole lot 'company tax'. Then we - the workers, equityholders, creditors and landlords who derive income from a company's productive activities - would stop talking about taxes paid by the company as if they were part of our personal income.
All factor incomes would be net of tax, as they already are in reality, given PAYE, withholding tax and dividend imputations. Personal income tax would become a thing of the past.
I must emphasise here that, so far, I have not suggested any new taxes or the removal of existing taxes. This exercise has been one of re-labelling, of redefining income tax, and of re-education re the popular concept of 'personal income tax'.
The Basic Income / Flat Tax Approach
One problem with accounting for income tax as production tax is that at present, different people pay 'personal' income tax at different rates. A production tax could not work that way. A garment producer should pay the same amount of production tax as a chartered accountancy firm if its contribution to GDP is the same. The employers would not care that the workers in the garment factory were, on average, on much lower incomes than the accountants. A proper production tax is a flat tax - ie a proportional tax.
This need not mean that garment workers become worse off than at present. Society should care, though in a technically different way from the way in which society cares today. All it means is that we account for existing tax concessions as tax credits, and pay them, transparently, as tax credits.
Of course, as we do this, we might start to question the pattern of tax credits payable. In particular we would ask why the employees of the accountancy firm get more tax credits per person than the employees of the garment manufacturer. (The impact of the present 15% and 21% concessionary tax rates is greater in dollar terms for high income recipients. These concessions however represent a greater proportion of the total income paid to low income recipients such as garment machinists.)
The best way to understand this point is to consider the tax scale that existed in New Zealand from July 1998 to March 2000. The rates were 15%, 21% and 33%. The top rate became effective at annual 'gross' incomes of $38,000. And the top rate was also the company tax rate (which remains at 33% today). We should note that the concept of 'gross' earnings disappears once we formally account for all income tax as company tax.
For any persons receiving an annual 'gross' salary in excess of $38,000, their net incomes could be calculated by the following simple formula:
Net Income = 67% of Gross Income plus $5,130
Under production tax accounting, an accountant 'grossing' $50,000 would in fact be earning $33,500 and receiving, in addition, a tax credit of $5,130. Total disposable income would remain at $38,630.
For any persons receiving an annual 'gross' wage of less than $38,000, their net incomes could be calculated by the following simple formula:
Net Income = 67% of Gross Income plus $5,130 minus X
Under production tax accounting, a garment worker 'grossing' $25,000 would in fact be earning $16,750 and receiving, in addition, a tax credit of $3,570. (In this example, X = $1,560.) Total disposable income would be unchanged at $20,320.
All employers would pay 33% of their value-added as production tax, but they might retain some of that to fund their employees' tax credits. (Pay slips would say "earnings" and "tax credits"; there would be no "deductions".) Alternatively, the tax credits could be paid in the same way that Independent Family Tax Credits are paid today, by direct credit from the IRD.
The most obvious question that arises from these examples is to ask just why the garment workers should receive smaller tax credits than accountants. The short answer would be because that's what happens now. The longer answer is that the new accounting process that I've outlined exposes this as a major anomaly; a significant breech of the principle of horizontal equity.
The sensible solution is to adopt the Basic Income / Flat Tax approach (Atkinson 1995). Thus every adult (including full-time caregivers and students) would receive a 'universal tax credit' of $5,130 ($99 per week). Obviously this would require some additional tax revenue, but probably not nearly as much as we might at first think. For workers presently 'grossing' $25,000 per annum, it would mean an additional annual tax credit of at most $1,560 (ie 'X'). (In most cases, existing Accommodation Supplements and Family Support tax credits would substantially reduce if not eliminate this X-amount shortfall in their tax credits.) For full-time beneficiaries, we would simply re-label the first $5,130 of their benefits as their universal tax credit. (More contentiously, such a universal tax credit could be labelled a 'universal basic income' [ref Rankin 1997, van Parijs 2000]. )
How would we fund the additional tax credits without increasing or decreasing the government's budget surplus? In part, we can fund them from growth. Indeed, once made universal, the obvious idea would be to adjust our universal tax credits to the size of the economy - they could be reset each year as a fixed percentage of GDP per capita. Alternatively we could fund additional tax credits by raising the flat rate of production/income tax by a few percentage points. Or we could adopt a new indirect tax (a carbon tax? a Tobin tax? a financial transactions tax?). We could raise an existing consumption tax; eg goods and services tax (GST) or tobacco tax. Or we could simply settle, in the short run, for a universal tax credit fractionally lower than the present $5,130 per annum (eg $5,000 per annum).
A second question that arises from my analysis relates to the modification of the tax scale in 2000; the addition of a 39% marginal tax rate for incomes over $60,000. We could account for a 'personal' tax rate in excess of the company tax rate as a "high income personal tax surcharge". My preference though would be to abolish such tax steps in favour of a flat rate of production tax somewhere in the range of 34-39 cents per dollar of value-added.
Future Implications of the Reinvention of Income Tax as Production Tax
Firstly, I'll address the problem that I started with. All personal income - including interest and rent - would be paid net of tax. Banks of course would pay tax as a proportion of their contribution to GDP. Likewise landlords would only pay tax to the extent that they produce something.
The wider question relates to our understanding of public property rights.
The Physiocrats (the world's first school of economists, started in France by François Quesnay in the 1750s) believed that one-third of the net product (what Karl Marx called surplus value) belonged to the sovereign (ref Einaudi 1933) who was seen as a super-landlord. Hence the commons could be analysed as if they were the exclusive property of the sovereign. The sovereign would fund his nation's public administration and also, if so inclined, a social wage from 'his' income. (The Physiocrats believed that the entire net product was due to the natural powers of the land, whereas Marx believed that surplus value was entirely due to the efforts of workers.)
Thomas Paine believed that in the process of economic development, much public (ie common) property was (and had to have been) appropriated as private property. The corollary of this appropriation, he argued in Agrarian Justice (1796), was that part of the income stream that had been diverted into private hands still belonged to society as a whole. Paine argued that it should be paid out as two pensions to each citizen; as a capital sum on their 21st birthdays, and when they reach retirement age.
We can think of a universal tax credit as a modern equivalent of Paine's pensions, payable weekly or fortnightly rather than just twice in a lifetime.
Much as these 18th century proponents of laissez-faire and liberty did, we can think of public revenue as a simple fraction of the national income. Public revenue arises in part from the public ownership of businesses (such as Mighty River Power or Television New Zealand). In part it arises from public lands or fisheries that might be leased to private producers. But in larger part, public property rights relate to that huge domain of productive resources than are by their very nature indivisible: the physical environment (including the oxygen in the air and the water that all production depends on); the physical and legal infrastructures that we inherited collectively from our parents, grandparents and perhaps their grandparents before that; the public knowledge that drives any knowledge economy; the social capital and other cultural attributes of a successful economy; the outputs of all levels of government; the institutions of any society.
This public domain is an inventory of "free" resources - of public capital goods - the most recent of which is the Internet. Because part of our economic nature is to give, this inventory of public resources and goodwill expands on a daily basis; eg through the outputs of the voluntary sector, unpaid creative/intellectual endeavours (such as submissions to public taskforces), the parents who raise new generations on a voluntary basis. However, these commons are also subject to depreciation: the pollution of the environment, the loss of biodiversity, the pressures placed on people to do ever more hours of paid work thereby limiting their output of unpaid work.
We can now think of the largest component of public revenue, not as an appropriation of private wealth by the state, but as a royalty or rental payment by firms for their 'free' use of the many productive resources that are in the public domain.
This only requires a minor extension of the existing economic theory of public goods and of private property. Theorems such as those of Ronald Coase work best at creating economic efficiency when both private and public property rights are well defined.
Following on from this quite straightforward adaptation of neoclassical economics, we can understand taxation as a factor payment much like wages and interest. It represents the public "half" of the capitalist economy (refer Heilbroner & Milberg 1995). Indeed the capitalist economy is like a two-sided coin. We overstate the private side and ignore the public side at our long-run peril. Production/income tax is the rent private capitalists pay to their public landlords.
I favour the term social wage to describe the entire fund of public revenue (eg including the profits of state owned enterprises [SOEs] and local authority trading enterprises [LATEs]). Universal tax credits can now be understood to be one of a number of things that are funded from the social wage.
The social wage is thus the equal property of every member of a society. How we distribute social wage goods, though, is a collective decision, a social contract, made through the processes of democratic central and local government. It seems quite natural to pay out a part of the social wage as an individual cash dividend. (Arguments do spring up about the entitlement of children. One argument is that universal tax credits that might otherwise be paid into children's bank accounts should instead be used to part-fund pre-school, primary and secondary education.)
Once we properly account for our social wage, we can then start to ask the pertinant questions about how, in future, we should spend it.
For now, the most important point that needs to be made is that the equal attribution of the social wage (ie of public revenue that is the equal property of all) is distribution, not redistribution.
We can nevertheless understand our social welfare system as a redistributive mechanism; a redistribution within the social wage. We choose to pay more than $5,130 per annum in cash to certain groups of people: the retired, sole parents, persons with disabilities etc. This, and not graduated personal tax scales, is our means of providing vertical equity, of treating different people with different needs differently. We should bear in mind that only those benefits that exceed $5,130 per annum should be classed as transfer payments. And they would only be transfers to the extent that they exceed $5,130. By drawing an individual tax credit out of our new way of accounting for income tax, we can fund the universalisation of that tax credit in part by significantly reducing the amount of transfer payments that we make.
In 1991 in a policy discussion paper that I wrote (Rankin 1991) - and that received widespread publicity - I introduced the term Universal Basic Income as a name for a tax reform along BI / FT lines. Such an approach to fiscal reform, as I then advocated and still do (eg Rankin 1997), is by definition fiscally neutral. Among other things, it complements and simplifies social welfare reform. Understood as I meant it to be understood, it neither replaces social welfare, nor is it a fiscal extravagance. It is not a rigid "one-size-fits-all" approach to fiscal reform. Nor is it "something for nothing". On that last point, a universal tax credit (or UBI) represents a form of public property income - a "patrimony" (Simon 2000). It is no more "something for nothing" than is the receipt of interest from a private inheritance.
A reform of the income tax system that moves towards a proportional (ie flat) tax scale can be neither equitable nor efficient unless it is accompanied by a basic income. In the absence of an explicit basic income (or the implicit but variable basic income that we do have at present) the economic surpluses gained from the use of public domain resources are in effect looted by the strongest private interests, be they capital or labour or a mixture of both.
History shows that the growth of the public inputs to production has outstripped the growth of private inputs. We produce much more per person today in part because of increased physical health and private capital (including human capital). But by far and away, as any exercise in growth accounting shows, it's the increase in knowledge and our collective institutions and civilised habits that enable the extraordinary growth of productivity and average per capita incomes that have come with economic development.
It is likely to be the same in the future. This suggests that the social wage - and hence the rate of production tax - should be increasing as a proportion of GDP, as indeed it did for the first 85 percent of the twentieth century. I believe that the trend to lower tax rates since c.1985 is an historical anomaly. Indeed I hope so, or else the public wealth upon which modern economic growth depends will become substantially degraded, leading to a reversal of the process of economic development.
We only have to do a thought experiment to understand the link between productivity and publicness. Imagine that, in another 200 years, productivity is so high that, on average, we only have to do a few hours paid work in order to sustain an average standard of living equal to today's average. The majority of people would only be able to enjoy a reasonable standard of living if say 90% of the world's GDP was distributed by way of a social wage that included a cash dividend. Even then - ie despite a flat tax of perhaps 90 cents in the dollar - those with access to private income would be fabulously wealthy.
Nobel laureate Herbert Simon (2000) suggests that already as much as 90% of our productive power is in the public domain. If so, the royalties we receive at present by way of income tax are pitifully low. (Of course any increases in production/income tax that we might want to make on account of this new perspective on capitalism would have to be incremental in practice. A huge one-off increase in income tax cannot be countenanced.)
The "International Competitiveness" Problem
My final comments relate to the problems of managing public wealth in an anarchic global economy. By anarchic, I am simply stating the obvious; that there is no de jure world government to provide global public goods, to regulate and otherwise address problems of global market failure, to levy global taxes.
What happens if we increase our tax rate in line with future productivity gains, but other countries do not? If we follow a pure free trade model, we will lay off workers and import goods from those countries which pay far too small a social wage. We would be forced to cut our social wage too, in order to compete with those countries. (This indeed could become a kind of global "race to the bottom", "prisoners dilemma" or "tragedy of the commons" like that which led to the Great Depression of the 1930s.)
While free trade is part of a general solution to maximise global economic efficiency, under certain conditions of global market failure, pure free trade can aggravate the problem while the introduction of trade restrictions can enable a country to find its own fiscal solution, to find its own private-public balance. To this end we can learn from Great Britain's recovery from the Great Depression. A number of economic historians (eg Glynn and Booth 1987,Tim Rooth 1993) have shown just how necessary the introduction of protective tariffs was to the consumer led recovery of Britain's economy.
Today the economics profession accepts, by and large, that it is appropriate for a nation to protect itself from imports of products that are cheap only because the labour force of the supplying country was exploited (eg child labour), the environment was degraded in order to make such products (eg clear-felling of rain forests), or inappropriate export subsidies were paid. In such cases the consumers of the imports benefit from the trade at the expense of the efficiency of the global economy. A responsible country does require its producers to bear costs that their foreign competitors may choose to evade, but does not allow its producers to fail on account of such competition.
Fiscal reform is built first and foremost upon the principle of horizontal equity; that equals are treated equally. The global trading system only works properly if competitors from all countries pay all the resource costs that they incur. Ultimately, to have an economically free global society, the principles of sound and just public finance must be applied globally; albeit sensitively.
In the meantime, if some countries are lead the way by reinterpreting income tax as production tax and by raising their production tax rates as total factor productivity increases, they may need to adopt some modest forms of import protection, such as tariffs or a Tobin tax.
Many of the intractable problems of public finance can be resolved by re-interpreting income tax as production tax. In the short run, the incidence of taxation need not change at all.
A production tax should be set at a flat rate; a rate that would represent a proportion of the value of production. Progressivity depends on the conversion of existing tax concessions into a universal tax credit or basic income. The overriding principle of both efficiency and equity in public finance is "basic income / flat tax".
Atkinson, A.B. [Tony] (1995) Public Economics in Action: the Basic Income/Flat Tax Proposal, Oxford: Clarendon Press.
Einaudi, Luigi (1933) "The Physiocratic Theory of Taxation" in Economic Essays in Honour of Gustav Cassel, London: Allen & Unwin.
Glynn S. and A. Booth eds. (1987) The Road to Full Employment, London: Allen & Unwin
Heilbroner, Robert & W. Milberg (1995) The Crisis of Vision in Modern Economic Thought, New York: Cambridge University Press.
Levi, Margaret (1988) Of Rule and Revenue University of California Press, Berkeley
Paine, Thomas (1796) Agrarian Justice (http://www.geolib.pair.com/essays/paine.tom/agjst.html)
Rankin, Keith (1991) "The Universal Welfare State; incorporating proposals for a Universal Basic Income"; Policy Discussion Paper No.12, Department of Economics, University of Auckland.
Rankin, Keith (1997) "A New Fiscal Contract? Constructing a Universal Basic Income and a Social Wage", Social Policy Journal of New Zealand 9:55-65.
Rooth Tim (1993) British Protectionism and the International Economy, Cambridge University Press.
Simon, J. Herbert A. (2000) "UBI and the Flat Tax" (http://bostonreview.mit.edu/BR25.5/simon.html), Boston Review October/November
New Democracy Forum (http://bostonreview.mit.edu/ndf.html#Income)
van Parijs, Philippe (2000) "A Basic Income for All" (http://bostonreview.mit.edu/BR25.5/vanparijs.html), Boston Review October/November
New Democracy Forum (http://bostonreview.mit.edu/ndf.html#Income)