The official poverty measure has a lot of problems, and we'll have to wait until November for the better alternative measure from the census.
For one thing, the official measure totally ignores the big impact of government programs on the poverty rate. But yesterday's income, poverty and health insurance report also included a few adjusted measures that can be informative. In particular, the report estimated what the poverty rate would be if you adjusted peoples' income to not count Social Security benefits and unemployment insurance (which are included in the official poverty measure), and to take into account the Earned Income Tax Credit and food stamps (which are not included in the official measure). Here's how these programs have affected the poverty rate, from 1981 to 2011 (the gray dashed line is the official measure; SNAP is food stamps):
What happens if you exclude retired people? Presumably that would blunt the effects of Social Security on poverty, given that, in general, only people on disability draw benefits before age 62. It does, but the effects of the government programs are still large. The programs reduced poverty for children under 18 by 8.8 percent (or 6.5 million children) and for people 18-64 by 6.1 percent (or 11.8 million people). The biggest program for working-age people is still Social Security, whereas the Earned Income Tax Credit is the most effective tool against child poverty, which it reduces by 4.3 percent. But the really startling numbers are for people 65 and older. Government programs reduce poverty among seniors by 36 percent, and 34.9 percent of that decrease is due to Social Security. Were it not for Social Security, 43.6 percent of seniors would be poor. That's 14.5 million seniors that one program is keeping afloat.
Maybe there are better ways to reduce poverty than through government programs like these. But there's no doubting that they're keeping millions of people above the poverty line.