Protecting undocumented immigrant mothers from deportation can dramatically improve the mental health of their US citizen children, according to a new study. The research shows that policies targeting undocumented immigrants affect more than the single individual, and carry important implications in the long-term for society as a whole.

The study, published today in Science, focuses on an immigration policy issued by the Obama administration in 2012 called Deferred Action for Childhood Arrivals, or DACA. This program temporarily protected over 780,000 undocumented immigrants from deportation. Researchers analyzed Medicaid claims data from Oregon and found that in that state, the incidence of mental illness dropped by half among children whose mothers were eligible for DACA.

There are an estimated 11 million undocumented immigrants in the US, some of whom are parents to the more than 4 million children who are US citizens by birth. Previous studies have shown that children of undocumented immigrants face challenges such as fear of separation and stress. Jens Hainmueller, a co-director of the Stanford Immigration Policy Lab and one of the lead authors of today’s study, wanted to see how one particular policy affected immigrants and their children. He and his colleague focused on DACA because, “it’s basically the most significant policy to affect undocumented immigrants in the last two decades,” Hainmueller tells The Verge. The Trump administration is also currently deciding whether to cancel or extend DACA.

DACA protected from deportation and gave temporary work authorization to undocumented immigrants who were brought to the US as children, and were under age 31 as of June 15th, 2012, when DACA was adopted. At the time, it was estimated that 200,000 US citizen children had parents who were eligible for DACA. So how did the policy affect these children?

The challenge is that there isn’t a lot of data on undocumented immigrants, Hainmueller says. They usually don’t answer government surveys, and prefer to stay off the radar. To overcome this problem, Hainmueller and his colleagues decided to delve through data from Oregon’s Emergency Medicaid. This is a government program that covers emergencies, as well as labor and delivery services, for low-income individuals who are not eligible for Medicaid. In states like California and North Carolina — the only states that have estimates — between 90 and 99 percent of mothers who use Emergency Medicaid to give birth are undocumented immigrants. So the team assumes that the Oregon Emergency Medicaid data they analyzed is also mostly from undocumented mothers. (Oregon was the first state to give the researchers the data they requested.)

Because DACA allowed only immigrants born after June 15th, 1981 to apply, the researchers were able to find out whether the mothers were eligible for DACA — identifying over 5,600 of them. Then, the researchers analyzed Oregon’s Medicaid claims to identify the children born to these mothers from 2003 to 2015. (Because the children were US citizens by birth, they were eligible for Medicaid.) They found more than 8,000 of them, six-and-a-half years old on average. Through the Medicaid billing records, they could track the kids’ mental health, and check whether they were diagnosed with illnesses such as anxiety disorder and adjustment disorder, a condition that occurs when people are dealing with great stress.

The researchers compared kids whose mothers were eligible for DACA with kids whose mothers weren’t eligible, before and after DACA was passed. They found that before DACA was passed, the two groups of children were diagnosed with these disorders basically at the same rate. But after DACA was passed, among the children whose mothers were eligible for DACA, the mental health diagnoses basically dropped by about 50 percent.

“You get these dramatic improvements in the mental health of the children as a result of DACA eligibility,” Hainmueller says. “It provides the first real causal evidence that DACA protection really has significant consequences for the health of the children of undocumented immigrants.”

And the consequences spread even wider than that. Research shows that children with mental health disorders are more likely to have depression, abuse drugs, and do poorly in school as they grow older. And that, in turn, can lead to unemployment and poverty. “There are significant social costs that come from these early mental health disorders,” Hainmueller says. So DACA isn’t just about the individual immigrant family that’s affected, he says, it’s about all of society.

The study has some obvious limitations: it’s impossible to know whether the Emergency Medicaid data analyzed actually refers to undocumented immigrant mothers, and we don’t know for sure if those mothers actually applied for DACA. Also, we don’t know exactly why the children’s mental health improves. It could be because protecting a parent from deportation decreases stress and anxiety. But it could also be because DACA gave those parents a work authorization, which might have allowed them to find a higher-paying job, which in turn improves quality of life overall.

Hainmueller wants to analyze Medicaid data from other states to see if he can replicate the results he got for Oregon. He also wants to see if the same effects can be seen in older kids, like teenagers.

In the meantime, this is one piece of evidence that shows that immigration policies have important consequences. Hainmueller hopes the research can inform the current discussion around whether DACA should be phased out, changing the life of thousands of immigrants in the US — and that of their US citizen children.