Brian Fallow: Family help struggles in sea of complexity

For a programme that will benefit 350,000 families and lift one New Zealand child in 10 above the poverty line, Working for Families is strangely unloved.

Perhaps that is because it was so long coming and is still not fully implemented. It was not until the fifth Labour Government's Budget last year that we saw a major income redistribution policy.

Fully implemented it will cost about $1.2 billion a year, which would rise to $1.6 billion if the extension Labour wants to implement is adopted.

But family assistance remains a nightmare of complexity, especially compared with the simplicity of tax cuts. Entitlements are festooned with ifs and untils and unlesses.

How much you get depends on how many children you care for and how old they are; on your income and your partner's income; whether it is from a benefit or from working; and, if the latter, how many hours you work.

Add the accommodation supplement and child-care subsidies and you have a system as complicated as three dimensional chess.

In part, the complexity reflects an attempt to balance three different objectives: reducing poverty, boosting the supply of labour by "making work pay" and keeping the fiscal cost low.

Keeping the cost down requires targeting that, in turn, creates a boundary zone where people face high effective marginal tax rates (EMTRs): the combined effect of income tax and the whittling down of their entitlements means they lose a high proportion of every extra dollar they earn.

High EMTRs impose high deadweight costs, an ugly name for a bad thing, limiting options, distorting choices, erecting barriers between what people want to do and what it makes financial sense for them to do.

Working for Families sets out to demolish EMTRs that trap people in poverty but erects them afresh further up the income scale, creating disincentives for middle-class mothers to work.

The policy already enacted will push up in April next year the threshold at which family assistance starts to abate.

Labour's policy announced last month would push it up further still, to $35,000, and reduce the rate at which it abates from 30 per cent to 20 per cent as in Australia.

The flipside is to move the high EMTRs further up the income distribution.

Patrick Nolan, in a PhD dissertation recently submitted to Victoria University of Wellington, studies the use of tax credits to deliver family assistance in New Zealand and other Anglo economies.

He found there is a trade-off between increasing the labour supply from sole parents and decreasing the labour supply from secondary earners.

The latter reflects the use of a couple's joint income as the basis for assessing entitlement.

"Basing entitlement on joint income means secondary earners generally face higher effective marginal tax rates from their first dollar of income than primary earners and when secondary earners reduce their earned incomes there is little reduction (or an increase) in joint income.

"Some secondary earners are effectively subsidised to stay at home."

If the object of the exercise is to increase the labour supply by getting more mothers back to work this is a perverse outcome.

Nolan said that while New Zealand used income as the basis for targeting, in other Anglo countries the focus was more on the number of children and their ages.

"We have given the most emphasis to fiscal cost. We are the cheapest in a sense," he said.

"Canada and even the United States have a child tax credit that goes well up the income distribution.

"The UK and Australia do as well and the UK also has a universal child benefit. We don't because of the cost."

So why not return to a universal child tax credit?

After all the old family benefit was a universal entitlement for the parents of dependent children until the Mother of All Budgets in 1991.

But that was against a background of straitened fiscal circumstances and high unemployment, the reverse of the present position.

If it makes sense for the tax system to distinguish between the parents of dependent children and other taxpayers then why not just do that?

Why attach all the other conditions about income, hours worked and so on?

Some of that tax break would be poorly targeted, that is, wasted on people on higher incomes but they will be paying more tax anyway.

Does it matter if Graeme Hart and John Key get a tax credit for having dependent children if under our progressive tax scale or even under a flat one it is dwarfed (one would hope) by the flow of money from them to the Inland Revenue?

The problem is that while replacing targeted assistance with a universal tax credit (dependent only on the number and perhaps ages of the children) would ease the problem of high EMTRs it would push up the fiscal cost and, therefore, the tax burden on other people, all else being equal.

The less efficiently you target, the more expensive it becomes to reduce poverty, Nolan says, and that ultimately was the priority.

If you define poverty as belonging to a household with an income less than half the median household income, then about 15 per cent of children are below the poverty line.

Officials estimate that the Working for Families package will reduce that to 4 or 5 per cent by 2008.

If you set the bar a little higher, defining poverty as 60 per cent of the median income means the child poverty rate would fall from about 29 per cent to 20 per cent.

But Gregory Dwyer, in a paper commissioned by the Business Roundtable, says the reduction in forecast poverty arises because the incomes of many families are just below the lower poverty benchmark or between the two property lines.

He argues that the package does too much for families with children who are not poor.

Nolan cites Treasury data that show 61 per cent of the spending on Working for Families goes to the poorer half of all households and only 15 per cent to the richest four deciles.

He says spill-over to the non-poor is a feature of social assistance provided through the tax system.

But there has been a trend among Anglo countries to deliver family assistance more through the tax system than the welfare system despite all the problems that arise from trying to integrate the two.

The National Party says the whole structure of Working for Families is fundamentally flawed.

Nonetheless, it would keep it (minus the planned extra $10 a child a week scheduled for 2007) until it could be replaced with a family taxation system that left no one worse off.

Easier said than done.

The inherent trilemma remains: the triangular trade-off between providing adequate income support where it is needed, minimising the cost to other taxpayers by targeting that support and then minimising the disincentives to work arising from high effective marginal tax rates at the edge of the target zone.

And the liberality of National's tax cuts package would leave it with little room to manoeuvre.

How much you get depends on:

How many children you care for and how old they are.
On your income and your partner's income.
Whether it is from a benefit or from working.
And, if the latter, how many hours you work.