The Poorest Norwegian Children Are Twice As Rich as the Poorest American Children
Previously, I used LIS summary data to show that Americans in the bottom half of the income distribution are worse off than individuals in the bottom half of the income distribution in other developed countries. Here, I use LIS microdata to similarly compare the plight of children across 15 developed countries, as the first part of a multi-part series on the subject.
The income data for the following countries all comes from 2003, 2004, or 2005, with the exception of Belgium, which comes from 2000. There is more recent data for some of the countries, but I have selected these years to avoid the effects of the Great Recession and to ensure maximum comparability.
The first income metric I used for this comparison is per capita PPP-adjusted disposable household income. Disposable household income refers to the amount of ultimate income a household has on hand to spend. In formulaic terms, it can be expressed as: Market Income (labor and capital income) + Transfer Income (public benefits) - Taxes. PPP-adjusted refers to a method of converting different national currency units to a common value unit that is defined by how much those national currency units can buy in local goods. Here, I use 2005 PPP conversion ratios. Per capita means that I have taken each household's total income and divided it by the number of people in the household. For example, a child in a four-person household with a disposable household income of $10,000 is recorded as having a per capita income of $2,500 ($10,000/4).
After assigning a per capita income to each child (persons aged 0-15), I then determine what the per capita income is at various percentile lines. The "5th percentile" refers to the line at which 4% of children have lower per capita incomes and 94% have higher per capita incomes. The "10th percentile" refers to the line at which 9% of childen have lower per capita incomes and 89% have higher per capita incomes. And so on.
Here is a table ranking the countries from highest to lowest income using this metric for various percentiles in the bottom half (click on image to get bigger version):
The percentage beside each country is how much more or less per capita income children have at each percentile relative to US children at the same percentile. So the 97% beside Norway under 5th percentile means that Norwegian children at that percentile have 97% more income than US children at that percentile.
The second income metric I used is equivalized PPP-adjusted disposable household income. Equivalized refers to a process of adjusting income for household size in such a way that accounts for the economies of scale that larger households enjoy. Here, I use the same equivalization method that the LIS uses internally (which is somewhat aggressive as far as these things go). Under this equivalized measure, instead of dividing household income by the number of members in the household, you divide household income by the square root of the number of members in the household. So, you divide by four to get the per capita income of a four-person household while you divide by 2 (square root of 4) to get the equivalized income of a four-person household.
Here is the same table as above using this metric (click on it to get a bigger version):
As you can see, using this metric makes very little difference. The percentages bump around a bit, but not much.
The poor showing of the US on the bottom is not a function of the fact that many of our benefit programs use in-kind rather than cash benefits. The LIS microdata includes these in-kind benefits in their income figures. So Section 8 (housing assistance), LIHEAP (energy assistance), food stamps, subsidized school lunch, and so on are all included. The only major item not included is Medicaid, but LIS doesn't include the health care benefits in the other countries either, which are better than Medicaid.
Also, despite popular sentiments to the contrary, I don't actually think these absolute income measures are the best economic indicators. Among developed countries at the technological frontier, I think relative measures (e.g. gini, relative poverty rates, P90/P10 ratios, etc.) more often do a better job of reflecting the success of various economic institutional regimes. Different countries have different capacities to produce income and make different labor/leisure tradeoffs as well. Because absolute income measures ignore these important differences (among others), they can generate strange conclusions on certain things that relative measures do not.
These considerations are perhaps not so important here because the US trails in both relative and absolute measures, but it deserves mentioning that relative measures would show the US doing worse than is suggested by the data here.
Sign up for our emails to stay updated on what we're doing and how you can help.