Chapter 3 - How income splitting could work

3.1 This chapter outlines a possible approach to the use of income splitting in New Zealand and poses a number of questions in relation to its possible introduction.

3.2 Income splitting for tax purposes is allowed in a number of countries and takes a variety of forms, as discussed in chapter 2. In most countries, income splitting is a matter of allowing couples to lower their total tax liability by allocating some of the higher earning partner’s income to the lower earning partner, thus mitigating the effects of the progressive nature of tax rates.

3.3 There are two broad ways to achieve income splitting. The first is to aggregate family income and (as the name suggests) split it evenly between the two partners. Total tax liability is then determined by applying the tax rate schedule to the split level of income, and multiplying the result by two. The second achieves the same effect by providing a separate tax rate schedule for families, one in which the tax rate thresholds are twice as high as the thresholds that apply to individuals. Effectively, allowing income to be split 50/50 would change the taxable unit from the individual to the couple or family.

New Zealand’s past experience with income splitting

3.4 New Zealand currently taxes on an individual basis. Family-based taxation is not, however, an entirely new concept to New Zealand. Between 1939 and 1962 New Zealand required the aggregation of a married couple’s incomes if it exceeded a moderately high level in aggregate. However, this measure was not targeted at families with children. In fact, because tax rate thresholds were not raised when aggregation was required, it acted to increase wealthier families’ tax burdens rather than decrease them.

3.5 In contrast, the 1982 Report of the Task Force on Tax Reform (the McCaw Report) strongly recommended income splitting be allowed in New Zealand as a means of reducing the tax liability of many families. The recommendation did not require a family to have children to be able to split income. The rationale behind the recommendation was the concern about a lack of recognition of the costs associated with the family unit in the tax system at the time. Essentially, the differences in tax liability between one-income and two-income families with similar abilities to pay tax were seen as unfair.

3.6 Although the McCaw Report recommended a specific, limited form of income splitting, it can be designed in a number of ways. The remainder of this chapter looks at the main considerations in adopting income splitting for families with children.

Who would be able to split income?

3.7 A “family” could be defined as including married, civil union and de facto partners, a definition that would be consistent with eligibility for the Working for Families tax credits.

3.8 One of the first questions to be considered is what age restrictions there should be for the children of the families involved. Should it apply to families with children up to the age of 18? That is the age for which the Working for Families tax credits are available, and it is also the age at which parents are no longer responsible as guardians for the day-to-day care of their children.

3.9 The child would also need to be dependent on the principal caregiver. For purposes of Working for Families tax credits, a dependent child is defined as a child 18 years and under that is financially dependent on the principal caregiver. To be considered financially dependent, the child must be financially supported by the principal caregiver and not work more than 30 hours a week or receive a student allowance, benefit or other government assistance. For simplicity and consistency across government policy, it would be desirable to use the same test if an 18-year age limit were to be adopted.

3.10 An alternative is to limit income splitting to families with a younger child – say, one who has not yet started school.

3.11 The question of restrictions on the ages of the children concerned in families that split their income is a fundamental one, and the government is particularly interested in readers’ views on this matter.

In what way would income be split?

3.12 The standard model of income splitting would be to allow a straight 50/50 income split. However, a number of variations are possible. These include allowing income to be split amongst children as well as parents (as in France) or allowing a restricted form of income splitting such as a 70/30 split (as in Belgium). A further restricted option would be to adopt the transferable personal allowance approach taken in Denmark.

3.13 The McCaw Report favoured a limited form of income splitting similar to the Belgian 70/30 approach. The report recommended (with one member of the Task Force dissenting) that the government introduce:

“a voluntary scheme of ‘partial income splitting’ for married couples, whereby couples would have the option of notionally dividing their aggregate income by some divisor of between 1.3 and 1.8, with their total income being taxed at the average rate applicable to a single individual whose income equals that quotient.” (p101)

3.14 Dividing by two would provide full 50/50 income splitting, so in effect the Task Force’s recommendation would allow only the majority of income to be split between partners, but would still provide a significant tax advantage to most couples with differing incomes. The Task Force favoured this restricted approach over a 50/50 split, to ensure that couples were not overly advantaged relative to individuals.

3.15 The Task Force considered that, in principle, income splitting should be allowed amongst children of the family as well, but for administrative ease, it recommended increased family benefit support instead. Similar administrative difficulties are likely to preclude a broader approach here also.

3.16 While not ruling out variations such as these, this discussion document concentrates on the standard 50/50 version, to illustrate the questions that would arise if income splitting were to be introduced in New Zealand.

Should income splitting be compulsory?

3.17 While no one would be financially worse off through 50/50 income splitting, the government’s view is that, if adopted, income splitting should be voluntary. Partners who currently have very similar incomes may gain very little from income splitting, and they should have the choice not to participate if they think the compliance costs involved outweigh any benefit they would receive.

3.18 If a 70/30 split were to be adopted, it would result in families that are already closer to a 50/50 split being made worse off. In this case, there is even stronger justification for income splitting to be voluntary.

Fiscal cost of income splitting for families with children

3.19 The fiscal cost of allowing standard 50/50 income splitting for families with children who are under five years of age would be around $160 million a year. Extending it to families with children who are 18 years and under would increase the cost to around $370 million.

Administrative considerations

3.20 Two important issues to be considered in relation to the possible introduction of income splitting are determining how it should be administered and how the tax benefit would be paid out.

3.21 Because income splitting would be tied to the presence of children, the most administratively manageable, and compliance cost minimising, approach is likely to be for Inland Revenue to administer income splitting through the Working for Families tax credit system. This way Inland Revenue could adapt the current Working for Families tax credit system that is already in place, rather than develop entirely new systems and procedures.

3.22 It may be possible for Inland Revenue to pay the tax benefit out during the year through the PAYE system by use of a special tax code. Further work would be necessary to determine if that were feasible. It would place some additional compliance costs on taxpayers by requiring them to estimate their correct tax rate having regard for the impact of their partner’s income.

3.23 The alternative would be for the tax benefit to be paid at the end of the year as an “income splitting tax credit” through the standard tax assessment process. Most people who receive the Working for Families tax credits are already required to have an end of year tax assessment, so income splitting would not impose additional compliance costs on the vast majority of them. The tax credit could be paid to either partner, or split and paid to both.

3.24 Further administrative questions that would need to be considered in more detail if income splitting were to be implemented include how Inland Revenue would deal with couples who file tax returns (especially provisional taxpayers), with relationship changes, and with Working for Families and other payments such as child support, student loans and certain rebates that are currently linked to individual earnings. The implications of shared care of a child would also need further consideration, particularly for determining when a family is considered to have a child, and therefore is able to split its income.