The sad, sad story of the UNICEF Child Poverty Report and its critics

David Morely, UNICEF Canada’s Executive Director, has just issued a bold challenge. “It is clearly time for Canada to prioritize children when planning budgets and spending our nation’s resources, even in tough economic times,” says a press release announcing the publication of a report on child poverty.

In fact, the UNICEF Innocenti Report Card released today is the 10th in a regular series on child poverty in rich countries, each report hitting the headlines every second year or so.

Sadly, when it comes to discussions of child poverty kick-started by these reports there are two things that are not new: the conclusions; and the reaction of pundits and many policy makers. I say “sadly” because the two are not linked, and public policy discussion is not the better.

UNICEF documents that at 13.3% Canada’s child poverty rate is almost two percentage points higher than the overall national rate. And of the 35 countries studied, Canada ranks 24th, in the bottom third.

You think that’s bad, check out the US figures: at 23.1%, and 34th out of 35.

It is just these sort of cross-country comparisons the structure of the report encourages: Iceland, Finland and Cyprus finish 1st, 2nd, and 3rd; the US, 34th, just ahead of—wait for it—Romania.

Here is the sad story of how the critics will respond.

I suspect the standard response will be to throw mud at the numbers. Most reports on poverty are greeted by snickers of derision because the calculations are based upon a relative rather than an absolute poverty line. Relative poverty rates are not poverty at all, they are measures of inequality, the critique continues, and as such can never be eliminated.

An “absolute” poverty line is often said to be a line determined by the cost of buying a basket of goods associated with a minimally necessary standard of living; a “relative’ poverty line is tied to some typical level of income, and it changes through time according to what is considered typical.

The UNICEF report takes these concerns head-on, and argues that it is absolutely right to be using a relative poverty line, in this case one-half of the median individual income.

In rich, and for that matter in many not-so-rich, countries “poverty” means not being able to fully participate in society, not just being free of hunger and other basic necessities required for survival, whatever that could mean.

The best economic theorists have taught us this over and over again. Here is the greatest of them, Adam Smith (whose book The Wealth of Nations is often said to have given birth to the science of economics), writing in 1776:

By necessaries I understand not only the commodities which are indispensably necessary for the support of life, but whatever the custom of the country renders it indecent for creditable people, even of the lowest order, to be without. A linen shirt, for example, is, strictly speaking, not a necessary of life. The Greeks and Romans lived, I suppose, very comfortably though they had no linen. But in the present times, through the greater part of Europe, a creditable day-labourer would be ashamed to appear in public without a linen shirt, the want of which would be supposed to denote that disgraceful degree of poverty which, it is presumed, nobody can well fall into without extreme bad conduct. Custom, in the same manner, has rendered leather shoes a necessary of life in England.

More recently Amartya Sen reached the same conclusion, along with a host of other theorists during the 200 years separating Sen from Smith and highlighted in the UNICEF Report.

The distinction between “absolute” and “relative” poverty lines is a false one, there are only relative lines with people disagreeing on how rapidly they should be updated.

Not only theory, but also the conduct of public policy makes all of this clear.

I review the current state of poverty statistics in this post. I also offer more detail and suggest that it is helpful to talk about the difference between “fixed” and “moving” poverty lines in a paper called “Principles and Practicalities for Measuring Child Poverty in Rich Countries.”

But the best example is the long debate over the official poverty line set by the United States. Rebecca Blank—a noted academic expert on poverty—gives a particularly good presentation of the issues in a paper she wrote just before accepting a position with the Obama administration.

In the early 1960s the US administration produced a poverty line based in part on food consumption patterns during the late 1950s to aid it in monitoring Lyndon Johnson’s “War on Poverty”.

At that time the line amounted to about one-half of median income. But for decades it was not updated, and continues to be produced as the “official US poverty line”: in spite of the fact that it is much lower than 50% of median income; in spite of the fact that consumption patterns have changed dramatically; in spite of the fact that all sorts of other government supports to the poor don’t get figured into the calculation; in spite of the fact that what it takes to fully participate in American society has changed dramatically.

Those who want to minimize poverty have adopted a relative line for the 1950s, called it “absolute”, and thrown a lot of muck to block attempts at updating it as society has evolved.

Further, the argument that poverty can’t be eliminated when it is measured in relative terms is just plain wrong, reflecting at best a misunderstanding of basic statistics.

Margaret Thatcher held this view, and in fact, if I remember correctly, it was actually voiced by a UNICEF spokesperson in the press coverage following the Innocenti Report Card number 6 that I was involved in writing.

Report Card 10 very clearly explains that the poverty rate calculated from a line based upon the median income, say one-half of median income, is not the mirror image of income inequality.

If all income earners were ordered from lowest to highest, the income earned by the individual who had as many people ranked below as above would be the median income. Give a whole bunch more money to Mr. High Tech who is already at the very top, and inequality will increase dramatically, but the median will not change one bit because the rankings have not changed. The ‘median’ is the individual exactly in the middle, and that person is still exactly in the middle.

Poverty defined on the basis of one-half of the median is about inequality in the bottom half of the income distribution. It can be completely eliminated if the poorest do not fall too far below those in the middle, with the median representing what it takes to participate normally in society.

The UNICEF report explains all these issues in detail, offers a complementary indicator of poverty based on deprivation, and comes to the conclusion that the “whole idea of defining child poverty in an absolute sense rests on shaky grounds. Unless we wish to argue that the threshold should be set at the minimum income necessary for sheer physical survival then there can be no such thing as an absolute poverty line.”

That the report spends so many pages on measurement issues is a sad statement that these same old arguments will likely once again be paraded out by critics to dull any momentum to make child poverty a public policy priority.

The more time spent on this discussion, the less that will be devoted to examining what the Report actually tells us and how UNICEF should improve future versions.

So here is the sad story of the Report: it paints the same picture as it did five years ago. I mean this in two ways.

Part of this is a reflection of reality: it is sad that child poverty rates are where they are, and that they have not changed. Compare the charts in the current report with those in Report Card 6, which was released in 2005. Many of the countries in the top third of the league table are pretty well unchanged, as are those at the bottom.

But the second part of what is sad is the structure and content of these reports. If UNICEF wishes to make a deeper contribution to debates on child poverty, then maybe it is time to change the format, de-emphasize country rankings, and highlight the progress or lack of it in places of most concern.

Country rankings are a valuable communication device, destined certainly to get headlines, and perhaps to shame some into action. But there is a counterproductive element to them as well.

The United States, for example, perennially appears at the bottom of these comparisons.

With a child poverty rate of 23.1% the country is ranked just ahead of last place Romania at 25.5%. What are we to make of this? The message that roughly one-quarter of children live in households with incomes significantly below a typical income gets lost in the totally inappropriate juxtaposition of two countries differing so dramatically in what is typical.

Cross-country comparisons are certainly helpful when the pairing represents a potentially relevant counterfactual.

The best example in the Report is the contrast between Canada and the US: child poverty rates based upon market incomes are the same in the two countries, but are much lower in Canada when taxes and transfers are taken into account.

But care is needed even in comparing these similar countries. The US  poverty rate is likely overstated, probably by a significant amount.

In a paper published in 2006,  Irwin Garfinkel, Lee Rainwater, and Timothy Smeeding find “that counting in-kind transfers … and taking account of indirect taxes on the value of cash benefits substantially shrinks cross-national differences and alters country rankings in respectively the aggregate value to recipients of welfare state spending and inequality at the bottom of the income distribution.”

Value-added taxes and in-kind transfers are not taken into account by the underlying data used by UNICEF.

For all these reasons is not an equally appropriate counterfactual for a country, that country itself; where it stood roughly four or five years ago, or even better where it stood when the current administration took office?

Over the course of five years the child poverty rate in the United States has not changed at all. Isn’t that just as relevant as knowing that the country ranked 34th out of 35?

Indeed, given all of the attention that this report pays to “monitoring” government progress, the most obvious indicator is missing: how has child poverty changed in each country since the last report focused on this issue?

All countries could learn from examples of progress and regression, but most importantly this would come closer to the so-called monitoring that the organization is calling for, and in fact would be a clear reflection of the tone in the Convention on the Rights of the Child: the ‘progressive’ realization of child rights.

Try as policy makers might, but they will never change a United States into an Iceland, or for that matter into a Canada. However, they might just improve the situation of children in the US compared to where it was five years ago.

In a sense the best aspect of these UNICEF reports is not so much the cross-country comparisons, but rather that they ask us to reflect on ourselves, where we have been, where we are going on this important public policy issue. This is the perspective most important for monitoring each country.

Cross-country comparisons help us to do that to a degree, but changes over time do so as well. What the Report needs is a league table in changes, not just in levels.

The Canadian office of UNICEF suggests: that child benefits and tax credits could be improved in Canada; that the government should establish a national poverty reduction strategy with a focus on children; and finally that it should adopt an official definition of poverty as a way of highlighting the severity of the problem and monitoring progress in addressing it.

The most relevant piece of information to support these recommendations is the finding that the child poverty rate in Canada is greater than the overall national rate (something that also takes place in 23 of the other 34 countries).

But at the same time Canada’s child poverty rate of 13.3% is apparently a bit lower than the 14.9% published by UNICEF seven years ago. Contrast that with the much more significant drop in the United Kingdom, a country that over this period did adopt the type of policies Mr. Morely is calling for.

Alas, we can’t even be certain of this. A troubling footnote in Report Card 10 tells us, without explanation, that the data are not “strictly” comparable between the two reports. So we don’t even have a clear answer to a simple question necessary to monitor a government’s progress: how has child poverty changed during the last five years?

UNICEF is asking Canadians to adopt an “official” poverty line to gauge progress in making child poverty a priority.

But given the US experience, this might lead to more grief than it is worth. It is not so important that a government set an official poverty line, but rather only that it agree to be evaluated by an accepted standard.

It might be best to have this established and updated by an independent and credible agency. As governments come and go the measure will have a life of its own, it will not be changed for political reasons, nor will a constituency develop to stymie updating when it is appropriate.

It would be all the preferable if that agency had the mandate and credibility to speak for children. What better challenge for UNICEF to take on in its next Report Card!