Over at Slate, both Jamelle Bouie and Jordan Weissmann argue against Representative Paul Ryan’s new anti-poverty plan on the grounds that it is unnecessarily concerned with the small percentage of the population that is chronically poor. Here’s Weissmann:
But here’s the fundamental issue with the plan: There’s not much evidence that the vast majority of the poor need this kind of hand-holding. As I’ve written before, for most Americans who experience it -- and you’d be shocked how many do -- actual poverty is a temporary detour through life. It happens, and you recover. Actual chronic poverty is relatively rare.
In 2011, according to the Survey of Income and Program Participation, the annual U.S. poverty rate was 14 percent. But only 3.5 percent of Americans were chronically poor, meaning they had been impoverished for three straight years. According to research by Mark Rank of Washington University in St. Louis, 38.9 percent of Americans experience at least one year of poverty between the ages of 25 and 60; 11.6 percent experience five years or more. During those same years, 44.8 percent use a means-tested safety net program, such as food stamps, but only 16.4 take advantage of one for at least half a decade.
And here is Bouie, in a piece subtitled “America’s poor need bigger checks, not a 'life plan'”:
In his response to Ryan’s anti-poverty plan, Jared Bernstein -- former chief economist for Vice President Joe Biden -- writes, “The main problem faced by the American poor is not that there’s something fundamentally wrong with the safety net. It’s that they lack the employment and earnings opportunities necessary to work their way out of poverty.”
Bingo. At some point in their lives, millions of Americans will experience a short spell of poverty. Not because they don’t have a plan to fix their lives or lack the skills to move forward, but because our economy isn’t run to create demand for labor, isn’t equipped to deliver stable work to everyone who wants it, and wasn’t built to address the distributive needs of everyone who works.
The best way to confront this problem for most people is to just address those needs. Yes, on the margins, there will be Americans who need an intensive approach, and I endorse government support for voluntary life coaching. (For example, look at the Center for Urban Families in Baltimore.) But by and large, the easiest solution is to mail larger checks to more people. In other words, we need more solutions like Ryan’s expansion of the Earned Income Tax Credit -- the best part of his plan -- and fewer life coaches for the poor.
None of this is wrong, exactly -- as far as I know, the figures they quote are correct. But I nonetheless think the figures are misleading. We do, in fact, spend a lot of our entitlement money on the chronically poor -- much more than you would think from hearing these statistics. That’s why Ryan is right to make that sort of relatively intractable poverty the focus of his plan.
There’s an interesting paradox in statistics and operations research: Something can be a relatively small portion of your overall flow yet a very large portion of your stock at any given moment. To see what I mean, think about street parking. If you counted the number of cars that park on a residential urban street during the course of a month, you’d probably find that more of them came from outside the neighborhood than inside it. Yet if you counted the cars parked on a residential urban street at any particular moment, you would probably find that the majority of them belong to people who live on that street.
How can this be? Because the residents' vehicles spend a lot more time parked on the street than those of deliverymen, construction workers, repairmen, friends or relatives. Each resident spends 12 to 14 hours a day parked on their street; each visitor probably averages an hour or two. So even if you have five times as many visitors parking as residents, the residents rack up a lot more hours in aggregate -- which means that at any given hour you examine, more cars will belong to residents than to visitors.
The longer the timeframe you examine, the bigger this distortion gets. My street has 60 houses, averaging one to two cars a house. That means that over the course of a day, a month or a year, the number of resident cars stays fixed at around 75 to 90, while the number of nonresident vehicles that park here temporarily keeps growing every time someone joins a new book club or decides to order from a different Thai delivery place.
In the early 1980s, two researchers named Mary Jo Bane and David Ellwood pointed out something very similar about welfare recipients. Social scientists in the 1970s had noted -- correctly -- that the majority of people who applied for welfare were likely to be on it for a relatively short time. (When conservatives complained about chronic welfare dependency, this was a common rejoinder.) But Bane and Ellwood pointed out that on any given day, the majority of people on welfare were long-term recipients who would be on welfare for at least eight years -- and that this group accounted for most of the Aid to Families with Dependent Children's budget.
This is not just true for the U.S.; you can see similar effects in this striking graph from an 18-year analysis of Swedish poverty rates:
I am unable to find anyone who has done a comprehensive update of Bane and Ellwood’s work for the U.S. in the new millennium. Certainly, a lot has changed since then, though I am not sure which way that ultimately cuts. On the one hand, the 1996 welfare reform has made it harder to stay on welfare forever; on the other hand, fewer of the chronic poor are found on Temporary Assistance for Needy Families (the successor to AFDC), and more are found on Social Security disability, which has even worse exit rates than the AFDC did.
Nonetheless, it is mathematically inevitable that the chronically poor are consuming a much greater share of poverty benefits than simple fractions would suggest. Consider a theoretical pool of seven needy families, two of whom are poor for the entire year and the rest poor for nine, five, three, two and two months, respectively. In all but one month, our two chronically poor families will claim at least 50 percent of the benefits, and over the course of the year, those two households will have received more than half the total payouts.
And this assumes that everyone is getting the same amount of money. But, of course, that will not be the case. By the definition of monthly poverty used by the Survey of Income and Program Participation, I have been poor for months at a stretch. By any reasonable definition, however, I was not poor; I simply had no income coming in for a bit. I certainly never resorted to poverty benefits; at worst, I would have resorted to parents with a well-located spare bedroom.
Before the comment trolls start in on that last paragraph, let me note the obvious: I am not suggesting that poor people should just move into the spare bedroom at Mom and Dad’s. My point is that I probably didn’t belong in any measure of poverty that you would want to use to guide government policy.
But my point is also that chronically poor people are more likely to require extra government benefits because they don’t have any of the assets that the temporarily poor bring with them from the middle class: reliable cars, houses, savings accounts, credit cards, friends and family who have spare cash to help out. The chronically poor will need more help, for longer, than folks who are struggling through a temporary job loss or divorce. Which means that, at the very least, they take up a disproportionate share of resources. It seems entirely possible -- perhaps even likely -- that the chronically poor still account for the majority of spending in many programs.
So, mathematically, I think the argument being made by Bouie and Weissman fails; it obviously makes a lot of sense to focus on the group that generates a disproportionate share of our entitlement spending. At the very least, we should consider the strong possibility that those struggling with chronic poverty might need very different kinds of help than those dealing with a temporary income problem -- rather than suggesting, as Bouie does, that we should obviously focus on doing whatever is best for people having an acute poverty episode because they’re the majority.
None of this necessarily means that Ryan’s anti-poverty plan is a good idea. You could still argue that it won’t work: that the chronically poor are basically beyond help, that we should just give them money without expecting that they will ever be able to wean themselves off substantial public assistance. Or that the particular sort of help that Ryan proposes won’t work and we should do something else instead.
I certainly hope that at least some of the chronically poor can, with help, haul themselves into the middle class. Empirically, I don’t know whether the Ryan plan will work -- but this is exactly what makes its focus on experimentation and rigorous evaluation so valuable.
Politically, however, I think it would be disastrous for Democrats to once again start arguing that there’s not much we can do about poverty except give everyone lots more money with no strings attached. “Ending welfare as we know it” was unpopular with much of the left -- but “everything is fine about welfare except that the checks need to be bigger” was wildly unpopular with everyone else. It also seems to have been empirically rebutted by the experience of welfare reform; poverty rates dropped sharply and stayed down even after the dot-com boom ended. By 2012, after the worst recession in 70 years, the poverty rate was no higher than it had been after the mild recession of the early 1990s.
Things may have changed, of course; perhaps our economy no longer has a place for marginal workers. But I’m not yet ready to take the counsel of despair. Maybe it will come to that eventually. But first, I’d like to see us try something like Ryan’s plan and see if we can’t open a few doors out of poverty and into the stable middle class.
To contact the writer of this article: Megan McArdle at email@example.com.
To contact the editor responsible for this article: Brooke Sample at firstname.lastname@example.org.