My new article just got released in Science Advances today. The article is open access. I try to both describe the nature of America's "systemically high poverty" and review explanations of it.science.org/doi/full/10.11…
Here’s the abstract.
By "systemically high" I mean U.S. poverty is (i) a huge share of the pop., (ii) a perennial outlier among rich democ's; (iii) staggeringly high for certain groups, (iv) unexpectedly high even among those who “play by the rules,” & (v) pervasive across various groups & places.
I critique 3 prevailing approaches focused on the poor not poverty. First, behavioralists aim to "fix" the poor, but cannot explain macro-level variation, the causality bw behaviors & poverty is questionable, & wrongly focus on prevalences but neglect the more salient penalties.
Second, "dramatizing the poor" aims to elicit emotion and compassion through humanizing narratives. But, this approach overemphasizes unrepresentative groups of the poor, disproportionately focuses on symptoms rather than underlying causes, & downplays effective social policies.
Third, "culturalists" "fix" AND "dramatize" - claiming pathological culture -> counterproductive behavior -> poverty. This is "hopelessly endogenous", & suffers from selection biases & subjective biases, & lacks contrast against rival explanations & comparison groups.
Rather than these prevailing approaches focused on the poor, I advance political explanations aiming to explain America's systemically high poverty. According to political explanations, power, policies, and institutions are the pivotal cause of poverty.
Political explanations emphasize: (1) the essential role of social policy, (2) political choices to penalize risks, (3) power resources of collective political actors, and (4) institutions.
Over the past couple of years, I've been posting drafts of some of the figures that ended up in this piece. Here are a few of the ones I'd highlight now:
• • •
Missing some Tweet in this thread? You can try to
force a refresh
The CDC's Social Vulnerability Index uses "minority status" as an *indicator* of "vulnerability." Even tho racial composition surely predicts vulnerability (i.e. racism), isn't there something problematic to equating vulnerability with having racial minorities in a Census tract?
To be clear, I've only learned of these SVI measures from my interdisciplinary colleagues. These are apparently some sort of industry standard in e.g. natural disaster research. But, I was coming from sociological lit on n'hood disadvantage, which omits these (justifiably IMO).
The CDC then treats being a Black like being an unemployed or poor person. Disadv. n'hoods have more poor & unemployed. TBC, of course, Black people disproportionately reside in disadv. n'hoods. But, seems wrong to treat their presence as an indicator of the disadvantage, right?
This is an impressive paper that makes some valuable contributions. I’m curious about this decision to allocate public healthcare expenditures as a lump sum to everyone (equally). This treats the extremely expensive US healthcare system as one gigantic universal social policy.
US healthcare is very expensive partly bc it is most privatized (bradydave.files.wordpress.com/2016/03/bradye…) & bc US is very unhealthy. This dramatically inflates health spending as % of GDP. This paper takes public spending on healthcare (inflated by above) & distributes it *EQUALLY* to everyone?
P.10 WP: “public spending proportionally to posttax disposable income (17% of national income), with the exception of public health expenditure (8%), which we distribute in a lump-sum way, considering that the insurance value provided by health systems is similar for everyone.”
Lots of people say there is an intergenerational "cycle" of poverty - as poverty is inherited. BUT, there is surprisingly little actual empirical evidence of this with high quality income & poverty measures and long panel data. This thread reports some estimates w/ @umpsid
All these estimates use Cross-National Equivalent File, which has high quality post-fisc income data. From that, I calculate relative poverty as less than 50% of median post-fisc equivalized household income. These choices are essential to doing this properly but rarely done.
Proportion of years in poverty during childhood (min 3 obsvs) has coefficient of 0.337 for respondent being currently poor as 25+ year old adult in 2015 (n=6237; R2 0.08).
Our article on deep & extreme poverty is now available @Demography (w/ @ZParolin). We provide estimates of the levels & trends in both deep and extreme poverty 1993-2016. IMO, we substantially advance beyond past research on this topic. link.springer.com/article/10.100…
Our most important finding is that there are disturbingly high levels of deep & extreme poverty in the US. In 2016, we estimate that 5.2 to 7.2 million Americans (1.6% to 2.2%) were deeply poor and 2.6 to 3.7 million (0.8% to 1.2%) were extremely poor.
Throughout, we demonstrate the crucial role of careful measurement. It is essential to follow leading international standards in income & poverty measurement, to correct data for survey underreporting, and especially to include ALL transfers & taxes.