If you actually follow the link to the data (ftp://ftp.bls.gov/pub/special.requests/ce/standard/2009/income.txt) you can see clearly that the $0-5k group spends more than twice as much as all of the other income groups on education, suggesting that there is a much larger proportion of students in this group. Clearly, students shouldn't automatically be considered "poor" just because they do not have income.
The analysis here assumes (explicitly) that everyone would rather not work than work. While that is an easy assumption to make, I tend to believe that all things being equal, most people prefer to be productive than unproductive. JK Galbraith took that thesis to its logical conclusion, suggesting paying people who don't work (some very minimal stipend) in order to prevent the social cost of poverty. He assumed that the percentage of people that chose to freeload would be negligibly small in the grand scheme of things. I would like to believe this is true, though executing on such a social experiment would be a little tricky.
This article compares non-workers to people who work and have very low incomes. It describes a situation where non-workers spend more money than low-income workers.
I'm confused here: where are the non-workers getting this money that they spend?
The article's point is that for many poor people getting a shitty job that only pays enough for rent and food is pointless when you can get rent or food from your parents, or from welfare, or by thrift alone.
You can't spend via thrift. Welfare, yes. Parents' gifts would be income if properly accounted for. It seems unlikely that these people have savings sufficient to support such expenditures.
My point is that these people have income equivalents that are being elided in this discussion, to its detriment.
