Increasing the Supply of Labor
Through Immigration

Panel Discussion Transcript

May 4, 2004
National Press Club
Washington, D.C.

Moderator:
Steven Camarota, Director of Research, Center for Immigration Studies

Panelists:
George Borjas, Robert W. Shriver Professor of Economics and Social Policy, The John F. Kennedy School of Government, Harvard University

Jared Bernstein, The Economic Policy Institute

Robert Lerman, Senior Fellow, Labor and Social Policy, The Urban Institute

STEVEN CAMAROTA: I’m Steven Camarota, director of research at the Center for Immigration Studies. The reason we’ve gathered you all here today is to discuss a new report that the Center is putting out, published by Harvard economist George Borjas.

The report is an important piece of research because it examines what I think we’d all agree is one of the central issues in the immigration debate – how does it affect American workers; how much, if at all; and how does it do that?

When I came in to work today and I drove by the AFL-CIO’s headquarters on 16th Street I saw a large sign that said, “America needs good jobs.” It seems to me that one of the central attributes of a good job is how much it pays. And this report is so important because it addresses probably one of the most central features of jobs: in addition to having one, what kind of pay does it offer you?

Now, Dr. Borjas, for those who are familiar with the immigration debate, needs no introduction, but let me try anyway. He is the Robert W. Shriver Professor of Economics and Social Policy at the John F. Kennedy School of Government at Harvard University. He is also a research associate at the National Bureau of Economic Research.

Dr. Borjas received his Ph.D. at Columbia University in Economics in 1975, and as I said, he has written, of course, extensively on the labor market impact of immigration. And he’s the author of several books on the topic, including most recently, “Heaven’s Door: Immigration Policy and the American Economy.” In fact, Dr. Borjas has been described by both The Wall Street Journal and Business Week Magazine as, quote, “the nation’s leading immigration economist.” In addition to all that, he still finds time also to serve on the board of directors of the Center for Immigration Studies.

Now, joining him to discuss his paper, to my far left, is Jared Bernstein of the Economic Policy Institute here in Washington. Dr. Bernstein joined the Economic Policy Institute in 1992. His areas of research include income and wage equality, technology’s impact on wages and employment, and more generally, the low-wage labor market and poverty. Between 1995 and 1996, he held the post of deputy chief economist at the U.S. Department of Labor, and he has published extensively in popular and economic journals on the U.S. labor market. Dr. Bernstein holds a Ph.D. in economics from Columbia University.

Joining us to our right is Robert Lerman. Dr. Lerman is a senior fellow in labor and social policy at the Urban Institute here in Washington. Prior to becoming a senior fellow, he was director of the institute’s Labor and Social Policy Center from 1995 to 2003. He holds a doctorate in economics from the Massachusetts Institute of Technology and is the author of more than 100 articles, monographs, reports, and conference papers. He has served on a variety of panels and commissions, including the National Academy of Science, and the board of the National Fatherhood Initiative.

With that, I’d like to turn it over to Dr. Borjas.

GEORGE BORJAS: Thank you very much, Steve. I want to stand up and do a little presentation on the overhead. The research I’m talking about today is basically addressing a very simple question, which is: Do immigrants alter the employment opportunities of native workers?

The answer to that, given by economic theorists is actually very simple. What I did when I began this research a couple of years ago was to go back to Paul Samuelson’s textbook of economics, and back to the 1964 edition for a particular reason, and a view which I’ll mention in a minute. If you read what Samuelson said back in 1964 he basically said, after World War I, laws were passed that limited immigration. Only a trickle of immigrants were admitted since then—speaking in 1964—and by keeping labor supply down, immigration policy tends to keep wages high.

Now, the interesting thing about the date is this was one year before the enactment of the 1965 amendment to the Immigration Nationality Act, the one policy that sparked the resurgence of large-scale immigration to the U.S. So he’s basically talking about the mirror image of where we are today. Economic theory predicts that when immigration policy is restrictive, it keeps wages up. It also predicts a mirror image of that, which is when labor supply goes up, like we’ve had in the last 30 years, the wages should go down.

So the prediction of the theory is actually very simple. Remarkably, it’s been incredibly difficult to find that empirically. And part of the reason has been the way that these things have been done in the past. Let me start with a well-known fact that most of you will be aware of, which is that immigrants in the U.S. tend to cluster in a very small number of areas. As you can see from the data that I have here, almost 40 percent of immigrants live in four cities. That clustering, that very extensive geographic clustering, has motivated many, many economists to try to estimate the labor market’s impact on immigration by comparing natives and – across different cities basically. It is that approach that has led to the current conventional wisdom that immigrants seem to have little impact on the employment of the communities of native workers.

In that approach, what is being done basically is that -- the physical study will say, well, if it’s really true, as Samuelson sort of implied, that when immigration goes up, wages go down, we should have been able to compare native wages, say in San Diego, a city with many immigrants, with a city like Pittsburgh, which is a city with relatively few immigrants, and holding everything else constant, it should be evident that wages for native workers in San Diego are lower than wages for native workers in Pittsburgh. And this is basically the approach that’s being used in much of the literature, to see if indeed the prediction of the theory is correct.

When people have done that, and many people have done that, the result has been pretty striking. And to sort of show you the net result, I quote here the 1997 National Academy of Sciences study that basically looked at that whole literature and concluded that overall it seems as if, if you compare wages in different cities—heavy immigrant cities with light immigration cities—you don’t tend to find much of an impact. The problem with that conclusion, which I point out in the report, is there are several problems, actually but there are two key problems. One is that immigrants do not tend to be randomly thrown into cities. They will tend to take particular cities and avoid other cities.

So, for example, suppose that immigrants chose to immigrate to cities that have high wages, that have thriving economies. Well, that will tend to build in a positive correlation that might outweigh whatever negative impact there might have been. So that’s one potential problem. Equally important is the possibility that when immigrants go into a city, natives respond. And how would natives respond to that? Well, just think about it. Suppose that immigrants going to San Diego – (unintelligible) – all the wages. Well, then the natives in San Diego will say, gee, I might as well go to Pittsburgh. And then immigration flows out of San Diego to Pittsburgh, and thereby will tend to equilibrate the labor market across the national economy. So you would tend to find very little difference between San Diego and Pittsburgh when you look at these two cities

The point of all this is that there has been a slow recognition by economists that perhaps the city is not geographic unit one should be looking at to measure the labor market impact of immigration. Because there are all these equilibrating flows, one might want to go to a higher level, like the national level. And that is precisely what I have done in this study.

I basically moved the analysis to try to look at what is the impact of immigration on the national wage structure? And the reason that is important is because by switching to the national level and by realizing the fact that immigration is not—that immigrants don’t have the same age distribution as natives—you can actually make a lot of headway in trying to estimate the labor market impact of immigration.

To give you an idea: suppose that immigrants in any given year tend to be quite young. Suppose that today we’re receiving a lot of young high school dropouts in the country. Well, that implies that it will be young high school dropout natives who will be most affected, not all the high school dropout natives. And using that, and by looking at these trends at the national level, then you can actually see very clearly what the impact of immigration is on the wage structure.

So let me show you some of the data [that] underline this approach. The data I’ve used, by the way, just to make everything clear, are data from the 1960 through 2000 census. So we’re looking at 40 years of Census data. These data contain millions of workers, so we actually have working characteristics from millions of native workers over the last 40 years, and that is the underlying data that I will be reporting the results on.

And to show you how the immigration has changed the labor market for some workers, look at what happened to the labor market for high school dropouts over the last 40 years. Here I have years of experience, basically age: how long have you been in the labor market? And on the vertical [graph] I have the immigrant share—the percent of the workforce in that year in that experience group that is foreign born. [Editor's Note: Please refer to the Report for tables.]

What this basically indicates, for example, right here, is in 2000 roughly 50 percent of all workers who were between—with 10 to 20 years of experience, roughly 25 to 35 years old—half of those workers were foreign born. In other words, immigration increased the supply of young high school dropouts substantially in 2000. If you look in previous years you can basically see the pattern, but in some years immigration didn’t have that big of an impact on the younger workforce. They had more of an impact on the older work force. So, depending on the year, and depending on education level obviously—this is for high school dropouts—it actually does matter; it’s important to determine which particular group of native-born high school dropouts is most affected.

And the same thing is true at the other extreme of the education distribution, which is college graduates. You can sort of see here that in 2000 if you were a younger college graduate you would have been in the labor market where 15 percent of the workforce is foreign born. In other words, immigrants came in and increased substantially the number of workers in that particular skill group. In other years, it would have been the older college graduates who would have been most affected by that.

That means that if you look at data from 1960 through 2000, you can then disentangle the effect of what happened to the wages of particular workers, as in some years it is a particular group of, say, highly educated workers affected. In others years, another group, you might be able to see something about what the impact of immigration is at the national level. That’s exactly what this report does.

And just to summarize the conclusions, this is basically one of the key numbers in the report. By looking at the actual earnings of workers over the last 40 years and trying to see what happens to the earnings of a particular age group or a particular type of education—workers at a given point in time—you can basically discover that immigration increased the workforce by 10 percent. And by that I mean the following: suppose there are a million workers who are, say, young and are high school graduates. And suppose that increased by 10 percent so that there would be 100,000 more people— 100,000 immigrants, in other words. If immigration increased the number of workers in that skill group by 10 percent, the weekly earnings of that group will go down by almost 4 percent.

Now, that is not a small impact; it’s not a huge impact either. It’s sort of a normal impact consistent with what economists have long suspected about the response of wages to supply. So it’s really not out of the ordinary in terms of what economists have long known about the structure of demand for labor in the U.S. economy. Every 10 percent increase in the number of workers in a skill group will tend to lower wages by between 3 and 4 percent. And that’s sort of the net result that I want you to remember from this study. It’s sort of a stylized fact. That is an important statistic for those of us who care about labor market impacts of supply shocks.

Now, that 3 or 4 percent change is what happens to their own skill group. In other words, what happens to high school graduates when more high school graduates come in? The fact of the matter is that when immigrants come in they also affect other skill groups. So, for example, you’re going to have a huge influx of high school dropouts. Well, that will affect the wage of other people in the economy—of college graduates, for example. And what the report shows is that these cross effects tend to be very, very, very small. In other words, they’re nothing like on the order of the 3 or 4 percent decline that I’ve shown you earlier. So is entirely true that when young high school dropouts come into the labor market, college graduates came, but they gained very little: only half of one percent.

So, if you want to get a complete picture of what these numbers mean for the wage effect of immigration, what I did in the report was actually to simulate a particular supply shock. And what I simulated was: what happened to the labor market as a result of the fact that over the last 20 years, roughly speaking, immigrants came in? So what I did was to say, what happened as a result of the 1980-2000 immigrant influx?

Now, let me describe that in a little more detail. Start the world up in 1980. In 1980, the U.S.—everybody in the U.S. at that point in time is a native. So that’s the beginning world. What happened is a result of the fact that over the next 20 years we’ve admitted X million of immigrants? You take the statistics I showed you earlier and this is basically what happens. For the typical native man in the labor market, the typical person present in the U.S. back in 1980, the wages went down by 3.7 percent, even if they’re accounting for all these cross effects. But if you look by education group, you can basically see that some groups lost a lot more and some groups lost a lot less.

Why is that? Because immigration, as many of you might know, tends to be very bimodal in a sense. Immigration, to a large extent, is either composed of a lot of low-skill workers or of high-skill workers. There are very few immigrants sort of in the middle of the distribution. Ok? As a result of that, at the very bottom end of the distribution, you find that immigrants lower the wage of high school dropouts, and by that I mean the immigrants who came in in the last 20 years, 82,000, lowered the wage of high school dropouts by 7 percent. They lowered the wages of college graduates by almost 4 percent. But for high school graduates and for workers with some college, there was a very small impact, only about 2 percent. And again, that reflects the fact that a lot of immigrants tend to be highly concentrated either in the low-skill or in the high-skill group, not really in the middle of the distribution.

Now, an important thing that one can do with these statistics is to look at the impact of immigrants on different groups of native workers. This 3.7 percent, this number I just showed you, is for all native workers in the U.S. as of 1980. But it turns out if you look by race you see very striking differences. And the reason is that because immigrants tend to be disproportionately low-skill; it is low-skill natives who lose the most. And it’s when you look at by race that you tend to see groups that have a lot of low-skill workers being affected the most adversely by immigration.

So, for example, if you look at the data, you will find that for the typical white, native worker in the U.S., that immigrant influx, the ’80 through 2000 influx, only lowered the wage by 3.5 percentage points. For Asians it was even lower, only by 3 percentage points. But for blacks and Hispanics, you see the impact being much greater. Why is that? Because black and Hispanic natives tend to have disproportionately larger numbers of low-skill workers, who are the workers competing directly with the low-skill immigrants who were admitted to the U.S. in the ’80 through 2000 period.

Now, let me conclude by doing something a little counterfactual in terms of the Bush proposal that was proposed about three months ago. As you will recall, President Bush proposed to legalize, or provide some mechanism for legalizing illegal immigrants in the U.S. already, and also providing some kind of guest worker program for people now abroad who want to enter the U.S.

Now, if you believe the results that I’ve just proposed before you, you would have to conclude that any kind of expansion of visas in that fashion would have a really substantial impact on the labor market. And by that I mean if one were to enact, as written, the proposal that we are now going to admit a lot of foreign workers as guest workers into the U.S.—now, in the proposal itself the numbers aren’t specified, but if it’s a large number, given this 3-4 percentage wage drop for every 10 percent increase in supply—we could be talking about a really radical rearrangement of labor market opportunities for native workers in the U.S. Ok?

In any case, what I did, just to show you the impact of illegal immigration, is the following: We don’t know how many, who, in the Census data that I’m using in the report, we don’t know who are the illegal immigrants. So what I just did was the following: Let’s suppose there had been no Mexican immigration over the last 20 years—in other words between 1980 and 2000. What would you have seen as a result of that? Now, clearly many of these Mexican immigrants that I’m taking out of the data, of the analysis right now, are legal immigrants. And in fact, about half are. If you look at the data of how many Mexican immigrants are in the U.S., it turns out roughly there were 9 million or so of them in 2000. And we know from INS predictions, or projections, that about half of those tend to be illegal immigrants.

So I’m just saying, let’s just suppose there had been no Mexican immigration of any type. What would we have seen? We would have seen that for college graduates, the impact would have been exactly the same: 3.6 percent. In other words, since so few Mexican immigrants are college graduates, we would not have seen any kind of shift on the impact on college graduates. But for high school dropouts, basically Mexican immigration explains for us the whole impact. Since Mexican immigration is about half illegal, about half of the impact that we see is actually due to legal immigration. So that will be one statistic to remember: roughly speaking, illegal immigration of Mexican immigrants in the last for 20 years or so accounted for about 3.5 percent drop in the wages of high school dropouts.

Now, whether the president’s proposal is enacted or not, this impact is already here. The illegal aliens—the illegal immigrants are already admitted. Therefore, their labor market impact is said and done with. The real concern over the Bush proposal will be what would happen in the future if in fact this guest worker program goes into effect as proposed, because as I said before, if it goes into effect as proposed, what that basically does is to open up many, many more jobs in the United States to competition from abroad. And that could have a really substantial labor market impact.

Now, let me conclude by basically re-stating a couple of things and making a couple of points. What I showed in this report is really going back to the Samuelson quote at the beginning. Pure common sense tells you that if you have more workers competing in the labor market, market conditions will be soft and the wages will go down. What I’ve done is basically put a number to that, a number that happens to be in line with what economists have long suspected about the response of wage to supply. For every 10 percent increase in supply—in total supply of a skill group, wages tend to go down by 3 or 4 percent.

Now, that does not mean that immigration is a net loss for the U.S. economy, because that wage loss doesn’t disappear into the thin air? It goes someplace. And where does it go? It goes to employers. And employers are able to save that wage loss that goes to workers, and therefore they have higher profits, and in the long run some of those higher profits will trickle down to consumers in the form of lower prices.

So, both employers and, in the long run, consumers, will gain. So one way to interpret what immigration really does is really as a redistribution program. Immigration is basically a redistribution program that at least in the short run redistributes wealth from workers to employers. It’s as simple as that, which is why, when you look at the debate over immigration policy, employers will always say something along the following lines: We have a shortage; we need more workers. Why? Because more workers implies lower wages and higher profits.

Now, the question that remains is whether that kind of transfer is acceptable or not. And that’s really what’s the immigration debate—one of the key questions of the immigration debate.

Thank you.

(Applause.)

MR. CAMAROTA: Thank you.

Now Jared Bernstein.

JARED BERNSTEIN: It was a pleasure—I’m going to sit down for the most part. I may use some slides; I may not. That was a pleasure to listen to George talk so clearly about his work. So I’ve been asked to speak for about 10 minutes about George’s paper, and it’s a pleasure to be here with George. As introduced, he’s a premier authority in this area and obviously knowledgeable and uncharacteristically able to explain this stuff. It’s nice to listen to him.

Now, I don’t think anyone—in tandem with the Samuelson quote—I don’t think anyone should find it a reach to entertain the possibility that, all else equal, immigration could lead to lower wages for those workers with whom immigrants compete. I think the notion is certainly commonsensical. The questions are: is all else equal, and what’s the magnitude of the effect, if there is one, and what, if anything, should we do about it? George posed some answers to those questions and I’m going to pose some different ones.

First of all, if nothing but immigration-induced supply shocks were going on, there would be little to argue about because you would be moving down demand curves as supply curves moved out. One of George’s slides showed that. And that would lower wages. I’m particularly sympathetic to that contention because I’ve been very active in the minimum wage debate. And in this debate, economists have really puzzled over why is it that when we raise the minimum wage it’s pretty difficult to find any evidence of significant job losses? And one of the reasons, many economists have concluded, is that demand curves for low-age workers are quite inelastic, meaning they’re pretty steep. So when you push out the supply in the case of immigration, you know, you can actually generate some pretty big wage losses, again holding all else equal. But it’s often tricky to figure out how these models, like George’s, are handling trends in labor demand. So, so far I’ve tried to make the point that the supply trends are clear, but demand trends are just as important as supply, of course, and they tend to be more volatile.

Now, labor economists have been very focused on the impact of demand shifts, and what concerns some of us about the type of findings George just presented is that it’s difficult to separate the impact of demand shifts that hurt low-wage workers regardless of their nativity—whether they’re immigrants or natives—from the impact of immigration. And to the extent that those immigration variables in George’s models are correlated with these demand shifts, it’s really difficult to pull one out from the other. And that’s one of the, I think, broadly made critiques of this literature.

My work has emphasized the very dramatic impact of the positive demand shifts in the latter ‘90s, in which—and by the latter ‘90s I’m really thinking ’95 to 2000— where the demand for low-wage workers increased more than it had in decades and appeared to swamp any immigration effect. Over this period we saw immigration increasing quite quickly the supply of low-skill workers. We also saw welfare reform—and Bob Lerman’s an expert on that—we also saw welfare reform contributing to a large increase in the supply of low-wage workers, yet the low-wage market never did better. And so, appreciating the impact of demand shifts and trying to understand their role here is important, and I think somewhat overlooked in this literature.

And so now let’s talk about magnitude. So my first point is that all else is unequal. My second point is: what’s the magnitude?

Now, most of the literature, as George suggested—and knows better than anyone here—most of the literature finds smaller effects than George’s. And these effects—it’s important to remember that these effects tend to be negative and they tend to be significant for the least skilled. They’re not nearly as significant for the more highly skilled workers as they are in George’s study, but they tend to be significant and negative for low-wage workers. However, they tend to be considerably smaller. And so now we have a magnitude discussion which is going to keep people like George and David Carr and others busy for the rest of their lives, and God bless them, but we need to try to sort out who’s right.

Now, as George pointed out, many of the studies that find the smaller effects tap spatial variation, and George and others critique this. I’m just much less convinced than he is—and I say that at great risk because he knows more about this than me—but I am a great peruser of the literature if not a creator, and I really am not convinced that these studies are as wrong as I think George [thinks they are]. He makes two contentions. One is that since there is mobility of native workers, that that can dampen the effect of the supply effect—that can dampen the supply effect. But in fact, that’s an empirically testable statement and numerous economists—there’s, I think, a good paper by Card and DiNardo and some others—have looked at the extent to which native displacements actually occur. I mean, do natives move from San Diego to Pittsburgh when immigrants locate there? And they don’t find it.

Just as one example, there’s a recent study by a couple of Fed economists, Araneus (ph) and Zavodny, and I think they used a nice technique—actually it’s a variation on something that Steve Camarota once presented at the Economic Policy Institute, to examine the impact of wages by occupation. So they actually look at the occupations that immigrants tend to congregate in and they try—and I think in an imitative way, using some very unique INS data—to control for where immigrants actually locate, so to really try to take account of the critique that George raised.

And just to give you an example from the literature that’s more in keeping with these smaller findings, “We find wage impacts that vary from half a percent to about, you know, less than 1 percent, with negative wage impacts that amount to about less than 1 percent for a 10-percent increase in the corresponding immigrant share. We find no evidence of adverse wage impacts on meeting the high-skilled native workers; in fact, increases in newly arriving immigrant shares of work within professional jobs actually have slightly positive wage effects,” et cetera, et cetera. So I’m just trying to present to you that there is another strain of this literature that finds smaller magnitudes than these.

I think it’s interesting to think about why you find a larger effect—why you find a lack of a larger effect, because I agree with Samuelson and with George’s intuitions that you’d expect to find a larger one. Well, first of all, they do find a negative effect, especially among the less skilled. But I think that the real story is that at different points in time the demand effects—the effects of the demand for workers regardless of nativity can swamp the supply effects.

And finally, there’s an interesting paper by Lewis that looks at how immigrants do get absorbed in the way they do. In other words, if they’re not having these large wage effects they must be absorbed into the economy somehow in a way that precludes that from happening. And he writes—it sounds somewhat cryptic—“Markets adapt production technology to local factor supplies,” which is sort of Greenspan-speak for industries actually find new ways to employ these low-skilled workers and they boost their production of low-wage goods in that regard.

Let me just conclude with a discussion of policy implications. And George, more than anyone—and I think this is a real attribute to his view on this—is very clear that none of what he shows, or others show, real obvious policy conclusions. Obviously it’s a complicated issue. Many of us value ethnic diversity and many economists believe it’s inefficient to restrict the flow of factors of production across borders, including labor. I mean, unions—you mentioned the AFL sign; that’s an interesting comment. Unions used to be dead-set against higher levels of immigration due to labor competition and now they want to organize immigrants. So the politics are all over the map. There are conservatives who are xenophobic and employers who believe, as George says, that they should have access to all the cheap labor they need. Advocates on the left support high levels of immigration in solidarity with ethnic minority populations and are in some denial about this negative coefficient that we’ve been talking about thus far.

Despite the negative correlation between immigration and wages, I don’t see an obvious problem with the current level of immigration, though I think one area where CIS has made a great contribution is in pointing out what looks like a pretty dysfunctional Immigration Service. So just enforcing current law would be a goal.

I’m particularly concerned about the plight of illegals who are heavily exploited in the labor market. My initial reaction to the Bush plan—I share George’s concern about labor supply—but my initial reaction was a positive one in the sense that, as he mentioned, these workers are already in the system and are heavily exploited. Bringing them under the protection of labor law seemed to me to be a good idea, at least in that one piece of the plan.

But just because I don’t think there are very large wage effects, as large as we’ve seen thus far, that doesn’t mean that they’re not out there and that they couldn’t become of much greater magnitude if the supply shocks increase. For example, if there were an amnesty program, I have heard folks from CIS argue convincingly that that sends a message to immigrants who are considering coming here to do so.

Let me just—I have one minute—let me actually go for maybe one minute extra and then I swear I’ll get out of the way.

One issue that I have with this work on immigration and the impact of immigration is that it may not be the low-skilled native men who are the most affected. It may be that the foreign born who have been here for a while are facing competition from newly arrived immigrants. That’s where there is the most head-to-head competition, I’d wage—I’d be interested in George’s comments on that—and that the persons who are actually competing most intensely with new immigrants are old immigrants.

Let me close with a word about trade and outsourcing. Many of the same economists who find wage effects from immigration fail to look for similar effects from economic developments in trade policy which create greater competition with lower-wage workers in other countries. This, too, creates greater labor supply. In fact, the wage effects on less-skilled workers, which you would attribute to our persistent trade imbalance—when we have year-after-year, decade-after-decade imbalances in manufactured goods trade—that’s the same thing as a supply shock. And the research shows that those wage effects are consistently larger than those from immigration. And now, with the outsourcing problem getting started, our skilled workers are facing competition from countries with seemingly endless supplies of skilled labor.

So, at any rate, you know, in this picture I think one can see immigration as another form of globalization. Now, let me just point out, since 2000, if you look at the employment rates of high school dropouts, that’s the share of high school dropouts who are employed, it’s a good proxy for labor demand. Between 2000 and 2003: unchanged. They haven’t changed one bit. They haven’t fallen; they haven’t gone up. The employment rates of college grads, on the other hand, have fallen 2 percentage points—are at their lowest level in 25 years. So there does appear to be—that’s not just off-shoring by any means—in fact, I would say off-shoring plays a very small part in that. But there are supply shocks that go beyond immigration and that are embodied in trade flows as well.

Thank you.

ROBERT LERMAN: Thanks very much. It’s nice to be here and it’s always a pleasure to read papers by George. The one that you read is a very easy read, but then I was supposed to read the more complicated one in the QJE and that is much more extensive but very engaging and worth reading.

George amasses a rich array of data to examine these issues, and he begins with the basic—I mean one of the things that when I teach, I teach at American U—one of the things I try to remind my elementary students is that if you’re at a cocktail party and somebody asks you about economics issues, you just say, well, it’s a matter of supply and demand. And that’s . . . you know, no economist can move away from the basics of supply and demand and the notion of competition, which is embedded in the analysis. But there is another basic notion that you also learn in economics, which is the fact that inputs are really resources, that it expands the capability; that is, more resources to produce, more factors of production to produce expand our capability to produce, and therefore, all you really need then is to increase demand sufficiently so that all of this added potential gets realized in added production.

Now, what does this mean about immigration? Well, one way of thinking about the problem is first to step back and ask, well, suppose that what we had instead of what we actually experienced—which was, as George mentions, a bimodal inflow—suppose instead we had a proportional inflow of workers as well as other factors of production—they brought with them some savings and we had a proportional inflow of all factors of production. Well, what would that mean? Well, that should mean that we just have a bigger pie, a bigger possibility of what we can produce. It shouldn’t necessarily affect the pattern of factor returns, and therefore the issue would end up turning on things like, is it better to operate at this higher scale; are there economies of scale, or are their diseconomies of scale? The economies of scale might be that people are sharing more, let’s say, in the cost of public goods, like the armed services. The diseconomies of scale might be that certain things get crowded.

And then you have other issues, like cultural diversity and heterogeneity of people, but overall you might expect those things not to be quite as significant. Now, of course, we did not have, as I mentioned before, a proportional increase in the factors of production, and we had indeed what George has alluded to, these various subgroups coming in.

So let me first say a couple of things about George’s paper specifically, and then I’m going to talk a little bit about growth issues and where we go in the future.

Well, what struck me about George’s paper was that he—it’s a kind of a static analysis in the sense that there’s no embedded economic growth in the analysis. One of the things you might expect is that maybe in the short run there’s some change; for example—just to give you one example, suppose there is an increase in the number of immigrants with low education, and that has an initial drop in the wage of unskilled workers; that should set off an incentive that should raise the rate of return for those workers to invest in skill and shift the share of those workers and move of them into higher-educated positions. Just one example. Another example is the role of capital flows. When we see this increase, what happens to the inflow of capital, and are firms more willing to invest in the presence of the growth in the workforce?

Now, these are very complicated matters, and I don’t pretend to be able to solve all of them for you, especially in my remaining four minutes. But there are papers that look at sort of the growth aspect, look at immigration in the context of an economic growth model. And what they tend to show is that you do get an initial change, especially in the group—let’s say if you have an increase in even, say, skilled worker immigrants, there is an initial drop in their wages. But then, because skilled workers work together with capital, that increases the likelihood of capital inflows, capital investment, which again increases, as Jared would say, the demand. The increased in skilled immigration raises a return of capital, so capital owners benefit. Some of the capital owners are the skilled workers themselves. And, therefore, it’s not obvious that, in this case, even skilled workers might benefit from the rise in inflow of skilled workers.

Certainly, in this model that I’m referring to, a paper by Michael Bengad called “Capital-Skill Complimentarities and the Immigration Surplus,” he finds that certainly less-skilled immigrants would benefit from inflow of skilled workers, and even skilled workers end up benefiting.

So this is one of the things that I would try to focus on, and George has focused on this in the past, which is the composition. Maybe it’s not so much the level but maybe it’s the composition, and maybe we ought to be paying more attention to avoiding this situation, which I agree is a very bad situation, where the crowding occurs almost entirely in the very low-educated groups.

One point though, and Jared talked about the ‘90s; I’ll just make one very quick point about the ‘90s. Since 1992 until today, we’ve had a kind of interesting demographic shift. Now, this doesn’t take into account all the experience components that George Borjas mentioned, but we’ve had an interesting demographic shift, the result of which has been the absolute level—the absolute level of high school dropouts, adult high school dropouts, 25 and older, has declined quite substantially. In fact, there was a decline between ’91 and 2001 as the growth in the U.S. labor—over-25 labor force of 16 million people, but there was an absolute decline of 5 million in high school graduate or less category. And you might ask, well how is that possible? Aren’t we churning out more high school dropouts, and aren’t high school dropouts immigrating? And the answer is, yes, but it’s also the case that older—the difference in the educational structure between older workers and younger workers—is so great that as older workers retire you get a big drop off in the numbers of high, in the absolute numbers of high school drop outs. So that may be one reason why in the 1990s we were able to absorb the high school dropouts more effectively.

Two more points; one point which I forgot to mention, which is about George’s paper. It’s sort of small technical point but it may drive some of the results. I noticed in the broader paper, where you have the wage trends, that there is essentially almost no real wage growth in 40 years. And this is hard to believe. And it might possibly be because of two factors: one, the price index that George uses may overstate the growth in prices, and the second is the non-wage compensation has increased quite substantially and that’s not taken into account.

Okay, where do we go for the future? Where do we go to for the future? Well, one of the things that we have not talked about is how immigration is interacting with the broader demographic changes in the society. If you look at what we were doing in the 1970s and the 1980s, we had labor force growth of 2 percent per year, partly because of internal demographic factors, partly because of immigration, and partly because of a big inflow of women into the workforce, which has largely peaked. Over the next 20-30 years, we’re going to have a much, much smaller—in fact, half of the rate of growth in the labor force, only about 1 percent per year. This, again, includes immigrants. And in some groups, such as prime-age workers—workers 25 to 54—we have virtually no absolute growth in the workforce.

Now, you might say, well, this is good; it’s going to increase the wages for that group, and that’s probably true. But again, it gives us a very unbalanced pattern, and I think that when we think about immigration we have to think about it in the context of what else is going on in the labor force. We have to think about it in terms of compositional factors and we have to think about it in terms of economic growth.

One way to perhaps compensate—and in fact we have a program, the H1B program there is a $1,000 fee for employers bringing in an H1B person. That fee goes into training programs and those training programs actually are among our more flexible training programs that the Labor Department sponsors. If we moved more in that direction I think we could have a win-win situation.

Thank you.

MR. CAMAROTA: Well, thank you to all of the panelists. I’d like to open it up for questions, but also like to exercise the chair’s prerogative and ask the first one.
 

Immigrants on average are poorer than natives, so that when immigrants come in —on the other hand, a large share of immigrants are workers, so when immigrants come in the economy is certainly larger—you’ve got more people working, and per capita income, or per capita GDP, is certainly lower when you count the immigrants because they’re poorer on average, so the economy is bigger and the average person in America is poorer.

But neither of those two facts tell us where there’s some net gain from immigration to natives. And I wonder if any of the panelists, but especially Dr. Borjas, could talk to the question of what does the research show? I know there was a recent NBR paper on this—that’s the National Bureau of Economic Research, the National Academy of Science. What do we think; is there a big net gain from immigration, at least to the economy? There might be other net gains and losses.

MR. BORJAS: First we need to define what we mean by gains. It’s pretty clear that when more workers come in from abroad, say, GDP grows. Part of that growth in GDP actually goes to pay immigrants themselves. In other words, you have to pay them to work and to produce whatever they produce. So, the way that economists have tended to define the gains happens to be as what remains of this increasing GDP as the result of more workers after you pay off the immigrants their wage and salary. In other words, how much is left over for everybody else after you pay them for the fact that they have to work for a living? And what the research tends to show, and summarized in the National Academy report, is that that net gain, the part of the GDP increase that accrues to natives, is clearly positive. There’s no doubt about it. There’s a net gain for the U.S. economy. But it happens to be very small.

Now, remarkably enough, in economic theory it actually happens to be the case, the size of the net gain depends on the number I was talking about before. In other words, what happens to the wage as the supply goes up? And the estimates that one gave with the 3-4 percentage point declining wage that I’ve documented here are, roughly speaking, that the gain is on the order of perhaps around 21 percent of GDP, okay? So it’s one-tenth of 1 percent of GDP. GDP in the U.S. today is roughly $10 (trillion), $11 trillion, so we’re talking about a net gain of maybe $10 (billion), $11 billion dollars, roughly speaking, okay?

So there’s clearly gain. It’s not a huge number but it is, nevertheless, an important number. And one important thing too about that gain is that it’s really masking two separate effects on the capital gain. One is the wage loss from workers, which can be substantial, and the other is the net gain to capitalists and employers and consumers, which could be substantial. So these flows could be very big, but on net, it’s about $10 (billion) $11 billion net gain, basically.

MR. CAMAROTA: So let me see if I understand, though, that what you’re saying is that sometimes people say, look, immigration doesn’t have any effect on wages but it creates a big net gain for natives. You’re saying that is not possible based on economic theory.

MR. BORJAS: I’m not saying—no, that’s not what I’m saying. What I’m saying is, based on what people—on the evidence that we have so far—nobody has ever documented a really big gain. Their documentation has been on the order of .1 percent of GDP. That’s what I’m saying.

MR. LERMAN: I’d like to comment quickly.

MR. CAMAROTA: Sure.

MR. LERMAN: The issue of . . . yes, it’s true that we have to pay the immigrants, and that’s a big part of the production that they generate, but there is a big gain for those immigrants. Those immigrants have a huge increase in wages.

Now, that is perhaps one reason why I believe that if you look at the political support—and George was talking about this earlier, so I’m going to comment on it, the issue of employers and so on—a lot of the political support for the immigration comes from prior immigrants who may be able to bring family members in, and in any event, even if they aren’t bringing family members in, they have a certain sympathy for groups that want to do what they did, which is to move from a very low-wage place to a substantially higher wage place, for them.

And I think it’s important, when we discuss these issues, when we’re talking about—and I certainly agree with George that it has distributional issues, it is a factor in affecting the earnings and equality increases, but there are people who feel that they’re going to benefit a lot, and their supporters in this country are among the big backers of immigration, even if to some extent some of them may indirectly be competing with them.

MR. CAMAROTA: Let me just comment very briefly. Actually, I guess I would disagree just as little bit in that what one finds in the surveys is that immigrants are not nearly as enthusiastic about immigration, further immigration, because they generally fear the job competition. But people who claim to speak for immigrants—that is, the Hispanic elite in the United States and the immigrant elite—is much more in unanimity on the need for more immigration. The immigrants themselves are much more circumspect about it, and we see that in poll responses to the ideas of amnesty or increases. It’s a more complex picture than that.

But anyway, let me open up the . . .

MR. BORJAS: I want to say one more thing about the gains, okay, and that is the following: One can think of the world as being composed of three populations: the U.S. population—let’s suppose that everybody here today is the U.S. population—the immigrants we’re going to accept, and everybody else left behind. We haven’t talked about that group at all. By natives we mean “us,” more or less, and some of us will gain; some of us will lose. Immigrants completely agree: they’re going to gain substantially. A minimum wage job in the U.S., as bad as it might be to most of us, is far, far better than the opportunities for many people elsewhere.

But what about the people left behind: that’s a group that we tend to forget about, who also suffer or gain as a result of immigration to the U.S.? For example, let’s consider immigration policy, like the H-1B program, in which we suddenly decide to import every single high-skill worker in India. Okay? Like every single engineer in India we’re going to import. Engineers in the U.S. wouldn’t like that but the U.S. as a whole might like that a lot. We’d be able to buy a lot more software, build a lot more bridges, and whatever, at much lower prices. But what about the Indians left behind? Is it really beneficial for India for a country like the U.S. to have so much pull and so earnings power, in a sense, that it can pull completely the skilled workforce from many other countries? And that gain also should be brought into account in any discussion of the gains from immigration. Yes, immigrants themselves benefit a lot, but people left behind may benefit or may lose substantially, and that’s a group we shouldn’t forget about.

MR. CAMAROTA: The best statistic I ever heard on that topic was once that half of Ethiopia’s doctors now live in America, and that’s a society desperately in need of doctors. And so that’s an interesting question to also think about.

Anyway, let me open it up to general questions. Who’d like to ask a question? Go ahead, and if you could identify yourself.

Q: Ralph Smith, the Congressional Budget Office. I’d like to get your response to a couple of things that Bob said, that in your analysis it’s not taking into account sort of dynamic feedback. I think you provide a good criticism of the problem with the cross-section, but there do seem to be some – (inaudible).

MR. CAMAROTA: Let me just repeat the question. The question had to do with the dynamic feedbacks in the economy that Dr. Lerman talked about.

MR. BORJAS: What I did was basically—I’m talking about the technical paper that came out in the QJE last November, okay? What I did was basically look at data from 1960, ‘70, ’80, ’90 through 2000 Census. So really what I’m doing is really differencing data at a decade level. In other words, I’m looking at the change in economic opportunities for native workers between ’60 and ’70, or ’70 and ’80, ’80 and ’90. Anything that occurs within that decade are more or less accounted, but I would think a decade is more or less in the short run in some sense. The analysis does not take into account anything that might happen after the decade perspective. So it may well be the case that after 10, 15, 20, 30 years, all kinds of things begin to happen that may show up in the data.

The best way to interpret my results is really to say, this is what happens within a 10-year period with increased supply of a particular skill group. That’s the wage change you will see within that 10-year period.

MR. CAMAROTA: Go ahead. Identify yourself, please.

Q: Yes, my name is (inaudible), member of the House of Representatives in the Dutch parliament. That’s quite interesting input from the panelists. First of all, I was quite struck by your one-to-one relation wage on the one hand and an increase of immigration on the other hand. But as we know, the results of wages are also influenced by, let’s say, international economics and governments, the role of the trade union, as know in the Netherlands, being strong. So I don’t get a clear picture of the one-to-one relation, and say if we increase the supply so the wages will go down.

A second remark is that what we have, for example, nowadays in Europe, and especially in the Netherlands, is that going down with the wages is not so bad in that sense because our economic position is not so well nowadays, so that we keep the wages at the minimum to improve our economic competition, one half, and the other half is it’s also in the benefit of the immigrants themselves because it improves their integration, because to get jobs, to work, is one of the prime things to integrate with society. (Inaudible.)

MR. CAMAROTA: The question again was, what are the other factors that have impacts on wages and also the integration of immigrants into the economy?

MR. BORJAS: The technical analysis—again referring to the version of the paper that’s much more technical, at the decade level—in other words within a ’60 to ’70 period, or ’70 to ’80, or something like that, actually takes into account almost everything one can think of, in some sense, that could happen to a particular group over a decade. In other words, what I’m trying to do is more or less follow, holding other things equal for a particular group. So for example, I am looking at what the relation is between the wage change between, say, 1980 and 1990 for high school dropouts, after controlling for all of the things that could have affected high school dropouts between 1980 and 1990.

So in the technical analysis, I do make a point—in the technical jargon we use, we call them fixed effects. I control for the fixed effects that could have affected the wage of particular skill groups in that decade, okay? So that’s clearly done.

In terms of integration and so on, it’s actually a very curious question because most European countries are very concerned with integration policy. In the U.S., it’s quite remarkable that we have something called immigration policy, but more or less, immigration policy ends the day that people get into the U.S. We have no such thing as an integration policy, roughly—at least at the federal level. And so we basically have these people coming to the marketplace; there’s very little government programmatic assistance of any type, any systematic type, that would help any of these workers integrate because immigration policy ended the day they got the visa, more or less. And what I’m doing more or less is just catching what happens in the labor market as the result of entrance.

So, unlike most European countries—and actually Australia and New Zealand, or Canada—we really don’t have a persistent follow-up of immigrants once they get into the U.S. The only other time that immigrants, once they get into the U.S. legally, ever get into the system is when they want to get naturalized, at which point they have to pass a very, very simple test. But that’s about the only extent to which we have any kind of policy of integration for immigrants.

MR. CAMAROTA: Go ahead.

Q: Ricardo Alonzo-Zaldivar with the LA Times, and I have questions for two of the panelists. For Mr. Borjas, today the Democrats are going to roll out their immigration plan, and basically it’s a legalization plan for those who are already here, coupled with a restrictive guest worker program that’s capped as opposed to open-ended. Can we get your thoughts on that, on the effects of that?

And for Mr. Lerman, could you tell us what share of the labor force—I was interested in what you said about the share of the labor force that’s high school dropouts and how that has changed over whatever period you have . . . give us the numbers.

MR. CAMAROTA: The two questions, one was a question for Dr. Borjas about the Democratic plan; the Democratic Party is releasing a plan today to legalize illegal aliens and also institute a guest worker program with some kind of cap. And then there was another question for Dr. Lerman about the declining share of the workforce represented by high school dropouts.

MR. BORJAS: Okay, let me start. I have not read the Democratic plan so I’m basically talking from what I read in the paper last night, okay, about what they want to propose. If, in fact, they proposed to legalize illegal immigrants already in the U.S. without doing anything about solving the illegal immigration problem, I think that has the same basic flaw as the Bush proposal, which is that we’re trying to address a problem without really addressing the underlying issue, which is the entry of illegal immigrants.

By legalizing the 10 million or so illegal immigrants already here, that plan, nor does the Bush plan, do anything about preventing the problem from being revisited 10 years from now. In other words, we might have to revisit the issue 10 years from now and give a third amnesty to 10 million more people.

I think that it would be a great mistake for the U.S. in general to address the problem of what to do with illegal immigrants already here until we prevent the problem from being addressed ever again. In other words, let’s actually control the border, prevent the illegal population from increasing, and then we can say, let’s worry about what to do with people already here. Any other thing we do would just create more incentives for another amnesty five years, 10 years, down the line. That’s number one.

In terms of what it would do in the labor market, as I said in my talk, illegal aliens are already here; they already had the labor market impact they’re going to have. So the 10 million already here already had the impact.

To extent that these programs of repeated amnesties sort of make a mockery, really, of our legal immigration policy, that just creates more incentives for more illegal immigrants to come in, having a larger labor market impact in the end. So I would not be in favor of any program, whether Democratic or Republican, to legalize current illegal immigrants until the legal immigration problem is actually addressed at its root level.

MR. LERMAN: I don’t have in my head the exact proportions, but I know that it’s fallen quite substantially, especially in the last 10 to 15 years, especially among the adult 25 and over population.

I’m just looking at some other data here. Again, this is on the 25 and over population. The less-than-high-school group went from 17 million out of 80 million in—I don’t have a calculator in my head, but that’s something on the order of 20 percent, I think—in 1980. In the year 2000, it was 12 million—this is adults 25 and over in the labor force—went to 12 million out of 120 million, so there is a big drop.

Q: Can you repeat those years again?

MR. LERMAN: 1980 to 2000. But one of the things I mentioned, I mentioned there was this big absolute decline in high school dropouts, adult high school dropouts in the workforce. Part of that decline is, as I mentioned, because of this difference in education between the people retiring and the people coming into the workforce. That is not going to continue to happen. That is not going to continue, and it will make controlling the border with respect to low-skill workers all the more important, because in the future, as the educational structure, the people retiring over the next 20 years, is going to be much closer to the educational structure of the new entrants, and therefore you will not get the same kind of shift that we had in the ’90’s.

MR. CAMAROTA: One point on this. Jared and I actually disagree about the statistics on employment for high school dropouts 2000 to 2003, and we were looking at the same data, so I don’t know why –

MR. BERNSTEIN: No, I was talking about employment.

MR. CAMAROTA: Oh, the number employed?

MR. BERNSTEIN: Yes, the employment rate—the share of the population that –

MR. CAMAROTA: Because unemployment rate; that is, those who say they’re looking for work, went up pretty substantially between 2000 and 2003. Okay. Well, good. So we don’t disagree. We were both looking at the same data, that’s why –

Go ahead, Sergio?

Q: Sergio Bustos with Gannett News Service. I just have something to clarify and then—I could point out more data on this, but on the front page of the Backgrounder you write that—at the top of the Backgrounder findings – “When increasing the supply of labor between ’80 and 2000, immigration reduced the average annual earnings of native men by and estimated – (inaudible) – or up to 4 percent.”

My question there is, did you get the same—what was the date for native women during that time? And secondly, is the chart on page six—when you say “all workers,” are we talking about men and women or just men?

MR. CAMAROTA: The question had to do with gender. Did the study look only at men or did it look at men and women?

MR. BORJAS: The CIS study looked only at men. If you look at the more technical version of this study that came out in the 2003 QJE, there are a couple of tables that actually looks at women, and the results are pretty similar for women, okay, but in this study it was only for men, for various reasons.

Why? Because a lot of women, especially earlier in the period—I’m looking from 1960 through 2000, right? A lot of women, back in 1960, 1970, were not in the labor force. So we’re having a large sample of women who clearly have chosen to stay out of the labor market. And the question is, what wage do you assign to women who are not in the labor market? It’s not clearly zero, because the fact that—it doesn’t mean they have zero earnings potential, just the fact they don’t work.

Because of that reason, bringing women into the labor market analysis often creates a lot of problems. This is not the only study that actually focused on men. Labor economics is not, this actually is not—it’s more or less a standard thing to sort of look at men initially and then extend the analysis to consider what happens when you bring women into the picture.

MR. CAMAROTA: It’s a problem actually with the data. For some reason the Department of Labor will not ask a series of questions on how long have you been at your current job, how long have you worked? It would be really nice if the Census or CPS would do that. It would make the analysis of this kind easier, but they don’t. So that’s why people look at men.

Q: (Inaudible) – subject of the impact of immigration on native-born citizens. I have statistics from the Bureau of Labor Statistics that break out non-Hispanic white, non-Hispanic black, Hispanic, and another category, “non-Hispanic other.” And what it shows is the three years, January 2001 to January 2004, roughly the period of the Bush administration, employment of non-Hispanic whites decline by 1,400,000; non-Hispanic blacks by 251,000. Hispanic employment grew by 1,289,000 and “non-Hispanic other” employment grew by 1,134,000. And there was actually a total growth in workforce— employed force of 744,000.

If you believe that non-Hispanic whites and blacks are basically native-born Americans, it seems to me the impact is very great, and I think it’s because of a shift in labor force. We are losing the high-paying jobs and gaining the low-paying jobs.

MR. CAMAROTA: Anyone want to comment?

Q: What do you have to say to that?

MR. CAMAROTA: I guess I should repeat the —Let me repeat the question has to do with, does it look like during the current recession that all the employment gains seem to be going to the immigrants? And the answer to that question is, yes. I mean, I’ve done that analysis. We have a report called “Immigration in a Time of Recession.” Between 2000 and 2003 the number of natives working in the United States fell and the number of immigrants holding a job increased all of the employment gains at the aggregate level went to immigrants.

So if that answers the question . . . that’s 2000 to 2003. Without question, that’s what happened. Now, what wage impact that has, I don’t know, and what displacement was going on isn’t measured. But, yes, if you want to know, the number of natives working went down and the number of immigrants working went up, so all the employment gains went to the immigrants in that time period.

MR. BORJAS: Well, first of all, my study really ends in 2000, my study focusing on wages as opposed to employment changes, so it’s really hard—I think it would be a little irresponsible of me to extend what I found to employment changes beyond the sample period, so I really don’t know . . . I don’t have any particular expertise saying that this study implies the following should have happened for employment rates in the first few years between 2001 and 2004.

Q: I believe Mr. Bernstein and Mr. Lerman have addressed this question to some extent.

MR. CAMAROTA: What? I’m sorry?

Q: I believe that Mr. Lerman and Mr. Bernstein have addressed this question to some extent.

MR. CAMAROTA: Would you like to make any comments?

MR. BERNSTEIN: The only thing – this is not particularly the comment you’re looking for, but it does always strike me —

Q: I’m looking for the truth.

MR. BERNSTEIN: No, this is the truth but it’s very kind of a nerdy point. You know, you’re using these numbers that the BLS says you shouldn’t use; you’re doing something the Bureau says that you ought not do because you’re comparing employment counts over a period where they have to have gone in and changed what they call the population control. They need to weight the data up to be nationally represented, and they have a lot of trouble doing that, and their biggest problem—Steve, you probably know about this—their biggest problem has been accounting for immigration flows. So they keep rebasing these weights, and when you cross those rebasing periods you’re actually adding—you’re comparing apples and oranges—so . . .

Q: (Inaudible) – to kind of fit into that area.

MR. BERNSTEIN: Yes, in my study, looking at that, I did use the new 2000 base weights and take them back to 2000, so hopefully we’re getting a nice apples to apples comparison, but there’s always some question.

Did you want to say anything?

MR. LERMAN: I just wanted to say something on the skill issue and losing skill jobs. I mean, the fact is that we have—our occupational structure has definitely moved toward the more skill occupations. When I talked about the decline, the absolute decline in adult high school dropouts and relatively stagnant high school graduates only, all of the increase in the adult workforce between ’92 and 2003 were among people with at least some college. And for the most part we absorbed them into reasonable jobs. Could we do better? Yeah. But it’s not true that all the good jobs are going away.

MR. CAMAROTA: Well, I want to thank you all for coming, but if you have more questions I’m sure the panelists wouldn’t mind staying a few more minutes and letting you ask some questions. Thank you again for coming to the Center for Immigration Studies panel.

(Applause.)