For comparability across countries with different income levels, the report uses a working definition of "poverty" as half of the median income in that country. This graph shows the share of children in each country who are growing up in families where the income is less than half the median income for that country. The U.S., with 23.1% of its children in such families, ranks 34th out of 35 countries.
It's important to be clear on what this graph doesn't show. It doesn't show that U.S. children are more deprived in absolute terms than children in other countries. The U.S. has a higher level if incomes than these other countries--much higher than some of them. Indeed, the UNICEF report notes that half of the median income for the 10 richest countries in this table is more than the actual median income for the 10 poorest countries in the table.
In addition, income is more unequally distributed in the U.S. than in many of these other countries. As a result, the U.S. will have a larger share of its population living in households that earn less than 50% of the median income compared with countries that have a much more equal distribution of income like Iceland or Finland. The UNICEF report offers an extended discussion of poverty lines set in absolute terms and those set in relative terms, and the report offers data and examples for both approaches. The argument that it can be useful to look at a relative poverty line, like 50% of median income in a country, goes like this:
"In sum, a relative poverty line drawn at 50% of median income is an attempt to define a concept of poverty on which there is widespread agreement in principle – a concept which says that the poor are those who do not have access to the possessions, amenities, activities and opportunities that are considered normal by most people in the society in which they live ... 50% of the median is a plausible measure of what it is intended to measure – the sense of falling so far behind the norms of one’s society as to be at risk of social exclusion. ...Thus, this argument holds that the reason why the proportion of children in households below 50% of median income matters is that it represents the share of children missing the "possessions, amenities, activities and opportunities that are considered normal by most people in the society in which they live." As an example of how these forces play out, I posted on May 23 about "Dimensions of College Attendance," One figure in that post shows that for Americans born between 1979 and 1982, of those born into families in the bottom quarter of the income distribution, 9% completed a four-year college degree by age 25, and of those born into families in the top quarter of the income distribution, 54% completed a four-year college degree by age 25. I strongly suspect that this enormous gap has little to do with the cost of college or the availability of loans, but instead is closely linked to how those born into lower-income families get on average less support from family, local community, and the K-12 education system to prepare them for a college degree.
The UNICEF report also looks at the share of children living in households with less than 50% of median income before and after government taxes and transfers are taken into account. The darker blue bars show the child poverty rates from the preceding figure--that is, after government taxes and transfers are taken into account. The lighter blue bars show what the poverty rate among children would have been, if those taxes and transfers had not occurred. For countries at the top of the list, like Ireland, Hungary, the United Kingdom, Finland, and Australia, the overall effect of government taxes and transfers is a dramatic reduction in the proportion of children that would have been living in households below half of the median income. In the U.S., in contrast, the overall pattern of government taxes and transfers lead to a relatively small reduction in the number of children in such households. In Greece, remarkably enough, the overall pattern of government transfers actually increases the share of children living in households below 50% of the median income--which can happen if taxes are tilted toward families with kids and spending is focused on retirees.
The politics of designing government policies that affect children is complex, because only adults can vote. In many countries, including the United States, children are a decreasing share of the population. For example, U.S. Census data from 2010 showed that those under 18 years of age were 24% of the U.S. population--an all-time low. The U.S. Census Bureau (Table AVG1) estimates that America had 118 million households in 2011, which it divides into 78 million "family" households with an average of 3.25 people each and 40 million "nonfamily" households with 1.25 people each. Of the subset of "family" households, only about 46% have children--also an all-time low. No politician attuned to re-election ever says anything negative about "the children," but as the proportion of children in the population and the share of households with children drops, it becomes politically harder to focus spending and tax policy on the concerns of families with children.
Of course, it is always delicate to design government policy in support of children, because a sensible policymaker needs to be concerned about the incentives created when resources typically flow through their parents. But along with thinking about how U.S. tax and spending policy might support families with children in direct ways, it's also useful to think about how schools, libraries, community organizations, and public areas like parks and sidewalks can be supportive to the children.
I ran across the UNICEF report at Miles Corak's blog here. In a post called "The sad, sad story of the UNICEF Child Poverty Report and its critics," Corak offers some pointed commentary on how such reports have been received in the past.