Get access to over 1,300 articles and reader Q&As
Many U.S. family policy proposals — and a widely held belief more generally — rest on a straightforward idea: more money in a household translates into better outcomes for kids.
But in recent weeks, new research has poked holes in that theory. Researchers were surprised to see that income subsidies to poor mothers didn’t improve their children’s development at the age of 4.
This result, coming out of an enormous, multi-year project called Baby’s First Years, is likely to be seized upon by some, questioned, and rejected by others. But before we do either, let’s dive into the data, walk through what the study found, and consider where this leaves policy aimed at closing gaps for low-income families.
What is the question? (And why is it so hard to answer?)
The core question this research asks is: Does having more money, on its own, improve outcomes for children and families?
As background, there is a lot of evidence that children who come from families with more income do better on health and cognitive development measures, and their mothers and families fare better, too, on mental and physical health. The problem is that it is not clear if that evidence is causal or not a result of other differences. Households with more income also tend to have parents with more education, are more likely to have two parents, and tend to live in wealthier neighborhoods. Even though it seems like common sense, it is difficult to isolate the potential power of income alone in the data.
Knowing the impact money has on families is important, though. Many policies focus on providing funds directly to parents, with the idea that it will trickle down to their children. One example of this is payments per birth, which the Trump administration has proposed. There is, therefore, a significant need for good causal evidence of whether this approach works.
The researchers in the Baby’s First Years project aimed to do this through a first-of-its-kind randomized controlled trial. In brief, they recruited 1,000 mothers from hospitals in four areas of the country (New York, New Orleans, Omaha, and the Twin Cities). Mothers were recruited at the hospital when they were giving birth. To be in the sample, women had to report an income in the prior year that put them below the federal poverty line, which is about $20,000 a year for an adult with two children.
Women were allocated randomly to one of two groups. The treatment group got a cash payment of $333 per month, and the control group got a payment of $20 per month. Each woman and her child were followed for four years. It is worth noting that the pandemic interrupted some of the data collection, but not the payments.
One of the criticisms of this trial is that the amount of money is too small to matter. In the scheme of things, this amount is small, but there was reason to think it would matter. For one thing, this is a very low-income population. The cash transfers represented an 18% increase in yearly income over baseline in the first year; inflation made this count somewhat less in later years. Second, transfers of similar amounts through the Earned Income Tax Credit and welfare-to-work programs have shown impacts on children.
The simple fact is that, prior to this research being done, the researchers and many others believed that transfers at this level would show at least some long-lasting health and developmental impacts on families.
What did the researchers find?
The researchers produced many papers over the past several years, looking at a wide variety of outcomes, from the time spent on parent-child activities to infant brain activity and child sleep. You can see a full list of papers here, and I have captured the outcomes below.
When the researchers looked specifically at spending habits, they found that the extra income does make a difference. Mothers report spending more on their children, including toys and books, and there is a reduction in the chance that the family is living below the poverty line. This is not that surprising, but it does suggest mothers do not, for example, respond to the additional income by working a lot less.
It is important not to dismiss this impact. Families were less likely to live in poverty, used fewer social services, and bought more books for their kids; this is all extremely positive! It has gotten less attention, though, because it is more mechanical; we already had a sense that giving people money would make them less likely to live in poverty (since the definition of living in poverty relates to income), so this is more of a validation of the experiment than a true result.
In addition, in a paper published early on within the group, researchers found a significant increase in brain activity among infants born to mothers who received the larger cash payments. This prompted a lot of excitement about the possible longer-term outcomes.
However, subsequent papers have shown no difference in any of a huge variety of outcomes. This includes developmental measures for children (both objectively measured by researchers using validated scales and parent-reported), health measures for children, health measures for mothers (both mental and physical), substance abuse, and intimate partner violence.
The researchers will continue to follow these families through at least age 8, but thus far, the evidence points pretty convincingly to the cash transfers not making a significant, measured difference.
What explains this?
This issue — how much income matters for families, the role of money in improving outcomes for children — prompts a lot of strong feelings, feelings which often break down on political lines. When confronted with evidence like this, it’s tempting to default to those feelings.
On one side, people who have been skeptical about the role of child care or family subsidies may trumpet this research as the end of the debate. See, it doesn’t work, so let’s forget about these kinds of support.
On the other side, those who believe in these policies may look for reasons to dismiss the results. The amount wasn’t enough. The pandemic interfered. You didn’t look at the right outcomes.
Neither of these reactions respects science. When we do really good science, which this is, we must come into it with a willingness to accept what it tells us (otherwise, what is the point of spending millions of dollars if you already knew the answer?). But we also must come in with humility to understand that conclusions are built out of many pieces of data, not just one.
This research is one really strong piece of evidence, and I believe it should affect our views about the role of income transfers alone, at least at this level, in affecting child outcomes. And this is not the same thing as taking support away from families.
There is much more to be done. It’s still very clear that children from families with more income, even on the scale provided by this study, have better health and development outcomes, often appearing at very young ages. Why exactly this happens — and how to reduce inequality among lower-income families — is still something we need to understand, and this research suggests we need to look at other ways to address it.
The bottom line
- Researchers tested whether giving low-income mothers $333/month (versus $20/month) after birth improves family and child outcomes, using a randomized controlled trial with 1,000 participants in four U.S. cities.
- The study aimed to provide causal evidence on the impact of modest but meaningful income increases, building on prior findings that similar transfers have benefited children in other programs.
- Extra income led mothers to spend more on children and reduced poverty rates, but aside from an early finding of increased infant brain activity, follow-up research has found no lasting effects on child development, health, or family well-being so far.
- The findings suggest that modest income transfers alone may not significantly improve child outcomes, pointing to the need for exploring other strategies to reduce inequality.
Log in
I appreciate this breakdown of the research, as well as the clear-eyed scientific takeaways it provides, but I still think this article could do more to clarify that this study doesn’t tell us that we shouldn’t do these kinds of money transfers, just that they don’t affect the specific outcomes studied. Providing families that don’t have enough money with money isn’t just a matter of whether or not these child outcomes improve, it’s also a matter of making any attempt whatsoever to start to even the playing field. When some families have millions of dollars and some are barely getting by on $20,000/year it is a matter of basic ethics to provide those families with more money. That it’s insufficient to do so can tell us that even more is needed. As the article does point to, there is often a lifetime of poverty and its attendant suffering affecting these families, as well as a lack of community resources, as well as other structural inequalities (of language, of health care, of opportunities, etc.). I just worry when I read an article like this: when things are so entirely stacked against people, the language of neutrality can sometimes point us away from truth.
Can you speak at all to findings on the international development side, say from GiveDirectly,or other NGOs that study cash transfers? My understanding was that in developing countries similar small (20% of income) transfers do yield more school enrollment and healthier child outcomes. Perhaps the context of those countries is so different that it’s inappropriate to use any of that evidence in the US?
I’d also like to know Emily’s thoughts on this! I have lived and worked in several developing countries and my sense – as a non-expert outsider to both situations – is that rural poverty in developing countries is massively different to urban poverty in the US. For example, in the rural African contexts I have known, barriers to school attendance include(d) modest school fees (really trivial amounts to us but impossible for some families) and the cost of school uniforms, as well as parents keeping the children home to help with farming.
When they covered this study on The Daily they devoted a lot of time to the fact that families got much larger cash payments due to the pandemic, which greatly reduced the value of the smaller payments in this study. Do you see this as a confounding factor, given that both groups got these large payments? Or would you say since they both got them it isn’t a confounding factor?