Are H-1B Workers Displacing Americans with Cheaper Labor?

- Thank you for joining us. I'm Paula Sapienza. I'm the JP Conte Senior Fellow at the Hub Institution at Stanford University, and I'm here with Amy Nice. Who is the Distinguished Immigration Council at the Institute for Progress and Distinguished Immigration Fellow at Cornell University. We are here to ask a very simple question, which is why

- Do you think it's not gonna be simple, but Okay.

- Ah, we'll see the question is simple. The answer is probably difficult. Okay. What the question is, are immigrants hired because they're being paid less? And specifically we're gonna focus on H one B visa. There is this misgiving around the country that international workers hired through H one B visas are only hired because they're gonna be paid less. And by doing this business defacto displace American workers. So the first question we're gonna ask is, are they really paid less? And then we're gonna try to delve into the more complex details.

- Well, I think the thing is that a lot of the people that have been writing about and talking about H one B workers being paid less are relying on concepts or data that are not answering the actual question. So there's a lot of reliance on the idea that H one B workers' salaries should be compared to the occupational median salary that's paid to anybody across the country or in, in that geography where the person's working in the occupation. But we all know, 'cause we all got jobs when we were early career professionals, when you're a newly minted graduate, you don't get the paid the same as the average person in your entire occupation. H one B workers have turned out to be primarily individuals who are earning F international students who are earning degrees in the United States and the primarily advanced STEM degree graduates. So a newly minted STEM masters or a newly minted STEM PhD is not gonna get the median wage for similar workers in their occupation. So we should compare it to the median wage of Americans, the same age education and experience in that occupation in that geography, but not the occupational median for everybody working. That that's one source of confusion. Another source of confusion is that so far people who are writing about this have been looking at data from the labor condition application that's required for H one B.

- What's labor con? What's a labor condition application?

- The labor condition application is an attestation by all H one B employers stating that they are going to pay a foreign worker the same as similar Americans that they employ. What

- The actual, but what's in debt in terms, why is it lower value? So if people are looking at that, what do they get them misled.

- Okay. So the labor condition application identifies what the prevailing wage is and the minimum wage that the,

- So it's minimum the question.

- Yep. And they often identify that because the labor condition application has to be a public record, the employers often express that minimum wage that they're gonna pay as a range. And so the range has to be such that the low end of the range is consistent with on this same labor condition application the prevailing wage. But the high end of the range can be considering a variety of legitimate business factors of how they pay similar people who they employ. So the range, the low end of the range can't be too low, but the high end can be quite significantly different.

- Quite significantly higher is gonna be higher. Yeah. But so if it's higher, this doesn't add up to me because, you know, we're trying to explain why, you know, people think that they're paid less, so they must be seeing a loan numbers,

- So, right. But the, the, the in, in fairness, until recently we didn't have public access to the data on this different government form that has been released through Bloomberg in a FOIA request to actually analyze the actual salaries that are paid. And so the LCA was the labor condition application was this approximation. And I just think because people are looking at the median for the occupation and people are looking at the else, the labor condition application numbers, that it's been misleading and imprecise. So, so what

- What I understand from all these is that so far people have been looking at the wrong data, which were the only data available for this.

- It's not wrong if that's all you have access to. Right.

- It's limited data. And then recently there has been some new data released, thanks to some journalist acquiring the actual compensation that this H one B recipient are receiving. And that would allow us to compare the average salary for a similar experience level of domestic workers. And you know, this H one B applicant

- Probably, it's actually a little more complicated than that because if you, it is key to compare to Americans that are, have similar education and experience. So in order to do that, and the reason why this hasn't been done regularly is that you have to be an economist and you have to link that data to the ace to the census data to compare it to Americans. So in the past, people found it easier to just compare to the occupational median salary that's as reg reg regularly available. And those of us who aren't data analysts or economists can easily see what is the median wage for the occupation.

- Yeah,

- I

- Understand this is a sort of a digression, but now we can do this. I'm, I'm just saying.

- Yes.

- So now that we can do this. Yes. So, so let me just go back one second. So it was like in the past there were not sufficient data Yes. To do this comparison. Yes. And people use the data available. That's the best they could do. And those data really were showing a partial picture and especially they were showing the application data, but not the actual ultimate salary being paid. That's right. Now we have the possibility of doing this comparison and I understand that there are some differences between professional services and consultant and other businesses. So let's discuss those differences separately. So, okay, let's first think about the average professional, sorry, let's, let's first think about exclude, let's first exclude the professional services and consultants and let's talk about, you know, this very large fraction of business that use H one B. What do we learn from that?

- Well, the companies, regardless of size, regardless of sector, regardless of geography across the country that are not the professional services or consulting firms, largely use the H one B program to recruit individuals that are here in the country already in a temporary lawful status. Something around 78% of the time. H one B petitions by those employers are for people that are already here. And most of those are international students. And when you're on campus recruiting, you're hiring Americans, you're mostly hiring Americans, but you, especially if you're looking for an advanced STEM degree holder, you're gonna have a 50 50 chance that the people that you're interviewing happen to not be born here. Since roughly across all the science and technology and engineering fields, about 50% of those graduate students are not Americans. And the employer has been told by Congress that the idea is that if a US employer decides that the best, the best candidate for a job, the best fit for their team is someone that doesn't happen to be a citizen, feel free to employ them here. As long as the numerical limits haven't been met already and you are offering the same terms and conditions, you comply with this labor condition application. And so the idea is that the government is not gonna micromanage the decision making of industry as to who they wanna hire, but that there's gonna be an element of protection to the US labor market by these numerical limits and by, you have to make sure you pay the right wages.

- So, so what you're saying is that excluding professional services, okay. That we will maybe tackle later when we look at the data coming from this FOIA request where there is a comparison, you could do the comparison between this international H one B and Department of Labor statistics related to similar occupation. We actually do find the opposite of what people believe, which is that the international students are generally paid sometime the same, but most

- Time a little bit more. I mean, I think one reason why that you might see that pattern in the data is that if an employer wants to hire a foreign national, they have to think ahead. So what's the cost gonna be if I hire this individual who's not an American? So there's like the actual dollars out of the pocket of the company and then there's the work that has to be done to com to be a compliant employer to attest to this, to file that, to do it in a timely fashion. So the cost as far as like dollars out of your pocket, they're very significant filing fees to do an H one B. You either have to have hired in-house to your organization, lawyers and paralegals to do these work or you have to hire outside counsel. As an economist

- Though, I'm thinking there is another very large cost, which is uncertainty. Uncertainty.

- Yes, absolutely. Because you hire the person, if, if you do hire someone who's an international student, which is roughly like 62% of the time of these H one Bs for the industry players that are not in the professional services or consulting world, you can have that STEM grad for up to three years. You may think first, do I, maybe the tie to the higher compensation is you're already making the calculation. Do I want this person for just three years? Is it worthwhile? 'cause I don't know, it's uncertain if I'm gonna have it.

- So okay, wait three years. I think we should explain

- Oh,

- Okay. To like, you know, so I'm, I'm an employer. I'm thinking I would like to hire the best talent. I have the opportunity to hire an international student. We'll take, I can have these people for three years on a temporary basis, but then I have to transition them to a different program. What's the uncertainty? What's the source of uncertainty at that point?

- The H one B category has a numerical limit of 85,000 divided between 20,000. That is a reserved lottery for only people who have a US masters or above and 65,000. That's an unreserved for anybody who has filed lottery for access. And it's a lottery because more than 85,000 petitions are interested in being filed.

- So there is a lottery. So I'm here is the uncertainty. Yes. So I'm a business, I really would like to hire talent, but I also know that when I hire someone, I'm gonna spend a lot of resources training them. And the best possible return that I could have is that if this trained individual after they've spent a sufficient time with us being trained, will actually stay along. And the uncertainty comes from the fact that there is a lottery. What's the probability of winning this lottery for the typical employer? That is a big cost.

- Yes. Well, it's changed over time as the demand has increased. But the, you know, when the, when the, when the CAPS were set in 1990, then we added in 2004, this 20,000, our GDP was half of what it is. Our, our economy was smaller. Our population was three quarters of what it is. So the odds have changed over time because the cap has stayed the same, but the economy and the country have not. But I think the odds, which could be another reason why you see compensation being equal to or more in the BLS data is that part of the calculation is, so if I am hiring people that have a US masters or above, my odds are better. 'cause I get two bites at the lottery. All of the potential H one B petitions under the current system go into the unreserved lottery for the 65,000. And then only the US Masters and above that weren't selected get to go again lottery. Oh that's very interesting. Second lottery. And so when they changed the order of the lottery, which is under the first Trump administration, you saw an increase in the percentage of H one B petitions that go to individuals that have a US masters or above.

- So that's very interesting. Yeah. Employers are doing this calculation, they're paying some fees to employ this. These are legal fees mostly.

- Yeah.

- And then of course there is time involved in this and so on. But then is one part you're adding the

- Uncertainty.

- But then there is this uncertainty and this uncertainty depends on the fact that along the way, after three years after having trained these employees, what they face is a lottery, like the standard lottery. And this lottery means that with some probability they will indeed be able to retain this worker. And therefore the investment that they made in training this investment, that's this worker as is gonna pay off. But with some of the probability they're gonna lose their investment. And so that's a real important calculation.

- Yes. But that can be related. That calculation can be related to what you're seeing in the data that the individuals who are holding H one B for industry in America, outside of the professional services and consulting are employees that are paid well because the employer has sort of baked in that calculation. Like it's only worth it if it's these people. It's something like 29,000 unique employers file petitions, 60%. I'm I'm talking about the employers for cap subject, the numerically limited H one Bs, not universities that are not limited by the numbers. 60% file one petition, 35% file, two to four petitions. So employer, and some of those employers are small, some of 'em are startups, but some of 'em are big companies and they're like, I don't wanna deal with this. Oh I need to deal with this one person 'cause this person is a value I see has high potential to provide valuable input to our efforts to innovate, to be more productive, to make progress on critical fields. So that's why as far as if you look at the number of employers that participate, most employers aren't filing very many H one B petitions. 'cause this decision is already baked into their calculus. So you're flipping the

- Equation for me completely. So we started with this idea of displacing domestic worker and to some extent to find international just because they're cheaper. And instead we found out that they're not cheaper, they're actually more expensive. And probably there's slightly more expensive or more expensive in some cases because they come with the real talent that is in great need. But this is also a rare event within what you're telling me is that from these numbers that this, this is not a very common for most employer employers. Most employer for most employers. There is another group of employer though that we left aside here at the beginning. And those are the professional services often described as consultant. Can you explain me what's the difference between this type of services versus other type of employer and what is the equation that is being playing there?

- Sure. So one sort of starting point for comparison, when we were talking about industry and you know, US employers that are not in the professional services or consulting, we said that 77 or 78% of the time they're hiring in the H one B program, people that are already here, professional services. It's like the ex, it's the exact opposite. 80% of the petitions for the large outsourcing firms who are providing professional services and business processes, outsourcing are petitions in the H one B category for individuals who are outside the United States who are already working for that firm, typically in India and who are being sponsored. So they can come in later for consulting work. So

- The really, we're looking at this program, which is the same program. Yes. But you know, in the, the really big divide is firms that are mostly have workers in a foreign country. You're saying India, those are the consulting business, the outsourcer that, that kind of business, right? Mm.

- Yes. - And they disproportionately use the H one B to bring inside the US worker that are already working for them.

- Yes. - And then you have, you know, the other group that are doing, you know, they're really looking for talent.

- Yes.

- They're following a completely different procedure.

- Yes. I think one thing that's important to, to say is that the professional services firms are also looking for talent, but they're not engaging in a talent search in the US labor market largely to find their, find their stuff. Oh, that's very, so there's a search, they're trying to find the right skills, but they're not looking, they're not doing an, an, an active search across the US labor market to find those people. Always. In the

- Case of consultant, what's different in terms of bottom line? Why there is so much concern about how the, these firms are using this program?

- Yeah, so I think that there's two things I think we should emphasize. One is the outsourcing firms, the large outsourcing firms, I mean, they're, they know that they're under the microscope. I I actually think that probably they're the most compliant employers in the H one B ecosystem. And I don't think that there's, you know, a situation where these employers are skirting the letter of the law. They're not

- Doing anything illegal.

- Yeah, no. And the, the, so the second point is the distinction is the business model. So if your business is that I am selling services to the Fortune 500 wage arbitrage is part of your business. It's not that those people are underpaid necessarily. They're being hired in jobs that are more basic.

- Yeah. So here is a fundamental difference from what you described before, which is the previous set of employers that go to universities, they're looking for exceptional talent sometime. Yeah. And really specific skills. These employers by definition, just for the business they're in, they do more basic stuff. Yeah. But they still tap on the H one

- B Yep.

- Program. And so as a consequence, they're gonna, we're gonna see people with lower skills going, coming through that channel. Yeah. Which is also a channel where not many domestic US worker are considered.

- I mean, I think that there, that the large outsourcers increasingly over the last years, over the last decade specifically have increased their presence in US recruiting. And so this is like a very changing landscape, but it's still the case that, as we said, you know, 80% of the, the data show that 80% of the H one B petitions for a large outsource are for people who are not in the United States and who are typically their employees outside of the country. And to, I just wanna respond to your point about there are these two different very, these two very different types of employers in this one program. When the employer, when this category, the, the H one category has existed since 1952, there was no outsourcing 1990, it was separated into this H one B category for the specialty occupation professionals. And then for the people who were really distinguished and extraordinary, this new O one A category and the H one B category in 1990 still was not based on any kind of recognition that, oh, by the way, there could be outsourcers who are using this category. So we haven't changed the law. It is a system that is designed for, we don't wanna micromanage your hiring, you're an US employer. So of course you're gonna be engaging in a talent search in the United States to find a particular person with particular skills. And the entire construct isn't really built for all of the different types of employers, including startups, including entrepreneurs and including outsourcers. And so we, part of the problem is that people, we have

- One big program

- Yeah.

- For many different needs. And it was not designed

- That's exactly right. To, to, to, and we need this program because we want there to be a way that any US employer in any sector, in any geography of any size, could wake up tomorrow and say, the best person for my team is this particular person who doesn't happen to be an American. We want there to be a way that employers do have access to talent. And we don't want to be limited to only certain employers.

- Thank you, Emmy. This was a really exceptional, anything else you want to add? No, that's great. My pleasure. It was great to be here.