Income splitting is a relatively novel concept in Canada's system of income taxation. Forms of income splitting have been used in several other countries, but in Canada income splitting was introduced for the first time for pensioneers in 2007. Known as pension income splitting, a couple who receives eligible pension income from annuities or a retirement income fund can split up to half of their income. Recently, the federal government announced that they will introduce spousal income splitting for families with children. This announcement has stirred considerable controversy, however.

Income splitting is one of various forms of taxation regimes for married couples, with or without children. Consider a couple with incomes \(A\) and \(B\), and a progressive taxation system with tax schedule \(\tau(x)\) for income level \(x\). The marginal rate of taxation is positive \(\tau'(x)>0\), and the marginal rate increases with income: \(\tau''(x)>0\). A tax schedule where the marginal rate of taxation increases is called progressive. Progressivity is usually bounded at some maximum level for the marginal rate: \(\tau'(x)\le\tau'_{\max}\). Progressiveness is meant to create fairness by taxing individuals and households according to their ability to pay. Let \(T\) denote the total tax payable by a household. Then there are various options of joint taxation and attributing income:

• With Cumulative Taxation the income of both spouses is added up, and the total household income is subject to income tax. \[ T= \tau(A+B) \] This regime penalizes marriage because the joint amount reaches higher marginal rates than each amount taxed separately. In fact, this policy was used in Germany for some time to actively discourage women from seeking employment, and rewarding single-earner households.
• Individual Taxation: this has been the norm in Canada until recently. Each spouse pays their taxes separately, although small allowances are made through tax deductions. \[ T=\tau(A)+\tau(B) \]
• Spousal Income Splitting involves taxing the household income as if it was divided equally between spouses. The income from both spouses is added and divided by two. \[ T=2\cdot\tau\left(\frac{A+B}{2}\right) \] This determines the tax rate, and this tax amount is doubled.
• Family Income Splitting goes a step further by allowing for the presence of \(n\) dependent children in a household. Each child increases the denominator in the tax calculation by an amount \( 0<\lambda\le 1 \), often \(\lambda=0.5\). For example, a family with one child would have a denominator of 2.5, and a family with two children would have a denominator of 3 instead of 2. \[ T=(2+\lambda\cdot n)\cdot\tau\left(\frac{A+B}{2+\lambda\cdot n}\right) \] Once the tax rate has been determined, the payable tax is determined by multiplying it with the same divisor.
• Income Transfer is a variation of income splitting where a fixed amount \(R\) can be transferred from the higher-earning spouse to the lower-earning spouse. Assuming that \(A>B\), then \[ T= \tau(A-R)+\tau(B+R) \] There are usually constraints on this procedure. The transfer amount cannot reduce an income below nil. It must also not exceed full equalization, and thus \(R<(A-B)/2\).

In Europe, Germany and Poland have forms of spousal income splitting, and France and Portugal have forms of family splitting. In France, the quotient familial is \(\lambda=0.5\) for each child except for the third child, which is favoured with \(\lambda=1\). The French government wants to encourage two-children families to have a third!

There are two rather different motivations for income splitting: supporting marriage and supporting children. The first motivation concerns the support of marriage as a societal institution. As a household is formed as a community of mutual support, it should not matter who earns the income. The argument is that two households living next door to each other making the same total income of, say $150,000, should be taxed the same regardless of whether in one household both spouses earn $75,000 or in the other household one spouse earns $150,000 and the other nothing. In Canada, the idea of income splitting has gathered significant support for the case of pensioneers who draw on their savings. As pensioneers are usually no longer participating in the labour force, income splitting has no or negligible impacts on labour market participation. This is rather different when income splitting is applied to spouses who participate or could participate in the labour force. Income splitting provides a strong disincentive to labour market participation for a non-earning stay-at-home spouse.

The labour market distortion from income splitting may not be obvious. Changes in marginal tax rates affect people differently. Under individual taxation, the marginal tax rate of the low-income earner is low, while under income splitting it will be much higher. Consider a household where income \(A\) is much larger than income \(B\) and the latter income earner obtains an increase in income. Then the marginal tax rate for the lower income \(B\) is \(\tau'(B)\) under individual taxation but \(\tau'((A+B)/2)\) under income splitting. Because of progressivity, it holds that \(\tau'((A+B)/2)>\tau'(B)\). Entering the labour force also subjects the new income to EI and CPP premia. Thus the effective marginal tax may be quite substantial. Low-income earners are also more sensitive to labour market participation than high-income earners. This makes income splitting inefficient economically.

Equitable treatment of single-earner and dual-earner households with the same joint income is complex. The single-earner household has greater earning capacity than the double-earner household because the non-working spouse can decided to enter the labour market. Thus it can be said that the single-earner household has greater ability to pay taxes than the dual-earner household with the same income. In addition, the stay-at-home spouse produces household services (e.g., raising children) in the dual-earner household, while the single-earner household may need to pay for extra services. Arguably, the dual-earner household should pay lower taxes than the single-earner household at the same income level. The question then becomes: how much less?

A policy proposal by Jack Mintz and Matt Krzepkowski attempts to fix the sharp edge of income splitting. Currently, one spouse can transfer the unused portion of the basic personal exemption to the other spouse. If both spouses were required to be earning income in order to qualify for the transfer, it would reduce the benefits of income splitting for single-earner households and acknowledge the larger provision of households services by the single-earner household.

While income splitting—with or without the Mintz-Krzepkowski adjustment—may somewhat improve horizontal equity (or fairness) between households earning the same income, it may exacerbate vertical equity between rich and poor households. Most of the benefits of income splitting go to affluent households—those in the highest tax bracket. Income splitting can improve horizontal fairness but comes at the cost of worsening vertical fairness.

A second policy concern the support for raising children. This is the reason why France has opted to go even a step further and introduce family income splitting rather than just spousal income splitting. Income splitting that is limited to families with children under 17 would benefit families, although families with higher income would benefit more from the policy. Thus the question becomes: which tax policy is best-suited to help families with children?

There are different ways to help families with children. One way is through the direct provision of services, such as child care. Another is through direct payments to parents of children, a fixed amount for each child. The federal government currently pays a Universal Child Care Benefit (UUCB) of $100 per month per child, for each child under the age of 6. In addition, the a Canada Child Tax Benefit (CCTB) is paid to eligible low-income parents of children under 18 years of age. The third option is to grant a tax deduction for each child (or through specific other tax credits), and the fourth option is to subsidize families with children through income splitting. The first two options are more direct than the last two options. Direct provision and UUCB treat all children equally. The CCTB helps poorer families specifically, while income splitting and tax credits benefit affluent families more.

Canada also has a number of so-called boutique tax credits for families with children, namely the young children amount ($2,234), the children's fitness amount ($500), and the children's arts amount ($500). These amounts mostly benefit more affluent families. Such an approach to taxation is often criticized by economists because boutique tax credits complicate the tax system, lead to higher costs of tax administration, and ultimately are often second-best approaches compared to other policy interventions aimed at the same objective.

The federal government's original proposal for income splitting was tampered down after strong criticism from the late Jim Flaherty, the former finance minister. The proposal calls for income transfers of up to $50,000 among spouses, but cap the tax savings at $2,000. Interestingly, because of difference in family structure across provinces, more families in Alberta will gain from this policy than families in Quebec. The government also proposes to raise the UCCB to $160 for the first five years and to $60 through age 18.

In my view, the modified proposal is better than the original proposal in particular in light of the increase in the UCCB and capping the benefits of income splitting for affluent households. However, the federal government has saddled itself with a policy promise that they made in the last election, and effectively has backed itself into a corner. Politically, the watered-down proposal allows them to keep their election promise and save face. As a public policy to help children, the available funds could have been used more effectively—by targeting child poverty directly, where it is most needed and helps the less affluent in our society. Universal daycare has been proposed by the Opposition in parliament, but that is a very expensive proposition as well. Rather than just helping those who need daycare and can't afford it (low income families, single parents), universal daycare would also benefit affluent dual-income households. If solely intended as a tax cut, the available funds would lead to greater equity in Canadian society if they were channeled entirely through the UCCB. All Canadian children should be worth the same to the government, regardless of how rich their parents are. The UCCB helps poorer families more than affluent families, and it helps families with more children more than those with fewer children. And have I mentioned that the public school system is poorly funded and could need some help? Of course, that's a provincial responsibility, but like health care there could be federal funds for meeting federal standards or for school infrastructure.

As Canada is heading into an election in 2015, there is little doubt that income splitting will become a key electoral topic. As journalist Andrew Coyne has put it: is the tax inequity for single-income households the worst inequity to be found in our tax system? Are there other inequities that deserve more attention, or other priorities? If the policy objective is to help children, let us start with the neediest children in our society. The Conference Board of Canada reports that Canada ranks 15th out of 17 peer countries when it comes to child poverty. More than one in seven Canadian children live in poverty. Canada's performance in this domain is dismal. But poor children don't vote, and their parents tend to vote less than affluent parents. For now, I concur with my Queen's University colleague Kathleen Lahey, whose contribution in the Globe and Mail was titled Income splitting won't help parents who really need a tax break.