In a new National Bureau of Economic Research (NBER) working paper, Richard Burkhauser and his coauthors argue that only 2.3 percent of Americans lived in poverty in 2017, if we define poverty in the way they claim “President Johnson defined it.” By comparison, the OECD put the US poverty rate at 17.8 percent using the standard international measure of poverty, which sets the poverty line at half of median disposable income (about $34,000 a year for a family of four in 2017).
To reach this conclusion, Burkhauser et al. make a long list of changes to the official poverty measure. Some of these adjustments are sensible. For example, they follow the international practice and use disposable income (including taxes and near-cash, in-kind benefits) to measure poverty. At least two other changes they make are controversial and problematic. First, they make inflation adjustments that have the effect of pushing down the value of the poverty line, compared to other relevant poverty lines, over the last half-century (to about $18,800 in 2017 for a family of four). Second, they count the dollar value of public and employer-sponsored health insurance as income.
What does it mean to say only 2.3 percent of Americans live in poverty today, as LBJ supposedly defined it? Imagine two adults raising two children in 2017 on a disposable income of $19,000 without receiving any public or employer-provided health insurance. According to most Americans and any modern poverty measure, this family would not only be poor, they would fall substantially below the poverty line. Yet, according to Burkhauser et al., they are not living in poverty, as Johnson defined it.
Suppose the same family had even less annual income, say $15,000, and received Medicaid; they would still not be counted as “LBJ poor” because the value of Medicaid would be counted as income and put them over the line. Average Medicaid spending per enrollee in 2017 was $8,015; it’s lower for children and non-elderly/non-disabled adults, but it’s still enough to push this family over Burkhauser et al.’s poverty line.
In both cases, it seems like a real stretch to classify these extremely low-income families as not poor, even in LBJ’s terms. In the 1964 Economic Report of the President, Johnson’s Council of Economic Advisors defined the poor as people who are not “maintaining a decent standard of living—those whose basic needs exceed their means to satisfy them.” It is not credible to think that a family of four living on $19,000 today has a “decent standard of living” or the means to satisfy all of their basic needs without going deep into debt or worse.
And, if the same family’s income is significantly lower than $19,000 a year, it’s even less credible to think that their receiving Medicaid is enough to shift their standard of living from non-decent (poor) to decent (not poor). Don’t get me wrong, Medicaid is valuable and important, but it won’t pay the rent or any of the other non-health care basics that are part of a minimally decent living standard for a family of four. As an aside, there is sophisticated research being done to develop a “health-inclusive poverty measure,” but Burkhauser et al. have little to say about it.
How many people LBJ would think are poor today may be an interesting academic question. Still, it has little relevance to policymaking or policy evaluation today, and it’s not something we’re likely to reach consensus on anytime soon. Burkhauser et al. say we need to figure it out in order to evaluate “progress in President Johnson’s War on Poverty,” but President Johnson has been out of office for half a century, and we’ve had nine different Presidents since then.
While certain programs that Johnson proposed (like Medicaid) have been expanded over time, other major policies that he saw as central to fighting poverty have deteriorated. Most notably, the value of the minimum wage has declined by more than 30 percent since 1968, when it reached a still record-high value of $10.54 or about $22,000 per year in today’s dollars. Similarly, as the Economic Policy Institute has documented, union membership density has declined steadily since the 1950s, from 33.2 percent in 1956 to 10.5 percent today, while the share of income going to the top 10 percent has increased. The top marginal tax rate has been cut from 91 percent in 1963 to 37 percent today. Other policies that are important today, like the Earned Income Tax Credit, have helped offset declines in the minimum wage and other labor standards (at least for parents), but these were never part of LBJ’s vision for fighting poverty.
At the end of the paper, Burkhauser et al. recommend that “policymakers might consider setting new poverty thresholds that reflect modern-day expectations for what it means to be impoverished.” This is correct, regardless of whether one agrees with the rest of their paper.
When it comes to measuring poverty today, the important questions include 1) how much higher to set modern poverty thresholds; 2) how to update thresholds over short and long periods of time; 3) whether and how to take care-related needs and benefits into account (including health care, child care, and children’s development needs beyond housing, food, and clothing); and, 4) whether and how to take assets and debt, including debt repayment, into account. (On this last question, which has received little attention in the United States, see the new poverty-measurement framework developed by the Social Metrics Commission in the United Kingdom.)
Where to set a modern poverty threshold has both normative and technical aspects. Nonetheless, there can be little question that our current official poverty thresholds have defined deprivation down and have little relevance today. In focus groups conducted in 2013, the center-left Center for American Progress found that, “most respondents were shocked to hear that the official poverty line [$23,550 for a family of four at the time] was as low as it is,” and many viewed the measure as “disconnected from reality.”
That same year, CAP conducted a nationally representative survey and found that Americans thought a family of four needed $30,000, on average, to not be poor in 2013. Similarly, in a subsequent pollconducted by the conservative think tank American Enterprise Institute, Americans thought $33,300 was needed, on average, to not be poor in 2016. If the same questions were asked today, the average amount Americans think is necessary to avoid poverty would almost certainly be several thousand dollars more. A modern poverty measure should reflect this reality.
This article first appeared on the CEPR blog.