The anniversary of LBJ’s declaration of a “War on Poverty” has prompted much reflection. Florida Sen. Marco Rubio has just given a speech at the American Enterprise Institute, which draws on some of the idea Oren Cass introduced in his essential National Review article on reforming anti-poverty efforts, “The Height of the Net.” I’ll have more to say about the speech, but for now I want to address a larger conceptual question. Last week, Mike Konczal of the Roosevelt Institute observed that while there is disagreement regarding whether a modest increase in statutory wage floors will generate employment losses, there appears to be a broad consensus that it will reduce the number of people living below the poverty line. Konczal’s argument is that even if a higher minimum wage results in (modest) employment losses, this decrease in poverty is reason enough to favor a higher minimum wage, as the (anti-poverty) benefits outweigh the cost (in job losses, or reduced work hours). Rather than rehash the argument over whether the minimum wage is preferable to low-wage employment subsidies (Greg Mankiw offers the latest salvo in favor of the latter position), Konczal’s column brings to mind larger questions about poverty: what is it, how do we define it, and why does it matter?
For those who believe that the dignity of labor is an important value in itself, the fact that a higher minimum wage might mean job losses is of no small significance. It is also true, however, that a higher minimum wage can be seen as a means of enhancing the bargaining power of low-wage workers relative to their employers, and that the same is true of offering other alternatives to low-wage work via something like an unconditional basic income, like non-market work (household production, etc.) or idleness. [A friend notes that it could be the exact opposite: "There are probably a lot of people who would be willing to do certain jobs for $10 or $15/hr that would not be willing to do so for $7.25. It'd also be harder to find a new job. So the bargaining power shifts, in fact, to the employer -- the worker can be replaced with someone from the long line of eager applicants. The boss doesn’t have to be very nice to you (allowing for swapped shifts, time off, paid lunches, etc) if there’s 20 people dying to take your job. Note that this happens irrespective of whether the number of jobs is affected by a higher minimum wage. "] But the argument from poverty reduction is a powerful one, despite the fact that the concept of poverty is poorly understood. Indeed, Dube doesn’t just rely on the federal poverty level as his yardstick (it is also true, however, that the standard errors in his paper appear to be quite large to my untrained eye)*:
The findings for the gap and squared gap measures show that minimum wage increases do not reduce poverty by merely pushing some families above the poverty line, but rather by increasing incomes substantially and further below the poverty line. This finding is consistent with sizable reductions in the proportion below 50 and 75 percent of the poverty line.
One of the most important challenges facing policymakers is the fact that families that climb out of poverty often fall back into it as various personal disasters. Raising incomes, and specifically raising market incomes, seems like a worthy goal, but it’s also important to emphasize “stickiness.” That is, we don’t want people experiencing big income drops after achieving income gains; we want them to build resources, economic and social, that can keep them from falling down the ladder. This is where concerns about worklessness come in.
To make a brief tangent, Roger Pielke Jr. of the University of Colorado-Boulder, a scholar best known for his work on environmental policy, describes the origins of the federal poverty rate:
The official poverty rate was initially developed in the early 1960s, based on a long history of research and debates over the measurement of poverty (this paper by Gordon M. Fisher provides a detailed look at that history). The more proximate history of the early 1960s involved several analyses by Mollie Orshansky see this in PDF) of the Social Security Administration which arrived at a quantification of a poverty threshold. Lyndon Johnson declared a “war on poverty” soon after Mollie Orshansky published her first paper on poverty thresholds.
Orshansky came up with a threshold based on how much it cost to feed a family (this history is detailed in this paper in PDF by Fischer). Quantification of how much it cost to feed a family was based on a survey done by the US Department of Agriculture, initially in 1933 and extended in 1961 based on data from 1955 to include an “economy food plan.” Orshansky assumed that a family would spend one third of its income on food. The poverty threshold was then determined to be that point at which family income was low enough such that the amount spent on food equaled that determined to be necessary under the “economy food plan” calculated based on food prices of 1964.
And as Pielke goes on to argue, though the poverty rate is in some sense arbitrary, it has important implications for how Americans think about improving the well-being of people at the bottom of the income distribution. He offers one minor suggestion for improving how we think about poverty — shifting to a dollars per day framework:
One quick way to improve such debates would be to do as the UN does and focus on poverty in terms of an income measured in dollars per day. So instead of talking about a percentage below a “poverty threshold” whose provenance date to the 1950s, we would discuss the number of people below a daily income threshold — which in this case is $32.11 per day (recognizing that the official numbers do not fully represent what poor people actually live on, as mentioned above).
This doesn’t help us answer the more difficult normative questions, but it gives us a more accessible number to work with.
I am increasingly convinced that the social indicators that merit more attention relate to neighborhood effects, like the concentration of poverty and the concentration of worklessness, as an individual- or even a household-centered framework misses the various ways in which people are shaped by their environments. Patrick Sharkey’s Stuck in Place offers an alarming look at life high-poverty neighborhoods:
In both periods, high-poverty neighborhoods had extremely high rates of joblessness and idleness. For instance, in 1970 high-poverty neighborhoods had 8 percent unemployment, 19 percent welfare receipt, and a high school dropout rate of 28 percent among youth aged 16– 19. By 2000 these figures had changed somewhat due to the growing educational attainment of the population as a whole and the worsening economic climate in such neighborhoods— in 2000, high-poverty neighborhoods had unemployment rates of 17 percent and welfare receipt rates of 23 percent, while 18 percent of youth aged 16– 21 were high school dropouts. Compared to the rest of the nation, these neighborhoods featured a remarkable concentration of jobless adults— in 2000, the unemployment rate in high-poverty neighborhoods was more than three times as high as in the rest of the nation.
People who work full-time, full-year are very unlikely to be poor, and they are relatively unlikely to live in high-poverty neighborhoods. So high-poverty neighborhoods are environments in which people are often isolated from the world of work. And since social networks are the main way information about employment opportunities is transmitted, poor people living in poor neighborhoods find themselves caught in a trap. When you look at the income of an individual or a household, we miss valuable information about the social context in which they are embedded. It is the social context that tells us whether they will be subject to churn, falling above and below the poverty line, or if they will be in a position to climb out of poverty in a lasting, durable way.
The anti-poverty agenda of the future will involve things like streamlining wage subsidies and work supports. But it will also have to involve reorienting the networks of mutual support that already exist in poor communities towards mutual uplift and self-help, and rethinking crime control and housing policies that exacerbate the concentration of poverty and worklessness.
*I’ve revised this post to reflect Dube’s work on alternative poverty measures.