There's a problem in how poverty is defined, Knoll reports. "The Census Bureau’s Official Poverty Measure is based on metrics that are little changed from the 1960s, and it has been criticized for using outdated poverty thresholds and neglecting government policies that include non-cash transfers. In response, the U.S. Census Bureau in 2011 began formally experimenting with a new poverty paradigm, the Supplemental Poverty Measure."
A study by the National Bureau of Economic Research used this measure to examine anti-poverty policies since 1967, and found that government programs were more effective in reducing poverty than previously thought. Without government programs, poverty would have risen from 25 percent in 1967 to 31 percent in 2012; the rates were only 19 percent in 1967 and 16 percent in 2012, Stephanie Knoll writes for Journalist's Resource, a service of the Shorenstein Center at Harvard University.
Government programs reduced child poverty "by 12 percentage points and deep child poverty by 11 percentage points, versus 3 and 5 percentage points respectively in 1967," the study's authors write. "Without government programs, deep child poverty rates would be as high as 20 percent during economic downturns, as opposed to the 4 percent to 6 percent rates we observe with government programs.” The study also found that tax credits and food and nutrition programs have been especially helpful in reducing poverty. (Read more) (National Bureau of Economic Research graphic)