Jamie Whyte claims that poverty statistics based on relative measures of poverty are misleading. I explain why his argument is unpersuasive.

Former ACT party leader Jamie Whyte recently wrote that:

There is no poverty in New Zealand. Misery, depravity, hopelessness, yes; but no poverty.

The poorest in New Zealand are the unemployed. They receive free medical care, free education for their children and enough cash to pay for basic food, clothing and (subsidised) housing. Most have televisions, refrigerators and ovens. Many even own cars. That isn't poverty.

Why then do we keep hearing that more than 20 per cent of New Zealand children live in poverty? Those who tell us this do not mean by "poverty" what most people do. They have a statistical definition…

Poverty statistics based on this measure are misleading anyone who believes them.

The sentence that I have put in italics plays a vital and somewhat overlooked role in Whyte’s argument. I will set out Whyte’s argument and some of the responses to it, and then give my comments.

Whyte’s argument

Whyte argues that:

1. By “poverty”, most people mean X.

2. When people claim that more than 20 per cent of New Zealand children live in poverty, they do so on the basis that “poverty” means Y.

3. Definition X and definition Y are different. Meeting definition X is not a good predictor of meeting definition Y.

The conclusion that Whyte appears to draw is:

4. The practice of making the claims described at 2. is misleading. It seeks to engage the sympathy of the public towards poverty (in the sense of X) by using the word “poverty” to make claims which are true if poverty means Y but inaccurate if poverty has the meaning most members of the public give it (ie X).

The purpose of misleading the public in such a way is presumably to gain support for policies that reduce poverty (in the sense of Y) but do not really help with poverty (in the sense of X). These sorts of policies generally turn out to be the sorts of things that the ACT party is not so keen on. For example, Susan St John suggests that we should be spending an extra $1 billion per annum on Working for Families payments. Or, Jess Berenton-Shaw argues that “cash transfers with no strings attached are the best bet for reducing family poverty.”

Responses to Whyte’s article

Dave Armstrong writes in the Dominion Post that:

[I]t's been a pleasant surprise to see some journalists discussing poverty, which the Oxford Dictionary defines as the "condition of having little or no wealth or few material possessions".

Organisations around the world define poverty differently, often to do with relative incomes within a society. Using these or the Oxford definition, poverty definitely exists here. Meanwhile, the poverty deniers, of whom New Zealand seems to have an increasing number, maintain that there is no real poverty in New Zealand, especially compared to places like India.

Danyl McLauchlan blogs in the Dim-Post that:

David Farrar also talks about this all the time. ‘Why is poverty relative? Why not have an absolute measure of poverty?’ Well, the reason is that poverty is an intrinsically relative measure. It’s like ‘warmth’ or ‘tallness’. Yeah, you can say that if you compare New Zealand poverty to Indonesian poverty, say, then there is no poverty in New Zealand. But you could also just as validly compare us to the residents of Mayfair or the Île Saint-Louis or East Hampton, compared to whom pretty much all of us live in abject poverty. Because those are both stupid things to do we don’t actually do that and there’s an official measure of poverty which has broad political consensus.

Deborah Hill Cone writes in the Herald that:

Dr Whyte claims there is no poverty in New Zealand because the measure we use to judge poverty is relative: you are deemed to live in poverty if your household's income is less than 50 per cent of the national median.

This is a disingenuous argument. We are all wealthy compared to homeless people in the slums of Sao Paulo and all paupers compared to mansion dwellers in the Hamptons. Poverty is relative to the lives of others in our community. 

None of these responses quite engage with Whyte’s claim about what most people think poverty means. Armstrong refers to the Oxford English Dictionary and notes that organisations around the world define the term in different ways. McLauchlan and Cone both state that poverty is an intrinsically relative measure. McLauchlan also says that there’s an official measure which has broad political consensus. None of these responses rule out the possibility that there’s a common everyday meaning of “poverty” that regular people use, which is substantially different from that used by academics or politicians talking about “poverty.”

My response to Whyte’s argument

1. By “poverty”, most people mean X

I have said that the responses I have noted do not rule out the possibility that there is an everyday definition of poverty. However, I do not think that Whyte does a particularly good job of convincing us that there is one. Whyte hasn’t given us any particular reason to believe that there is a common everyday definition of poverty. He mostly tells us what poverty isn’t.

I would think that if you are trying to make a case about the meaning of a word, there are two things which you should try and do. First of all, you need to actually set out your definition of the word. Whyte doesn’t do this. He does state that “[a]ny sensible definition of poverty must be relative” and that “poverty is a matter of consumption”. That’s about it. Second, if you are making a specific claim about how people use the word (as Whyte is doing), you need to provide some evidence of that usage.

Regarding Whyte’s definition of poverty, I assume he has a definition in mind along the lines of that offered by David Farrar in his blog post on Whyte’s article:

[T]he best definition of poverty in NZ is whether people are in material hardship, as measured by whether they can afford things such as shoes for their kids.

This is an “absolute poverty” approach to defining poverty, which assesses whether or not someone is in poverty by where they fall relative to the absolute basic needs to participate in society. This is in contrast to “relative poverty” which assesses whether or not someone is in poverty by where they fall relative to other people in society.

Turning to the second thing - whether or not most people mean “material hardship” or something along those lines when they say “poverty”, I do not know. Whyte has certainly not given us much evidence about that. I suspect that perhaps what Whyte is going on is that when he, and the people he knows, say “poverty”, they do not mean relative poverty. That is evidence for what people generally mean by poverty to the extent that Whyte and the people he knows are representative of people generally.

As I have said, I do not know if most people share the same definition of “poverty” and, if so, what that definition is. I suspect that there is no generally accepted definition, at least not in the way that Whyte seems to think there is. At the end of the day, though, the onus is on Whyte to make out this claim, since it is vital to his argument, and he has not done so. Because he has not given us a good reason to accept the first premise of his argument, I could say that Whyte’s argument fails before it gets off the ground and leave it at that. However, I will go on and consider the rest of his points, supposing that (i) most people share a definition of poverty and (ii) that definition is an absolute poverty/material hardship measure along the lines of the one proposed by Farrar.

2. When people claim that more than 20 per cent of New Zealand children live in poverty, they do so on the basis that “poverty” means Y

This is fair enough – when people make such claims they do so on the basis that “poverty” means some relative measure of poverty.

3. Definition Y and definition X are different. Meeting definition Y is not a good predictor of meeting definition X

Material hardship and relative poverty are clearly different measures in the sense that they look at different things. However, it is arguable that there is a strong correlation between the two. McLauchlan states that:

If these guys [Whyte and Farrar] can get that consensus changed so that only people whose living standards are comparable with homeless people living in the slums of Jakarta then yeah, great, there will be no ‘poverty’ in New Zealand any more. But there will be a big group of people who used to be classified as living in poverty, who have poor health and poor educational outcomes and higher rates of crime and unemployment and plenty of other metrics that strongly correlate with our current definition of poverty.

Whyte argues that relative poverty measures are silly because doubling everyone’s income would have no effect on the amount of poverty in New Zealand. That might be true in the abstract, but I think we can safely rule out incomes doubling in the near future. If our starting point is the actual New Zealand that we live in, relative poverty can be said to have a reasonably strong correlation with material hardship. It also happens to be more straight-forward to measure: it’s a lot easier to get information on household incomes than survey everyone on shoe-buying. If there is a strong correlation, then claims about poverty in New Zealand based on relative poverty are not especially misleading.

4. The practice of making the claims described at 2. is misleading.

Even if we accept premises 1-3, the conclusion does not necessarily follow. The public are only going to be misled by claims about “poverty” (in the sense of relative poverty) if they misunderstand them as being claims about material hardship.

Different people can use the same words to mean different things. The same person can even use the same word to mean different things at different times.  Fortunately, this does not make communication impossible, because we use context to work out which meaning a speaker was intending to give to a particular word.

The public are quite capable of understanding that a speaker might use a word to mean one thing, even though people normally mean something else by that word. The claim that more than 20 per cent of children in New Zealand are in poverty would only be misleading if the public do not realise that “poverty” is not referring to absolute poverty. Perhaps there are some people for whom this is a surprise. However, there has been substantial ongoing dialogue in poverty in New Zealand for some time, and I think it would be difficult to have avoided the fact that some people mean relative poverty when they say “poverty”. From there, it is generally not difficult to work out what a particular speaker is meaning especially if, like Susan St John, they refer explicitly to the measure they are using. As a result, Whyte has failed to make the case that the public are being misled about poverty. Whyte himself cannot claim to be misled, because he knows the meaning that statements on poverty are intended to convey. And he has failed to identify anyone else who is being misled.

Conclusion

Whyte argues that poverty statistics based on relative poverty are “misleading anyone who believes them.” He fails to justify this claim. First of all, he does not substantiate his claim that most people think “poverty” refers to absolute poverty. Even if people did adopt such a meaning, it is not clear that this makes the poverty statistics based on relative poverty misleading. Relative poverty measures in New Zealand may be a reasonable proxy for the kind of material hardship that Whyte thinks people think counts as poverty in New Zealand. Finally, members of the public are likely to be aware that there are different measures of “poverty“ and to be able to work out what measure a particular speaker is meaning.