Immigration and the Education of US-born Children

>> Steven Davis: How do immigrant students in the classroom affect the educational outcomes of US-born students? Stay tuned and find out.

>> Steven Davis: Welcome to Economics Applied, a Hoover Institution podcast. I'm Steven Davis, a senior fellow at the Hoover Institution. Joining me today in person are two highly accomplished researchers. Over to my left here we have Paola Sapienza, she's a senior fellow at the Hoover Institution and she co directs the JP Conti Initiative on immigration at the Hoover Institution.

Directly in front of me we have David Figlio, he's professor of economics and education at University of Rochester, and he's a visiting fellow at the Hoover Institution right now. Welcome to both of you, it's great to have you here.

>> Paola Sapienza: Great to be here.

>> David Figlio: Great to be here.

>> Steven Davis: So, you two are among five authors of an article recently published in the Review of Economic Studies, that's one of the top journals in the economics profession. And the goal of your study, as I understand it, is to assess how more or less exposure to immigrant children in the classroom affects the educational performance of US-born students.

So many parents, indeed many Americans, worry that the presence of immigrant children in the classroom might detract from from the quality of the education or the educational performance of American students. And I think that's a reasonable concern. But as you will explain, if it's not evident already to people listening.

It's actually quite challenging to confidently assess how the presence of immigrant children in the classroom affect the performance of US-born students. And your study is gonna try to address that question. So we're gonna get into the details a bit, but do you wanna say more first about just the motivation for your study?

What led you to write the study and, what was in your thinking?

>> Paola Sapienza: Yes, yes, thanks, Stephen. It's great to be here. As you said, there is a lot of interest about the impact of immigration in public education, and this is mostly because in some schools there is a high concentration of immigrant children.

And on one side the worry has always been lots of resources going to these incoming immigrant children, maybe English is a second language, the don't know the system and so on. And so we need remedial help. And that would take away resources from the US-born students, from the local students.

But on the other hand, we had a previous research, similar team combination of the authors, where we noted that, especially some group of immigrants, and there is differences. We can go into that later. They actually outperform in US schools controlling for the same socioeconomic status compared to US-born students.

So we wonder what that additional factors, which is a combination of attitudes, performance, and so forth, could compensate or could.

>> Steven Davis: Okay, so there's concerns that extra resources would flow to the immigrant children, take those resources away from Americans, the US-born children, and that might negatively affect their educational performance.

But you're pointing out, as I understand it, that kind of controlling for income and socioeconomic background and so on, immigrant students often perform very well in the classroom. And maybe that might somehow pull up the performance of the American students, that's the idea. Or the US-born students, okay.

>> Paola Sapienza: They also tend to behave, controlling for socioeconomic status better, they have less incidents. They set a good example. And so maybe there is less disruption in the class. It's not about, yeah.

>> Steven Davis: Okay, so there's these two competing forces that might be at work, and you've already been very clear to point out, and we will come back to it.

Of course, it depends on the nature of the immigrants we're looking at. And different immigrant groups, different groups of immigrant children might have very different characteristics with respect to their behavioral tendencies, their interest in academic subject, and so on. So we'll get into the particular setting of your study.

But that's basically the landscape or the reason there's a real question here. Have I got it right?

>> David Figlio: Yeah, and I mean, one of the things that we were trying to speak to is, we're by far not the first people to have studied this question. There've been hundreds and hundreds of papers written on the subject of how do immigrant kids affect native-born kids in the classroom.

But there's this fundamental challenge to actually studying that, if you look just cross sectionally, right? The classrooms with more immigrant kids in the classroom, native-born kids are doing worse. So it's easy for you to think just looking at that, that more immigrant kids in the classroom are making native-born kids do worse.

>> Steven Davis: Okay, so that's kind of, seems like sense. Well, what's wrong with taking that approach? Why doesn't that work?

>> David Figlio: What's wrong with that is that immigrant kids are not randomly dropped into schools, and native-born kids are not randomly dropped into schools.

>> Steven Davis: Okay, so we're not just randomly parachuting immigrant children more in some schools less than others?

>> David Figlio: Exactly.

>> Steven Davis: Okay.

>> David Figlio: Yeah, immigrant kids tend to end up, because immigrants in the United States are disproportionately relatively low income. They tend to be in neighborhoods with more low income native-born kids who tend to do worse in school. And likewise, native-born families sometimes see more immigrant kids in the neighborhood and decide they wanna move to another neighborhood or different school.

>> Steven Davis: Okay, so there's two different things going on, as I understand what you're saying. The immigrant children are non-randomly distributed across schools in a way that might affect average performance outcomes at those schools among the US-born students. But also, the parents of these US-born students might respond to the presence of more immigrants by taking some of their kids out of the schools.

And my guess, you can tell me if I'm wrong, it's largely the more affluent parents who would take their kids out of school. And well, it's a fact that the students of more affluent parents tend to perform better in schools other things equal. So there's two different sources of confusion here as to why you can't just take the simple minded approach and look to see whether schools with more immigrant children perform better or worse.

And your study, just to make sure we, before we get into the details, your study, I don't know if it's unique, but one of its strengths is that it's simultaneously addressing both of Of those challenges. Is that correct?

>> David Figlio: That's right, that's really the unique feature of our studies.

That's why your studies in a top journal. In a top journal. So for years people have been dealing with that first source of confusion, the fact that immigrant kids are not randomly just dropped into schools. And we're certainly not the first people to worry about the fact that native born kids are also not randomly assigned to schools either but nobody had the data to deal with it.

>> Steven Davis: Okay.

>> David Figlio: But we had the insight.

>> Steven Davis: Okay, I think we've laid out the main challenges to confident assessment here. Let's start to get into the details of your setting because as we've already said, the context is likely to matter in terms of the immigrant groups at issue and where they go to school and so on. So tell us about the setting for your study.

>> Paola Sapienza: We have two major advantages vis a vis the rest of the literature. The first one is we can see the students over time.

>> Steven Davis: Okay.

>> Paola Sapienza: So we're gonna have a very precise measure of exposure because we're gonna measure cumulative exposure over time.

That's gonna be a huge advantage. But the second one, which really address what David was saying, is we have birth information of each one of these students and so we can actually connect sibling within each other. And as you were pointing out, a key determinant, a key correlation with academic performance is socioeconomic status.

By comparing two siblings that come from the same family, same socioeconomic status, but one was exposed to more immigrants over time just because of the flow of immigrants affected one more than the other. By comparing the two of them, we can actually tease out what is really happening in terms of correlation without confounding effect like, socioeconomic status or this phenomenon of White House.

>> Steven Davis: So this is a key part of your empirical design that you know that families differ in many respects that are probably correlated with the educational performance of children. And so you're going to be comparing say two brothers or brother and sister within the same family that because they went to school at different times, happen to be exposed to different degrees of exposure to immigrant children in the classroom. Is that the idea?

>> Paola Sapienza: Yes, in fact, all our effect is gonna be measured exactly by comparing siblings.

>> Steven Davis: Okay, that's how you're kind of controlling or conditioning out in statistical language these differences across families which we know matter a lot, but which are kind of hard to get a handle on directly.

Okay, so where are we looking here? I think it's, it's the state of Florida, as I understand, it's K through 12 public schools, it's 2002 to 2012 I've got down here in my notes. And you've been working with the Florida data for a long time in many studies and that's linked to records on where these kids were born, right?

So we know who was born in the US, who was not, we follow them over time. This is the cumulative exposure measure that Paula mentioned. So that's the setting, right?

>> David Figlio: That's the setting, so Florida, third biggest state in the country.

>> Steven Davis: Lots of immigrants.

>> David Figlio: Lots of immigrants.

In fact, Florida has a very diverse set of immigrants. I mean, overwhelming majority of the immigrants in Florida are from Central and South America or from the Caribbean region. But there are also a good number of immigrants from Africa, from eastern and southern Asia, from Europe, for example.

And the Florida state administrative data are really very rich. So we know the country of origin of every single immigrant in the state through the school records. And thanks to the match with the birth records, we know whose siblings with whom. So for the US Born children, or I should say the Florida born children, we can match kids into families as well.

>> Steven Davis: Okay, and how should I think about the background of these immigrant children in terms of how affluent they are, whether they have good facility with English language when they come to school, the cultural views of their parents, whether their parents really aspire for them to succeed at school, and so on?

Because as we've already mentioned a couple times before, immigrant groups differ in these respects. Give us the profile or the distribution of the profiles for the immigrant students who are in your study.

>> David Figlio: Sure, so there are hundreds of thousands of immigrant kids, and they're very diverse in all of these different settings.

They're diverse in terms of their country of origin, they're diverse in terms of parental income and parental education. I'll mention a little bit, and I bet Paola will join in a little bit in terms of maybe some things about the aspirations of the families. Disproportionately, immigrants in Florida, just like in the rest of the United States, are relatively low income.

So I don't remember the precise number, but I'm going to ballpark estimate somewhere between 2/3 and 75% of the immigrant kids in our sample are eligible for subsidized school meals. So those are people at or near the poverty line. For example, the majority of the immigrant kids in the data are limited English proficient.

They come from households where different language other than English is the language spoken at home, overwhelmingly, Spanish, but not exclusively Spanish. So this actually gets into some of the concerns that Paolo was mentioning at the beginning. When people are worried about, what's gonna happen in classrooms with finite resources, only limited teacher attention, limited amount of money.

You have all these kids coming in overwhelmingly low income, overwhelmingly from non English speaking households with parents who have limited understanding of the U.S. educational or social systems. You might be worried that these are kids who would ordinarily do pretty poorly in school and maybe bring down the classroom.

>> Steven Davis: Okay, so that's a key point. So the particular setting you're looking at is not one that has an atypically rich or well educated set of parents for the immigrant backgrounds. These are exactly the kinds. The kind of immigrants that give rise to concerns about resources being diverted to immigrant children rather than US born children.

So, that's an important feature of the setting for your study.

>> David Figlio: Yeah, the only difference I would say, the biggest difference between Florida and, say, California or Texas, is that Florida has a smaller fraction of immigrants from Mexico, for example, than California and Texas do. But in terms of things like language proficiency, socioeconomic status, there's not large differences between.

>> Steven Davis: Okay.

>> Paola Sapienza: But you don't want to imagine, since we are here in the Silicon Valley, the average public school in Palo Alto that has a lot of immigrants, but not from the same background.

>> Steven Davis: Children of people here on H-1B visas to work for tech firms. That's not the predominant immigrant child in your study.

>> David Figlio: That's correct.

>> Steven Davis: I think, as I think about your study and how it relates to concerns about immigration, I think there's one other important aspect of your setting that is important to emphasize, and that is your study and your setting. This is not a criticism, this is just an observation.

It is not, and you can tell me if I've got it wrong, but it's not designed to assess the impact of a sudden large influx of immigration and immigrant children on the school system. Rather, you're looking at a more stable situation. But when you look across school districts and across classrooms, there's wide variation in exposure to immigrant children.

So, have I got that right?

>> Paola Sapienza: Yeah, you're right, I mean, we have variation over time, but we don't think about jumps, we have a lot of cross sectional variation. We have some schools with 60% immigrants, and then we have a lot of variation, but we don't have, with some exception, I think David studied refugees specifically, resettlement of refugees.

And there the result could be certainly different, and maybe David.

>> Steven Davis: We think about what's happened in the United States in the last four years. It can certainly happen that certain school districts get overwhelmed with an influx, an often unanticipated influx of immigrant children, that's not the setting here.

And again, that's not a criticism cuz your paper is speaking to the question. Well, over the longer term, if we have, say, twice as many immigrants in a controlled flow or half as many as we've had in the past, how might we affect that to affect the performance of US born children? That's how I think of your study.

>> David Figlio: I agree with one small caveat. So, it's absolutely the case that this is not a setting where there's massive influx. However, the cross sibling comparison that we're doing means that just for dumb luck, right? It could be that 30% of the students in the third grade happen to be foreign born and 10% of the kids in the fourth grade happen to be foreign born.

And that type of comparison actually happens surprisingly frequently. And so, you could imagine a case in which a third grade is actually disproportionately overwhelmed with immigration.

>> Steven Davis: Yeah, I take that point. I guess what what I'm trying to point out is, the TV image of a school that has suddenly doubled in size and it has a bunch of temporary shelters on the outside where kids are learning, that's not the setting here.

>> David Figlio: No, yeah.

>> Steven Davis: And so, we should keep that in mind as we think about the study. So, you've given us the basic idea behind your empirical design, but maybe you wanna flesh that out more, maybe you just wanna jump into what you found. What did you find?

>> Paola Sapienza: So, we look at, remember, at the effect on academic performance of US born as a function of the cumulative exposure over time from the time they enter the school, typically kindergarten, to the time we're observing. So, think about a fourth grader. We're gonna look at how much exposure this fourth grader had from the time entered in school, in kindergarten to fourth grade.

And compare this academic performance to the academic performance controlling for the grade and the year of his or her sibling at the time, who is in a different grade and has been exposed in a different way.

>> Steven Davis: And because you have many, many thousands of these sibling pairs, you have a large dataset, you can look at the average outcome across.

>> Paola Sapienza: And we look at standardized math scores, standardized reading scores, and then we look at this variable, which is disciplinary incidence, which are large enough that they're being noted in the.

>> Steven Davis: You're suspended from school or something like that, or.

>> Paola Sapienza: Right.

>> Steven Davis: Okay.

>> Paola Sapienza: So, we look at these three variables and what we see is that when we control for the environment, the school, but we don't compare just directly to siblings, we find what the rest of the literature had found.

So, a negative correlation between size of the immigrant exposure.

>> Steven Davis: And this is where you're not controlling for the influence of family.

>> Paola Sapienza: Exactly, once we control for the family, we see that the result flips. It's slightly larger in math than in reading.

>> Steven Davis: Flips in meaning what exactly?

>> Paola Sapienza: The correlation between the size of the exposure to the cumulative exposure to immigrants, and the academic performance, is gonna become positive, so more.

>> Steven Davis: Okay, so it looks like the immigrant children are actually pulling up the performance of the American children in the same classroom.

>> Paola Sapienza: Yeah, we are a little bit more cautious and we call it a correlation.

Now, we haven't proven the causation because we observe that the two things are positively correlated. In economics, there is a lot of debate about correlation.

>> Steven Davis: Okay, but this seems like better evidence than most of the evidence that we have to speak to the causal effects of immigrant children in the classroom.

>> Paola Sapienza: Especially because we're looking over time, we're really measuring the lagged the exposure.

>> Steven Davis: And again, your main result, I just wanna make sure this is very clear to everybody. Your main result, where you say it looks, and I'm gonna use my language, which you guys might find a little too strong.

It looks as if the presence of more immigrant children in the classroom, pulls up the performance of the US born students, especially in math. And that's the result that comes out when you're looking at these between sibling comparisons. And it overturns the more conventional results that kinda confounds that effect with differences across families, in how likely their kids are to perform well in the classroom anyway.

So that's really a key innovation in what you're doing.

>> Paola Sapienza: Suggesting that indeed this concept of white flight, so this idea that if you're affluent and you can, you move away. And so there is a correlation between the fraction of students in a school and the socioeconomic status of the families is indeed one of the main mechanisms to which the traditional result has been found.

>> Steven Davis: Okay, just correct me if I'm wrong, but my understanding is your study is consistent with this idea of white flight, but maybe it's perhaps more accurate to think of flight of affluent.

>> Paola Sapienza: That's correct, yeah.

>> Steven Davis: It's not just whites. The flight of the affluent, that is a phenomenon that's been uncovered in many studies. Is that correct?

>> David Figlio: That's right.

>> Steven Davis: Okay, so we know this is happening, and we know other things equal, that, that kind of flight of the affluent will tend to pull down the performance, the average educational performance of students in the schools from which they are fleeing. And it looks like they're more likely to flee when there's more immigrants.

I wanna just digress for a minute here cuz that's an interesting phenomenon all by itself, and there's many potential explanations for it. It might be that there's bigotry among these affluent people, but it doesn't have to be that. It could be that they think that the educational performance of their students will be undermined by many immigrants.

Your study suggests otherwise. But here's what I wanted to get to. And I don't think you talk about this in the study, but maybe I missed it. Is it that there's something else about the quality of the educational outcomes that is being affected by the presence of immigrants that is not captured in your particular performance measures?

In other words, is there a rational reason for the affluent to flee? Why are they fleeing? And is there a rationale?

>> Paola Sapienza: Well, there could be many reasons because it's not necessarily just about education. Remember, most parents send their kids to the school nearby, and so is the amenities.

As you become more affluent, maybe you like a different type of neighborhood. And that doesn't mean that you don't like, you may like a neighborhood that has more parks or some other things. And so we don't know what, we call it flight, but in reality, we don't see people leaving. We just see fewer people who are affluent in the school where there are more immigrants.

>> Steven Davis: Instead of flight, you could just call it sorting.

>> Paola Sapienza: Exactly.

>> Steven Davis: We know people sort on many, many characteristics. The homophily idea that people tend to be attracted to that which is like them. And it involves amenities. And the amenities can be like what kinda restaurant do you like?

>> Paola Sapienza: Yeah, but the other thing is, if you have means, you can choose. So what we study, at one point we separate the group of US-born students between the affluent and non-affluent and minority students.

And what we notice is that while the result really flips in the first case, so we see this negative correlation that becomes positive later on, when we look at the non-affluent, we see that the result is always positive, why? Because there isn't distorting, as you said.

>> Steven Davis: Okay, that's a little subtle, but what you're, so let me try to restate it. I'm not sure I actually understand as I try to restate, this white flight phenomenon is obviously gonna be concentrated among the more affluent. So the non-affluent aren't leaving.

>> Paola Sapienza: Yeah.

>> Steven Davis: Okay.

>> Paola Sapienza: That's very simple, yeah.

>> Steven Davis: Yeah, okay. So if you focus on the non-affluent to start with, then even the simpler approach used in the previous literature tends to find more positive effects which is consistent with the so-called selection story that you're telling.

Okay, so that's interesting, but let me just ask you a related question. Does your study say anything about whether your best estimate of a causal effect of immigrant children on US-born students varies with the socioeconomic background of the US-born students? So you hinted earlier these disciplinary issues, behavioral issues.

Well, you can tell me if you disagree, but I think it's kinda well understood that behavioral difficulties among students in school are more common for students who come from lower income, less advantaged backgrounds. And so you might think they would especially benefit from being exposed to well-behaved immigrant groups.

Does that come out of your study or can you speak to that?

>> David Figlio: Absolutely, so in fact we do stratify our US-born families along two different dimensions. So we look racial dimensions, so minority versus non-minority. And then also comparatively affluent to comparatively less affluent native born families or students from families that are relatively affluent or relatively less affluent.

Importantly, we always find positive or at least non-negative relationships between more immigrant exposure and US-born kids test scores and behavior. Even for the affluent families and for the white families, for example. But our results are particularly pronounced and positive for exactly the groups of students that you just suggested, Steve.

>> Steven Davis: Okay.

>> David Figlio: Those students who are in schools that tend to have bigger behavioral challenges, for example.

>> Paola Sapienza: Twice as large.

>> Steven Davis: Twice as large, okay, big difference. So in simple terms, again, maybe you won't be comfortable with this summary, but in simple terms, the segments of American families, the types of American families for which behavioral issues are most pronounced, showing up as disciplinary infractions in school. It's those discipline-challenged students that seem to benefit most from exposure to the immigrant children and maybe because the immigrant children, on average are better behaved.

>> David Figlio: Yeah.

>> Steven Davis: Is that a fair summary?

>> Paola Sapienza: Yeah, we actually try to look at the mechanism.

>> Steven Davis: Okay.

>> Paola Sapienza: So we can go the reason we can look at them later, but your summary is perfect.

The question is why? And so we try to look at several mechanisms, and one of them is looking at absolute performance of these immigrants. So the question is, is it because more immigrants maybe is correlated with higher absolute performance of the immigrant groups? The answer is not, by the way, is not the case.

We see when we try to measure, we add absolute performance and see what the correlation is with the performance of the US born, it's always positive for every group. But our measure of the presence of immigrants remains significant, mostly for the non affluent and the minority kids. So that absolute performance doesn't seem to explain, relative performance is an interesting one.

You seem to be going into that direction that we spend a lot of time thinking about that. And this is complicated from an econometric point of view, but let me give you

>> Steven Davis: What do you mean by relative performance?

>> Paola Sapienza: Yeah, so let me give you a simple way, think about, we divide the schools in four groups. Think about the schools where there is a predominantly, non affluent kids and affluent one. For the moment, let's think about these two groups. And let's look at the characteristics of the immigrants in these two schools, and how do they compare with the average performance of the kids in those schools?

And interesting, remember what David said early on also the immigrants self select, right? So, it's possible that immigrants are different going to these schools depending on where they are. And so what we find is that, in the schools that are predominantly dominated by non affluent kids, when you compare the average non affluent kids with the immigrants, the one that they are facing, the immigrants do better in test scores and along the disciplinary incident measure.

And that could be one of the reasons, so it's relatively speaking, it's not the case that immigrants are over-performing all the US born students. But maybe they pull up the students that are less performing than them. When we look at the affluent and we make the same comparison, typically the affluent US born students are always over-performing vis a vis the immigrants, okay?

And those are the one that are not affected that much positively, but not affected negatively. So it's not a clear relationship.

>> Steven Davis: It's a somewhat complicated story. But as I understand it, if you put a bunch of immigrants into a mostly affluent school first, those immigrants might be more affluent than the average immigrants.

They don't end up outperforming the affluent US born students, so there's not like this higher bar being set or this example that you hadn't achieved before. But when you put a bunch of immigrant students, even poor immigrant students into a less affluent school with a lot of less affluent US born students, the immigrants performed very well.

Presumably the US born students or their parents see that and they say, well, maybe there's room to do better here. Again, I realize your study isn't dis positive on exactly why, but one possibility is you just see by example. One other possibility is that there are fewer disciplinary incidents, because there is many more immigrants and the teacher can teach better.

>> Paola Sapienza: This could be a sympathetic.

>> Steven Davis: I sympathize with that. I spent some years in schools with a fair number of discipline challenged classmates, and they were extremely disruptive to the classroom. So that resonates with my own personal experience, but I wanted to ask you about a different issue. 

So we've talked about, there's definitely some evidence in your study of different effects across US born students, who are more versus less affluent. It's the less affluent one, who are most positively affected by the immigrant groups. But going back to something we said earlier, the immigrant groups themselves can also differ a lot.

Do you have the capacity in your study to measure, to say, look at the intensity of parental aspirations for educational achievement among students. That's something that differs across cultures, across ethnicities and nationalities. Can you measure that and then see whether that actually matters for the performance of the, not just the immigrant children, but the US born students? Can you speak to that?

>> David Figlio: So we can get part of the way there, so Paola was mentioning earlier that. So our first joint paper together was actually studied just about immigrants themselves. We were curious about, we and two other co authors were curious about cultural differences across countries.

And so there are scholars who have characterized different countries, based on the level of long term orientation.

>> Steven Davis: Long term orientation, okay, that's a good way to say it.

>> David Figlio: Which is getting kind of what you were describing. So our first paper together that was published in American economics journal Economic Policy, showed that immigrant kids in Florida from cultures of origin with higher long term orientation did better in the classroom.

Turns out that again, as Paula mentioned, this doesn't explain everything, but it does explain a portion of our findings. That if you have more peers, if you're a native US born kid, with more immigrant peers from higher long term orientation cultures, you do better than if you had more immigrant peers from lower long term orientation cultures.

So that's a smoking gun.

>> Steven Davis: This resonates with common sense.

>> Paola Sapienza: Actually it's even a bit like, if we divide between the median long term orientation below, the one that have the immigrants, the fraction of immigrants that have low long temporal orientation, they have zero effect. So this is a hint that, and remember we are at Stanford, this is the famous story with marshmallow.

With the marshmallow, maybe, so how long, in that study, what they did, they said, okay, how long the kids can can wait for the immediate reward, and they correlate it later on with activities.

>> Steven Davis: If you're going beyond the the marshmallow study says, as I recall, kids who demonstrate greater capacity to wait for gratification do better over the longer term.

You're saying now, or at least your evidence suggests that it's better for your kid to be in a classroom with lots of other kids, who can wait to eat the marshmallow, right? And I think most parents will say, well, duh, of course I want my kid to be with other kids who have good habits.

I don't want not everybody, but most parents will think I'd like my kid to have friends, who help cultivate a forward looking outlook on life, good habits, good behavior and so on. Because these things tend to spread across peers within the classroom, right? That's I realize you don't fully prove that, but that's consistent.

And I think it's also relevant to thinking about immigration policy in the United States. When we think about, there's obvious concerns about, we don't wanna allow criminal elements to enter the United States. Some people have concerns about low income people entering the United States and pulling down employment and wage opportunities for US-born students.

But one of the things I take from your studies is, well, first, immigrant students have some behavioral characteristics that have positive spillovers on US-born students and that's beneficial to society. But also, when we think about what makes an immigrant who's likely to have positive effects on US society, we shouldn't just be looking at income.

We should be looking at these things like long term orientation and whether that's something they wanna cultivate in their children and whether that will affect the attitudes and behaviors and outcomes of US-born children. Yeah, to some extent this result is consistent with the work of Abramin that looked at the second generation immigrants and why they do much better controlling for initial conditions than the US-born.

>> Paola Sapienza: And so maybe there is this grit and of course, one of the key things that we need always to take into account is that typically immigrants are selected as well, right? Just we should not ignore that, they've decided to do a long journey, sometime a very long one, and with some uncertainty, they uproot themselves.

And so maybe that really is a measure of willingness to invest in the future.

>> Steven Davis: Right, yes. It takes a lot of gumption and dedication, especially as a parent, to uproot yourself, put yourself in another society that you're not very familiar with in the hopes that your child will have a better life.

And that's a powerful motivator. And I agree with you fully, we should think about that when we're thinking about, well, who are the people who want to come here and which ones do we want?

>> Paola Sapienza: This transmission to the children must be really powerful because the kids observe their parents going through these, sometimes sacrificing.

And so maybe they get something out of it, we hope.

>> Steven Davis: That's an old story about the immigrant child feeling guilt about the sacrifice their parents made. And by gosh, I've got to succeed, I've got to apply myself because my parents did a lot for me, I need to show them some reward.

David, do you wanna-

>> David Figlio: And it is important to note too, and this actually gets to this other paper of mine with one of our co-authors, Umut Ozek that was actually looking at this influx of Haitian refugees into Florida too. Even in the paper we're talking about today where we think about immigrants from very low long-term orientation cultures.

As well as then Haitian earthquake induced migrants, we never, in that paper with Umut about the Haitian migrants, in this paper we never find anything negative. So it's only a matter of degree of going from zero to positive estimated relationships with-

>> Steven Davis: Right, well, so even going back to the earlier point about immigrants are selected.

Even the Haitian immigrants, you're not getting a random sample of Haitians, you're getting the ones who had sufficient motivation and push and pull that they actually got here. It's not so easy if you're a poor Haitian to do that.

>> David Figlio: Precisely.

>> Steven Davis: Okay, so I think that's an important thing to keep in mind about immigrants.

And why, by the way, these kind of actual changes we're looking at in the flow of immigrants into the United States, we're talking that's kind of marginal changes in some sense within the scope of the data we have. Doesn't necessarily say it would be a good thing to allow anyone to come to the United States, because there might be a very different selection process for those people.

So, are there other things you wanna say about how I should think about the larger lessons of your study for US Immigration Policy. Is there more you wanna say on that question? I'm inviting you now to tell us what are the broader messages, we talked about some of them already, but are there additional broad messages?

>> Paola Sapienza: One important broad message is, when you allow immigrants to come in and you want to measure their contribution to some extent, family is very, very important. And so to some extent just we often hear is, but are they positive net contributor and so on. And longitudinal research should allow us to think about not just the first generation, but also the second generation.

Because often the second generation is the one that really contributes enormously to society.

>> Steven Davis: Yeah, you made that point a couple times and I agree with it and I wanna reinforce it. You see this, for example, among inventors. We have a lot of immigrant inventors, but we also have a lot of second generation immigrants who were very much overrepresented among people who have patents, people who go into STEM fields.

Who we think based on a number of studies contribute disproportionately to the economic and social success of the United States. And I think that's a really important message to keep in mind and just to put my own cards on the table, I think it's one of the sources of American greatness.

We can use that term over many, many decades, even centuries now, is this positive selection of people who really have a strong desire, through their own efforts and through their children, to contribute to society and to advance upward mobility. It's that immigration refueling which has played a big role in the success of the United States, in my judgment.

>> David Figlio: I'll add one more thing to underscore, which is that language is not as important as I think many people may think or fear. For one thing, immigrant kids learn English very rapidly. But even when you're looking at immigrant kids in the classroom who come from home, say Spanish or Mandarin is the only language spoken.

And who come to the United States and they themselves don't speak any English yet, there's not this negative spillover. It's amazing how quickly kids can pick up language, and it's amazing how easily kids can communicate with one another. So that's, I guess the other thing I'd like to underscore is I'll echo your American greatness point here.

The English language and American English is extraordinary in terms of how we absorb all sorts of concepts and ideas and words from all different types of cultures. And this is something I think it's nice to underscore here as well.

>> Steven Davis: Okay, I think this is a good note to end it on, I wanna thank you both very much for the paper you wrote, I enjoyed reading it, and for a great conversation.

>> Paola Sapienza: Thank you, Steve.

>> Steven Davis: And I'll see you around in the hallways.

>> David Figlio: Thanks, Steve.

>> Steven Davis: Okay, thanks so much.