Teleconference Transcript: Immigration and the American Worker

By George Borjas and Steven A. Camarota on April 12, 2013

Related Publications: Backgrounder


Steven Camarota, CIS Director of Research

George Borjas, Professor, Harvard University
Transcript by Federal News Service Washington, D.C.


STEVEN CAMAROTA: My name is Steven Camarota. I am director of research here at the Center for Immigration Studies. We are located in Washington, D.C. The center is a think tank that examines and critiques the impact of immigration on American society. In case any of you don’t know, our website is

Today, we are delighted to have on the phone with us Dr. George Borjas, who is joining us from Harvard University. He will be discussing a new report he wrote for the Center for Immigration Studies. That report is available at our website and it is entitled “Immigration and the American Worker.”

Dr. Borjas has been described by both the Wall Street Journal and Business Week magazine as the nation’s leading immigration economist. His new report summarizes and synthesizes the academic literature on the issue of immigration.

I think probably the most important point or one of the most important findings of the report is that immigration creates both losers and winners. This is an important point, but it is often not generally acknowledged, and certainly not in the current debate over immigration, that some Americans are made significantly worse off by immigration and at the same time some people are made better off.

With that, I will turn it over to Professor Borjas.

GEORGE BORJAS: OK. Hello everybody. Let me describe to you the birth of the report a little bit and tell you why I think it’s important. So all of you know that the debate over immigration has a few central questions in every single immigrant-receiving country, including the U.S. And one of those central questions is what happens to people already here in the labor market when immigrants come into the labor market?

The academic literature is quite extensive and for those of you who followed the debate over the years, you have probably become aware that the literature has become even more technical, more mathematical over time so that most non-economists right now have a really hard time trying to figure out what it is actually that economists are trying to say about this.

So my goal in writing this report was to try to provide an objective and a non-technical description of what it is that economists know as of right now, what do we need to assume to get these findings, and sort of summarize what it is that the labor market impact of immigration is in the U.S. There are other impacts of immigration, such as the fiscal impact of immigration, for example. Those are not in the report. So all the numbers in the report focus exclusively on what is happening to the economy through the fact that immigrants come into the U.S. and participate in the labor market and create competition and create benefits along the same time. And I’m trying to do this in a fair, and easy to explain way as possible, given that the literature has become much more complicated and technical over time.

Now, one reason why we want to know what the labor market impact is is because the same study of the labor market impact is necessary to answer two crucial questions. One is, what is the distributional impact? In other words, are there winners and losers? How much do the winners and the losers each lose or benefit?

And the second is the parameter that determines the distributional impact; in other words, once we know what the wage impact is, it’s a very simple step to go to the next stage, which is to calculate the benefits. So when one talks about costs and benefits in the economic model of immigration, one is not talking about two different things. One is talking about the same thing: What is the wage impact? Once we know that number, we can get both who loses and who gains and how much the economy gains on average.

Now, one way to do this is by basically looking at what’s happened over the last 50 years, basically, to the wages of skill groups in the U.S. By a skill group, for example, I mean something like a high school dropout in the early 20s, a high school graduate in the late 30s, and so on.

One can define skill groups and just track them over time and see what happened to them, and see what happened to their wage, and see what happened to the wage of immigrants either entered that group or did not enter into the group. Immigration is actually very bi-modal in the U.S. in the sense that we have a lot of highly skilled immigrants but we have many, many more low-skill immigrants. And there’s actually few immigrants in the middle of the skill distribution. So it affects particular groups at particular points in time.

If you look at the evolution of wages for specific groups over time, you will tend to find this sort of negative correlation between the wage growth of a group and how many immigrants into the group. And this sort of – that scattered diagram that’s in the report – and you can basically see it visually – that those groups that receive most immigrants are the groups where the wage grew the least over the last few years.

From that, one can then develop – one can then sort of write down a theory, an economic theory, which is very technical, trying to figure out what is the labor market impact on all the groups. In other words, if immigrants come into – suppose a lot of high school dropouts in the early 20s come in, what happens to the wage of high school dropouts in the early 20s, what happens to the wage of college graduates in the late 40s, and so on. See, there are all these class effects. That’s a very theoretical exercise which requires a lot of assumptions.

When you do the theoretical exercise, you tend to find that the main impact of immigration is through lower – relatively speaking, the wage of high school dropouts. Why? Because immigrants, by and large, tend to increase the supply of high school dropouts the most. So you have a huge supply increase at the very low end of the skill distribution and that translates into substantial wage impact of high school dropouts.

How big is the wage impact on high school dropouts? One can argue – one can develop slightly different assumptions, but it’s basically somewhere between two, three, four, five, six percentage points. I mean, it really all depends mostly on whether you look at it from a short-run point of view or a long-run point of view.

To imagine what I mean by that, just visualize a short run, and suppose that immigrants come in overnight tonight, land all over the United States, what happens tomorrow morning? That’s the short run. In the long run, the economy adjusts. All kinds of people adjust. People move around jobs, move around capital flows. If after every adjustment that takes place, has taken place, that’s what we call the long run. And that’s when the labor market impact on high-school dropouts is probably around 2 or 3 percent, depending on which study you sort of look at.

Now, there are some studies in the literature – one more thing I should add about the long run, short run. Nobody knows for a fact what the right metric should be because we don’t know how long the long run takes to actually appear in the U.S. It could be that the long run is achieved by tomorrow night, or next year, or the next decade or, like Keynes said, after we’ll all dead. We just do not have any information on how long those adjustments take.

So the way to interpret the data really should be in terms of bounds. The short run gives the higher – the upper end estimate of the wage impact. The long run gives the lower end estimate of the wage impact and it’s probably somewhere in the middle.

Now, there’s some studies that have been appearing recently that claim that due to various reasons, these wage impacts on the low skill are much lower and perhaps even zero, or maybe even positive. Those studies tend to depend on two arguments.

One is that there is a kind of complementarity between high school dropouts in the early 20s who are foreign born, and high school dropouts in the early 20s who are native born. In other words, even though they look alike to us, somehow they do things differently and they – the immigrant speaks native something, and within that skill group, people learn. That hypothesis made a big splash about five or six years ago. In the latest – in the public version of the paper, you actually find that the hypothesis sort of tends to be much weaker and, in fact, disappear. So that’s actually not that important.

What’s really replaced the hypothesis is what you mean by a low skill worker. And what we mean by low skill – and that actually matters quite a bit. And the kind of research that argues low impact is trying to define the low-skill workforce as a grouping of high school dropouts and high school graduates. If you view all these people sort of the same way, then the fact that you have a lot of low-skilled immigrants or high-school dropouts is important but doesn’t really increase supply of low-skill workers all that much because there are millions upon millions of natives who are high school graduates. So that sharp the size of the supply increase is not that big and that’s why you tend to get a small impact. As I discuss in the report, the data underlying that hypothesis is actually very, very weak.

Now, once you know the wage impact, you can then calculate the gains. It – there’s no doubt about the fact that immigration increases total GDP by a lot, in the order of $1.5, $1.6 trillion. It is also no doubt about the fact that most of that money goes to the immigrants themselves. After all, when they produce output, it gets counted into GDP, they have to get paid for that work. And what remains for natives is actually a very small number, something on the order of $30 (billion) to $40 billion.

Now, a lot of people cite that number, including the Bush CEA a few years ago. What people don’t tend to cite is that if you believe the model that leads to the number of $30 (billion) to $40 billion, that same model implies a huge distribution, on the order of like $400 billion. So that net number of $35 billion actually comes around from the fact that workers lose on $400 billion, and firms or users of immigrants gain another $400 whatever billion. And the difference is 35 billion (dollars). So the same model that generates the 35 billion (dollars) gains also generates the prediction of a huge redistribution of wealth on the order of $100 billion.

And that’s where the winners and losers argument comes from. The only way you get a net gain is if you have this redistribution going on. Some people win, some people lose. And it’s certainly the case that the winners win by more than the losers lose by focusing on the labor market impact.

One last thing I should say about undocumented immigration is that a lot of the wage impact that we see in the lower end of the skill distribution is actually due to undocumented immigrants. And the reason is that a lot of undocumented immigrants we see in this country are in fact high school dropouts. So a huge part of the supply increase at the very low end of the skill distribution is due to undocumented immigration.

So let me just conclude by basically saying what I think is the main lesson from economics. If you concentrate on the labor market, there is a net gain. But that net gain only arises because there’s a wage loss. It is the wage loss that creates incentives for people to adjust and to – and for the economy to expand in a way where winners win by a lot, losers lose by just a little below that and the net gain is on the order of $35 billion a year. And I’ll leave it at that.

MR. CAMAROTA: I want to thank everyone for signing in. And, again, I want to very much thank Professor Borjas for both writing the report and participating in the teleconference. And the whole report is available at our website. And there is an executive summary, of course, if you want to get just, you know, the main points.