The research director of Columbia University’s Center on Poverty and Social Policy discusses a staggering new report on child poverty in the United States.

In February, the Center on Poverty and Social Policy, at Columbia University, released a staggering new report on child poverty in the United States. The study found that the child-poverty rate had increased dramatically in the space of just a month, with seventeen per cent of children living in poverty in January, 2022, compared with around twelve per cent at the end of 2021. The reason for the sharp rise was the expiration of the expanded child tax credit, which President Biden signed into law last spring, and which Congress has not agreed to renew. In the absence of the monthly cash payments guaranteed by the expanded credit, about 3.7 million more American children are living in poverty.

The child tax credit has existed in several different forms since it first became law, in 1997, and Biden’s expansion of it was the biggest in its history: the payments became bigger, and were distributed through monthly checks instead of a single annual lump sum. Perhaps most crucially, Biden made the benefit available to families with little to no income, who were previously ineligible for the credit. I discussed the Columbia report with one of its three authors, Sophie Collyer, the research director at the center. During our conversation, which has been edited for length and clarity, we discussed what made the child tax credit so effective, the impact of its disappearance on Black and Latino children, and what the pandemic has taught us about how to improve the social safety net.

What did your study find at the most basic level?

We found that 3.7 million more children are in poverty as a result of rolling back the child tax credit between December and January. We’ve been measuring poverty over all and among children since the beginning of the pandemic, but using a monthly framework. Traditionally, poverty’s measured in an annual framework, where you take people’s incomes across the year and evaluate it against a poverty threshold that’s calculated for that year. What we’ve seen though, with the pandemic, is much volatility in terms of income that families are receiving from month to month, because different government policies have been in effect for part of the year and not in effect for other parts of the year. Some examples would be things like stimulus checks and unemployment benefits.

So it’s been a really useful tool for evaluating the impact of the expanded child tax credit in real time. Over the summer, with the initial payments, we saw an immediate reduction in the rate of child poverty, particularly because it was being paid out monthly for the first time and reaching about one in three children who were previously ineligible for the credit. So over the summer you’re seeing both the initial payments being paid monthly for the first time but also a much greater share of the child population being eligible for the credit. But it was only in effect in terms of monthly payments through the end of 2021. And the expansion of the child tax credit was only in effect for the calendar year 2021. With the change in the year from December to January, January was the first month where families would not receive these monthly payments of either three hundred dollars or two hundred and fifty dollars per child, depending on age. So the exact reverse of what we saw over the summer happened, and we saw a spike in child poverty as a direct result.

Intuitively, it seems very likely that the end of the credit is responsible for the spike we saw from December to January. But are there ways to control for the increase so that we know it was due to the child-tax-credit expiration?

That’s a great question. Within December, for example, you could see the number of children moved out of poverty by the credit in that month. In December, we saw about 3.7 million children being moved out of poverty as a direct result of the credit. Then when we move to January, we’re seeing 3.7 million more children in poverty. I think the difference between those two months lends more evidence to the fact that this is a direct result of the absence of the credit.

When you’re studying this, obviously you’re not talking to 3.7 million people, but you’re studying individuals. Are you having people tell you directly that the credit is why they have less money, or something else?

This framework is developed using data from something called the Current Population Survey, which is run by the Census Bureau. The Census Bureau every year releases an annual poverty rate based on data collected in the March version of this C.P.S. survey. But there’s actually a survey done every month. And there’s a methodology that allows us to use the most recent data from the C.P.S. to estimate or project what the monthly poverty rate is within a month, and also account for these variations in income from different policies.

What did you find about poverty rates for Black and Latino children specifically in this December-to-January change?

A big piece of the expanded child tax credit last year was that the earnings requirement associated with the credit was removed for 2021, and it was made fully refundable. The amount that you received was independent of your earnings. This was the first time this has ever happened. Before the expansion, many Black and Latino children were left behind when it came to receiving the full benefits of this credit—one out of two Black children and one out of two Latino children were ineligible for the full credit because of family income levels.

Then, with the expansion, you saw that gap being filled in. Now, with the absence of the credit, we saw a 5.5-percentage-point increase in the poverty rate of Black children and 7.1 percentage points for Latino children, translating to about nearly seven hundred thousand Black children falling into poverty and 1.3 million Latino children falling into poverty as a direct result of the credit being removed. It gets tricky when you are talking between these relative increases and percentage-point increases, but over all Black and Latino children saw the largest percentage-point increase in their poverty rates.

You said that you’ve been studying child poverty since the beginning of the pandemic. I just wanted to take a step back. What did child poverty look like for the first year-plus of the pandemic, before this expanded child tax credit went into effect?

Child poverty in 2020 was actually lower than you might have expected. And that actually had to do with the stimulus payments and the unemployment-insurance benefits that many families received. Counter to what many people probably think, 2020 actually had one of the lowest poverty rates on record as a direct result of the policies in place in that year. That said, child poverty has been higher in the United States than nearly any other wealthy nation for decades, or at least for the past decade. We had a child-poverty rate of about thirteen per cent before the pandemic, which again was extremely high, and it was much higher in various areas, and by ethnicity there were also dramatic variations. Before the pandemic, people were pushing for this expansion to the child tax credit, knowing that it would address the problem that we had, and had for a long time, which is strikingly high rates of child poverty.

So it was introduced and included as part of the American Rescue Plan, in part because of the knowledge that it would help address child poverty at the same time as levelling income in response to the pandemic. But it wasn’t necessarily thought of as a specifically pandemic-related policy. In 2019, the National Academy of Sciences released a report on policy packages that would reduce rates of child poverty, and the expansion of the child tax credit was found to have the greatest potential impact on the child-poverty rate.

So it seems like what you’re saying is that before the pandemic, we had these terrible child-poverty rates compared with other rich countries, and then things actually got better than people expected during the pandemic. But it appears, from the latest report, that child poverty has now risen above what it was prior to the pandemic. Is that accurate?

Before the pandemic, it was about fourteen per cent. Now it’s higher than it was before the pandemic. All of the government support, in terms of what families were receiving to stabilize their incomes, is basically gone at this point, with maybe the exception of some continued expansions to food stamps. That’s the only piece of pandemic-related income support that is still being made. But also the seventeen per cent is based on a monthly framework, and it’s not necessarily apples to apples because it jumps around from month to month, and you tend to see dips around March and April.

Why is it that the child tax credit is so effective, do you think?

Well, it’s giving families money directly. When you’re thinking about poverty, you’re evaluating a family’s income relative to a line, and this is a direct way of increasing family income. That’s why it’s so effective. It’s also been extremely effective because it fills in so many of those gaps that existed beforehand. Children in poverty and low-income children benefitted disproportionately from this expansion because they were ineligible for the full value of the credit beforehand, so they were seeing greater gains than middle-income families as a direct result of the expansion. But basically all children were receiving this credit. More than ninety per cent of children in the country were eligible for either the full thirty-six-hundred-dollar credit or the three-thousand-dollar credit in 2021.

Does the incredibly steep rise from December to January change any of your prior assumptions about the social safety net?

I think it wasn’t that surprising, and that’s an unfortunate thing. We knew that providing this money to families had a direct impact, and so the inverse of that is the same, where it has a direct impact below the poverty line or further below the poverty line. I wouldn’t say it was something unexpected, but what I think about is the volatility that families are facing. We’re both seeing an increase in the poverty rate, and that it’s connected to families having had a stable flow of income for six months in 2021, and then that just disappearing. We’re seeing the impact of it in terms of the poverty rate, but it’s also creating more instability for families and children across the country.

There has been a lot of well-done qualitative work on what families have been using the money for. What is a theme across all data sets and many different sources of information is that many families have been using this to purchase food. Just as we’re seeing this increase in poverty, one might expect—we don’t know yet—an increase in the rate of food insufficiency and hunger among households with kids and among children, which is deeply troubling.

But another very interesting aspect of the program is that it’s cash-based. So many social policies and social programs in the United States consist of in-kind transfers—housing subsidies, food stamps—and they’re infrequently cash. But with this you saw families receiving a cash payment, and cash is fungible. In one month, you might need it to fill in a food budget, but for the next month, it might be used to fix a car. Another month, it might help with child care. That flexibility is also something that comes out of the data, with families using it to meet needs that vary from month to month. I think that’s a really important takeaway in terms of the importance of cash in families’ lives and the importance of the monthly aspect of this. As I said, it used to be this tax refund that would come at tax time in a lump-sum payment for those who were eligible for the full credit. Bills often come once a month. They don’t just pile up and come at tax time, when you might be able to use a refund to pay them back. The monthly aspect of this also contributes to that stabilizing effect.

I don’t want to put a rosy spin on any of this, but it seems like, from what you’re saying, the one positive that might come out of this, even if it is not renewed, is that there are some new ideas about smart ways to do social policy.

Yeah. I think policymakers right now know that they have a lever at their hands that, if pulled, would move millions of children out of poverty. I think that’s just an important thing for everyone to know: that there’s a choice at hand that people can make, and it will have dramatic impacts on the poverty rate. I think that the pandemic also over all has revealed how effective policy can be at stabilizing family income. We did not see a sharp rise in the poverty rate in 2020, and I don’t think anybody in March of 2020 thought that would be the case. Every family experienced a lot of financial strain, but it also revealed that policies can be effective at least at keeping people out of poverty.