Mad Libs: Bruenig v. Piper

No subtweeting allowed...

Welcome to Mad Libs. This is an irregular debate column where our columnists, contributors, staff writers, (or even you, dear reader) can duke it out over the big ideas we’re discussing in the metaphorical pages of this magazine.

Today’s Mad Libs was inspired by The Argument Staff Writer Kelsey Piper’s Tuesday column “Giving people money helped less than I thought it would.” Piper reviewed the recent disappointing evidence on basic income and concluded: "No randomized controlled trial is really going to convince me that money does not improve people’s lives. But I do think cash as an intervention is best used in emergencies, for pregnant women, domestic violence victims or in other, narrower, contexts where study is still ongoing."

The Argument columnist Matt Bruenig took great umbrage at these arguments, as you can read below. And Piper responds to his column in kind. Future editions of Mad Libs will be paywalled, so subscribe to get more content like this in your inbox! And if you’re interested in pitching us, go here.

-Jerusalem

Matt Bruenig, The Argument columnist

In 1969, the President’s Commission on Income Maintenance Programs released a report titled “Poverty Amid Plenty” that remains perhaps the best writing on the subject ever produced in America. On page 19, the authors note that:

Although the poor family may not starve, and although it has a roof over its head, its lack of buying power in a world where expenditures are a part of all social relationships subjects the poor family to social, if not physical, starvation.

They go on to quote from a National Association of Social Workers Newsletter published five years earlier:

To go to school costs money — books, notebooks, pencils, gym shoes, and ice cream with the other kids. Without these the child begins to be an outcast.

To go to church costs money — some Sunday clothes, carfare to get there, a little offering. Without these one cannot go.

To belong to the Boy Scouts costs money — uniforms, occasional dues, shared costs of a picnic. Without these, no Scouts.

To have friends into the house costs money — for a bit of food, a drink.

To visit relatives costs money — for travel, a gift for the kids. These people cannot afford to visit their relatives.

For a teenager to join his friends on the corner he must have some money — for a coke, a show.

How does a fellow take a girl out on a date without some money? And how does a girl pretty herself for a fellow without some money?

How do you join a club? Buy a book, a magazine, a newspaper?

Poverty settles like an impenetrable cell over the lives of the very poor, shutting them off from every social contact, killing the spirit, casting them out from the community of human life.

These days, this sort of writing sounds like it is from another planet. The idea that class difference and alienation are themselves menaces to society has given way to research aimed at determining whether increasing incomes at the bottom makes kids less fat at age 4.

It is hard to say when this devolution in our understanding of poverty began, but it seems to have been driven in recent years by liberal policy thinkers obsessed with “human capital.” In the late 2000s, James Heckman led this charge by publishing research showing that public spending on things like high-quality early childhood education so dramatically improves kids’ brains that it pays for itself in the long run. In this new telling, the social benefit system should be understood, not strictly as an egalitarian project aimed at smoothing and compressing incomes and consumption, but rather as an investment in the future.

Heckman himself was so committed to his human capital obsession that, in 2016, he wrote a paper titled “The Scandinavian Fantasy” in which he demonstrated that Denmark’s low levels of inequality and high levels of social mobility are almost entirely attributable to union-driven wage compression and the nation’s generous welfare state, not his preferred approach of human capital development. Rather than reflecting on whether these findings suggest that his entire life project may have been a mistake, Heckman bizarrely concluded that they send “a cautionary note to the many enthusiasts endorsing the Scandinavian welfare state.”

It is in this tradition that Kelsey Piper found herself earlier this week when she wrote here at The Argument that giving people money does not actually help that much and that therefore “ending the war on poverty will take more than cash transfers.” In her piece, Piper doesn’t bother with pesky questions like what the purpose of the welfare state should be. Instead, she unreflectively adopts the Heckman-inflected approach to this whole area of policy, gives summaries of a few recent studies finding that more cash income did not improve things like “mental health, stress levels, physical health, child development outcomes or employment,” and then concludes from this that they don’t work.

There is so much wrong with Piper’s piece that it is difficult to know where to begin.

Perhaps the best place to start is her apparent conclusion, which is that “cash as an intervention is best used in emergencies, for pregnant women, domestic violence victims or in other, narrower, contexts where study is still ongoing” but that “it is not going to deliver significant changes for chronic poverty.”

In reality, cash is the key part of every welfare state in the developed world and absolutely critical for keeping poverty down. Well-designed welfare states systematically provide cash to elderly people through an old-age pension, disabled people through disability benefits, unemployed people through unemployment benefits, children through child allowances, and caregivers through paid leave and home care allowances.

What all these beneficiary populations have in common is that they are not working and therefore must get their income from someplace other than the labor market. Provisioning out income to the half of the population that exists outside the labor market at any given time¹ is the central task of the welfare state and the central challenge of keeping poverty low.

It is amusing to imagine someone conducting this sort of research on typical welfare recipients. After checking in on the sample for the fourth year in a row, the foundation-funded researcher studiously writes that, despite receiving tens of thousands of cash dollars over the period, the retiree still has no job, the paralyzed woman still can’t walk, and the man with dementia has actually seen his cognitive abilities decline rather than improve.

Historically, simple cash programs aimed at nonworkers have worked very well. When the U.S. began expanding the Social Security old-age pension in the middle of last century, the elderly poverty rate declined from 35% to 10%, around where it remains today. When the U.S. briefly expanded its primary cash benefit for children, the Child Tax Credit, in 2021, it cut child poverty in half. A year later, when that expansion expired, child poverty predictably doubled.

Piper’s conclusions are wrong because her method of arriving at them is wrong. Her piece and subsequent comments under the piece make it clear that Piper first became aware of cash benefits in the context of development programs in low-income countries. Those programs are designed as alternatives to other development strategies and thus focus on various developmental measures. This has nothing at all to do with cash benefit programs in high-income welfare states, which exist for completely different purposes.

Had Piper actually looked into the long literature on Western welfare states, rather than clumsily apply certain effective altruist ideas about the developing world to America, she probably would have approached her argument much differently.

As recounted in Gøsta Esping-Andersen’s The Three Worlds of Welfare Capitalism, the social-democratic welfare state, which exists in the low-poverty Nordic nations, focuses on providing universal benefits — including cash — to its population in order to maximize equality and level living standards, provide workers refuge from the labor market, and promote individual autonomy by making people less reliant on family transfers to get by. Cash benefits do all of those things, even if they don’t lead to “human capital improvements.”

I would accuse Piper of diminishing or dismissing these other goals, but that would be unfair. It’s more likely that she simply has never heard of them.

Another baffling thing about Piper’s presentation is trying to figure out what exactly she means by “ending the war on poverty” or “winning the war on poverty.” It sounds as if she has an understanding of poverty as afflicting a specific group of people who we hope to lift above poverty and then subsequently be done with anti-poverty policy.

But the poor are not a static group of people. In the most recent census data, covering 2022 and 2023, we can see that only 42% of people who were in poverty in 2022 were also poor in 2023.²

This is because poverty, for the most part, is about weak spots in our income-distribution system, not specific individuals. People churn in and out of those weak spots when they do things like lose their job, become too old or sick to work, or have a kid. These weak spots can be patched, but only through a permanent stream of cash transfers directed at those spots, which is what a good welfare state does.

Even if we put aside these and other conceptual errors, Piper fails to offer a convincing argument for the main thesis of her piece, which is that cash welfare programs fail to result in improvements on a variety of cognitive and health measures and that there is a conspiracy of silence around this in the media.

The latter point is hard to take seriously when many of the same studies Piper discusses in her piece were also discussed in a July 28 New York Times article titled “Study May Undercut Idea That Cash Payments to Poor Families Help Child Development.” The New York Times is America’s largest newspaper with 11.7 million subscribers.

There are some studies where certain cognitive and health measures did not improve after cash increases, but there are also some that reach the opposite conclusion. Indeed, all one needs to do to find some of this contrary research is to read the literature reviews in the more recent papers with the null results. One could even look to other countries, like Finland, where rigorous randomized research was recently done on a cash transfer program that showed positive results for various measures of well-being, including self-reported trust, satisfaction with life, confidence in one’s future, health, ability to concentrate, financial well-being, and stress. One quirk of the Finnish study is that the government ran it and was therefore able to literally force the randomly selected individuals to participate in it. The studies Piper cites rely on volunteers, which technically makes them nonrandom, despite the fact that they are called “randomized control trials.”

Other than noting in one sentence that we need to build and improve “institutions that provide education, health care and housing,” Piper does not offer any policy alternatives in her piece.

Without further elaboration, it is hard to figure out what exactly Piper means by this.

As far as education goes, the United States clearly has the best colleges and universities in the entire world, whether measured by the amount of top schools or the breadth of quality across its mostly public higher education institutions. The U.S. also provides universal schooling for children, 87% of whom attend public schools. Pundits love to complain about the K-12 schooling, but in international assessments, American students perform similarly to their developed-country peers (usually a little better in reading, a little worse in math).

The United States also has cutting-edge health care institutions and already spends radically more on its health care system than comparable countries. The U.S. health care system is an inefficient mess, but fixing it would reduce, not increase, the amount of money spent on it, which would ostensibly free up more money for useless cash benefits.

What is ultimately disappointing about Piper’s piece is that it is yet another effort to make poverty, inequality, the welfare state, and issues of distributive justice seem much more complicated than they really are. As a policy matter, these are mostly solved problems. There are countries in the world right now with very low levels of poverty and inequality, and they follow a very well-documented policy approach called the Nordic model to get there. Acting like this is all very inscrutable and we don’t really know and we need more study can make you sound like you are overflowing with intellectual honesty, but in reality it is a kind of pseudo-profundity that papers over the moral atrocity of poverty amid plenty.

---

Kelsey Piper, The Argument staff writer:

I want to contest in the strongest possible terms the claim that poverty is, as a policy matter, a mostly solved question and the related claim that we do not need to conduct studies on what works because we know what works — the “Nordic model.”

“Be Finland” is not a law that you can pass. Indeed, as Bruenig observes, in Finland, cash transfers seem to have lots of large and measurable effects on life satisfaction, stress, and health — all areas where I expected to see improvements from cash in the United States, but did not.³

Since various Finnish policies do not seem to have the same effects here, I do not consider the success of Finnish policies in Finland sufficient grounds to reach Bruenig's conclusion: That actually having a just and effective welfare state isn’t complicated and doesn’t need further study.

But even if you think we could or should just adopt Finland’s whole safety net, the question remains whether, in the United States, with our social insurance programs, the marginal dollar might be better spent on housing, health care, transportation, or a dozen other options. That question seems important, perhaps even worthy of study — if you don’t reject all study as pseudo-profundity meant to obscure the obvious fact of what we should do.

Bruenig says that cash must be a part of any developed nation's welfare system — which I don’t even disagree with. What he does not seem to find an interesting question is whether cash is the best use of limited resources in the United States today. Does Bruenig think spending money on basic income for the homeless is better than other homelessness programs? Does he think giving cash to recent mothers is better than expanding their access to prenatal care clinics?

I expect he rejects the trade-offs (a theme with Bruenig). Many of the people I’ve spoken to tell me they worry that what limits our ability to aid the poor is political will, not material limits, and so every criticism of a poverty program does not improve or redirect the safety net but simply saps it. Therefore, checking whether an anti-poverty program is having the desired effects is a betrayal — indeed, we must reject the whole framework in which it can have desired effects at all.⁴

I don’t think this is an accurate model of what holds back expansions of the social safety net. Instead, I suspect the opposite: Many people who would support the safety net if it were to succeed, oppose it precisely because they perceive it as failing — and perceive its defenders as lying. If we treat every conversation as primarily about advertising for programs rather than for figuring out what works and what doesn’t, we fail to get our politicians good information about which priorities they should pursue when in power.

Even putting aside the financial trade-offs, passing major legislation takes time. The Obama administration had one big swing: The Affordable Care Act; Trump’s first term had the Trump tax cuts; Biden had the IRA. Is Bruenig’s play UBI? The next time an anti-poverty Congress and president come into power, what is the first thing we should ask them to do?

I don’t know the answer to that question, which is why I indulge sometimes in the pseudo-profundity of trying to figure it out.

What I mean by the war on poverty

Bruenig laments the move in policy and development circles toward measuring outcomes and running studies. But I think he misses the reason that this happened: People became interested in what could deliver meaningful outcomes because our existing policies were not delivering the hoped-for outcomes.

Since 1969, our GDP in real terms has quadrupled. Transfer payments to individuals have risen, in 2019 dollars, from $417 billion (in 2024 dollars) in 1969, or 4.8% of GDP, to $3.2 trillion in 2024 — or about 12% of GDP. (This only counts federal spending, not any state or local programs.)

When a society gets four times richer and spends eight times as much on transfers to the poor, you would expect fairly dramatic results. The results are … well, they’re not zero.

The official poverty rate — introduced in the 1960s and adjusted for inflation but not otherwise changed at all — says that 12.2% of Americans lived below the poverty line when it was established in 1969, and now it’s 11.1%. But that metric isn’t very good⁵ — people I respect prefer the anchored supplemental poverty measure, which better incorporates the effects of programs like SNAP, and which looks like this:

Like I said, it’s not nothing. You can see the Great Society on there — it did genuinely help! In 1969, 24% of people lived below the poverty line under this measure before transfers and 22.5% after them. In 2023, 22% lived below the poverty line before transfers and 11% after.

Those are real gains. But let’s put them in context: The amount we spend on combating poverty today is close to the entirety of GDP in 1969. The share of Americans whose pretransfer income places them in absolute poverty has barely fallen, and a massive increase in social spending has made posttransfer poverty meaningfully rarer but still extremely widespread. These are worse results than we hoped for, and people got interested in the technocratic project of measuring which programs work best because we think it is probably possible to get better results than this.

Here’s my take on why all this money hasn’t improved things more: We have made the parts of the country where there is the most opportunity completely unaffordable to the poor. Yes, $1,000 a month might be transformative in a world where an apartment for rent in San Francisco goes for $1,500. So, I suggest we build so much more housing that the price of housing craters in every major metro area.

Bruenig says American schools are great. I think school reform after school reform has served every conceivable interest group except students (who do not vote) and so have failed to meaningfully increase literacy and numeracy, even though we now have a road map for how to genuinely let every child thrive. So, I suggest our schools drop the so-called “innovations” they’ve introduced and teach phonics and basic arithmetic in small groups at each student’s level.⁶

But I’m really not very ideological about this: I want to spend more money, by all means, on any program, at all, that consistently gets good results, whether it concords with my preexisting ideology or not.

In that spirit, I have absolutely not written off cash transfers (I even said as much in the original piece). I am open to learning that the metrics these studies employ are entirely the wrong ones, that there is a better way to measure whether someone’s life is getting better. I am not entirely a fan of subjective well-being metrics, and time-use surveys have their limits, and maybe you need to do community-level basic income grants (as are often done in developing countries) to really capture the benefits of cash that flow to people outside the recipient’s household. Fine.

But I reject the perspective that there is nothing to be learned from high-quality studies that find we have spent millions of dollars and people don’t even tell us they are happier because of it. Or findings that demonstrate giving homeless people thousands of dollars does not much alter their odds of finding housing.

If we do not have in mind a measurable criterion for success, a point at which we will say, “this didn’t work as well as we expected, and so we should shake up the formula and do something else which might work better” — then my prediction is that we will do the entire Great Society and have almost the same share of people living on very low incomes 56 years later.

You cannot persuade me that all we need is a vision and an increased percentage of GDP spent on alleviating poverty, because those have not, in fact, sufficed. Ending poverty is not a solved problem and we are not going to achieve it without studying the effects of the programs we’re running and trying to prioritize somehow among them.

1. About 63% of the population over 15 is in the labor force, according to the Bureau of Labor Statistics. Once you include children, this figure drops to around 50%.

2. Author’s calculation of the Current Population Survey.

3. Bruenig seems to suggest that the reason these results show up in the study he cites, but not in U.S. RCTs, is that Finland can force people to participate in studies and the U.S. cannot. I do not think this explains the null results in U.S. studies.

4. I don’t think anyone actually consistently rejects the idea that anti-poverty programs ought to have effects and we should measure them; Bruenig himself resorts to it to defend the Scandinavian model. Instead, when a program does not have the desired effects, they argue that the philosophy of measuring effects is technocratic (guilty) and oblivious to the real objectives (which I would argue are generally also possible to measure and to succeed or fail at).

5. In particular, it does a poor job of capturing in-kind benefits, which are an important element of our efforts to fight poverty and have changed a lot since the 1960s.

6. I agree that we sometimes hold up fine in international comparisons, but I don’t think we hold up fine compared to the results that we know are possible to achieve. I’m less impressed with Finland than Bruenig and I do not really regard them as the bar to measure our schools against. We have seen intensively focused, well-run programs with unselected overwhelmingly poor and minority student bodies achieve 90%+ literacy, so that’s my bar. We aren’t meeting it.