According to an economic study released this month, poverty-reducing programs such as food stamps and housing vouchers are almost twice as effective as originally estimated.

Using government administrative records for the state of New York, University of Chicago economist Bruce Meyer and CERGE-EI’s Nikolas Mittag were able to test the accuracy of the Current Population Survey’s (CPS) Annual Social and Economic Supplement (ASEC).

“The survey… doesn’t do a very good job of capturing income because people are reluctant to respond to it,” Meyer says. “They’re reluctant to answer the questions about income if they do respond, and if they do give answers they’re often inaccurate. More than 60 percent of people getting the programs for really low-income people, TANF [welfare] and state-level general assistance don’t report it at all.”

Before we go into the effects of under-reporting assistance, a brief discussion on defining poverty is necessary. The census measured the New York’s poverty rate between the years of 2008 and 2011 as 13.6 percent, but strangely, the way the census measures poverty doesn’t capture the effectiveness of government programs to lift people out of poverty because it explicitly leaves the means of assistance out of the calculation.

“It’s foolish consistency,” Meyer says. “It’s because so many benefits didn’t exist in the 60s when the poverty measure was first devised.”

The poverty thresholds for different family sizes are based on a calculation of “the cost a particular food plan from the FDA in the 1960s,” Meyer says. Since then, the thresholds have been adjusted according to the consumer price index. “The main problem with [using income to quantify poverty] is it doesn’t account for taxes,” he says.

“At the time, food stamps [and Medicaid were] just getting started, housing benefits were pretty small,” he continues. “We didn’t have the Earned Income Tax Credit or the Child Tax Credit in the 1960s, and now they’re some of the biggest anti-poverty programs we have, so the world has changed since this poverty measure was devised.”

Meyer and James X. Sullivan from Notre Dame advocate an alternative method of measuring poverty based on consumption rather than income. “We find that using what people are able to spend on food, housing, utilities, transportation, and other things give us a better measure of someone’s material living standards than looking at their income,” says Meyer. “Low consumption is more likely associated with ill health, not having health insurance, low education, not having air conditioning, dishwashers, or other amenities.”

A good example is retired folks who might not have much income, but live comfortably on their savings, and if they already own their cars or homes they’re not paying them off any more. A 2012 paper by Meyer and Sullivan (PDF) maps the progress of this alternative measure of poverty from 1964 (31 percent) to 2010 (5 percent), offering insights into the effectiveness of government programs to fight poverty.

Based on the results of the ASEC survey and under this consumption-based measurement of poverty, the study estimates the poverty rate in New York from 2008 to 2011 as 10.8 percent—but by using the numbers inferred from government administrative records, that number reduces to 8.3 percent, suggesting that government assistance programs are helping even more than was originally estimated.