Dodgy statistical work to generate headlines supporting higher taxation risks eroding the credibility of Oxfam’s aid work

Opinion: If someone told you that America’s tax system is light-years ahead of either Sweden or New Zealand’s in reducing inequality, you’d be right to think they’re either misguided or lying to you.

So how the heck did Oxfam manage to create a ranking of tax system fairness that put the US 10th best in the world, the UK 24th, Sweden 104th, and New Zealand 136th?

Last week, Newshub reported that Oxfam ranked New Zealand’s tax system as 136th fairest of 161 countries. Greens co-leader Marama Davidson, responding to Oxfam’s report, said New Zealand’s tax system is “really crappy” at reducing inequality.

The New Zealand Herald followed it up with a podcast interview with Oxfam’s New Zealand advocacy director Dr Jo Spratt.

Dr Spratt noted that while New Zealand’s overall ranking in Oxfam’s report was high, the sub-indices that make up the overall ranking provide detail on where New Zealand could do better. She claimed New Zealand’s GST is quite regressive and suggested the tax system doesn’t work well in reducing inequality. And that New Zealand’s tax system is the 136th most fair.

The claim about GST was, and is, at odds with how New Zealand’s tax experts see things. The rest has larger problems, but let’s start with GST.

A background paper produced for Labour’s 2018 Tax Working Group said, “In New Zealand, the most recent research on the issue has described the lifetime impact of GST as “either proportional or at worst slightly regressive”.”

So it is hardly accurate to say that ours is “quite a regressive GST.”

But more importantly, GST is part of a tax system.

When I lectured Public Finance, I told the students that a tax system can achieve any desired level of progressivity by combining a proportional GST and a progressive income tax schedule. This is bog-standard Principles of Public Finance material. Looking at GST in isolation is a serious mistake.

Even more importantly, looking at tax in isolation from transfers is at least as big an error in evaluating overall system progressivity.

But I was particularly curious to see where that ranking came from. I follow data on inequality reasonably closely, and I’d never seen anything putting New Zealand in the very worst global tier. New Zealand is more typically rather middling.

My go-to source on income inequality, and the extent to which tax systems reduce it, is the OECD’s Income Distribution database. You can there compare income inequality in market incomes and in after-tax-and-transfer incomes, as measured by the Gini coefficient.

Both before- and after-tax-and-transfer incomes matter. If a country starts with less income inequality than others, its voters may prefer that its tax system do a bit less to reduce inequality. And if market incomes are more uneven, voters might prefer a bit more redistribution.

The OECD tables include data for very few countries in 2020 and 2021. Only 44 countries are in the overall dataset. 2019 includes data from 27 countries. More recent data would include any effect of the 39 percent tax rate that took effect in April 2021, as well as substantial increases to benefit levels from 2020.

In those tables, New Zealand is very middling. Before taxes and transfers, New Zealand’s Gini measure of income inequality is 17th highest of 27, and comparable to Hungary and Israel. After taxes and transfers, it’s 10th highest of 27, and comparable to Romania, Korea, and Spain.

Differences between middling-ranked countries are not really all that large. The gap between the United States, which has the third highest after-tax-and-transfer inequality, and the United Kingdom, which is fourth, is larger than the gap between New Zealand at 10th, and Canada at 17th.

How much does a tax-and-transfer system reduce income inequality? You have to compare inequality in market incomes before taxes and transfers with inequality in incomes after taxes and transfers.

New Zealand has the 18th largest reduction, if measured by percentage reduction, or 19th largest reduction, if measured by Gini-point reduction. By percentage reduction, New Zealand sits between Canada and the United Kingdom. By Gini-points, we’re between Lithuania and Latvia.

And either way, on the OECD 2019 data, New Zealand’s tax and transfer system does a fair bit more than America’s to reduce income inequality.

Inequality (Gini) in market incomes and after-tax-and-transfer incomes. Source: OECD (2019)

New Zealand is very far from the bottom of the most credible ranking of which I’m aware. Finding New Zealand to be 136th would seem to require that almost every country that didn’t make it into the OECD’s tables ranked ahead of New Zealand in Oxfam’s ranking.

The Oxfam report was little help; the sub-indices that made up the overall rankings were not included.

So I asked Oxfam how the numbers were put together.

Dr Spratt very helpfully pointed me to the underlying data, available online.

New Zealand ranked 136th on Indicator T3: Impact of Tax on income Gini.

The spreadsheet says the measure takes “the share of total tax raised from each of income, corporate and VAT taxes multiplied by the actual or predicted impact of each on the Gini coefficient.”

And Oxfam’s rankings for New Zealand bore little resemblance to the OECD rankings. But there were bigger problems. Among the 27 countries in the OECD tables, Oxfam ranked the US tax system as doing the second greatest amount to reduce inequality; more than the UK (8th), France (9th); Sweden (19th); and, more than New Zealand (25th).

According to the OECD tables, New Zealand’s tax and transfer system does more to reduce income inequality, measured as proportionate drop in inequality, than nine countries. But, according to Oxfam, eight of those nine countries have tax systems that do more to fight inequality than New Zealand’s.

And does it really seem plausible that America’s tax system should be ahead of Sweden’s in reducing inequality?

Reduction in inequality rankings: OECD v Oxfam

The difference is unlikely to be the three years between 2019 and 2022. In New Zealand’s case, benefits went up and a 39 percent top tax rate was introduced over the period.

It is difficult to tell just how Oxfam went wrong, but there are a few possibilities.

I noted that tax and transfer should be thought of as a system. The reason is simple.

Tax credits like the Minimum Family Tax Credit do not count against income tax revenue in our Budget documents. They are treated as government expenditures.

Imagine that, instead of having Working for Families, New Zealand had dozens of different income tax schedules with different income tax rates and thresholds that depended on whether you were part of a couple and on your number of children.

It would be complicated, and a bad idea, but it is eminently possible.

That system could be designed to have identical effects to the current combination of our simpler income tax system and redistribution provided through tax credits based on household circumstances and means.

If a country’s accounts work a lot of its redistribution into explicit tax schedules, that country’s tax system will appear to do more of the work in reducing inequality. A ranking like Oxfam’s would praise that tax system while potentially damning its benefit system as being inadequately redistributive.

Oxfam’s overall spreadsheet includes a different column tallying Social Protection spending as percentage of total spending. There, New Zealand ranks 34th of 161 countries. If tax expenditures like Working for Families wind up included in that column, as they should be, rather than counting towards progressivity in the income tax system, Oxfam’s rankings potentially start making sense. But they have to be counted together, as part of a combined system – like the OECD measure.

What really does not make sense are claims that New Zealand’s tax system is among the least fair in the world – based on a measure that excludes some of the more important ways that our system engages in redistribution.

Subindices add heat rather than light when they separately target things that only make sense when measured in combination.

Costa Rica’s ranking was particularly odd.

According to the OECD tables, Costa Rica has the highest levels of income inequality whether measured before or after taxes and transfers. It has the smallest reduction in income inequality of any of the 27 countries in the OECD table. Oxfam ranked Costa Rica 13th of 161 countries – more than a hundred places above New Zealand – and 4th among the OECD-ranked countries.

The single highlighted measure, on which New Zealand ranks 136th, simply does not seem to provide any valid reflection of differences across countries in tax system progressivity.

Also worrying is that one of Dr Spratt’s preferred remedies, wealth taxes, would not move New Zealand’s position on her index. Oxfam’s tax rankings do not include any effects of wealth taxes: only income, company, and consumption taxes. New Zealand could introduce a heavily punitive wealth tax tomorrow on top of existing taxes, and the next iteration of the Oxfam index would still generate press releases about New Zealand’s unfair tax system.

We could give Oxfam the benefit of the doubt here and chalk it up to honest error, but Oxfam does have a history.

Centre-left economist Noah Smith, earlier this year, in a column titled “Oxfam serves up a lot of dodgy statistics”, tried to make sense of Oxfam claims about global poverty. The numbers simply did not add up, and not just because there was a lot of double-counting.

But it isn’t just tax figures and poverty figures.

Every October, Credit Suisse puts out a report on Global Wealth. And Oxfam typically follows it up with its own gloss on the figures well-keyed to the headlines that journalists like to write, but not particularly reflective of the underlying data.

The 2019 Oxfam wealth report was fairly typical. Where Credit Suisse had found that wealth inequality in New Zealand had fallen from a Gini figure of 72.3 to 70.8, Oxfam highlighted wealth increases for two people on the Rich List for a Dominion Post headline claiming growing inequality. Meanwhile, work released by Treasury in 2021 notes that wealth inequality over the last decade “appears to have been slowly decreasing (becoming more equal), trending down by an average 0.5 percentage points per year since 2010”.

Noah Smith’s critique also highlighted Oxfam’s treatment of wealth statistics. He noted that “Oxfam is a serial repeat offender of dodgy statistics.”

It really is a shame. Decades ago, Oxfam was mainly known for good work in providing international aid to people in desperate need.

That historic reputation means their reports rarely receive appropriate scepticism.

Dodgy statistical work that seems aimed at generating headlines in support of higher levels of taxation and redistribution risks eroding the credibility of the organisation’s aid work.

Leave a comment