This is the ninth of ten articles in the Public Service Association's "Ten perspectives on tax" series.*
By Susan St. John*
One of the least tractable concepts in tax is that of the effective marginal tax rate (EMTR). The official income tax rate applied to each dollar earned is not the only payment a family will make on that dollar.
There may be other losses and government imposts that look and feel like taxes. Families with children are particularly affected. The overall effect is a high EMTR that leaves a parent little to show for the effort of earning that extra dollar, constituting a strong disincentive to earn it in the first place. It is one of the major reasons that child poverty rates are high in low income working families.
Let’s take the example of a parent on the minimum wage of $15.75 per hour, earning $36,350 per annum. There are some Working for Families and rent subsidies but increasingly this family relies on foodbanks and loan sharks.
Let’s say there is an opportunity to earn another $10,000. Once tax and ACC are paid (18.71%), Working for Families is abated (22.5%), student loan repayments are made (12%), Accommodation Supplement is reduced (25%) and KiwiSaver extracted (3%), the $10,000 has been effectively taxed at 81.21%.
Furthermore, there may be a sudden drop in child care subsidies and child support payments of between 18-30%. Every family is in a different set of circumstances, and few will understand what is actually happening. They will know that at the end of the year despite their extra work effort they are no better off, may actually be worse off and will undoubtedly feel despair.
The EMTR effect, arising from the tax-transfer interface, is always in the too hard basket. In 2010 the Tax Working Group felt it was outside their brief and passed the buck to a welfare working group. But when the Welfare Working Group (2011) was established, it was explicitly forbidden to examine this issue. Yet it is this interface that impacts intensely on the well-being of families with children, and their ability to work their way out of poverty.
It is important to understand how this debilitating problem has arisen. The high tax rates of the early 1980s were thought by economists to reduce the incentives of high income people to earn and save.
Enter Rogernomics and the low flat-tax broad base solution of the late 1980s. However, a comprehensive view of income was not achieved because the capital gains part of the package did not eventuate.
Nor did flat tax solve the high EMTR problem; it merely shifted it from the top end to low and middle income earners. Low flat tax under the ideology of the day was to be accompanied by more user pays of social provision. However a three-way iron rule applies.
The greater the amount of social assistance to be reduced as income rises, the longer the range of income before all assistance has been bled out unless a higher rate of abatement is imposed. Reducing the amount of assistance acts to intensify poverty, a long income range defeats the purpose of welfare only for the poor, while a high rate of abatement imposes severe disincentive effects.
In advice to the incoming government in 1990, Treasury warned:
As a general rule, the more people facing higher effective marginal tax rates over longer ranges of potential income, the greater the costs to society and the greater the probable loss of output.1
Treasury identified high levels of benefits as a major factor preventing a more gradual abatement system and benefits were cut significantly in 1991. The Change Team on Targeting Social Assistance in 1991 was tasked with designing a new ‘integrated’ system of targeted social assistance. Thus the 1991 budget announced a complex system of Family accounts based on aggregated family assistance and a constant bleed-out or abatement rate. 2
Unfortunately, while aggregating assistance onto a family-based smart card, abating at one rate worked in theory, but the technocrats could not make it work in practice. One of the problems was that the typical modern family did not resemble the assumed nuclear family model.
Another was that the scale of assistance to be targeted, even with the 1991 welfare benefit cuts, meant assistance would be paid well up the income scale even with a very high single rate of abatement.
The integrated solution that had been used to justify the low flat-tax user-pays approach had quietly disintegrated. Now all that was left was the welfare mess of a plethora of high and overlapping abatements.
While one might have expected a re-examination of the wisdom of the 1990s reforms, instead the welfare morass has been intensified
in the name of target efficiency. 3 Welfare ‘only for the poor’ and low top tax rates has been the means of achieving an implicit, if not explicit, objective: more wealth and income for the top earners and an ever-immiseration and indebtedness of low-income families.
An alarming part of this picture is way tax credits for children are treated. One of the requirements to moderate the regressivity of a flattish income tax system with a notably high GST on everything, is a well-designed set of family-based tax credits.
The New Zealand system, known as Working for Families (WFF) operates to offset taxes paid by the family, thus improving horizontal equity – at least for low income families – by acknowledging that children reduce the ability to pay tax. The progressivity of the overall tax structure is further enhanced because tax credits are ‘refundable’ when they exceed the taxes paid by a low income family.
WFF tax credits are a major mechanism to alleviate and prevent child poverty. Just as older people need tax-funded support, so do families, especially when on a low income. It is critical such child tax credits are understood, supported and enhanced. When policy confines the tax credits ever more closely to low income families the EMTR problems are intensified.
If there is no will to revisit the ideology that drove the 1991 changes of flat low tax and user pays, then mitigation of high EMTRs is the urgent task. Unfortunately, current policy, based on increasing target efficiency, is going in quite the wrong direction.
For WFF, this sees the threshold for abatement heading back to $35,000 – where it was in 2005 – and the rate of abatement going up to 25% over time. Student loans are replacing student allowances, so repayment of larger student loans applies for long periods of time, especially for women. The thresholds for loan repayment have been frozen, as has the threshold for the parental income test and the cap for the Accommodation Supplement.
An innovative set of policies that reverse the recent cuts and enhance the programmes that help families, including a debt forgiveness programme is urgently required. Better still, a revisiting of the low flat-tax broad-based dogma is
 Shipley, J. (1991). Social assistance: welfare that works. Wellington, Government Printer
* Susan St John is an Honorary Associate Professor in the Department of Economics University of Auckland where she directs the Retirement Policy and Research Centre. This is the ninth article in the PSA's "Progressive thinking series, Ten perspectives on tax."