The 2016 U.S. presidential campaign has come to a simmer, and candidates from both major parties are already talking up ways to help the working poor. Exactly how many people are in this socioeconomic group, however, has long been a subject of scholarly debate. Now, a trio of sociologists has taken a fresh look at the question and come up with more than 100 answers—the best of which are lower than official estimates.
“Measurement assumptions have important implications for the estimates of the size as well as the distribution of the problem across different racial and ethnic groups,” says Brian Thiede, a sociologist at Louisiana State University, Baton Rouge, and lead author of the study, published online before print in Work and Occupations. Differing assumptions—such as about who is poor and who is a worker, for example—have produced a wide range of answers, varying by up to 16%, write Thiede, Daniel Lichter of Cornell University, and Scott Sanders of Brigham Young University in Provo, Utah.
In a bid to do better, the researchers first set out to identify the assumptions used in previous studies that might swell or shrink the number of people considered to be working poor. One issue is who should be counted as a worker. For example, they found that studies typically didn’t count people who are past retirement age but still working to supplement their income.
They also examined how researchers defined who was poor. Some studies considered only the earnings of a working head of household, for instance, whereas others tallied the total earnings of all members of a household—which could shrink any estimate.
The definition of poverty has long been debated, the researchers note. The official U.S. definition of poverty, for example, is based on an absolute dollar figure. (In 2015, it is $11,770 for a household of one in the 48 contiguous states, and a bit more if you live elsewhere or count other family members.) But some studies define poverty differently, using methods recommended by the National Academy of Sciences’ Panel on Poverty and Family Assistance. Although these methods add in-kind income and support provided by government health care programs, they also subtract certain tax payments and work expenses. The net effect tends to increase estimates of the working poor.
Other studies use a relative benchmark for poverty: for example, placing the poverty line at about one-half of the median household income of a nation or region. As the median rises or falls, so does the poverty line.
This concept of relative poverty is not new, notes David Brady, a sociologist with the WZB Berlin Social Science Center. In 1776, Adam Smith, the pioneering 18th century Scottish economist, wrote in his The Wealth of Nations that “a linen shirt … is, strictly speaking, not a necessary of life. … But in the present times, through the greater part of Europe, a creditable day-labourer would be ashamed to appear in public without a linen shirt.” Brady says that “most international social science, and arguably the best social scientists of poverty even in the United States, are converging on [using] a relative measure of poverty.”
These two approaches produced very different results, Thiede notes, when the researchers applied them to the census data. When using the absolute poverty threshold, the researchers concluded that about 4.3% of working-age persons worked year-round yet lived in poverty in 2012. In contrast, using a relative poverty threshold of about one-half the median U.S. income, the poverty estimate for year-round workers increased to 11.7%. (The poverty rate among the total population, which includes nonworkers, is much higher.)
Once the researchers had identified the various assumptions, they began to experiment. They used different combinations of assumptions about work and poverty to calculate estimates of the working poor. Overall, they came up with 126 estimates, ranging from 2.5% to 18.5% in 2012.
Some measures that produce these estimates are better than others, however, the authors argue. Using statistical analyses and other methods, they ranked the combinations of factors, looking for those most likely to reflect the real world. The highest ranked combination for individuals, they found, defined a worker as someone who worked more than 17 hours a week; adjusted income to a full-time, full-year equivalent; and set the poverty threshold at 125% of the official threshold. Using those assumptions, the researchers estimated that 12.4% of American workers were poor.
In contrast, the lowest ranked combination of assumptions produced an estimate of 9.2%. It defined a worker as someone who worked more than 35 hours a week for more than 50 weeks per year, did not adjust income, and based the poverty threshold on 50% of median family income.
In a separate step, the researchers also calculated that between 9.3% and 11% of working family heads are poor. That translates to between 6.4 million and 8 million workers and—when their household members are included—between 20 million and 24 million persons living in poor families with a working head.
These numbers are smaller than the 10.6 million working poor estimated by the federal government in 2014. The gap is partly the result of a broader federal definition of who is a worker. Still, the study “seems very solid and robust scholarship to me,” Brady says.
The results highlight the fact that work alone is not enough to avoid poverty, Thiede says. “It’s a fairly significant problem and [the findings are] contrary to the ideological assumption that work is a universal, automatic path out of poverty,” he says.
That’s a topic that’s likely crop up again in coming months, as presidential candidates hit the 2016 campaign trail.