Universal Basic Income: a System of Tax Reform 

Keith Rankin, for the 2nd NZ Universal Basic Income Conference, Wellington, March 1998

"[Jenny] Shipley and [Winston] Peters deliberately distinguish between New Zealanders who pay taxes and those who receive social services. This division is a fallacy."

Rod Donald, Green Party MP, on the proposed Code of Social and Family Responsibility.



The term "Universal Basic Income" (UBI) represents in New Zealand both a scientific approach to tax-benefit reform and, for want of a better name, a "New Left" social policy brand. While it remains important that the UBI name retains its strong association with social liberalism (and hence its dissociation with the economic neoliberalism branded in New Zealand as the New Right), it is also very important that the political branding does not cause the advocates and potential advocates of UBI to lose sight of its core principles.

The name "Universal Basic Income" arises in New Zealand from a paper I wrote in 1991 ("The Universal Welfare State; incorporating proposals for a Universal Basic Income"); a paper that considers both the positive (ie analytical, value-free, scientific) basis for the introduction of a UBI and the normative values that underpin my preference for "the UBI way". As such, the UBI brand in New Zealand is linked irrevocably to my name. Therefore, it is of particular importance to me personally, as well as to the maintenance of clarity in the public debate on welfare issues, that the central theme of combining income tax concessions with explicit benefits to form a truly universal benefit is not lost sight of.


Language and Definition

The core of a "UBI system" derives from the idea that benefits contained within the income tax scale are no more nor less benefits than are benefits paid through the social welfare system, and that the integration of the two kinds of benefits forms the basis for a universal benefit; that is, an equal benefit paid to all adult tax residents of a nation. My 1991 paper mooted a universal benefit set at 20% of GDP per capita (then, $6,000 per annum) combined with an income tax rate of 48%. Such a benefit might or might not be sufficient to meet the basic needs of every such citizen.

The core formula is:

Net Income  =  Gross Income  less  Income Tax  plus  Universal Benefit

    eg Net Income  =  Gross Income  less  48%  plus  $6,000 per annum

    or Net Income  =  Gross Income  less  48%  plus  $115 per week


The core property of the UBI system is the trade off between the income tax rate and the amount of the universal benefit. All else unchanged (ceteris paribus), a higher universal benefit requires a higher income tax rate. This income tax rate must be proportional (or "flat" as it is often called1) because a proportional tax rate is the necessary consequence of splitting income taxation into (i) a gross revenue fund, and (ii) a set of benefits (ie concessions). In a UBI system, there is no ambiguity; increasing tax concessions is conceptually the same as increasing benefits.

Figure 1 makes explicit the implicit tax benefits that arise (in 1998) from taxing all income at the 33% corporate rate. The benefits abate sharply for low income recipients. The "tax cuts" in July 1998 (called "growth dividends" by Reserve Bank Governor Donald Brash [1996, p.16] and others) - really increases in tax benefits - are heavily weighted to upper income recipients.

The core idea behind the UBI is that of universality, of horizontal equity, which is the notion of treating equals as equals. The tax benefits depicted in Figure 1 clearly fail the horizontal equity criterion.


Citizenship is a universal concept that regards people as equals. The idea that public income should be distributed on an equitable basis is neither a left-wing nor a right-wing idea; rather it is intrinsic to the principle of horizontal equity. The essential difference between left and right lies in the division of total income between public and private. The right-wing view is that most of a nation's income is or at least should be private, and that therefore any divisions of public income will involve comparatively small amounts.

The core idea is known in the academic literature as "basic income / flat tax" (eg Atkinson 1995). I prefer to keep the word "basic" out of the discussion of core principles, because that word gives some impressions as to what level a universal benefit might be set. In the normative context of UBI, "basic" means "adequate"; ie sufficient to meet basic needs. In other contexts, it might mean "low". In other words, the term "basic" can mean either a floor or a ceiling.

The Universal Basic Income brand means more than its core principle of horizontal equity. It means more than an integration of the explicit benefits paid through the Social Welfare arm of government and the implicit benefits paid through the Inland Revenue. It is the "basic" element - understood as a social minimum - that marks the UBI out as a socially liberal brand. The underlying philosophy of the UBI movement includes the view that every person should receive public support sufficient to ensure their basic needs are met. Thus the UBI brand reflects a reaffirmation of the ideals of the "cradle to grave" welfare state; ideals strongly affirmed in for example the 1972 Royal Commission on Social Welfare (the McCarthy Commission). The brand contains the concept of a true "guaranteed minimum income" as well as the concept of a core universal benefit (eg "citizens income" or "social dividend").2 We can claim that this concept of "basicness" represents an essential feature that differentiates a UBI from a simple social dividend.

Recognising basicness as an essential feature of UBI does not mean that all individuals have the same basic needs. Any effective basic needs approach has to provide some element of "vertical equity"; a means of treating unequals unequally. Hence a UBI system must either provide for a two-tier system of benefits - one core tier for horizontal equity, and one essential tier for vertical equity - or the social dividend must be so high that the majority receive a social dividend set at a level higher than that of their basic needs.

To reiterate, the core issue is to create a universal benefit - a social dividend - by integrating tax benefits with social welfare benefits. The essential issue, over and above that, is to integrate supplementary benefits with social dividends. A UBI, as a benefit that meets basic needs, represents such a conflation of two benefit types; a conflation of a social dividend and a guaranteed minimum income. At its most elegant level of simplicity, supplementary (second tier) benefits are individually customised "supplementary basic incomes" (SBI) which retain the core property (that of a trade-off between a flat-tax and a fixed-benefit) of the UBI system.3 Whereas in my 1991 paper I treated the UBI as simply a social dividend, with the SBI as an addition to it, the UBI concept has evolved to the point where the UBI, as a benefit, is the sum of the social dividend (SD) and the supplementary benefit (SB). A social dividend set too low to meet basic needs is also known as a "partial UBI".

While a UBI system has an impartial SD core and a socially liberal SB essence, these are minimum requirements. Any variation which is faithful to both core and essence can be called a UBI system. Thus a UBI system may pay additional benefits and make use of additional sources of revenue. A UBI system may include universal benefits to children, or to caregivers on behalf of their children.

There is nothing in the core or essence of a UBI that suggests a departure from the longstanding social differentiation between adults and minors. It is, I believe, universally true that, while both adults and children (ie minors) are citizens, minors (by definition) are dependent on adults; individually dependent on parents and guardians and collectively dependent on adult society. As dependents, children are supported from the incomes of adults. Thus, while it is essential that a UBI system meets the basic needs of families of all types, the payment of benefits directly to children is a radical option that is consistent with UBI but not essential to it.


Core Samples

Tax benefits can be integrated with social welfare benefits in two distinct ways: the negative income tax approach which counterposes personal income taxes with benefits, and the production tax approach which eliminates personal income tax. The analysis in this section follows the negative income tax approach.

The "negative income tax" approach regards all benefits as tax credits (or, on the other hand, treats personal income taxes as negative benefits [or benefit debits]). We calculate average tax rates (which range from minus infinity to the official tax rate) through dividing net incomes by gross incomes:

Average Tax Rate = 1 - (Net Income Gross Income)


A progressive income tax scale is one in which the average tax rate rises as gross incomes rise. A core UBI system is always more progressive than is a graduated (ie stepped) tax scale, because, in a UBI system, average tax rates are much lower for persons on low incomes. (When comparisons are made between UBI systems and graduated tax scales, I am assuming that the corporate tax rate is the same in both cases, and that the top personal tax rate is the same as the corporate rate.)

Proportional income tax at the upper rate combined with a social dividend set at the maximum is always more equitable than the graduated tax scale from which it derives. Consider any graduated income tax scale, in which the upper tax rate is also equal to the corporate rate, as is the case in New Zealand. All persons whose last dollar is taxed at the corporate rate is already participating in a UBI-like system; they get social dividends in the form of tax benefits according to the formula:

SD = (b1-b0)*( tn-t0) + (b2-b1)*( tn-t1) + ... + (bn-1-bn-2)*( tn-tn-2) + (bn-bn-1)*( tn-tn-1)

or SD = å i=1..n (bi-bi-1)*( tn-ti-1)

where bi is the tax step (b0 = $0), ti is the associated marginal tax rate, and n represents the number of concessionary tax rates.

The formula is illustrated in Table 1, using the graduated tax scale being introduced in July 1998.



The problem that arises with the graduated scale approach is that people whose last dollar earned is taxed at less than the top rate receive lesser tax benefits; they receive abated social dividends. That is inequitable, making such people into lesser citizens. Converting to a UBI system simply requires that lower income recipients have their tax benefits topped up to the same level as those received by higher income recipients. (The " " sign in Table 1 becomes an "=" sign.) For any given corporate tax rate, a core UBI system must be always more equitable than a graduated income tax scale, because it is only lower income recipients who require their tax benefits to be topped up.

In a society with no other kinds of cash benefit, the topping-up of the tax benefits must be funded by raising the tax rate, or by introducing an additional tax. However, liberal western societies contain a complex array of explicit benefits, so the cost of ensuring that every adult receives a benefit of at least the SD (social dividend) level may be quite low. It all depends on how many people receive such benefits, bearing in mind that some people do not receive benefits that they are entitled to. A UBI system would always ensure that any person with an IRD number would get their entitlement; nobody would "fall through the cracks".

In New Zealand, most caregivers and low paid workers are entitled to receive "social welfare" benefits. Here I'm defining social welfare benefits as all explicit benefits targeted at lower income households (ie including family support tax credits). This contrasts with tax benefits, which are implicitly targeted at higher income households. Ordinary tax benefits arise from ordinary tax concessions; ie the concessionary steps in the graduated tax scale. Tax benefits also arise from other types of tax concession, such as allowances for children, for childcare, for housekeepers. These benefits all favour those who pay enough tax to get the full benefit. "Social welfare" children's, childcare and housekeeper benefits - where they exist - are higher for lower income recipients.




Many graduated income tax scales have a higher top personal rate than the corporate rate. Such a scale was Labour Party policy in 1996. A similar scale is depicted in Figure 2, and Table 2. The upper tax rate creates an abatement of the ordinary tax benefit for high income recipients as well as low income earners.

This is a viable option so long as the average tax rate remains lower than the corporate tax rate for almost everyone. With such a scale, the tax rates in excess of the corporate rate can be considered to be surcharges. Where high income abatements or surcharges apply, average tax rates on very high income earners are higher than in a UBI system. Therefore it is not possible to claim that a UBI system is always more progressive. Indeed, converting such a system to a UBI system is perhaps best done by raising the corporate tax rate.

Table 3 repeats the Table 2 example, but with corporate tax raised to 39%. This forms the basis of a centre-left UBI system that requires minimal supplementary benefits. But, the considerable top-ups required to remove the low income abatements require additional tax, such as a further rise in the corporate rate or an introduction of a new tax such as a financial transactions tax (FTT, Tobin tax).



An Essential Matter: Meeting Basic Needs

Supplementary benefits (SBs) are redundant if the social dividend (SD) is high enough to qualify as a UBI. This is a UBI system in its "purist" form.

A UBI system in which the SD is sufficiently high to cover everyone's basic needs contains a different kind of redundancy, however, in that it exceeds most people's basic needs. The purist approach can be criticised as expensive, inefficient and inequitable. From a "left-wing" point of view, the revenue required to fund such a high UBI might be better spent on social wage goods. From a "centrist" viewpoint, it would be better to have a lower SD, lower tax rates, and supplementary benefits. The trick is to design SBs is accordance with UBI philosophy, which is ant-poverty, anti- poverty traps, anti-bureaucracy, and in favour of "real freedom" (van Parijs 1995).

To create a two-tier UBI system, in the interest of vertical equity and maintaining tax rates no higher than necessary, the second tier can be thought of either as a single individually customised benefit with an abatement rate, or as an augmented social dividend matched to a higher rate of income tax. The two alternatives are identical in practice, but I prefer the former because it opens new options in the way we account for income tax; ie it is more compatible with the ideal of creating a fiscal contract based on re-appraising income tax as production tax. While the second tier benefit can be thought of as a supplementary basic income, I prefer simply supplementary benefit.

The kinds of circumstance that might qualify a person for an SB include: being retired, having children, having a disability, being sick, being without a paid job, or being in part-time work. There may also be a need to pay a higher SB in places where housing costs are high, but I favour public housing schemes over pure aid-in-rent provision because they create some discipline for landlords.

Simple means-testing removes the need for any subjective forms of conditionality, such as someone making judgements as to whether a person is seeking work, available for work, or making unpaid contributions. The only criteria for supplementary support are a lack of income and objective criteria such as having dependent children.

The kind of SB that I prefer moves away from the principle of individual payment that is fundamental to the core UBI system. Family means-testing ensures that the partner of a millionaire cannot qualify for an SB. Family means-testing for the SB is acceptable so long as SDs are sacrosanct.

For single adult families (which include single person households), there would be an automatic SB whenever the person's income falls below X. X would be raised where other objective factors exist; eg children, retirement, disability. Figure 3 shows the components of a single person's UBI. The SB is defined by the abatement rate (20% in this example) and X ($400 per week gross in the example). X would simply be raised to a higher amount (eg X') when special circumstances are present (eg the inclusion of children, disability, retirement age).



For two-adult families, there would be a shared payment whenever the family income exceeds Y. Y would be raised where the same factors apply. Figure 4 shows an example of UBI benefits payable to a couple, and to a family with one child. In the case of the child, the basic assumption is that the $40 per week increment is an addition to the SB. An alternative scenario, also depicted in Figure 4, is that the children's increment is an (unabated) addition to the caregiver's SD.

Rather than having a bureaucratic state decide who is family and who is not, the initial assumption might be that all members of a dwelling constitute a single family. It would be up to others - eg boarders and flatmates - to declare themselves as separate households.



The philosophy that brands the UBI is one of equity, of automatic entitlement, and of minimal bureaucracy. Vertical equity is an important component of equity. Different needs relate to family circumstance as well as individual circumstance. Economic efficiency is important, too. Once having established a two-tier UBI system, it is not clear that there are advantages in moving towards a single tier system. Such a move would require cutting SBs as the SD rises. Any relative growth of the SD over the SM nevertheless constitutes an asymptotic shift towards a single-tier UBI.


Replacing Personal Income Tax with Production Tax

The introduction of a UBI system opens a number of doors to more comprehensive reforms. One of those is the opportunity to pay "growth dividends" equally to all citizens, rather than "tax cuts" which inevitably favour the rich. Another is the opportunity for political parties to brand themselves in accordance with the UBI parameters they favour. Once in place, a UBI system is much more transparent to voters than is the present convoluted "welfare mess" (St. John 1996). Such transparency leads to clarity in the public's differentiation of political parties fiscal policies.

For me the most important opening door is the opportunity for a radical reinterpretation of income tax and the system of national accounts; radical enough to form the constitutional basis for an explicit fiscal contract (see Rankin 1997); a contract that recognises the economic sovereignty of citizens as owners of the social and natural inheritance that has created modern "civilisations". This reinterpretation can be called "Accounting for Social Profit".

At the centre of this "social profit" project is a return to an old interpretation of tax as interest attributed to the property of the sovereign; a mixture of economic rent and payment by producers for the use of public domain resources.

From this new (old?) point of view, income tax can be reinterpreted as "production tax". Taxes would be treated in the national accounts as a third fund, a return to the people in their capacity as sovereign proprietors. The two funds recognised in the modern national accounts are that paid to labour ("compensation of employees") and that paid to private capital ("operating surplus"). The new approach treats taxes just as it treats wages and profits; as a production cost. Because all costs represent someone's income, taxes represent a social profit, an income for the people rather than the government; a return to social capital.

Replacing graduated income tax scales with a proportionate scale in which all market income is taxed at the corporate rate renders the concept of personal tax meaningless. All income tax - corporate and personal - conflates into a simple production tax. All wages, dividends and interest come to be paid net of tax, and the personal marginal tax rate becomes zero. The tax system becomes incredibly simple, thereby enabling a slashing of the IRD bureaucracy as well as a decimation of the Social Welfare Department.

When we remove income tax entirely from the realm of personal finance, then the practice of netting income tax against benefits - ie by calling benefits "tax credits" - becomes impossible. Thus UBI payments simply come to represent social income. Having drawn social dividends out of the graduated tax scales, we have no need for the concept of income tax. Such SDs, like tax credits, can still be paid through the agency of employers. Employers would simply deduct employees' social dividends from the amounts of production tax they remit to Inland Revenue.

Under a UBI system, ACC payments would also become a levy on employers only. But the ACC bureaucracy would itself become redundant, able to be replaced by much more efficient "no fault" supplementary benefits.

Once we eliminate personal income tax, then the only benefit that is subject to abatement is the SB. A poverty trap can still occur as a result of quasi- income taxes such as student loan repayments, child support liable parent contributions, and compulsory savings/insurance schemes such as the late unlamented 1997 Compulsory Retirement Savings Scheme.

In the production tax approach, the SB is a quite different sort of benefit from the SD; indeed it is like any of the means-tested benefits we are familiar with. Nevertheless, it remains true that:

UBI  =  SD + SB

It is important to note that, once a production tax regime is in place, a tax increase would not be automatically followed by a wage cut, as it would be expected to be under the tax credit approach. The value added by a firm is, as income, divided three ways, between capital, labour and society. An increase in the share to society does not necessarily mean a fall in the share paid to labour. This is true even with retention of personal income tax; it is just less obvious.

The equivalence of the "negative personal income tax" approach and the "production tax" approach is revealed by Figure 5, which is essentially the same as Figure 3. The only difference is that the "gross weekly income" scale has been replaced by a "weekly market income" scale which is net of tax. A person's total weekly income can be calculated by adding the figure on the horizontal axis of Figure 5 (typically a wage) to the UBI on the vertical axis. For example, a person with no special needs who is earning $200 per week (equivalent to $300 gross in Figure 3) would receive a UBI of $120 per week, giving a total weekly income of $320. The SB in the example would abate at 30 cents per dollar of weekly market income (equivalent to 20 cents per dollar of gross income).



Rule of Thumb; how Big a Social Dividend?

If we assume that all government income constitutes a social fund to be allocated to all adults, and that income tax revenue is a simple proportion of "GDP at factor cost", then it is a comparatively simple matter to calculate the maximum value of the individual social dividend. If we fund all government consumption and government investment from indirect taxes and profits of state-owned enterprises (SOEs), then this social dividend can be fully paid out as a UBI.

If the income (or production) tax rate is 33%, then the social dividend will be 33% of GDP per adult. In the case of New Zealand at present, GDP at factor cost is about $85 billion. There are about 2.7 million adult residents. Thus, for this exercise, GDP per capita equals about $31,500, which, at 33%, equates to a maximum social dividend of $10,400.

It is very useful to note that the GDP per capita number used here - $31,500 - approximately equals the average annual fulltime wage in New Zealand. Thus a simple rule of thumb is to multiply the average wage by the tax rate to calculate the maximum affordable social dividend. (I have Brian Easton to thank for this insight.)

There are of course difficulties in achieving such a maximum. It may not be possible to collect the full 33% of GDP, and the funds from indirect taxation and SOEs may not be enough to fund the public goods we demand. Nevertheless, such a calculation represents a simple mental arithmetic way of checking the costings of UBI proposals.

The social dividend implicit in the new 1998 tax scale is $5,130 per resident adult, just half of the $10,400 maximum. Everyone grossing $38,000 or more will pay 33% income tax and get an effective rebate of $5,130. By setting a flat tax of 33%, we can afford an explicit social dividend of $5,130, while still funding a generous range of social wage goods out of income tax.



"Universal provision . . . improves resource allocation, minimises qualitative differentiation of service, is politically sustainable because of the wide spread of beneficiaries, and performs an important socially integrative function by underpinning rights of citizenship."

Geoff Bertram (1988, p.135)

The UBI solution is a genuine alternative to the seemingly intractable problems posed by existing fiscal conflicts, and the politics of division. It is a pragmatic, flexible solution with clear principles; a solution that does not seek to prescribe identical benefits to everyone regardless of need. The UBI way is affordable, equitable and efficient.

A UBI system comprises a core (social dividend, proportional tax) to establish horizontal equity, a socially liberal essence that provides for basic needs and an objective basis for positive discrimination (vertical equity) in favour of (for example) parents, the disabled and the retired. And a UBI system allows any number of expenditure or revenue add-ons, just so long as the core and essence are maintained.

Not only does a UBI system resolve socio-economic tension in a socially integrative way, but it creates conditions which can lead to a complete conceptual reappraisal of the central fiscal variable; namely income tax. Redefining income tax as production tax is a key step that makes it possible for citizens to reclaim their economic inheritance by way of a meaningful fiscal contract between citizens and producers; a contract that defines the public share of national income on a constitutional basis. Such a fiscal contract ensures that growth dividends, in the form of rising social dividends, are paid to all of a nation's sovereign citizens, and not just to those with exclusive private property.




  1. Unfortunately the term "flat tax" has been branded in New Zealand as a neoliberal (ie right-wing) concept. Neoliberals are in fact unattracted to a flat-rate income tax set at say 48%. Even 33% is high to most neoliberals, who are really interested in low taxes, not flat taxes. [return]
  2. There are 9 combinations of name that can be made by combining "universal", "social" or "citizens" with "income", "dividend" or "benefit". Each combination really means the same thing. Each person will have their preference of course, and that will depend on perceptions of branding associated with each of the six words. [return]
  3. This was the essential feature of my 1991 paper that set it apart from other basic income proposals at the 1991 University of Waikato seminar on Basic Income, and that led to an avalanche of public interest in the wake of a radio interview. [return]



Atkinson, A.B. [Tony] (1995) Public Economics in Action: the Basic Income/Flat Tax Proposal, Oxford: Clarendon Press.

Bertram, Geoff (1988) Middle Class Capture: A Brief Survey. In Report of the Royal Commission on Social Policy Volume III, Part Two, Future Directions Wellington, Government Printer.

Brash, Donald (1996) New Zealand's Remarkable Reforms; 1996 Hayek Lecture, IEA Occasional Paper 100, London.

Rankin, Keith (1991) "The Universal Welfare State"; 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.

St. John, Susan (1996) The Welfare Mess Revisited, Department of Economics Policy Discussion Paper, The University of Auckland, December 1996.

Van Parijs, Philippe (1995) Real Freedom for All, Oxford: Clarendon Press.


Rankin File