Poverty Is a Policy Choice

By Dylan Matthews, Vox

12 August 21

The lesson from the past year: Poverty is a policy choice.

n 2020 and 2021, the federal government gave Americans an unprecedented amount of money.

A big part of the US’s response to Covid-19 was strengthening the existing safety net, or providing relief specific to particular types of spending, like food or housing. Food stamp/SNAP benefits were raised. Evictions were barred nationwide for over a year.

But the core of the response was an unprecedentedly large and inclusive set of stimulus checks directly shoveling money to most Americans. In three rounds — March and December 2020 and March 2021 — most adult Americans got $3,200 each, and most American children got $2,500. Then, starting in July 2021, most American children started getting $250 every month, with young kids getting $300 a month.

These are broad-based policies with only quite wealthy Americans excluded from benefits. The first round of stimulus payments, for instance, were available to 93 percent of Americans, with only the richest 7 percent left out. But unlike stimulus checks passed during the 2001 and 2008 downturns, the 2020-2021 checks were universal at the bottom of the income scale. They had no work requirement or requirement that recipients paid federal taxes in the past.

That means that the stimulus checks should have had a profound effect on poverty this past year or so — and that’s exactly what researchers are finding.

In March, researchers at Columbia led by Zachary Parolin estimated that as a result of President Joe Biden’s stimulus package, the American Rescue Plan, the US poverty rate would fall to 8.5 percent, the lowest figure on record and well below 2018’s figure of 12.8 percent. This past month, researchers at the Urban Institute, using a slightly different means of measuring poverty, found that 2021 poverty will be around 7.7 percent, almost a halving relative to 2018’s rate of 13.9 percent per their methodology. (Official US Census poverty statistics for 2020 have not yet been released.)

The Columbia authors find that if you compare 2021 to every year for which the census does have data, from 1967 to 2019, and use a consistent poverty line, 2021 is projected to have the lowest poverty rate on record.

Considering that the US endured a pandemic and economic shock in 2020, these numbers are remarkable.

That’s the big news, and it’s good. It’s important, though, to dig a little deeper into what these studies are projecting — and what it could mean for future policymaking. Much of the credit for lower poverty in their models goes to the stimulus checks. And while the Biden administration hopes to keep the $250/$300 per month checks to parents going, the periodic rounds of stimulus payments during the height of the pandemic were meant to be temporary emergency measures.

That means that, absent further government action, poverty could be set to increase again in 2022.

The good news is that Americans learned a key lesson during the pandemic: Poverty is a policy choice, and it can be easily reduced through increased government support.

The bad news is that the government, particularly swing senators hesitant about spending too much money, might see that choice, and make the wrong one.

How we measure poverty, briefly explained

The convergence of research finding that poverty took a major hit is a big deal. One thing that’s important to understand is that measuring poverty is actually trickier than it might seem. How exactly do you measure it? What counts as poor? And are the metrics researchers set really showing the thing they want to capture?

Take, for instance, the official poverty measure that the government uses. This was developed by the Social Security Administration’s Mollie Orshansky in 1963. Orshanksy defined the income cutoff for poverty as three times the “subsistence food budget” for a family of a given size. That subsistence food budget, in turn, was derived from an “Economy Food Plan” developed by the USDA in 1961, based on data from the 1955 Household Consumption Survey.

In other words, the way we measure poverty in 2021 is based on an analysis from 1963 using data from 1955.

The official measure has other problems. The way it defines income leaves out in-kind benefits like food stamps or housing assistance. More relevant to this round of studies of poverty in 2020, the official measure only counts pre-tax income, meaning refundable credits like the child tax credit and the 2020-2021 stimulus payments (which were technically structured as tax credits) don’t count for poverty purposes.

The upshot? The official poverty measure shouldn’t move at all in response to the stimulus checks, or the strengthened child tax credit, or increased food stamp benefits. Even though all those measures make life materially easier for people in poverty, the official measure ignores them. It’s just another way in which the official measure is outdated, and highlights how limited this picture of poverty in America is.

That’s why these new studies are so important. Combined, they really start to give us an accurate picture of poverty in the current stimulus era.

What the new poverty studies say

Because the methods used in each study are so different, it’s striking that they reached a very similar conclusion: Poverty in 2021 will be much lower than it was in 2018, and it will be lower largely because of anti-poverty programs.

The Columbia researchers estimated that the December 2020 relief bill (which reintroduced supplemental unemployment checks and sent out $600 checks) reduced poverty in 2021 from 13.6 percent to 12.3 percent; the Biden stimulus, including $1,400 checks, additional unemployment support, and the enhanced child tax credit, cut it further to 8.5 percent. Without these interventions, poverty would’ve been higher than in 2018, not lower.

Meanwhile, the Urban Institute team broke down the anti-poverty effects program by program. The stimulus checks alone lifted some 12.4 million people from poverty this year, their research finds. Food stamps, including Covid-related improvements, removed another 7.9 million people from poverty, while unemployment insurance (both the base program and Covid-19 bonuses) lifted another 6.7 million people out.

Another Columbia project, for which Parolin is also the lead researcher, uses a different dataset to attempt to estimate poverty every month during the pandemic. This is a slightly different project than estimating annual poverty; for instance, it implies that for people who got the $1,400 checks in March 2021, the checks reduced poverty that month, but did nothing for poverty in April. But this data still underscores how important Covid-specific policies like those have been for reducing poverty. Without relief measures, poverty rates would be much higher throughout much of the pandemic, as a trend line comparison shows.

At the University of Chicago, Bruce Meyer and other policy researchers have their own monthly poverty measure that works slightly differently. Like the Columbia measure, it uses data from the monthly Current Population Survey — a census study that asks some 60,000 households every month about their income and other life conditions — to estimate the current poverty rate. It uses the official poverty measure threshold for income, but it includes sources of income the official measure excludes, in particular tax credits like the stimulus checks.

Their measure showed poverty falling substantially with the first round of stimulus checks in 2020. In February 2020, the last pre-pandemic month, poverty was 10.7 percent per their measure; in May, it bottomed out at 9.1 percent. But they then saw poverty steadily increase for the rest of the year, even continuing after the Biden stimulus, with June’s rate estimated at 11 percent, higher than before the pandemic.

That’s a striking finding, especially when you consider that the Columbia researchers also saw poverty increase in April and May after falling when stimulus checks began to go out in March. But the data from Meyer and his colleagues doesn’t account for some sources of financial support, like food stamps, which might partially explain why it finds higher poverty than other sources.

What the fall in poverty tells us

Poverty measurement is incredibly tricky, as the nuances above hopefully make clear. But the basic math of measuring poverty is rather simple. First, you pick a dollar amount. Then you find out how many people’s incomes fall below that dollar amount. Choosing what amount to use, and how to define income, is hard, but the basic concept requires nothing more than arithmetic.

This also implies that reducing poverty is pretty simple: Just put more money in people’s hands. That won’t necessarily be enough if, like the official poverty measure, your metric entirely ignores money the government gives out through tax credits. But by any normal poverty measure, handing out cash should reduce poverty.

This past year was a huge vindication of that insight. Years of research had suggested that cash programs don’t necessarily have the big downside their critics always highlight: discouraging work. If handing out cash led people to work dramatically fewer hours or to quit their jobs, then cash payments wouldn’t cut poverty by as much as they initially seem to.

Luckily, cash doesn’t seem to discourage work to that degree. In 2019, a group of economists and sociologists specializing in child poverty put together a major report for the National Academy of Sciences, and their estimate based on the research literature was that a cash benefit of $3,000 per year for all but the richest children would reduce work effort by about 1.15 hours a week on average — a fairly trivial amount that barely changes the antipoverty impact of such a program.

The effects of stimulus checks to adults, like those pursued in the past year, are surely different, but the evidence generally suggests that work disincentive effects of cash are small. University of Pennsylvania economist Ioana Marinescu, in a wide-ranging review of the effects of cash programs, concluded, “Our fear that people will quit their jobs en masse if provided with cash for free is false and misguided.”

All of which suggests that using cash to reduce poverty might really just be an arithmetic problem: give people enough money to escape poverty and they’ll escape poverty.

The US has been sending out a lot of cash during the pandemic. But that’s almost certainly coming to an end. The enhanced child tax credit is a policy many Democrats want to make permanent, or at least (as the Biden administration has proposed) extend for several more years. But the $1,200 and $600 and $1,400 stimulus checks were emergency measures, as were the $300/$600 weekly unemployment supplements.

All that implies that in 2022, when those measures are gone, poverty is likely to shoot back up again, even in a strong economy with robust job growth.

That doesn’t have to be the case. A permanent poverty reduction agenda could make sure we don’t only fight poverty in extreme conditions like pandemics.

The US could, for instance, boost unemployment benefits on a permanent basis, and implement triggers to increase benefits during downturns. Senate Finance Chair Ron Wyden is working on a plan for this currently, and Sen. Michael Bennet (D-CO) has a detailed outline of proposals along these lines as well.

We could also adopt a guaranteed income for adults to approximate the effect of the stimulus checks, as a permanent policy. Just as one example, a team at Ohio State’s Kirwan Institute for the Study of Race and Ethnicity, led by Naomi Zewde, have developed a proposal for a guaranteed income set at the national poverty line (currently $12,500) for adults, with $4,500 per year per child in extra benefits for families with kids.

Another, more modest, option would be to turn the standard deduction into a refundable tax credit. A single person could get monthly checks totaling, say, $2,761 per year — the same benefit someone in the 22 percent tax bracket gets from the standard deduction right now. That would make doing taxes simpler while reducing poverty.

But the broader point is not that we need any specific guaranteed income plan. The point is that poverty is a policy choice. The federal government can literally make the poverty rate whatever it likes. It could continue reducing poverty year after year, even after Covid-19 is over. It just needs to make that choice.