Not so much from a lack of nominal income but from rising mortgage interest rates

The just released Statistics New Zealand (SNZ) estimates child poverty for the year ending June 2023 show the proportions of children on nine different poverty measures are higher than they were in the June 2022 ending year. SNZ warns that the increases are not large enough to outweigh the sampling error but here I accept the conclusion as meaningful and discuss why.

(Each indicator in 2023 is below the 2018 measure, when the SNZ first began measuring. Even so, the data is not on track to hit the child poverty reduction targets set out in the 2018 Child Poverty Reduction Act.)

Some readers may find the next few paragraphs a bit tediously data (and definition) driven. If you trust the analysis – you shouldn’t without checking it – you might skip to the next italicised paragraph – eight paragraphs down – which provides a conclusion.

Average annual household equivalised disposable income (before housing costs) rose 7.0 percent between June 2022 and June 2023. The Consumer Price Index (CPI) rose 6.8 percent. Seems households are fractionally ahead?

I am now share with you a secret. The CPI does not reflect all the outgoings of households. Very importantly, it omits mortgage interest outlays. (There is a complicated story why it is omitted.)

Now mortgage interest rates rose sharply during the period. For instance, floating rates on new mortgages increased from 5.0% p.a. for the year ending June 2022 to 7.5% p.a. for the year end June 2023. According to SNZ, all household mortgage interest payments jumped 50 percent.

SNZ also produces a Housing Living Cost Price Index which includes mortgage interest payments, but it is rarely reported. According to the measure, between June year ending 2022 and June 2023 household prices rose 7.7 percent, almost 1 percentage point higher than the CPI rise. It is also greater than the 7.0 percent rise in incomes. The rise in poverty is now less surprising.

The interpretation of the household equivalised disposable income figure of a 7.0 percent rise in incomes assumes that mortgage interest payments are spread evenly through the income distribution. SNZ estimates that there are an extra 12,000 children in poverty between the two years. (This uses a poverty line based on 50 percent of median (equivalised) household incomes.)

But households with children are more likely to have mortgages, so they are likely to be hit more heavily by the rise in interest rates. SNZ also provides an estimate of the change in all household incomes if housing costs (including mortgage interest) are deducted. It showed an increase in disposable income after housing costs of 4.4 percent, well below the 7.0 percent increase of incomes before housing costs and the various price index increases. When SNZ allows for this, it estimates that there were an additional 36,000 children in poverty – the increase is as common sense would predict.

Working with a large data base sometimes results in conclusions inconsistent with common sense. Usually that is because one has made a mistake or an oversight. This column simplifies how tricky the analysis can be. I did a lot of cross-checking. I have put some of the results in an appendix.

SNZ also provides another estimate of poverty, based on asking households whether they are having to make serious cuts in their spending. On their preferred measure they think ‘hardship poverty’ among children rose 23,000, from 10.5 percent of all children to 12.5 percent. Different poverty line, different level, but a similar increase in the poverty rate.

The above analysis, and the appendix, shows that the rise in child poverty was not so much weakness in income growth. Rather the evidence points to rising mortgage interest payments.

We know why interest rates have risen. The Reserve Bank has been increasing them to restrain the economy in order to reduce inflationary pressures. (World interest rates rose in the same period for similar reasons.) I do not want here to get into the intricacies of this macroeconomic-monetary policy, but to draw attention to the way it impacts on the income distribution (as most policy changes do).

It is ironic that among the people most impacted by the policy are the most defenceless and innocent – children. John Kenneth Galbraith once said that the unemployed were employed fighting inflation; so, apparently, are children.

If you raised the issue with the Reserve Bank, I would expect it to say something like while it is aware of distributional issues, its legislation charges them with restraining inflation but has no mention of doing so fairly or taking the concerns of the weak into consideration. That, the Bank would probably say, is the responsibility of Treasury and fiscal policy; in any case it has no policy instruments to directly modify the income distribution..

We sort of recognised this when both National and Labour campaigned on raising Working for Family (WFF) payments. Note that WFF does not target families paying mortgages. It is hard to think of a viable fiscal instrument that would. WFF is intended to benefit many families with children, so a hike is likely to reduce child poverty to some extent. We shan’t know until February 2026 (and there may be other factors that will increase it).

What surprised me when I did the above analysis is that the evidence is that, until June 2023 anyway, there is not a special case that households need income tax relief. There are always demands for cuts, of course, and National campaigned on income tax cuts. It is quite likely that they had not looked at the evidence – who does? Their promise to cut income tax may be based on the ideological desire to reduce the scope of the state and increase the size of the private sector – an honourable political ideology (although not particularly mine). Thus their decision to stop indexing benefits to rising average wages, returning to indexing by the CPI, even though it will increase child poverty.

My guess as to what is currently going on in tax policy is that state sector spending is proving much harder to cut than promised and the economy seems to be deteriorating more than expected. That is going to make it much harder to deliver the promised income tax cuts within the coalition government’s promised public debt target.

If the public debt track is higher, it will be paid for in the future by today’s children. If the Reserve Bank decides that requires a higher interest rate track, it will be paid for by today’s children. If we get the macroeconomic stance wrong and unemployment rises, the income cuts will be paid for by today’s children – in part anyway. Funny, isn’t it, how often the frontline payees are the weak and innocent?

There are two appendices. One reports on the data I used in the above analysis; the second is some notes on disability and poverty.

Appendix I: Changes Between the June 2022 and the June 2023 Year

Consumer Price Index: up 6.8 percent.

Housing Living Cost Price Index: up 7.7 percent.

Household (Equivalised) Disposable Income: up 7.0 percent.

Household (Equivalised) Disposable Income after deducting housing costs: up 4.4 percent.

Ordinary-time hourly wages: up 7.3 percent.

The minimum wage up 10.8 percent (but not everyone on the minimum wage has children).

The net social security benefit for sole parent support: up 8.1 percent, and the family tax credit up by 11.4 percent (these are but representative of benefits generally).

The unemployment rate went up from 3.25 percent to 3.43 percent (or about 13,500 souls). (Some research I did many years ago suggested that incomes at the bottom of the distribution were very sensitive to unemployment, but my study focused on far bigger changes than have occurred recently.)

Appendix II: Disability and Poverty

According to the SNZ data, the proportion of children with a disability living below the poverty line (defined as 50 percent of median equivalised disposable household income) is much the same as the proportion for non-disabled children. (They are reported as about 1 percentage point higher over the last four years). However, on the measure of severe material hardship the proportion of disabled children in poverty is about two-and a half times that for the non-disabled children (5.7 percentage points).

This is no surprise. Disability generates additional expenses. We make no adjustment for this in the poverty-income measures.

We knew this when the poverty measurement paradigm was developed five decades ago. But we have hardly progressed our thinking since. Over the five decades we have never taken serious research on poverty seriously – it just has not been a research priority. The weak missing out again?