The equalizer - what government provides

Dr Eric Crampton
The National Business Review
31 August, 2018

People can argue the toss about whether the government should be more or less active in redistributing income but it’s good to have an accurate picture of how much the government already does to reduce inequalities in market income. It’s especially important when the government’s reviewing the tax system.

Basing calls for greater redistribution on wage differences never made much sense to begin with. But it’s worse than you might have thought.

We all know the drill. Income inequality has been stable since the early 1990s, whether measured as the Gini coefficient on market earnings or as the Gini coefficient on after-tax-and-transfer earnings. Pointing to differences in pay across people as an argument for increased redistribution forgets that the state is already heavily involved in income redistribution that must be taken into account.

Inequality in household incomes for those aged 18-65 drops by a little more a fifth once we adjust for existing redistribution and taxation.

But that misses the government’s role in providing universal services.

Statistics New Zealand this week released a new experimental data series looking at the redistributive effects of government spending on education and healthcare.

This kind of work is not completely new.

In 2012, Policy Quarterly published work by Treasury analysts that tried to figure out what inequality in final incomes, after tax-and-transfers, looks like if we count health and education services provided by the government in the overall measure. They assessed how much the government spends on health and education for households in each income decile.

This earlier work showed the share of overall education funding going to households in the bottom half of the income distribution increased from 49% in 1988/89 to 54% in 2009/10 while richer families benefited disproportionately from government spending on tertiary education.

When taxes, transfers, and universal spending programmes were put together, they found households in the poorest four income deciles received at least $20,000 more a year in redistribution and in services than they paid in taxes, households in the three middle deciles came out about even, and households in the top income decile contributed about $50,000 more in tax than they received in transfers or services in 2009/10.

Household spending
The latest Statistics New Zealand work does not let us provide an update to the earlier Treasury figures on the net fiscal impact of existing redistribution because it does not include all taxes paid. Statistics NZ provides the caveat that the data is experimental.

But it does give us a more current look at spending across richer and poorer households.

There are myriad ways that the education system fails poorer communities but that failure is not due to schools in rich communities having a lot more money to play with. There is a lot of redistribution built into the existing education system – and arguably rightly so.

The Ministry of Education even provides a simple funding calculator that lets you plug in the details of your local school and see the funding effects of changing the school’s decile. For my kids’ school, dropping from the highest decile to the lowest decile would increase its grant from about $1500 a student to about $2500 a student. Parents at higher-decile schools make up some of the difference in voluntary contributions and at the school fair.

Statistics New Zealand’s latest figures show that while households in the highest income quintile spend about $150 (or about 38%) more on education services than households in the poorest quintile, government spending on education services is about $1100 higher for households in the poorest quintile. Central government education spending in 2015-16 averaged just under $1500 for households in the highest income quintile and just over $2600 for households in the poorest quintile – despite tertiary education spending that skews toward richer households.

A more thorough analysis would and should adjust for age differences across households in different income quintiles; that work is impossible from the data released.

Government service effects
We can look, though, at the effect of government spending on in-kind transfers like health and education on one measure of inequality: the ratio between the highest and lowest quartile, or the Q5:Q1 Ratio. That ratio has a lot of problems as well, not least of which being that it takes annual snapshots of household incomes and misses how household income moves over an individual’s life cycle. But it does let us get a handle on the effects of government services.

Gross disposable income, which includes taxes and cash transfers, is about six times higher in the top 20% of households than in the bottom 20% of households. Adjusted disposable income adds in government spending on in-kind transfers like health and education. Once we adjust for those in-kind services, the ratio drops by a third.



Inequality in actual consumption is much lower: people smooth their incomes over the life-cycle, borrowing while young, saving while earning, then spending in retirement. The highest-earning households spend about twice as much as the lowest-earning households.

You can’t get an ought from an is. But if you do have an ought in mind, whether policy should change will depend on where is is. And when we’re thinking about redistribution, it’s too easy to forget how much of it the government already provides.




Stay in the loop: Subscribe to updates