Don't overthink poverty in the United States
Our welfare state programs work well, but they are relatively stingy
I found Matthew Desmond’s piece on poverty for the New York Times Magazine to be vexing, over-complicated, and in key respects, outright wrong.
Most things in life require a degree of nuance, but the truth about comparative poverty in the U.S. is genuinely very simple: the absolute poverty rate in the United States is unusually high for a rich country because our welfare state is stingy.
If we had a generous universal (or universal-ish) cash assistance program for parents and children, the poverty rate would be much lower. If our existing programs for elderly, disabled, and unemployed people were more generous, the poverty rate would be much lower. Describing a solution to the problem of poverty is actually very simple — what’s hard is that generous, universal programs cost a lot of money, which is politically challenging.
Now for a little nuance:
Poverty is often measured in relative rather than absolute terms, and relative poverty in the United States is incredibly high because we have both high absolute poverty and high median income.
Conservatives often want to say that it “doesn’t count” to solve poverty through the welfare state, and that a real solution to poverty would involve everyone earning high market wages. It’s true that insisting on a market solution generates a very analytically complicated problem — no country has successfully addressed poverty this way, and I doubt anyone ever will.
Fundamentally, though, I sincerely do not think this is a very nuanced issue. Taxes are much higher across the board in Europe than in the United States, allowing those countries to spend more money on programs that put a floor on people’s incomes. The typical American, who is not poor, would probably be very annoyed by a proposal to shift to that system. The United States does, though, put a quasi-floor on incomes for elderly people via Social Security. This is financed through payroll taxes that are both pretty high and somewhat regressive. Everyone could enjoy lower taxes if we eliminated FICA and Social Security, but the poverty rate for senior citizens would go up a lot. The typical American would hate that proposal, too. If veteran readers of Slow Boring know anything about American politics, I hope it’s that status quo bias is a big deal, and proposing large-scale controversial changes is almost always bad politics.
But I really do think the basic analytical issues are much simpler than Desmond makes them out to be:
The welfare state is highly effective at reducing poverty, and countries with lower poverty have more generous welfare states.
The United States has a relatively stingy welfare state by OECD standards because we have relatively low taxes by OECD standards.
There honestly just isn’t that much more to it. Telling specific stories about poor people and their lives is important work and can add a lot to our understanding of the human experience. But while individual human experiences are very complicated, the big picture on poverty really isn’t. Lane Kenworthy wrote a great book over a decade ago making these simple points forcefully. I think most people found his book a little dull and will find Desmond’s more interesting, which is unfortunate because I don’t think it’s helpful for Desmond to over-complicate the situation. And in the course of framing his inquiries into this, he’s elevating a piece of dangerous misinformation about the existing American welfare state.
American anti-poverty programs do a lot to reduce poverty
What shifts my read of Desmond’s work from “somewhat annoyed” to “actually angry” is his claim that the United States has made no progress against poverty in 50 years.
That’s just not true. It is true that there are flaws in the way the official poverty measurement is calculated, and I assume this is why Desmond mentions an additional metric called the Supplemental Poverty Measure. But then he implies, without quite saying so, that SPM poverty hasn’t fallen either, which also isn’t true:
In the past 50 years, scientists have mapped the entire human genome and eradicated smallpox. Here in the United States, infant-mortality rates and deaths from heart disease have fallen by roughly 70 percent, and the average American has gained almost a decade of life. Climate change was recognized as an existential threat. The internet was invented.
On the problem of poverty, though, there has been no real improvement — just a long stasis. As estimated by the federal government’s poverty line, 12.6 percent of the U.S. population was poor in 1970; two decades later, it was 13.5 percent; in 2010, it was 15.1 percent; and in 2019, it was 10.5 percent. To graph the share of Americans living in poverty over the past half-century amounts to drawing a line that resembles gently rolling hills. The line curves slightly up, then slightly down, then back up again over the years, staying steady through Democratic and Republican administrations, rising in recessions and falling in boom years.
What accounts for this lack of progress? It cannot be chalked up to how the poor are counted: Different measures spit out the same embarrassing result. When the government began reporting the Supplemental Poverty Measure in 2011, designed to overcome many of the flaws of the Official Poverty Measure, including not accounting for regional differences in costs of living and government benefits, the United States officially gained three million more poor people. Possible reductions in poverty from counting aid like food stamps and tax benefits were more than offset by recognizing how low-income people were burdened by rising housing and health care costs.
I’m concerned that this passage won’t come in for the vigorous criticism it deserves. Desmond is a big deal in scholarship and the New York Times Magazine is a big deal in journalism, and most people aren’t going to want to call them out for how irresponsible this is. It’s also not the kind of academic/journalistic error that conservative or “heterodox” people like to get mad about — Tucker Carlson isn’t going to do a segment about how irresponsible it is.
But I am.
Over the past 50 years, the United States has spent a considerable amount of money on new programs designed to lift the living standards of low-income people. The reason this hasn’t reduced OPM poverty is that OPM poverty excludes those benefits by definition. The SPM includes them, and the SPM does in fact show poverty falling over time. Desmond elides this by saying that SPM poverty indicates a higher rate than OPM poverty, but that’s neither here nor there. A time series of the SPM shows a Great Society drop, then a Clinton drop, and then an Obama drop. When the welfare state expands, SPM poverty goes down.
Bruce Meyer and James Sullivan argue that even SPM poverty is mismeasured and we should look at consumption levels of the poorest instead. Their metric shows an even bigger drop.
I don’t have a dog in the SPM vs. Meyer & Sullivan fight. But if you want to write a complicated, nuanced inquiry into the details of poor people’s lives, I think delving into the differences between those measures is an area where qualitative work would be very useful. You could talk about survey data vs. administrative data, hassle costs, and administrative burdens. People might find it a little tedious though, because broadly speaking, SPM and consumption-based measurements of poverty tell a similar story: poverty has fallen over time, thanks in large part to expansions of the welfare state. The idea that the welfare state has totally failed is a lie I’m used to hearing from Paul Ryan, not something I expect to hear from someone like Desmond.
Conservatives have weird ideas about poverty
On the right, ideas about poverty tend to be very confused.
Instead of seeing OPM poverty’s failure to account for the impact of the welfare state as a methodological flaw, they see it as correcting a moral error — it shouldn’t count to have your material living standards raised by redistribution rather than by the sweat of your own brow. And in a very abstract sense, one can see where they are coming from. But in a different, more true sense, SNAP and Medicaid recipients are hardly unique in having their living standards bolstered by public policy choices. Car dealership owners and beer distributors whose franchises are protected by regulation have genuine incomes. Donald Trump imposed protective tariffs on Korean home appliances to bolster the incomes of American appliance makers. One of the stated goals of conservative immigration policy is to raise wages by choking off labor supply. One can debate the wisdom of these policies, just as one can debate the wisdom of SNAP and Medicaid, but conservatives don’t employ complicated statistical moves to negate those methods of boosting income. And now that most elderly people vote Republican, the GOP has generally become very protective of current retirees’ Social Security and Medicare benefits. Programs’ upsides count unless Republicans decide they don’t like the program.
But essentially, their basic idea is that a proper anti-poverty policy should create a situation in which nobody has low labor market earnings.
This just doesn’t hold up to much scrutiny, though. Should 90-year-olds have high labor market earnings? The severely disabled? Wounded combat veterans? Surely that doesn’t count. What if you’re a stay-at-home mom with a two-year-old and you’re eight months pregnant, and your husband who was earning $55,000 a year in a steady job dies unexpectedly — time to get out there and start earning? I think most conservatives would agree that there is a range of sympathetic cases where they have no problem with the idea of a handout. What we’re really talking about, at the end of the day, is not whether a specific program “works,” but conservatives’ lack of sympathy for large classes of people.
This all leads to a slightly weird dialogue around the “success sequence” — the idea that married high school graduates who work full-time are statistically unlikely to be poor. And as life advice, I totally agree with the success sequence. Anyone out there reading this and wondering if it’s a good idea to drop out of high school, my answer is — no. Some people, like my dad, drop out of high school and end up doing fine in life. But finishing school is the better choice. So is getting married before you have kids. Working full-time is a good way to earn money. That’s all true and I don’t have a problem saying it or if people want to run a propaganda campaign with Facebook ads or whatever. But we’re still left with the question of what should we do for people who are poor, especially children whose parents may have made poor decisions or had bad luck in life.
An expansive welfare state is expensive
I think the best answer is that just as all kids in America are entitled to K-12 schooling whether or not we think their parents have made great life choices, all kids in America should be entitled to some basic material living standards. There should be a monthly child allowance for all kids, probably one that starts quite large, then tapers when the kid hits kindergarten and again when the kid turns 14 or 15. You should get this allowance whether you’re rich or poor or anything in between; we can make the system progressive through taxes.
A program like that would be relatively simple to get people signed up for because eligibility is easy to determine.
It would avoid any “poverty trap” or “welfare dependency” perverse incentives, and I would hope it would minimize the toxic pitting of people against each other based on their circumstances in life. Kids have needs — policy should account for that. And just as Social Security massively reduces elder poverty without being “an anti-poverty program,” a Social Security benefit for kids would massively reduce poverty among children.
The problem is that while it’s analytically simple and easy to describe, it comes with a hefty price tag. I don’t think it would be costly in economic terms to finance a program like this with broad-based taxes. Indeed, in a lot of ways I think it would be superior to the current practice of bolting together a jumble of different programs for families. But the numbers involved are big and scary, GOP elites think it’s a bad idea, and Democratic Party advocacy groups who like this idea generally have higher priorities.
So policy entrepreneurs are left trying to improve life for low-income people with a mix of means-tested programs that are either in-kind (SNAP) or work-linked (EITC) and really sweating the details to try to maximize the benefits of scarce program dollars. That work is very complicated and technical because these programs all have their own legislative histories and design parameters, and some of them interact with each other.
But to say that the state of American anti-poverty programs is complicated is very different from saying that the persistence of poverty itself is complicated. Poverty persists because straightforward, highly effective solutions are politically untenable in the short term, leaving people in the trenches to deal with a very complex situation. Part of the role of those of us in the article-writing community should be to clarify this, not to layer new levels of complexity onto it.
I have been a long time free subscriber, however after this article I decided to become paid. I am a right of center Southerner that thinks your obsession with license plates is weird, but really enjoy your balanced and thoughtful take on most subjects even when I don’t always agree with you. And just picking with you about the license plate. It is good to have a quirky thing we become known for.
Here's where I lose respect for Desmond:
"...12.6 percent of the U.S. population was poor in 1970; two decades later, it was 13.5 percent; in 2010, it was 15.1 percent; and in 2019, it was 10.5 percent..."
Saying that this run of numbers shows no change -- handwaving away the difference between 15% and 10.5% -- is malpractice. That is a huge difference. If we were talking about a medical intervention in eg cancer treatment, this would be a cause for celebration.
Any honest investigator would respond to this by saying, "holy shit! For every three people with Condition X in 2010 there were only two people with Condition X in 2019? Someone has stumbled on a powerful causal lever! Let's find that thing!"
So...what was the new thing in 2019 that produced this striking change in the numbers? Because we should probably keep doing that thing.