Migration, But Better: April 2026
More immigration op-eds I'm still getting grief for, the DEI failure as a policy design problem, and polls on whether I should post fewer hot takes.
Quick personal news before we get to the links. As this newsletter goes out, I'll be on my way to the Borders and Belonging conference at IAST in Toulouse, one of the more exciting migration gatherings in Europe this year. If you're there, come say hi. On the way, I've been trying to visit every Notre Dame cathedral I can reach, on the theory that this is what most of my non-American friends assume my employer looks like anyway.
At the actual University of Notre Dame, we just hosted the 2026 Midwest Migration Conference using a new workshop-format without presentation slides, gathering the new and existing work on the botched reception of Venezuelan asylum seekers in Chicago, the effects of anti-immigration policies, and the difference between various Chinese communities in Paris. Let me know if you want to participate next year.
Looking further out, I’ll be speaking at the Progress Conference 2026 in Berkeley on October 8-10 about immigration as a key progress issue. It’s one of my favorite conferences out there, both because the attendees are unusually thoughtful and motivated to make things better despite their disagreements, and because it’s one of the few places where people treat policy design as a genuine problem to be solved rather than an afterthought. If you’re going, come find me. And if you don’t, try to sign up and come anyway.
April was busy. At Migration Policy Institute, Caitlyn Yates and I traced 15 years of UK immigration policy swings and the salience cycles that drive them: salience rises, labor visas get cut, shortages follow, and visas quietly reopen. At The Atlantic, I published “How the Left Accidentally Bolstered the Nativist Right” (their title, not mine), on what we lost when the “I only oppose illegal immigration” norm collapsed. In the Frankfurter Allgemeine Zeitung, I explained why the German migration debate has gone off the rails, specifically how “well-intentioned disinformation” from the center-left feeds the far right. I released a Keough School policy brief, “To Protect Humanitarian Immigration, Build Public Trust First”, on how the sequential logic of my case applies to refugee and asylum policy.1 And, thanks to The Chronicle of Higher Education, I now have “AI Is a Better Researcher Than You” as a headline attached to my name forever.
OK, let’s move on to something a bit awkward. After two rounds of online pile-ons this spring, I’ve received some earnest private feedback from colleagues, with some worried about my mental health and some others worried about me becoming a “right-wing crank” for lack of a better word, all suggesting I should dial back how much I post on social media (I’m talking about short-form takes rather than long-form essays). I want to take that feedback seriously, so I’d like to also hear from the people actually reading this newsletter and following me on social media. Conditional on being factually accurate (which I still strive for believe it or not):
Here are the April links (linking does not imply endorsement):
Lauren Gilbert launched In Development this month, a new magazine on what works and what doesn’t in global development. If you liked her job market papers roundup and her thinking on evidence-based policy, this is the natural next step. The inaugural piece, a guest essay by GiveDirectly co-founder Paul Niehaus on how evidence supported the organization’s journey to delivering $1 billion in unconditional cash transfers, is worth reading in full. Immigration is a huge part of global development, so I’m glad Lauren is doing this. There is also a rumor that their next piece will be exactly on immigration.
Amy Nice and Paola Sapienza have a new Hoover Institution conversation on “Unlocking Global Talent: The J-1 and O-1A Visas”. It’s the clearest thing I’ve seen on the practical alternatives to the H-1B lottery for companies trying to hire international STEM talent. Amy is one of the sharpest practitioners working on legal immigration in Washington, and she’s unusually good at explaining the mechanics to non-specialists.
Luke Eure at No Idle Sitting argues Trump’s H-1B lottery policy kind of works. The $100,000 fee is clumsy and imprecise, but it does effectively raise the bar on employer-sponsored immigration in roughly the direction selection-based systems should go. Take our wins where we can get them.
We had some minor disagreements with Alex Nowrasteh lately, but when he’s right, he’s right. His new piece on the “culture crutch” is a banger. The problem with “culture” as an explanation is that it’s the most conceptually overstretched term still allowed in polite scholarly society. Whenever it’s actually useful, more precise concepts (norms, institutions, prices) usually do the work better.
Daniel Di Martino on selection, not origin, driving immigrant welfare use. When you control for pathway type, country-of-origin effects go away. The policy lever worth pulling is how people come (i.e., for needed work), not where they come from.
My friends and co-authors James Dennison and Andrew Geddes have a new book forthcoming soon, What Europeans Think about Immigration and Why It Matters (Oxford). It maps decades of European survey data and offers a proper framework for why immigration becomes politically salient when it does. I pre-ordered mine. Anyone serious about European immigration politics should.
Kelsey Piper at The Argument on whether a liberal society can do affirmative action, makes a point that generalizes well beyond DEI. As I’ve been beating this drum on immigration, the problem with unpopular policies isn’t usually that they’re sold poorly. It’s often that the policies themselves need to be different. Reframing only gets you so far when the underlying design is the thing voters are rejecting.
Andy Hall at Stanford has a must-read The Roots of Progress piece on how AI is already 10x-ing academic research. His line I keep repeating: “stop waiting for permission. The tools are here.”
Yamil Velez at New Instruments asks who actually answered this survey in a world where people routinely delegate tasks to AI. Survey research has to shift from detecting AI-generated responses to redesigning instruments that make authentic participation easier than delegation. Yamil was creatively using AI in polls before it was cool, and he’s worth following if you want to make sense of where public opinion research is headed.
Besides the aforementioned Progress Conference 2026, The Roots of Progress is still taking applications for the BBI fellowship (deadline: June 1). I’ll be advising and speaking on the human talent and potential track there. Highly recommended for any academic (or non-academic writer for that matter) seeking more serious public engagement.
As before, if you want me to write more about one of these or other related topics, let me know. Or if you agree that I should stay away from social media for some time before I turn into a complete crank, this is your chance to speak up, too :)
For those who likes listening to things, I also had two podcast conversations: Solutions with Henry Blodget on why 80% of Americans agree on immigration yet neither party can get it right, and WashU’s new Ideas Matter podcast with host Sandro Galea on immigration in a changing world.