Official, Supplemental, and Relative Poverty in 2014

Determining the poverty rate is always a somewhat arbitrary affair. There are dozens of potential combinations of income concepts, family concepts, and poverty thresholds, none of which stands out as the obviously correct one. Then, on top of that, there are known problems with income survey data that you have to decide how to "correct" for, which involves yet another contestable methodological choice.

If your goal is simply to figure out who occupies the bottom of the economic system, you can thankfully avoid much of the poverty definition debate. This is because basically every definition of poverty turns up similar conclusions about who poverty afflicts. That is, they tend to show that poverty afflicts those whose life status makes it difficult or impossible for them to directly collect labor incomes.

Official Poverty Metric

Here is who made up the poor under the Official Poverty Metric in 2014:

In total, children, the elderly, the disabled, and students make up 66% of the officially poor. If you add in carers and the unemployed (mostly temporarily unemployed), the number shoots to 84%.

Supplemental Poverty Metric

Here is who made up the poor under the Supplemental Poverty Metric in 2014:

Under this metric, 61% of the poor are childen, elderly, disabled, or students. Bringing in carers and the unemployed pushes the percentage to 78%.
Relative Poverty (Per Capita)
Under a relative poverty metric (described in technical note below) that uses per capita disposable income, poverty looks like this:
Children, elderly, disabled, and students make up 64% of the poor. Adding in carers and unemployed gets you to 81%.
Relative Poverty (Equivalized)
Under a relative poverty metric that uses equivalized disposable income, poverty looks like this:
Children, elderly, disabled, and students make up 65% of the poor. Bringing in carers and unemployed takes you to 81%.
As you can see, these popular poverty metrics generally tell the same story. Across the board, children, elderly, disabled, and students (CEDS block) make up around 60-65% of the poor. The CEDS block plus carers and those who faced a spell of unemployment (i.e. vulnerable populations) make up around 80-85% of the poor. Because these groups are not generally available to work, if you want to increase their incomes and pull them out of poverty (however defined), that's generally going to require an increase in welfare incomes. Some progress can be made with higher employment and higher pay, but labor incomes only reach the CEDS block and other vulnerable populations through diluted and indirect means, if at all.
You can also try to restructure all of society so that members of vulnerable populations wind up cleverly rearranged around those who currently capture labor income. That is, you could theoretically try to enact a redistribution of income through a redistribution of people among households. This is a popular idea in some conservative camps (at least for certain kinds of household member redistributions), but has proven to be a failure and also involves a kind of odd social engineering that betrays certain liberal values like individual freedom.

Technical Note.

The eight populations in the above graphs are defined at my personal blog (click here).

For relative per capita poverty, I take disposable income as defined by the supplemental poverty metric and then I add back money spent on child care and out-of-pocket medical expenses. From there, I divide each family's disposable income (so defined) by the number of people in the family, and assign the resulting income to each person in the family. After doing that, I find the median disposable income (weighted by person) and divide it by half to get the poverty line. All people whose disposable income (so defined) are below this povety line (so defined) are counted as being in poverty.

For equivalized per capita poverty, I do basically the same thing as relative per capita poverty. The only difference is that, instead of dividing disposable family income (so defined) by the number of people in the family, I divide disposable family income by the square root of the number of people in the family. Everything else is the same.