One Could Reduce Child Poverty At No Fiscal Cost
Following the Richardson/Shipley 1990 ‘redesign of the welfare state’ – which eliminated the universal Family Benefit and doubled the rate of child poverty – various income supplements for families have been added, the best known being ‘Working for Families’, introduced in 2005. The result of the various ad hoc incremental adjustments with confused objectives is a difficult-to-understand and poorly targeted system of family assistance.
As you might expect from such a Heath-Robinson arrangement, the outcome is inefficient in that it is both an expensive means of reducing family poverty and not very effective at reducing the worst child poverty. The clumsiness is well recognised but every attempt to get a better system of supporting children has failed because the approach has been incremental rather than a fundamental redesign – more strings to the rackety structure.
A simpler delivery of the same income support would markedly reduce child poverty without costing the state anything more – so inefficient are the current arrangements. Let’s call it a ‘Universal Family Benefit’ ((UFB). If differs from the one introduced in 1946, by its contribution reducing as the family earns other income. The bleed-out rate on this extra income is 39 percent (the top income tax rate).
To evaluate the proposal, I’ve used the Treasury's microsimulation model of the New Zealand personal tax and transfer system TAWA (Tax and Welfare Analysis), which they use for assessing tax and benefit changes. (The Treasury is not responsible for these results, but thankyou for their help.)
The system would work by families choosing to go onto a different tax code – let’s call it the Family Tax Code (FTC) – which involved their paying 39 percent on all their market income. Additionally, the family would receive for each child an untaxed benefit of $255p.w. (the precise calculations apply for the 2021/2 year). They would not get all the other family tax credits and benefits would be stripped out (except for the adult part of the benefit which goes to solo parents). (Some high-income families would be worse off if they went onto the FTC; they would choose to remain on the existing tax code instead.)
The level of the child benefit was chosen so that the package had net zero fiscal cost (overall cost to the government). Yet despite the fiscal balance, the proposed package would result in 64,000 fewer children being below the poverty line (using the ‘moving-line BHC50’ measure; the broad conclusion of a marked reduction in child poverty will apply for any sensible poverty line). TAWA thinks there are about 115,000 children below that poverty line, so the UFB package would reduce the numbers in poverty by over a half – the ambition of the Child Poverty Reduction Act – without any extra government spending. One might argue that the current system of income support for children is less than 50% efficient.
There are modifications which would reduce child poverty (and increase the efficiency) even further. For instance, there is a case for having the family benefit higher for the first child. Best Start, which provides additional support for recently born children, should be kept. The simulation also left the existing housing and early child education support in place; both can be rationalised, with further gains.
Since the UFB is fiscally balanced, if some children (and their parents) are better off, it follows that some others (and their parents) are worse off. In fact, the scheme pushes 6,000 children below the poverty line as well as lifting 70,000 above it (hence the net 64,000 children).
The UFB also reduces the incomes of many families, although they would still be above the poverty line. That poor targeting is where the inefficiency of the current scheme comes from. (Typically, these are smaller families; the current regime underfunds the poor in large families.) This is an example of Rabin’s Law – named after an American economist, Matthew Rabin – that all policy change makes somebody worse off.
The big challenge of introducing a new redistributive scheme is how to get from the current inefficient scheme to a more efficient one without causing too much pain to those who are made worse off. That is the trick of the different tax code. A family does not have to join the scheme. On the other hand the rule would be that no family could join the existing Heath-Robinson scheme. Moreover, the old scheme rates would not be increased, so that it would become less costly and eventually phase out over time over time as families dropped off. (So there would be transition costs, which would temporarily unbalance the budget. Most incremental redesigns assume that their new scheme would be more costly.)
What is the catch (once we have got through the transition phase)? First, it involves combining both parents’ income, but that happens already. But it does not involve any household paying more tax, since parents could leave the tax code and miss out on the accompanying family benefit.
The big issue is the 39 percent uniform income tax rate. Many people will judge that too high. A high rate is necessary in order to finance the scheme. The tax rate could be reduced, but that would mean a lower UFB and more children left in poverty (while well-off families would do better).
The UFB is an extension of minimum income support. Those over 65 already have it in New Zealand Superannuation. However, universal minimum income schemes require high tax rates.
It is possible that some families faced with the high rate will choose to reduce the hours they work. In effect, income inadequacy is forcing them to work longer hours than they judge prudent. Many will be able to spend more time with their children, which is no bad thing. The most likely reduction will be parents choosing to finish work early so they can be home in the afternoon or during school holidays.
In any case, there are families already facing marginal income taxes far in excess of 39 percent. It is just that the muddle of current rules makes it hard to identify when that happens; the UFB is more transparent and less confusing (it may even require fewer bureaucrats to run it).
The reason that the scheme does not eliminate all child poverty is because it does not provide a minimum income for parents – that would require much higher income tax rates. But it will markedly reduce income stress in most poor families below the poverty line.
So it is possible to reduce child poverty substantially with simpler, more transparent and better targeted income support at no extra cost to the taxpayer in the long run.
As a final point, observe that this is an example of how difficult it is to replace a badly designed policy by a better one – a consequence of Rabin’s law. A poor-quality policy makes some people inappropriately better off. A better policy will make them worse off and they will resist the change. There are many other examples of badly designed policies; sometimes I think New Zealand specialises in them.