Possible scenarios for child poverty
For this report, the Treasury has undertaken modelling of possible trends in child poverty.
A strong note of caution is required in relation to the detailed findings of this modelling. It is significantly harder than usual to estimate future child poverty levels due to both the uncertainty of economic forecasts and also the availability of survey data that reflects the current situation. Officials' view is that if anything the modelling results are more likely to underestimate the likely increases in poverty rates over the coming years.
Modelling on each primary measure is set out opposite. Note that the modelled reductions for the changes between the 2019 and 2020 years reflect changes in incomes from the 2018/19 to 2019/20 tax years.
Figure 4 - Children in households below the BHC poverty threshold (proportion of children %)
On the BHC moving line measure, rates are modelled to decrease from 2018/19 to 2020/21 - both from the impact of the Families Package, but also owing to a reduction in median incomes for households. Rates then steadily rise, reflecting the increase in the number of households supported by benefits or on lower incomes while employed.
Figure 5 - Children in households below the fixed line AHC poverty threshold (proportion of children %)
On the AHC fixed line measure, rates are modelled to also initially fall, though the reduction from the impact of the Families Package in 2020 is tempered by the economic impact of COVID-19. Rates are then projected to increase, particularly under scenarios with higher rates of unemployment - as on this measure the poverty threshold is held constant in real terms, and reduced employment and earnings flows through to reduced household incomes for some families.
While the expected trajectory for material hardship cannot be modelled, this is the measure that has been most sensitive to changes in economic conditions in the past, and which is most likely to show the largest rise in the coming years.
-  These show modelled estimates based on the BEFU forecast available at the end of March (the green “Budget 2020” line), as well as other projections based on the different economic scenarios published by the Treasury in early April. These scenarios show the impact of different levels of increase in the number of people supported by Jobseeker Support, and are intended for illustrative purposes - to show the sensitivity of the modelling to different economic outcomes. While the scenarios are unlikely to play out precisely as forecast, they are useful to show how the poverty measures could respond to different levels of unemployment. To make the charts readable, only three of the five scenarios were included. The modelled trends are consistent with the patterns outlined above. Relative low-income rates of poverty fall initially, reflecting the impact of the Families Package on incomes prior to COVID-19, and the impact of the falling median once the impact of the COVID-19 crisis takes effect. The fixed line rate of poverty also falls initially, though the impact of the Families Package is more tempered by the impact of COVID-19 on incomes. The rates on both measures then rise, with the rise stronger for the fixed line measure than for the moving line measure.
-  There are important differences between TAWA modelling and Stats NZ's reporting on rates under the Child Poverty Reduction Act 2018. The TAWA modelling estimates rates of poverty on incomes within a given tax year, whereas Stats NZ's reporting measures annual incomes through a survey carried out over a June financial year. In the graphs above, rates for the target year (2020/21) are based on modelled incomes for the 2020/21 tax year, whereas Stats NZ's reporting on that year will include household incomes from the previous 12 months for each household interviewed from July 2020 to June 2021. This means that some of the income in the Stats NZ data will go back as far as mid-2019. This difference is likely to impact some of the detail, but does not compromise the account given above.