“Berkeley Law Voices Carry,” hosted by Gwyneth Shaw, is a podcast about how the school’s faculty, students, and staff are making an impact — in California, across the country, and around the world — through pathbreaking scholarship, hands-on legal training, and advocacy.
This episode features UC Berkeley Law Professor David Hausman, the faculty director of the Deportation Data Project, the first centralized repository of individual-level U.S. government immigration enforcement data. This spring, the project obtained and made available online adataset including anonymized identifiers that correspond to individuals, allowing users to follow people through the enforcement process without learning their identities. The dataset is publicly available so scholars, journalists, and policymakers can analyze what actions Immigration and Customs Enforcement, or ICE, has been taking between mid-November 2023 and late June 2025. The project also has historical data on ICE actions.
The project recently released new data, which has already been used in dozens of news articles about ICE’s activities.
Hausman, who joined the UC Berkeley Law faculty in 2022, is an empirical researcher studying U.S. immigration enforcement and teaches Civil Procedure and Immigration Law. He has a J.D. and a Ph.D. in political science from Stanford and is a former attorney for the American Civil Liberties Union Immigrants’ Rights Project in New York.
What follows is an edited version of the conversation. Listen to the full episode below and visit the “Voices Carry” archive for all episodes.
GWYNETH SHAW: How did the Deportation Data project come about? And why is this information so important to have?
DAVID HAUSMAN: So back when I was working as a lawyer at the ACLU, oftentimes we wanted to count things. So for example, we brought a case challenging how long people had to wait after they were arrested by Immigration and Customs Enforcement before they got to challenge their detention before an immigration judge. And we wanted to count how many days that was for most people.
And for example, we wanted to show in another case that ICE had a policy of releasing nobody at all and that absolutely everyone ICE was arresting it was choosing to detain. And so in both those cases, we wanted to count things up. And luckily, the government actually maintains datasets on what it’s doing in immigration enforcement. And it has to do that because it has to know where people are. It has to know who it’s arresting where. It has to know who it’s booking into detention where, and it has to know when it’s running deportation flights to other countries.
So it’s collecting this information internally in order to administer its programs. And under the Freedom of Information Act, it’s possible to ask for those datasets that ICE keeps itself and then to analyze them.
And so back when I was a lawyer at the ACLU, we did this. It was useful in our cases. And that ended up forming my research agenda as an academic as well. And then over time, I’ve been wanting to systematize this and pull together all of the data that’s out there from the different government agencies that are involved in immigration enforcement and put it all in one place, and then also try to gather and make public the know how that we have what different parts of the dataset mean and how they can be used.
And so finally, now that I’ve been at Berkeley a couple of years, I decided the time was right to actually fundraise and start this project.
GWYNETH SHAW: How did you get access to this information? I know it’s through the Freedom of Information Act, or FOIA, but there are some other parts of this that had to come together to get this dataset as well, right?
DAVID HAUSMAN: So in starting this project, I was really lucky to be at Berkeley because I was able to gather together this team of folks who are already here. So a couple of law students have been really helpful.
I’ve been working with a predoctoral fellow here as well, also a wonderful staff member in the faculty support unit, and all of these different folks have helped in pulling together the project. And we’ve gotten some funding from an anonymous donor in order to hire our litigation director, Amber Qureshi.
And what we’re doing, together with the staff here at Berkeley and Amber and our deputy director, Graeme Blair, who’s a political scientist at UCLA, is we are seeking frequent updates of these datasets from the government.
We’re asking the government every month to update these datasets. And when we don’t hear back, we’re suing because the Freedom of Information Act has a provision that says, if you don’t hear back within 20 business days, you can go to court to enforce this law. So that’s the basic model. We’ve got this group of people, we know which datasets to ask for.
We are requesting them regularly. When we don’t get them, we’re suing the government to get them. And when we do get them, we’re putting them up on the website for everyone to see. Now, maybe the one thing to emphasize here is that what we’re after is the raw data that the government is keeping itself. These are the datasets that the government is actually using for its own internal purposes. And what that means is that these are individual-level datasets.
“Individual-level” just means there’s a row for each thing that happens. So there’s a row for each time ICE arrested someone in the spreadsheet, or there’s a row for each time somebody was booked into detention in the spreadsheet.
And it’s really important to get that data at that granular level, because what that means is you can do anything you want with it in terms of analysis. You can figure out exactly which questions you want to answer with it, and you don’t have to depend on the government aggregating it up in some particular way in order to determine which questions you ask.
GWYNETH SHAW: What are some of the insights and patterns you and others using this data have found so far?
DAVID HAUSMAN: Bloomberg News did an analysis where they looked at how often somebody who’s arrested is transferred somewhere far away in the country to a faraway detention center. We’ve heard about this happening in a couple of high-profile cases.
Somebody gets arrested, for example, in New York City, gets shipped to Louisiana, far from their family, far from their attorneys, and that obviously has huge consequences for their immigration case. And so Bloomberg News took the data, which has every single time somebody gets booked into a detention center anywhere in the United States, and it analyzed it and found that, yes, in fact, already in the first month of the Trump administration, these transfers to faraway detention centers were increasing.
So that’s one example. Another example is the Washington Post relied on the data in an article that showed that lots of the arrests that the Trump administration had boasted about in that first month were actually transfers from local jails, rather than arrests out on the street. And in fact, this is really typical of the way Immigration and Customs Enforcement, ICE, operates.
Mostly, it’s not very good at making arrests. What it does is it goes and finds people who are already locked up somewhere else, and it transfers them to ICE detention. And so the data was useful in showing that as well.
The sky’s the limit for this. And actually, in thinking about how this can be useful, I like to think back to the past. I think there’s an original sin in immigration data use, which is the way immigration data was used in the first case challenging prolonged mandatory immigration detention.
So in this case called Denmore v. Kim, back in 2003, the ACLU challenged detention without any access to a judge. So if you’ve been convicted of certain crimes, ICE can actually hold you without giving you any chance to show that you should be released on bond while you fight your immigration case.
And the ACLU challenged that law in 2003, and the Supreme Court upheld it. And a major part of the reason why the court upheld that law is that it relied on statistics that the government produced in the case that showed that, generally, immigration cases were over very quickly, and therefore, people weren’t being detained for very long.
And it turned out later, 15 years later, the Solicitor General discovered that, in fact, that analysis was mistaken. The government had just made data errors in its analysis. And it turned out that for people who had appealed their case, actually, they waited a really long time in detention, months and months, much longer than the government had said in its submission.
And because the other side, the immigration rights side, didn’t have access to the government data, there was no way for them to check. And so that’s an example of what motivates me here in getting this data that journalists, policymakers, immigrants, rights advocates should all have access to this data, so that we can all know what’s actually going on in terms of enforcement.
GWYNETH SHAW: And this can be helpful in terms of holding the Trump administration accountable in terms of the numbers that it’s putting out there now, in terms of numbers of arrests and things like that, right?
DAVID HAUSMAN: That’s exactly right. So for example, at one point, bizarrely, the new administration was claiming that all of its new arrests had taken place at large. And that’s just clearly not true and is contradicted by the data, because most ICE arrests happen in jails and prisons as I was saying.
Mostly what we call “ICE arrests” tend to in fact be transfers from one type of detention facility, a local jail, to another type, an ICE immigration detention facility.
GWYNETH SHAW: You’ve studied immigration enforcement now under both Trump administrations. What’s different this time around?
DAVID HAUSMAN: I think the differences between the first Trump administration and this one are enormous. And in fact, I think we’re learning in every sphere that the first Trump administration is a very poor guide to what’s happening in the second Trump administration.
So under the first Trump administration, in the interior of the United States, not that much happened, quantitatively. There were all sorts of changes in terms of law and policy, but the numbers just didn’t go up that much. Under the second Trump administration, it’s very clear that they are focused on increasing the numbers, and we’re seeing that now.
So under this new administration, it’s already the case that we’re getting close to the level of enforcement that we saw at the most recent peak, which was during President Obama’s first term. So lots and lots of arrests within the United States.
That’s one thing that changed. Another thing that’s changed is the political context. Millions of people arrived at the border seeking asylum under the Biden administration, and the politics of the border really shifted with a real increase in hostility towards migration in the U.S. population. And I think that’s part of the political context here.
So that’s the second thing, is that the Trump administration feels more empowered to pursue these policies. But the third thing is that I think we shouldn’t mistake some of Trump’s actions for immigration policies.
For example, with deportations under the Alien Enemies Act, or arrests and detentions of people for protest activity. These uses of immigration law are really only incidentally related to immigration. These are instead attacks on democracy and the rule of law that are clothed in immigration policy.
And so in a lot of ways, the immigration story under this new administration is often not about immigration.
GWYNETH SHAW: After Trump campaigned on the promise of mass deportations and you saw at the Republican National Convention people holding up signs saying, “Mass Deportations Now,” do you think what happened in Los Angeles is the beginning of a major enforcement escalation? Or do you think that it’s more related to what you just said, which is a show of force that is wrapped around immigration enforcement but that is aimed at a different group of people?
DAVID HAUSMAN: I really have no idea. But a couple of things we do know. ICE has really ramped up the number of arrests it’s making very considerably. What we don’t know is what will happen next.
So one of the reasons it’s so important to have data on what’s happening in the immigration apparatus is that I like to think of it sometimes as like a giant ball of string, where you pull out a thread, and you just often can’t get very far unless you’re able to actually have data that allows you to see what’s going on within that ball.
And the other way in which it’s like a ball of string is that from the government’s perspective, when it pulls on one thread, it may not be able to untangle things either. So here’s what I mean by that in terms of arrests.
When the government arrests people, it doesn’t mean they’re going to be deported. The government can arrest someone, and if that person has a good argument for why they can remain in the United States, then that person will be able to remain in the United States.
If that person is eligible for release from immigration detention, then that person can seek bond before the immigration judge and be released from immigration detention. If ICE doesn’t have enough immigration detention beds to hold all the people at arrests, then ICE will release those people as well.
So arrests themselves don’t tell us that much about how much the administration is able to ramp up its effort. Of course, it’s certainly a sign that the administration is trying to do that, and we absolutely know that it’s trying to do it.
And in fact, there are reports of widespread demoralization within ICE as administration leadership has been really dissatisfied with numbers from ICE. But just pulling at one thread, just increasing the number of arrests will not necessarily drastically increase the eventual number of deportations.
GWYNETH SHAW: As people around the country are dealing with their neighbors and friends and sometimes family members being pursued by ICE, what, if anything, can communities do to protect people? Is there anything that can be helpful in this situation as the administration pursues this policy?
DAVID HAUSMAN: I think the first thing to say is that these kinds of raids, these workplace raids, for example, at a Home Depot, are still really the exception rather than the rule in terms of how immigration arrests get made.
Again, most immigration arrests are actually transfers from a jail or a prison to ICE custody. And what we’re seeing right now is an effort from the administration to increase arrests and therefore to increase some of these other kinds of arrests, like worksite raids.
And what we’re seeing in response is outrage from people, because that’s such a visible kind of enforcement. We can actually see ICE going into communities and arresting people. And it’s also such an indiscriminate kind of enforcement. It’s a kind of enforcement that has ICE just sweeping up anybody it encounters and arresting those people. And that really doesn’t match the claims that the administration made that it wanted to somehow relate its enforcement to crime.
And so when the administration does these kinds of indiscriminate raids, I think the first thing we see is political opposition. And so that’s a first reason to think that if the administration really tries to get to mass deportations, to get to the kind of scale it’s threatened, I think lots of people are going to have a new understanding of what immigration enforcement entails and will discover that they really oppose it for the people they know in their communities who are affected by it.
So the first thing to say is that there may be a counter-reaction to this sort of sweeping enforcement. The second thing to say is that in terms of what people can do, locally, the most effective policies in terms of slowing down federal immigration enforcement have been so-called sanctuary policies.
What those policies don’t do is stop federal agents from conducting raids. So worksite raids are generally not affected by sanctuary policies. There’s nothing a local government can do to say to the federal government, “You can’t come and conduct raids in public places in our city.” Sanctuary policies don’t affect that. What sanctuary policies do affect is actually the more common type of ICE arrest, which is that transfer from one custody to another. So the way that ICE likes to do those transfers is that when it gets a report of somebody being booked into a local jail, which it gets in an automatic way from the FBI, then it sends a request to that local jail for a person to be held for 48 hours beyond when they otherwise would be released, so that ICE can come at its leisure and arrest that person while they’re still locked up.
Those requests are called detainer requests, and the most important element of sanctuary policy is a refusal to honor those requests. And so that refusal, it turns out, has been really effective. So actually, some empirical work that I’ve done with these datasets from ICE shows that when localities put in place these sanctuary policies, deportations tend to go down by about a third in those places.
GWYNETH SHAW: Ahead of Trump’s inauguration there was a lot of coverage about the details of the immigration system and how it wouldn’t be able to handle anything close to what most people would consider mass deportations. Is that still true? Has that been borne out, do you think, in the last few months? I know we’re still really short of data for some of these questions, but do you feel like the system is prepared to deal with what the Trump administration wants to do?
DAVID HAUSMAN: I think it’s true that for the Trump administration to reach the kind of numbers that it threatened in terms of mass deportations, we would have to see a radical change in the way immigration enforcement works. And I don’t think we’re seeing that change yet.
We are seeing really significant ramp-ups in enforcement, and that’s happening in two ways. One way we’ve seen that is actually in decreases in arrests along the Southern border of the United States, where because the Trump administration now simply expels everyone who had encounters along the border, it arrests many fewer people along the border than before.
Those numbers were actually already way down in the last six months or so of the Biden administration, but they’ve gone down still further under the Trump administration. So the first big change we saw under the new administration is increased enforcement in the sense of a more draconian policy at the Southern border and therefore, decreased arrests along the border.
And at the same time, we saw a big increase in the number of people being arrested in the interior of the United States. And that increase has continued, and the number of people being detained in immigration detention centers has also continued to go up and has just reached a peak in the last couple of weeks with well over 50,000 people being kept in immigration detention at a time.
And that’s a high number. And I think we’re also seeing deportations from the United States start to creep up closer to where they were during the first Obama administration. So we’re seeing a really significant ramp-up in enforcement in the United States, along with lots of other changes that are intended to cause immigrants to be afraid.
We’ve seen stripping of protections for large groups of immigrants who actually arrived here with permission and, of course, spectacular deportations of people to prison in El Salvador with no due process whatsoever. So huge changes in terms of immigration policy, a real intensification in terms of the numbers. But still, if the administration were to start deporting, as it’s threatened, 1 million people a year, that would mean deporting close to 100,000 people a month.
So it would need to really quadruple the number of people it’s deporting every month, and that would require a huge expansion of the administrative apparatus that it uses for deporting people.
Now, it’s taking steps to try to do that. It’s opened new detention centers. It’s sought new funds from Congress to increase detention. It has brought lots and lots of new localities into collaborative agreements, so-called 287(g) agreements, which might allow those localities to make immigration arrests.
It’s taken a bunch of steps in this direction, but I still think that scale of change is unlikely. And that further, if we saw that scale of change, we would really see a change in the politics of immigration enforcement, because at that point, so many people who people know would start being swept up in this dragnet.
GWYNETH SHAW: What other things do you hope to add or expand to this project moving forward?
DAVID HAUSMAN: The thing we want is an update to the dataset. And so we’re pursuing that in every possible way. Beyond getting updates to the data, we are trying to help people use it. So we’re working now on a code book that explains what each variable in the dataset means.
We’re working on trying to find ways to make it easier to download parts of the data or to download the data in different formats, so we’re working on that on the website. We recently produced a Spanish version of the website, which is now available.
And so we’re just trying to take steps to make the data easier for people to interact with. And luckily, those are things we can do ourselves. In terms of the updates, we’re a little bit at the mercy of the government, but the government is also subject to the Freedom of Information Act, so we are hopeful that we will get there with the updates.