This essay is a follow-up to my Washington Post piece and a companion to a recent talk I gave at the Progress Conference 2026. It is specifically addressed to people who already favor (skilled) immigration and care about economic growth (the abundance and progress communities) to make the case that they should be championing top talent immigration as a signature issue above anything else. The speech has been edited for clarity and lightly expanded, with sources and context in the footnotes.

Here is the number I want everyone who cares about progress and abundance to know: 80%. That is roughly how many American voters support bringing in more highly skilled immigrants. Remarkably, this number holds up no matter how you ask the question, no matter who you ask, no matter how you frame it. But this is also a pro-growth policy that neither party has made much of a priority. Nor, for that matter, have most of us here. Still, large majorities support it across party lines, including among those voters who are still angry about the border and skeptical about immigration in general.1

Subscribe now

Think about it. Eight out of 10 Americans actually agree on something, including most Trump voters. It is close to a prototypical “80/20 issue,” the kind of lopsided majority that political strategists spend their careers searching for and almost never find. For comparison, in recent surveys, expanding nuclear power polled at 59%. Many YIMBY proposals struggle to reach even majority support. Support for “deregulation” has never once hit 50% in a quarter-century of tracking. So, if you are building a coalition around growth, attracting talent is one of the most popular pro-growth policies available. Yet we treat it as too dangerous to touch.

So yes, I want to convince you all here to do more about immigration. Not any immigration, and not even the so-called “skilled” immigration generally, but at least the part where we bring in the world’s best people to work here. To be more precise, I mean people with exceptional abilities whom employers are competing to recruit, whose expertise they cannot readily find at home even when they are willing to pay for it. That includes leading scientists and engineers with highly specialized experience in AI and other key fields. I call this top talent immigration.

A lot of ink has been spilled on what immigration is or what effects it has had, and I have contributed to this discussion myself. But now, I am much more interested in what immigration could be. People can have all sorts of legitimate concerns about immigration, concerns that shouldn’t be dismissed, but still recognize that we could do better on the issue. We don’t have to agree about every migrant or even H-1B worker to agree that at least we want to get the top foreign talent here. Yet this broad support has done little to protect skilled immigrants from new restrictions and attacks.

So my argument for today has two parts. One is that top talent immigration can be unbundled from the rest of immigration policy, which now holds it hostage. We can make progress on the popular part without settling every other immigration dispute. And two is that just being okay with it is not enough. Attracting top talent should come first within the progress or abundance agendas,2 which means putting our money where our mouth is and sometimes prioritizing it over other issues. Talent is paramount, since the best people we recruit can help us accomplish almost everything else we want to do, starting with AI.

AI shows what we could gain or lose

Artificial intelligence—one thing I know for sure that a lot of us here actually care about—provides a particularly clear example of what we could gain from attracting top talent or lose by failing to do so. The vast majority of leading AI researchers in this country were born or educated abroad, with many coming from China. The American advantage in AI is, to a remarkable degree, an immigration advantage.3

Whether we care most about American leadership in AI or about AI safety, we should care about where the people actually building it choose to work. The US is still the leading destination for top AI and most other talent despite our convoluted immigration system, but China is becoming better at retaining its own researchers. Recruiting someone who would otherwise strengthen a lab in a competing country can improve America’s relative position. And since some of the pressure to build AI quickly comes from fear of what China will do, a stronger position can also give us more room to test systems before we release them. Any scientist educated in China or anywhere else abroad who wants to build a career here is a potential contributor to American research. With the right security checks, we should judge each person’s work on its merits and do our best to bring them here to stay.

But we also know that right now the system for bringing these people here too often works like an obstacle course. In a survey of AI researchers living in the United States, the vast majority identified visa and immigration problems as a serious impediment to research. A change in presidents or who controls Congress will not, by itself, solve this issue. Even a much friendlier administration in the future would still inherit visa caps and green-card queues, while employers would still have to work through the limited and convoluted routes available to a particular recruit.4 That is, unless we do something about it.

We can unbundle top talent from the rest

That brings me to the first part of my argument. I’m pretty sure most people here are already pro-immigration and certainly agree that we should bring in top scientists and engineers. The less obvious claim I want to argue is that we can pursue this top talent agenda without solving the entire immigration debate first.

Right now my sense is that the progress and abundance movements, alongside most people on the center-left and center-right, are hesitant to champion this issue. Let’s take, for instance, Ezra Klein and Derek Thompson’s Abundance, the book of the movement. It mentions skilled immigration but barely engages with it as a major growth lever. When I have voiced these concerns, progress-oriented writers and thinkers have told me privately, more than once, that immigration is all good but too polarizing to include explicitly in their growth pitch. Leading with it, the argument goes, could alienate allies and drag every other worthy reform from housing to energy abundance into a bigger culture-war fight.

I understand the instinct. I just think it concedes too much. It is also arguably counterproductive because if you don’t provide an affirmative case for some immigration that you need and voters already agree with, somebody else will provide an even more convincing case against most of it.

It is true that immigration in the abstract is toxic, because voters see illegal crossings and real visa abuse by people willing to break the rules. But the resulting backlash does not extend to welcoming top talent, especially if it comes via specific, well-designed, legal programs. In my recent book, I document at length that regular voters—including those who are not very informed about the issue—can still distinguish between different kinds of immigration. Someone can absolutely want the government to control the border and deport everyone who is here illegally, but still also want the world’s best scientists to work here—and not in China.

People who say we need order before more immigration have a point, and I have made the case for enforcing the rules myself. But what exactly are we waiting for when it comes to top talent? Border Patrol encounters at the southern border just fell to their lowest annual level in more than 50 years. We still need more administrative capacity, but the United States can screen and admit a limited group of scientists and engineers while we build it. We can make those channels work better without waiting for an agreement about asylum or legalization.

Yes, we have done this before

Immigration has always been divisive, yet America has been able to do this before, so we can do this again. In 1992, Congress passed the Soviet Scientists Immigration Act, giving hundreds of weapons scientists a route to permanent residence following the collapse of the Soviet Union. Despite its name, nobody thought of it as immigration legislation, and the bill cleared both chambers easily on a simple nonproliferation rationale. The idea was that bringing those researchers here would help keep their expertise from spreading to our geopolitical rivals. I’m glad that people in Congress could agree on something constructive related to immigration, at least back then, without agreeing on everything about the issue.5

The idea has since resurfaced across party lines in Congress and beyond.6 Most recently, for example, the Institute for Progress proposed an Office of Talent Assessment in the Defense or State Department to recruit leading foreign scientists. The proposal starts with 80 green cards a year for principal recruits in defense-relevant fields. I would personally favor something much broader, both in the number of people and the fields covered. But the general idea makes perfect sense: someone would have the official job of finding the people America needs for the sake of national security and helping them build careers here.

We know that some people and politicians will oppose expanding immigration of any kind regardless of how we describe it. The main historical lesson here is that unbundling is possible. In fact, you can design a specific immigration program around benefits both people and their representatives can see, and pass it despite all other disagreements.

Why broad support still loses

Now to the second part of my argument. So if most people want this, why does so little happen? That is, apart from the fact that Congress is not doing much of anything. In more than 15 years of studying immigration and public opinion, I keep coming back to the same finding: people who oppose immigration tend to consider it more important than people who support it.

The same research also found a difference within the issue. Immigration supporters are especially concerned about helping the less fortunate immigrants already here, while opponents focus more on preventing future arrivals. That leaves the talented people who might come here with few champions. They cannot vote, and the research they might do or the companies they might build do not yet have anyone depending on them.

I should admit my own field is no better. Migration researchers, especially outside of economics, tend to treat immigration as a humanitarian topic first. In my experience, advocates and academics like me are rarely moved by the plight of potential legal immigrants blocked by per-country caps or US businesses unable to recruit because of visa quotas. While presenting the research behind my first book, I was asked many times why anyone should care about unrealized skilled immigration when so many less fortunate immigrants are being deported and denied entry at the border right now.

It is an understandable question coming from a good place, but my answer has always been the same. We live in a democracy—so we need to do what’s good for us in our country and what is popular, even if it sometimes means prioritizing some migrants and policies over others. That also means being more attuned to trade-offs—if we want the world’s best scientists here, we should be willing to accept restrictions elsewhere, especially on distant family migration or asylum, in exchange. Besides, doing so can also help build a more trusted migration system that benefits everyone, including the most vulnerable.

This lack of sustained pressure on skilled immigration from either business or advocates helps explain how an administration can attack it despite its broad support. Over the past year, the Trump administration has imposed a six-figure fee on some H-1B petitions and then proposed a new version. Employers and states have challenged restrictions in court. But when supporters look elsewhere, restrictionists—who oppose most immigration and are now mostly in the Republican Party—have more room to act while in power.

What you can do now

Our Congress could and absolutely should expand legal routes for top talent, and there are some encouraging bipartisan proposals that still have a chance, from state-sponsored visas to keeping more American-trained STEM graduates here. But I have little optimism about major legislation in the current political environment, especially if it is anything comprehensive. Fortunately, folks here can actually do quite a lot without waiting for Congress.

If you are hiring or trying to get hired, make greater use of the O-1A extraordinary ability visa. The program has no numerical cap and no lottery. This visa exists, and many more companies and immigrants should use it. For some researchers, the national interest waiver, a green card you can apply for on your own if your work matters to the country, can work as well. Ask an immigration lawyer which route fits you or the people you want to recruit. Employers should pay for that assessment and the legal help that gets people through it.7

And if you or your business has access to the administration or even a large following on social media, this should be one of the things you raise. Explain which rules prevented you from hiring top talent you needed, especially if it was in the clear national interest. And push for better rules and faster processing while pushing back on clearly bad proposals like the newly proposed OPT fee or the elimination of the H-1B grace period.8

If you work at an AI company, you may be reasonably calculating that it is better not to be on the bad side of the current administration. You may also think that, with AI building and improving itself, you will not need as much top talent anymore. Even if that is true, getting the best people here as soon as possible becomes even more important. If AI systems will soon do much of their own research, the people who build and oversee them over the next few years will largely decide where that research goes. If the stakes are as high as you say, who gets to work on these systems here should matter more to you than staying on any administration’s good side.

If you are a philanthropist, you can sponsor the recruitment and legal help to process extraordinary ability and other visas that smaller employers cannot afford. The policy side needs people and money too, meaning analysts who read every rule the agencies propose and file comments on the record, and lawyers who challenge the ones that make top talent immigration harder. Of course, some organizations such as the Institute for Progress already do sustained work on this issue. But they still need more manpower and resources behind them to make politicians take it seriously.

Let me be clear. This year it is the proposed OPT and H-1B fees. Next year it will be something with a different name, so the people who care about this issue should have a standing capacity to respond, the way the housing and energy wings of this movement already do when some ruling moves against them. That newly built capacity needs to survive changes of administration.

If you work in government and want America to prosper by recruiting global talent, you do not have to concede this issue just because immigration in general is toxic. Whether you are a Democrat or a Republican, promising voters you will be tough on immigration does not oblige you to turn away a Chinese scientist who wants to be here and who is needed in this country. You can keep your popular commitment to enforcement and still improve the skilled pathways that voters like just as much. The government should also invest in the staff and systems that process applications. As I argued before, faster decisions on both admissions and removals reduce uncertainty for everyone, whether you think immigration in general is good or bad. Expanding paid premium processing for all sorts of immigration paperwork could fund more staff and shorter waits.9

And if you are a conservative, you know that you are a part of the coalition that is currently on the restrictive side of the immigration debate. But, if anything, this means that you should push for top talent immigration even more. Being selective about whom we admit is politically advantageous. So is showing voters, including the moderates still undecided between the two parties, that Republicans are not just hostile to all foreigners no matter what, but a party that wants what’s best for our country. Conservative parties around the world can attest to that.10

So, let me close with this. I’m telling you this both as an immigrant from Soviet Russia and as someone who recently came out in favor of American nationalism. I truly believe America is great because it is a place of freedom and opportunity where builders can turn their ideas into reality no matter where they come from. That is exactly why most of the world’s top talent still wants to be here despite all the hurdles we put in their way. This is our ultimate strength.

So let’s stop letting paperwork block top talent for no good reason. In the end, if the progress movement cannot win this fight, where both the evidence and the public are already on our side, we won’t be able to win any of the harder ones.

Share

I’m excited to welcome a new guest essay by Steven Denney, a political scientist at Leiden University who writes Pixels and Patterns. It continues our more practical series on professional writing with AI tools. Although this one might seem a little nerdy, I promise the payoff is worth your time. Instead of worrying that AI will make things up, Steven explains how you can use it as a superhuman fact-checker on your side to boost your human writing. It’s especially useful for anyone who already uploads sources to AI and wonders why they would need anything more.

Imagine you are doing some research, and you need to write something based on it. This could be a paper, a concept note, a literature review or something similar. In the process, you have been downloading papers and collecting them in a folder – or maybe they’re just sitting in the downloads. The naming of the papers is a bit all over the place. Some of them have readable names, others have really long strings of numbers that you got from whatever website you downloaded them from, and some are just called something like “Download (3).pdf”. Or maybe you’re a stickler for the FAIR data principles and are very organized and have been downloading and filing the papers as you go.

Subscribe now

However you organize them, if you’ve uploaded a few sources to ChatGPT or Claude, you know it helps in the moment. But you can’t be sure what it has read, and next week you start over. The fix is to keep a text copy of each source that the AI can search, in the same folder as your draft and notes. Then every session starts from the same sources, and the AI can do more than summarize them. It can check your writing against them, telling you whether a reference is real and whether a source says what you claim, and quoting the passage so you can judge for yourself. Below, I explain how I set this up and how you can try it.

Back to that folder of papers. At some point, out of curiosity, you drag a few of the papers into your AI tool of choice and start to ask questions about them. You spend a good amount of time interacting with it, discussing the papers regarding what they say, how they relate to what you are working on, how they compare to each other, and where certain claims show up in them. (I trust you’ve read the papers yourself, but it can be helpful to have the AI “read” them as well.)

When you come back to the piece, whether you start a new conversation or go back to the same one, you are going to have a hard time picking up where you left off. You’d have to figure out what you actually did last time, which papers were relevant, what you concluded about them, and so on. The problem is not really that you have to upload the PDF again. It’s that this way of working with sources in a conversation doesn’t help you stay organized as you read, bring in new sources, and check your claims. The last of these matters most and it’s not (just) because people are offloading the work of checking source claims to AI. Citing a source for something it doesn’t actually say is a problem pre-dating AI. A 2015 meta-analysis of 28 studies of medical journal articles found that about one in four citations had errors in how they used the cited source, minor ones included, and about one in eight had a major error. Perhaps AI is exacerbating this problem, but the problem is not new.

Some tools can help out a bit here. A Project in Claude keeps documents available across related conversations. Zotero lets you collect and organize your sources, take notes, and keep PDFs and bibliographic information together. I’m not arguing against either, to be sure. But neither quite works the way that is optimal for AI-integrated workflows. As Alexander wrote in August, uploading your papers to a project folder “changes what the chatbot knows; an agent changes what AI can practically do with the files you share.” So I came up with a way of organizing my sources inside a research project (where you also have the manuscript, notes and decisions) that an AI agent can inspect as well. I call this a “research knowledge base,” borrowing the terminology from the AI researcher Andrej Karpathy, whose own version, an “LLM wiki,” I come to below. It contains:

• The original sources, renamed but otherwise untouched.

• A plain-text copy of each source, which the AI can search.

• A bibliography, with one entry for each source.

• Instructions on how to work with the knowledge base, including a number of AI “skills.”

Figure 1 shows how these pieces fit together in a project folder, and how a citation in your draft leads back to its source.

Figure 1. The knowledge base inside a project folder
On the left, the project folder keeps the original PDFs, their searchable copies and the bibliography together. On the right, a citation key in your draft leads to its bibliography entry, then to the searchable copy and the original PDF. The example files are invented.

Where I part ways with Karpathy

Karpathy, whose insight into AI best practices and how new technology and education will interact is some of the best out there, recently wrote about a similar setup he has built, which he calls an LLM wiki. It uses AI to write summaries and other notes on top of a set of documents and to keep them up to date. He frames the problem much as I do above. Without the wiki, in his words, “the LLM is rediscovering knowledge from scratch on every question. There’s no accumulation.” I describe his setup in more detail below.

In Karpathy’s setup, the original documents are kept as the source of truth, and the LLM cannot change them. But questions are answered from the wiki, and it’s the wiki that grows. For my purposes, it seems better to work directly from the sources, with the AI taking me to them, rather than only reading AI-produced summaries and the connections between them. Summaries will always leave things out and run the risk, although smaller than it used to be, of hallucination.

So, I will expand on Karpathy’s idea a bit, to explain how I work with my sources. Next to each original source, I keep a plain-text copy (which LLMs can inspect and search) that is linked to bibliographic information about the source. Notes and connections can be made within the collection, but they stay separate from the source texts. As the copies can be “messy” (we will see why), I don’t recommend treating them as the only source of truth about a source. They are working copies.1

Building your knowledge base

To follow along with the instructions I’ll provide below, I strongly advise using Claude Code, an AI agent that can work with files and run other tools on your computer. I advise also installing a set of additional instructions that I have created, which I call Open Science Skills.2 These “skills” are instructions that can be used by an agent to perform certain tasks like setting up a research project, work with a source, check citations in a manuscript, review a manuscript, and so on. Some of them also have scripts or other files that help the agent perform the task. You don’t need to know computer programming to use them. They also need some other tools in order to work. You can read about how to set them up in the getting-started guide.

Setting everything up takes about half an hour, and you only have to do it once. You need a paid Claude plan (Pro or above), or, if you use OpenAI’s Codex instead of Claude Code, a ChatGPT plan that includes it. Then you install the agent, the skills and the PDF converter, which requires Java. The guide is written for someone who has never used the command line before, but recommends that they go this route. If you run into problems installing the skills or the converter, you can ask the agent to help you fix them – and even just “talk to your terminal.” Once it’s set up, it takes the agent a minute or two to add a new paper, and a few minutes for you to check that it named the file correctly and to look through the conversion. Even if you have never worked with agents before, it’s really quite simple. Checking the agent’s work takes some care, and the knowledge-base walk-through on the AI for Research site lets you try that on made-up sources first.

All my skills start with oss:, short for Open Science Skills. You run each skill by typing its name in Claude Code. For this post, you only need a few of them, plus one command the first one adds to your project:

• /oss:research-repo sets up a project folder for your sources.

• /process-source adds a new paper to it.

• /oss:citation-check checks that your references exist and match the publisher’s record.

• /oss:fact-check checks whether your sources say what you claim.

Then, here’s how the process works. I open Claude Code in a folder where I want to set up the project (it can be an empty folder) and type /oss:research-repo. The agent turns the folder into a repository, which is just a fancy word for a project folder that also keeps track of changes to the files in it. If there are any issues, you’ll be notified and asked how you want to deal with them. (I’ll say a little more about that below.)

Now, let’s say I have a paper that I want to add to the project. I drag it into a folder for incoming sources (sources/unprocessed/). The skill will tell you where this folder is if you need guidance. Say the paper is called “Download (3).pdf”. I then type /process-source and the agent will start to process the source. (This command has no oss: prefix because /oss:research-repo added it to your project. You can also just say “process sources” and the agentic system will know what to do.) It will first try to figure out the title, authors and year of the source. Check these, because every later step uses the name it gives the file. Then, it will move the source to a folder for sources, and rename it to something like ferreira-nair-2021-compulsory-voting.pdf. It will not change the content of the source.

Next, it will convert the source to a Markdown file, which is a plain text file that can be read by an LLM. The conversion is done by OpenDataLoader PDF, the open-source converter the skill currently uses. Finally, it will create an entry in a bibliography file, where it will store bibliographic information about the source. The entry will have a citation key, which is a short identifier for the source. In this case, the citation key is ‘ferreira-nair2021.’ The key links a citation in your manuscript to its entry in the bibliography. Because the key and the file name share the authors and year, the agent can also find the files for the source itself, as the right side of Figure 1 shows. This naming convention is also saved in an instruction file, CLAUDE.md, which the agent reads in future sessions to understand how the sources are organized. Figure 2 shows the whole sequence.

Figure 2. Adding a paper in five steps
You drop the paper in, and the agent does the rest. You can check its work at steps 2 and 4, where it reads the paper’s details and where the converter makes the searchable copy.

Converting a PDF to text is not always clean, so check the text copy against the original PDF. Two problems come up often. One type of problem is when the PDF is a scan, and there is no text layer in the PDF. In this case, the conversion will not work, and you will need to use a tool to perform OCR (optical character recognition) on the PDF first. The skill will recognize when this is the case, and will file the source but mark it as not yet converted, as with the scan in Figure 2. For a paper or two, you can just ask Claude Code (or Codex) to read the scanned pages and write them out as Markdown, which your subscription easily covers.3 But you should still check the OCR’d text against the original.

Another type of problem is when the PDF contains tables. The conversion may not be perfect, and the table may not be converted correctly. For example, in a small results table from the walk-through, all the numbers came through, but the column headings were merged into the section title, and the two rows ran together on a single line. So, the information is there, but it’s not very readable. In this case, the conversion log didn’t flag any problems, which is why you should check the conversion yourself.

Sometimes you’ll encounter a source that you don’t have, either because you couldn’t find it, don’t have access to it digitally (e.g., a book), or because you couldn’t convert it. In these cases, you should still keep it in your knowledge base in the form of a summary Markdown file, kept apart from the text copies and marked as not having been checked yet. That way, you can still track it in your knowledge base and can come back to it later if you need to. This kind of approach is a bit closer to Karpathy’s LLM wiki idea. And it’s better than having nothing. But for researchers who need to check things against full sources, it’s, of course, not ideal. This will likely require something old school – getting a physical copy of, say, a book and checking it yourself.

Keeping a record of your work

Having a knowledge base is one thing, but you also need a record of the work on the project itself. This includes your research question or writing goals, notes about the sources you’ve read and how they relate to your project, decisions you’ve made, and things you still need to do. This type of information should be kept separate from the sources, because it is your interpretation and plans, not part of the sources themselves.

One way to do this is to save notes from your conversations with the agent. For example, if you’ve had a conversation with the agent about a certain comparison between sources or some ideas about how to synthesize arguments, you can save that comparison as a note. Or, if you’ve made a decision about the project, you can write that decision down, along with the reason for it. I generally recommend that you ask the agent to update two types of notes at the end of a session:

• A handoff note, which is a note about where you are in the project and what you need to do next.

• A session log, which is a log of what you’ve done in the session. This log should not be updated, only added to.

I have a skill for that, /oss:finished, which you use to close a session. The skill tells the agent to keep the two apart, and also to keep work that’s been done (and verified) apart from work that’s only been tried or talked about. But always make sure the work has actually been done. It’s very easy for an agent to say work is done when it isn’t, and to update earlier entries in the log when it shouldn’t. So check that the log is correct before you save it.

Another way to keep track of your work is to use a version control system like Git, the tool behind GitHub. This will allow you to make “checkpoints” in your work, so you can see what has changed since the previous checkpoint and go back to previous versions if needed. These checkpoints are called “commits.” Also, if you upload your project to GitHub, make sure to set the project to private. Although the papers have been converted to text, they are still subject to the same copyright as the original papers.4

Now that you’ve saved your notes, the next time you start a session, all you’ll need to do is have the agent read the project instructions and the last handoff note (/oss:sitrep, the partner of /oss:finished, does this), and open any sources it needs. You still have to interpret what it tells you, but you won’t have to rebuild the project.

Checking your references and claims

You can run two kinds of checks. One is to have the AI agent check that the references you have listed actually exist and that the details in your reference list match those held by the publisher (including the DOI). Use the /oss:citation-check skill for this. This may become more important as more people use AI, since AI can make up references. The agent will flag any it cannot confirm or with which it has concerns, for you to look at.

The other, more important check is whether the sources you cite actually say what you claim. Use the /oss:fact-check skill for this. The agent can do this because it has the knowledge base, with text copies of the sources you cite. If it finds the part of a source you are basing a statement on, it will quote it to you and say whether or not it backs up the sentence you have written. You can then check the quote against the original yourself. Only you can decide if the source itself is correct.

Here’s an example from a paper of mine on public support for immigration policy, which asks whether the way a policy is made affects support for it. To answer that question, I drew on research on presidential power in the United States, an area outside my immediate area of expertise. I wrote the following sentence in a draft:

“Studies of unilateral policymaking similarly show that citizens penalize executive action relative to legislative action (Reeves & Rogowski 2016; Christenson & Kriner 2017).”

In other words, I was claiming that the two studies I cited showed that people react less favorably to a policy adopted through executive action than to one adopted through legislation.

I was a bit wrong there. The agent found the relevant passages in both studies (Figure 3). Reeves and Rogowski did not compare support for executive and legislative policymaking. They found generalized support for unilateral powers to be very low, with only about a quarter of respondents supporting them. Christenson and Kriner did compare the two, in experiments in which respondents read about the same policy but were told either that it came from executive action or from legislation, and there was no statistically significant difference in support. Support depended mostly on respondents’ partisanship and whether they agreed with the policy. So neither paper supports the claim that citizens penalize executive action, and the agent suggested narrowing the sentence to what each paper actually found. That part of the paper now reads:5

“Studies of unilateral policymaking find that Americans express low generalized support for unilateral powers (Reeves & Rogowski 2016) while judging particular unilateral acts largely by whether they agree with them (Christenson & Kriner 2017).”6

Figure 3. Checking a claim against its sources
For each paper, the agent quoted the relevant passage and said what it supports and what it does not. The bottom box is the corrected sentence that replaced my original claim.

The original claim does have support elsewhere. In a later paper, Reeves and Rogowski do show that people react more negatively to policies adopted through unilateral action than to policies adopted through legislation. I had cited the wrong paper, and looking at my reference list would not have caught that, because the reference itself was correct. Across the whole paper, the same check led to seventeen corrections to how I cited sources.

It’s a good idea to save these checks, including the date and the model used, so that you can run the check again later on the same text, perhaps with a newer model.

Try it yourself

If you are new to AI agents, start with the getting-started page on the AI for Research site. It shows you how to install Claude Code and the Open Science Skills.7 When you are set up, try the knowledge-base walk-through on the same site. It uses some made-up sources, including a scan that needs OCR and a table whose structure breaks on conversion.

Once you have gone through it, try setting up a knowledge base for one of your own research projects, and add a single paper that you want to use in it. At the end of the session, have the agent write a handoff note on what you have done and what you need to do next. Then open a new session, and ask the agent to find the passage in that paper that bears on a claim you have made in a draft, and whether it supports, qualifies or contradicts the claim.

Doing research and writing in 2026 means checking your claims with the best tools available and saying how you did it. In a short note posted in early October, the Yale political scientist Peter Aronow argues that, at least in quantitative research, a paper without a statement on whether the authors used LLMs is now cause for suspicion. If they used one and did not say so, they misrepresented how the research was done. If they did not use one, they “declined an inexpensive and effective check of their results without acknowledgment or justification.” I’d extend that to anyone who writes professionally and cites sources. In 2026, refusing to let AI check your claims against your sources, when the check is this cheap, is arguably a form of bad research practice.

P.S. Other ways to build a knowledge base

Karpathy uses his LLM wiki for personal knowledge bases on topics he is researching. The idea is to build a wiki on top of your original source documents. The wiki holds notes about those documents, such as summaries, comparisons, concepts and syntheses, in a format the machine can easily read (ideally Markdown). There is also a file of instructions, which the LLM reads, on how the wiki should be maintained. When you add a new source document, the LLM reads it and adds or updates notes in the wiki as needed. He writes that the idea is related in spirit to Vannevar Bush’s Memex of 1945, in which readers could build trails between documents. Using AI to maintain those trails is what is new, and for project-based researchers, it is the exciting part.

One feature of Karpathy’s approach that I like is that you can look at it. He reads his wiki in Obsidian, a note-taking app whose graph view, he writes, is “the best way to see the shape of your wiki.” He also has the AI answer with slides and charts. In May, he suggested asking for answers as web pages, with the prompt “structure your response as HTML,” and then opening the file in a browser. Christopher Kenny’s workshop demo for political scientists is likewise a set of linked notes meant to be opened in Obsidian. It would not be hard to do any of this on top of a research knowledge base, and it may be worth exploring. But a graph or a web page is another kind of summary. To check a claim, you still go back to the text copy, and from there to the original.

I am not the only one who goes back to the sources. The journalist Casey Newton built a wiki like this for his newsletter, with more than 1,440 pages. He describes using it to refresh his memory before writing, “while also opening up the original sources to make sure nothing I was about to say was hallucinated.”

In the latest development in knowledge base developments, Soyeong Jeong and colleagues at KAIST and Microsoft recently posted a preprint “Follow the Entities: A Corpus Map for Agentic Search,” on how to arrange a large collection of documents so that an AI agent can find its way around it. They arrange it as a map, with one page for each entity, such as a person or a product, that comes up in two or more documents. Each page links to every document that mentions it, and each fact is tagged with the document it came from. Tested on thousands of workplace documents, the map improved the agent’s answers and cut its costs. An AI-written wiki modeled on Karpathy’s did worse than plain search with three of the four models they tried. The map is like a research knowledge base in that it points back to the sources instead of replacing them. The problem is that it points you to entire documents, so you still have to open them and find the passage that bears on your claim. It is also geared toward much larger collections of documents than any individual project or paper would be. It’s worth exploring but I don’t find it well matched to a researcher’s way of using these kinds of systems.

For something as small as a few hundred papers, it makes sense for the agent to just search the whole thing every time. If I were to add something new drawing from more recent work and ideas, it would be a simple index of the people, parties, datasets and cases mentioned in my sources, and the files they appear in. I am tinkering with that approach at present.

Share

Some good news first. In Our Interest received the Best Migration Book Award from APSA at this year’s meeting in Boston. I am also grateful to colleagues who checked up on me and who appreciated my stances and provocations even when we disagreed. If you liked the book, please rate or review it on Amazon and spread the word.

This issue is going out a few days early because this is the last call for two chances to meet in person in Washington, DC next week. On Thursday, October 1, the Making Immigration Better workshop I am organizing at Notre Dame’s DC office (1400 16th St NW) opens with a public reception and a keynote from Johannes Haushofer, the Cornell development economist who founded Malengo. Drinks start at 5pm and the talk around 5:40, and the registration form stays open until we hit capacity. Then on Saturday, October 3, from 10am to noon, I am hosting the first ever Popular by Design reader meetup over coffee in Dupont Circle. RSVP on Luma and I will send the exact location to everyone who does. If you are in the Bay Area instead, I will be at the Progress Conference in Berkeley the following week (October 8-11), making the case that the progress movement and the abundance crowd should go all in on top talent migration.

Subscribe now

Later in October I am taking the show to the UK for a mini-tour to present my new book project, Immigration for Realists (I’ll be writing more on that soon). For those of you in London and nearby, here is where to find me:

• Tuesday, October 20, 5:30pm. London Migration Research Group seminar at LSE.

• Friday, October 23. Oxford, the Department of Politics and International Relations research colloquium, at lunchtime.

• Monday, October 26, 2pm. King’s College London, PPE research group seminar.

• Wednesday, October 28, 4pm. Brighton, the 13th SCMR-JEMS Annual International Migration Conference at The Old Courthouse. The conference is open to all, and you can register on Eventbrite.

• Friday, October 30. Newcastle University.

In case you missed it: Earlier this month I published How Australia Already Made Immigration Popular and How to Keep It That Way. I wrote the original version for an Australian magazine last month, but reposted it here with Germany in mind, where the AfD had just taken almost 44 percent of the vote in Saxony-Anhalt. Then, after the far right finished first in yet another German state while the mainstream right lost ground, I published Why Governments Don’t Have to Restrict Immigration to Save Democracy, an updated version of a research note I wrote in 2024 for the European University Institute’s Migration Dilemmas project, in response to Zsolt Kapelner.

Book mini review of the month:

Monica Prasad’s Problem-Solving Sociology: A Guide for Students (Oxford University Press) is written for sociologists who set out to solve problems but end up only describing them (or mostly complaining about capitalism and racism, to be frank). Prasad names three traps of doing such research as usual, chief among them describing the victims and critiquing other people’s solutions without offering anything yourself. Her fix is to study the people and institutions who could actually change the situation (who may even be the “villains”), and, in her words, to “spend less time cataloguing [victim] suffering and more time cataloguing what causes it.”

I thought it was a great book and would recommend it to any researcher, including folks outside of sociology and academia. The author is obviously left of center, but given that her main target audience is extremely left-wing sociologists, it is remarkable how much she pushes back on the field’s habits, from defending objectivity to insisting that refusing to solve problems is a normative position too. Her observation that a literature focused on describing the oppressed ends up feeding the very stereotypes it wants to fight describes the humanitarian focus of a lot of immigration research pretty well too. My only reservation is the audience fit: to non-sociologists it can read a bit too jargony. But I also really appreciate the humility of an author who admits she has not solved any social problems herself but is pushing this agenda anyway.

Here are the September links (linking does not imply endorsement):

• reads the AfD’s shocking near-44% in Saxony-Anhalt as the firewall breaking. Since many folks asked me about the result, I wrote a brief FAQ with links. The short version: people vote for anti-immigration parties because they don’t like immigration, particularly when it is a nationally salient issue like it is now. Voters care about what they think immigration is doing to Germany as a whole, and it is now almost impossible to deport anyone from Germany, even when their asylum application is denied or they commit a crime, as I wrote in July.

  • argues that liberalism needs a new philosophy of immigration, built on three principles: migration is not a human right, immigration law must be enforced, and immigration should benefit the citizens of the receiving country. My readers will find most of this familiar, but I am glad to see it stated this plainly from the liberal side.

  • Anton Jäger in the New York Times says there’s a reason Europe is so obsessed with immigration, and the reason is that “Europe doesn’t know what it is.” That’s a lot of smart words to conclude Europeans’ worries about immigration are actually about something else. Folks should consider the possibility that Europeans vote for anti-immigration parties because they genuinely don’t like migration or how their governments manage it.

  • The same week, a chart made the rounds showing that far-right support is highest where there are the fewest foreigners. I think folks should stop using the negative correlation between foreigner shares and far-right support as proof that immigration isn’t at fault. Cosmopolitans self-select into areas with more foreigners, and people care about migration as a national issue beyond their own locality.

  • replies to me directly: yes, voters who fear migrants are sociotropic, and? He grants the finding and argues it cuts the other way. Voters with no skin in the game have no incentive to check what they absorb from the media, so sociotropic concern makes them easier to mislead. Fair enough, but my response is that the same “rational ignorance argument also applies to pro-immigration cosmopolitans, even if they are on average better educated.”

• at Migration Weekly asks how Australia should manage partner visas. For my US readers who think literally everything we do is dysfunctional: Australia charges an A$11,000 upfront fee for a marriage green card, then grants only a set number a year, and that quasi-cap is why couples now wait two years or more. Doesn’t seem ideal. His fix is a demand-driven program, with the honest caveat that governments can always tighten eligibility through other means.

• ran economics papers through AI detectors to ask whether writing is AI-generated. Almost nothing was flagged in 2022, while roughly half of recent arXiv economics papers and about a quarter of NBER papers now have at least one section that reads as AI-written, and the newer detector version finds even more.

  • Incidentally, can someone start a prediction market on when academics will give up on shaming their colleagues for using AI to improve research? Possible resolution: a major academic publisher adopts an explicit policy that 100% AI-written papers are OK. My bet is within the next three years.

• at Securing America’s Promise argues that Democrats are still scared of immigration. Hakeem Jeffries’s governing agenda leaves it out entirely, while Republicans have outspent Democrats on immigration ads by more than three to one, $53 million across 88 races against $17 million across 20. makes the general version of the point in Don’t Try to Bait-and-Switch Voters: you cannot campaign on one thing and govern on another, and you certainly cannot talk voters out of the issue they have already ranked first. Similarly to what here argued about populist breakthroughs, mainstream parties need to deliver visible improvements in immigration policy.

• rounds up some recent and quite surprising fertility news. South Korea’s marriages and births are both rising after its 2022 to 2025 expansion of marriage subsidies, and Japan is creeping into positive birth territory for the first time in roughly a decade, against his own prediction.

As before, if you want me to write more about one of these or other related topics, let me know! And see you in DC, Berkeley, or London.

Share

I keep getting questions from folks asking whether I’ve become a reactionary conservative. Well, here is one of my more liberal takes for you, if you wish. I originally published this research note in 2024 for the European University Institute’s Ethics of Migration Policy Dilemmas project, in response to Zsolt Kapelner’s article (which he later responded to). My apologies for a bit of the extra academic language here.

Germany’s latest state elections are bringing yet another first-place finish for Alternative for Germany (AfD), this time in Mecklenburg–Western Pomerania. The debate keeps returning to a point I recently made: people vote for anti-immigration parties because they don’t like immigration. I still think that is a good place to start. But it doesn’t follow that mainstream parties should oppose immigration in general, or that admitting more immigrants necessarily puts democracy at risk.

It is true, though, that politicians who favor immigration have reasons to worry about the electoral consequences of letting more people in. They can believe in freer immigration and still fear that, by pursuing it, they would help bring their anti-immigration opponents to power (which is presumably something their pro-immigration voters would also want to avoid). If these opponents are also willing to undermine democracy, the case for restricting immigration becomes much more appealing even to people who otherwise find such restrictions objectionable. They may see it as a necessary compromise to preserve the institutions that allow them to pursue a more just immigration policy in the future.

Subscribe now

Probably the best conceptualization of this “democratic dilemma” that I know comes from the philosopher Zsolt Kapelner. He generally favors immigration justice, but allows that restrictions might be justified in exceptional cases if they help avert a democratic threat. His question is what governments should do when their obligations to would-be immigrants conflict with their obligation to protect democracy. I agree that empirical research alone cannot settle every such choice.

But the evidence can tell us how often we should expect that conflict. While counterproductive backlash to freer immigration is possible, the risk is limited to particular types of immigration and particular policies. The backlash argument is thus not a good justification for most existing immigration restrictions, and it should not be used to argue against programmatic1 pro-immigration reforms that demonstrably benefit citizens in receiving countries.

Whose backlash to what, when, and why?

Some commentators use the term “backlash” as simply another word for any unexpected political disagreement. Kapelner defines backlash more narrowly, focusing on support for anti-democratic politics in response to immigration people perceive as excessive or otherwise undesirable. But “backlash” is also not a synonym for any adverse political reaction to change among some people. As many authors including myself emphasize, for the concept to be fruitful, it must refer to a particularly strong and sustained adverse reaction among large swaths of the public to political advancements that may be counterproductive to these very advancements. Since the idea of backlash necessarily implies uncertainty and a counterfactual argument about what would have happened to the cause and related behaviors without a particular advancement, reasonable observers can disagree regarding whether the backlash has actually occurred.

What constitutes pro-immigration advancements that voters lash back against? The backlash argument is interesting because people can adversely react to various advancements, from increasing immigrant presence and ethnic diversity to specific immigrant behaviors, pro-immigration rhetoric, and policies. Scholars have primarily conceptualized immigration backlash as an adverse voter response to the rising physical presence of immigrants or “anti-immigrant backlash”, as Kapelner puts it. However, focusing solely on immigration-induced demographic change may miss a significant part of the issue, as immigrant presence is neither necessary nor sufficient for voter backlash.2 While rapid immigration flows may increase the salience of immigration and the likelihood of anti-immigration voting, they can only account for a small portion of the variation in these outcomes. Moreover, evidence regarding immigrant presence is not very helpful for policymakers, as migration flows have multiple complex causes beyond migration policy. Therefore, focusing on how voters react to immigration policy changes regarding would-be immigrants (or “anti-immigration backlash”) is more analytically fruitful, especially when considering how policymakers should incorporate these public responses into their decision-making.

There are also numerous ways in which pro-immigration advancements can be counterproductive, from increasing public opposition to pro-immigration policies and immigrant rights to contributing to populist radical right voting and democratic backsliding. While many empirical immigration scholars, including myself, have primarily focused on the immediate signs of anti-immigration backlash, such as changes in attitudes toward immigration, Kapelner, like many other generalists, is mainly concerned with the broader threat that immigration backlash poses to democracy. The general idea is simple—open immigration may be just, but it can undermine the democratic institutions that make immigrant-receiving countries desirable in the first place.

But anti-immigration or any other backlash is arguably most plausible when it concerns immediate behavioral causes and outcomes where the causal chain is clear: “if you advance this particular policy too forcefully now, people will oppose this policy more, making it more likely to be rescinded than if you had advanced it more slowly.” When arguing that backlash against immigration threatens democracy more generally, the causal chain becomes much more convoluted and thus less likely to hold up. After all, not every voter opposes or cares about immigration, not every politician is willing or able to use anti-immigration sentiment for their electoral benefit successfully, and not every context is susceptible to democratic backsliding.

Furthermore, since much of politics is dynamic and “thermostatic”, for every reaction of immigration opponents, there may be a counter-reaction from immigration supporters. When it comes to the more proximate outcome of immigration support, for instance, there also appears to be a reverse backlash to the rise of populist far-right parties, as James Dennison and I found. In this sense, one may even argue that Kapelner’s paper and related work are part of this ongoing pro-democratic and pro-immigration backlash to populism and backsliding among intellectuals.

How significant is the threat of anti-immigration backlash to democracy?

To his credit, Kapelner acknowledges that high-risk hypothetical scenarios of the democratic dilemma are rare. He also notes that anti-immigration parties are not always anti-democratic and that further immigration restrictions may not be sufficient to defuse democratic threats when they arise. Nonetheless, I believe it is also important to re-emphasize that immigration is just one of many policy areas that people may have strong opinions about and vote based on. Immigration is not unique compared to other issues like redistribution, trade, crime, education, and foreign policy, all areas in which people may lash back against policy advancements in a democracy-threatening way.

Still, what does the best evidence say about the relationship between immigration and democracy? In a comprehensive study of Europe, Christopher Claassen examines how people’s attitudes toward democracy have changed over several decades in relation to immigration increases. His results show that immigration does not undermine people’s trust in, satisfaction with, or support for democratic institutions.

My own (admittedly somewhat underpowered) study of reforms examined legal labor and family admission policies across 24 European countries. I compared changes in immigration policy with changes in populist voting and immigration concerns within countries. I found no relationship between either pro-immigration or anti-immigration reforms and changes in populist voting. The findings are about reforms to those legal channels, which make up a large part of immigration into rich democracies. They do not establish that every possible immigration policy is immune to backlash.

What about the impact on democratic institutions? In his 2023 book, Larry Bartels also shows that democratic backsliding is not a grassroots process but rather a top-down one. In countries where democracies have recently eroded, such as Hungary and Poland, this erosion was engineered by politicians without any mandate from voters, whether related to immigration issues or not. Overall, while inferring causality or the absence thereof is always tricky, these recent empirical studies indicate that countries that liberalize and have consistently liberalized their immigration policies are not the ones experiencing threats to democracy.

What immigration policies do voters lash back against? Can some policies legitimize migration?

Regardless of whether we focus on what happens to immigration or democracy more generally, it is important to be clear about what people lash back against when it comes to immigration. Curiously, as Kapelner notes, it is not about all immigration: “[w]hen wealthy Scandinavians migrate between Norway and Sweden, no tension between immigration and democracy seems to arise.” Similarly, Ryan Pevnick acknowledges that “the evidence of backlash does not appear to extend to programs that prioritize highly skilled migrants.”

This is all good, but I would argue that Kapelner, Pevnick, and most of us probably have it backwards. Namely, we don’t appreciate enough the facts that “immigration” is not just one thing and that some migration is already popular. In fact, to the extent that anti-immigration backlash is possible, I would argue that it is confined to narrow types of unauthorized and mismanaged unskilled and forced immigration. Indeed, it is notable that most backlash scenarios Kapelner depicts involve involuntary migrants, whereas the role of labor or other types of regular immigration is not discussed until the end of the paper.

At the same time, according to the growing literature on immigration public opinion, most people support skilled and other selective immigration because they believe it benefits their countries. Fortunately, uncontroversial migration is much broader than just attracting the best and the brightest, and some pro-immigration policies can even further legitimize freer immigration.

In my 2025 book, In Our Interest, I show that voters generally support even very liberal pro-migration policies when these policies explicitly and straightforwardly benefit their countries. These policies, which I call “demonstrably beneficial,” can also be about filling labor shortages, reuniting families, increasing population growth, revitalizing declining regions, or even improving national security. When governments open legal pathways for foreign workers and their families through policies that demonstrably benefit their countries, voters do not lash back.

On the contrary, relaxing immigration restrictions on skilled and other economically beneficial workers can legitimize international mobility and increase public support for immigration in general. When consistently high immigration flows of workers come to be viewed as legitimate by the electorate, and voters start trusting their government on the issue, there is also more room for humanitarian immigration and other responsible policies even when they are not quite responsive to public opinion. Overall, when representative democratic governments enact programmatic pro-immigration policies that align with the preferences of their citizens, it ensures that these governments have a sufficient public mandate to lead on the issue and help the most vulnerable involuntary immigrants when a crisis hits.

In his later response, Kapelner agrees that growing immigration does not automatically produce a threat to democracy. His remaining concern is that racial and ethnic anxieties may prevent acceptance of favored migrants from extending to other groups. Those anxieties may also affect which immigration people call “mismanaged.” I agree that support for a particular immigration policy does not settle every question about asylum. But the possibility of those hard cases does not justify treating popular legal pathways as a threat to democracy.

Immigration is a contentious issue on which many reasonable people disagree. Policymakers who otherwise believe in the merits of more open immigration policies understandably fear backlash from their voters and thus tread carefully. Given the complexity of the issue, even the best-intentioned and sound reforms can backfire. It is also unlikely that there will be an ultimate policy solution that satisfies all relevant parties, even among those who are already pro-immigration.

Mainstream parties in Germany can take immigration concerns seriously while supporting reforms that let more people enter legally and deport more people who are there illegally. They need to distinguish which policies voters oppose from which policies voters would welcome. Perhaps immigration debates would be more productive if we talked about the relevance and consequences of the more specific “backlash to mismanaged migration” or “backlash to large-scale unauthorized migration” rather than “anti-immigration backlash” or “anti-immigrant backlash” in general.

Share

As many of my readers know, Canada has done more than any other country to shape how I think about immigration done better, which is why I keep coming back to it. But the question I get almost every time I write about it is "what about Australia?" My honest answer for years was that I did not know enough to say. So when the good folks at Eureka Street asked me to write about what my research means for an Australian audience, I saw it as an excuse to finally learn about the other most successful immigration country in the world.

Almost one in three Australians was born overseas. This is the highest share since the 1890s, and roughly double the American figure. Although some people now believe this is too many, for decades that grand scale coexisted with exceptionally welcoming public attitudes. Australia was the rare standing rebuttal to the claim that high immigration must sooner or later provoke revolt. The past few years, however, have put that reputation to a hard test.

When the borders reopened after the pandemic, net migration surged to an all-time record, briefly adding more than 500,000 people in a single year—the equivalent of 2 per cent of the country’s population. The government has been cutting back ever since, and the intake has fallen by almost half. Meanwhile, public concern has moved the other way. In the 2025 Scanlon survey, just over half of Australians said the immigration intake is too high—the highest share recorded. Pauline Hanson’s One Nation, the right-wing populist party built around cutting immigration, doubled its Senate seats at the 2025 election. In a March 2026 survey, more voters said they trusted it than any other party to bring migration down.

Subscribe now

A closer look at the same surveys, though, shows this is no simple anti-immigrant turn. The same Scanlon survey found that some 80 per cent of Australians still agree both that multiculturalism is a good thing and that immigrants are good for the economy. So, what is going on?

Australia is living through an unfortunate series of events I have already watched unfold twice. I have spent years writing about Canada, where a nearly identical sequence played out two years earlier, and about Britain, where a version of it helped produce Brexit and then Reform UK. What I see is that a long pro-immigration consensus does not collapse because voters suddenly master the arithmetic of intake targets or get brainwashed by right-wing populists. Instead, it collapses when the system objectively stops looking like something anyone designed to benefit their country.

Indeed, the long-standing Australian support for immigration rested on a visible bargain: citizens could see who was being chosen and why, and they could see that their own government was doing the choosing. In my book I call this demonstrably beneficial immigration—a set of policies that explicitly and straightforwardly serves the national interest in ways ordinary voters can see for themselves, without any elaborate explanations or justifications from the experts. The poorly managed post-pandemic surge broke the second half of that bargain, and I suspect the reaction since has had much more to do with that loss of visible control, and with the non-selective streams the newcomers arrived through, than with any particular number of arrivals.

Immigration concerns are never just about the numbers or propaganda

Political scientists increasingly describe public opinion on immigration and other salient issues as thermostatic. That is, like a household thermostat, the public signals ‘too much’ when a policy it cares about overshoots its comfort zone and ‘more, please’ when it undershoots. I find this metaphor useful, but it hides two complications, both of which are on display in Australia right now.

First, the thermostat reacts with a lag, often long after the temperature has already started to fall, which is why concern can still be climbing a few years into a decline in actual arrivals. After all, people have no way of tracking migration statistics in real time, and they respond not just to the flow of new arrivals but to the stock of immigrants already here, the people they can actually see around them. And that stock, unlike the flow, is not declining. Second, the thermostat responds to immigration issues beyond the headline number, such as particular episodes of mismanagement and abuse. Those episodes may or may not be exaggerated by the media, but they almost always register the sense that no one is in charge, sending signals of chaos and visible dysfunction.

The clearest illustrations of this dynamic come from episodes where the numbers and the sense of control point in opposite directions. Australia’s own Tampa affair in 2001 involved a few hundred people rescued at sea, but it became central to that year’s federal election. Meanwhile, Israel absorbed more than a million people after the Soviet collapse with comparatively little uproar. These two cases differ in plenty of ways, but they show that small inflows that look uncontrolled can ignite politics, while huge inflows that look orderly often pass quietly. If you want to know whether a backlash is coming, it seems like counting arrivals is less useful than asking whether voters believe their government chose those arrivals on purpose.

Note that none of this means the anti-immigration backlash is simply whatever the right-wing media makes of it. Partisans certainly try to frame every immigration episode as either orderly or chaotic, but the framing that sticks is usually the one with at least some grain of truth behind it. The relatively small-scale migrant crisis in New York City a few years ago, for instance, was a staple of right-wing coverage and led to significant backlash in arguably one of the most pro-immigration cities in the world. But it also was a genuine crisis that carried far more than a grain of truth about government dysfunction.

By mid-2023, the city was sheltering around 40,000 newly arrived asylum seekers. Texas Gov. Greg Abbott had bussed nearly 8,000 migrants to New York for political reasons, part of a broader Republican campaign to show that the federal immigration system was dysfunctional. And the stunt worked because there was plenty of dysfunction to show. Many asylum seekers were barred from working legally, but New York was still responsible for sheltering them with little help from Washington. The Manhattan Institute estimated that this was costing the city more than $1.5 billion a year. In the end, even relatively cosmopolitan taxpayers reasonably felt it might not be sustainable to house tens of thousands of people they didn't invite who couldn't provide for themselves.

Why did everyone try to copy Australia’s immigration system?

Let’s consider what Australia actually did on immigration and what made it exceptional. Canada originally invented the famous points-based system in 1967, scoring would-be immigrants on achievable traits like education, skills, and language ability rather than ascribed traits like race or country of origin. Australia adopted the idea but pushed it further by building a capped permanent program, set at 185,000 places this year—all combining points-tested selection with employer-sponsored, state-nominated, and family routes, while reserving roughly 70 per cent of places for the skill stream. Its signature route, the Skilled Independent visa, requires no employer sponsor at all. Since 2012, hopefuls register an expression of interest in a system called SkillSelect, and the government invites the strongest candidates to apply. The innovation was apparently good enough that Canada itself borrowed it back for its Express Entry system in 2015.

Britain, meanwhile, spent the last two decades trying to copy the Australian brand without doing the actual hard work of adapting any substance underneath it. Politicians from Gordon Brown to Nigel Farage to Boris Johnson have kept promising an ‘Australian-style points-based system’, and as Madeleine Sumption and Peter Walsh have shown, British governments have been introducing and modifying points tests since 2002. The appeal, as they argue, was probably always more symbolic than practical: ‘Australian-style’ served as political shorthand for control, with focus-group participants associating the phrase with boats turned away at gunpoint by the Navy—something that has nothing to do with how a points test works. By the post-Brexit version of 2021, Sumption and Walsh conclude, the points had become ‘purely cosmetic’, shoehorned onto a traditional work permit system driven by employer sponsorship, the opposite of how Australia picks its skilled migrants. The rebranded system then presided over the largest migration wave in British history, peaking at 944,000 in a single year, roughly three times the pre-Brexit level. Voters unsurprisingly noticed the gap between the label and the delivery, and Reform UK has been collecting the proceeds ever since.

Meanwhile, the harder half of the bargain has always been control at the border and immigration law enforcement, including boat turnbacks and offshore processing under Operation Sovereign Borders. Researchers at the Kaldor Centre have specifically called offshore processing ‘cruel, costly and ineffective’, crediting whatever deterrence the package achieved to turnbacks rather than the detention camps. But the political record of these hard and uncomfortable policy decisions is difficult to ignore. Governments of both mainstream parties concluded that visibly stopping unauthorized arrivals was the price of sustaining one of the world’s most generous legal intakes, and for two decades the polling largely bore them out.

There is a more general lesson here that governments keep relearning—even in a liberal democracy, some migrants need to be stopped and deported. When the state visibly cannot do it, the politics only get uglier. Germany ended 2025 with about 232,000 people legally required to leave the country but fewer than 23,000 actually removed all year, and the far-right AfD, feeding on exactly this gap, now polls ahead of the governing center-right.

Canada, for what it is worth, has long run the Australian formula with quieter machinery, pairing its large legal intake with credible enforcement. In 2025 alone, for instance, its border agency carried out more than 23,000 removals with hardly any political drama to show for it. American readers, meanwhile, have watched the opposite sequence, where years of visibly lost control produced a mandate for an enforcement campaign that has since gone far overboard—all the way to the point where federal agents shot and killed two American citizens protesting ICE operations in Minneapolis this past January. When a government stops removing people in an orderly way, the eventual thermostatic correction tends to be harsher than anyone intended.

The pandemic shock no Anglosphere government handled well

Like virtually every other government in the world, Australia closed its border in March 2020. But it also closed it harder, recording in 2020-21 its first annual net migration loss since 1946. This radical closure did something remarkable to public opinion. In the 2022 Scanlon survey, fielded a few months after the border reopened, only 24 per cent of Australians said the intake was too high—the lowest reading ever recorded. Canada, Britain, and the United States recorded their own all-time lows in numbers and opposition around the same time. But when migration later restarted, the record surge had not yet appeared in the data or in anyone’s suburb. So the lagged thermostat everywhere called for more migration.

Australia famously began reopening later than much of the rest of the world, lifting restrictions in stages from November 2021 until most travelers could return in February 2022, all straight into an economy short of workers. In hindsight, this rebound overshot everything you could imagine—a record 538,000 net arrivals in 2022-23, more than double of the pre-pandemic pace. Meanwhile, Canada grew by a record 1.27 million people in 2023, its fastest rate since 1957, and Britain hit its 944,000 peak the same year. The United States, for its part, was living through what David Leonhardt of the New York Times called the largest immigration surge in US history with net migration averaging some 2.4 million people a year and the foreign-born share of the population surpassing its previous record from 1890.

To be fair to the policymakers all across the world, restarting a stalled migration system amid desperate employer demand, collapsing university finances, and major visa backlogs was genuinely difficult. In hindsight, basically no English-speaking government got it right. Yes, these surges were not inevitable but clear policy choices. But it is also true that these choices were all made by governments under pressure from business while simultaneously betting that the unusually positive public would not mind. Their mistake, of course, was not realizing that this positivity, while genuine, was a lagging—and effectively thermostatic—indicator.

With numbers rising, public opinion began moving back toward negativity within a year, each on its own country’s clock. Even the United States has followed the same logic in its own convulsive way. The share of American voters telling Gallup they wanted less immigration want from 28 per cent in June 2020 to 55 per cent in 2024. Then it fell back to 30 per cent in 2025 and 29 per cent earlier this year, as border crossings collapsed—and the Trump administration’s crackdown made the sense of border control unmistakable, whatever one thinks of their methods.

Where exactly the Australian plan broke down

For someone who wants immigration to succeed, it would be convenient to explain Australia's backlash as an overreaction to anti-immigration propaganda from the far right. I don't think that explanation holds up very well once we look at where the increase came from. Australia's celebrated points-tested program stayed capped throughout. The surge came through uncapped temporary migration, especially international students, and the government's own reviews subsequently documented serious problems with these pathways.

Now, don’t get me wrong, there is much room for well-managed international student and temporary migration pathways. If anything, temporary routes in particular are often unfairly derided, even by pro-immigration advocates, as second-class or abusive compared with permanent pathways. But it is also true the Australian government’s own Parkinson Review warned in 2023 of a growing “permanently temporary” underclass. This is basically the direct opposite of what their system built on selection has been promising for many year. The other Nixon Review, published the same year, found systematic exploitation of the student visa route, with non-genuine education providers and unscrupulous agents selling visa access rather than education in so-called “ghost colleges.”

Underneath all the individual scandals, however, there is a glaring policy design Australia had bolted an uncapped stream driven by provider revenue onto its famously selective system, effectively outsourcing migration decisions to institutions that are paid per enrolment. International education had become an export worth more than $50 billion a year, and institutions from strip-mall colleges to sandstone universities had a direct financial stake in maximizing the immigration numbers. Students also need somewhere to live. So, when you have a housing shortage every additional renter is visible to regular voters. This is true even when the more careful research eventually concludes that immigrants are not necessarily at fault.

In the end, Australia’s experience here fits a common Anglosphere-wide pattern all too well. Canada too let its private colleges balloon until its own immigration minister called them “the diploma equivalent of puppy mills” and imposed an emergency cap. Britain’s failure took a subtler form—its official review found no widespread abuse, but universities had become financially dependent on international fees to subsidize research and domestic teaching. So their enrollments (plus a boom in accompanying dependents) exploded accordingly without much voter consent.

In each country, gatekeeping drifted from the state to institutions with a revenue incentive, and in each country the public eventually refused to treat the result as chosen migration. To its credit, Canberra has since largely conceded the diagnosis. The December 2023 Migration Strategy brought a tougher genuine student test, integrity legislation aimed at the worst providers finally passed in late 2025. As a result, the latest international student enrollment numbers now run under a national planning level—steered through visa-processing priorities rather than hard caps—that sets their scale openly in advance.

How to keep immigration popular

The cautionary experiences of Canada and Britain now offer Australia at least two distinct lessons in what not to do. The Canadian lesson is about honestly acknowledging mistakes and overreach, but also about not slamming the brakes so hard that the population starts shrinking. Though, of course, it is true that braking that hard can buy some short-term political relief.

But it can also come at the cost of the system’s long-term growth. The British lesson, promising an Australian-style system while running something else entirely and ignoring the thermostatic nature of public opinion, is the fastest known route to a Reform-style party. Both would be strange things for Australia to import, given that both countries effectively spent years trying to import Australia’s success.

There is an American lesson here too, and it is simply that the more conservative Australians with legitimate concerns about immigration should resist the temptation to elect anti-immigration populists. No matter how bad the current elites seem and how much the populists promise, the results are almost always worse. On immigration specifically, the Trump administration has, to its credit, largely stopped illegal border crossings. But it has also cut legal immigration more than illegal immigration, hitting skilled and other popular streams hardest—all while offering no constructive solution for the millions of unauthorized immigrants already in the country.

The permanent program remained capped and largely uncontested through the crisis, and surveys still find Australians favoring skilled migration, with most not even wanting student numbers reduced. The backlash was about the streams that were neither planned nor capped nor selected.

There is a lot of hard policy design work to be done here, but extending the permanent program’s design logic to student and temporary migration, by setting their scale openly and in advance, is something that both immigration advocates and restrictionists should agree with. I am not holding my breath this might actually happen soon. But it is also worth remembering that Australians arguing bitterly about the right number are still arguing within one of the most open, yet selective systems on our planet.

Australian public opinion will probably lag the recovery, just as it lagged the surge. The good news is that Australia enters this fight with an advantage over the United States and most other immigrant-receiving countries like Germany, since it—just like Canada—already made immigration popular by design once. The bad news is that keeping it popular is not easy, since it means running the temporary streams the way it has always run the permanent ones, all the interest group pressures notwithstanding.

But the goal of keeping immigration popular is absolutely worth it. The government just needs to make sure that anyone—whether on the left or right—looking at the system can once again see immigration working in their national interest.

A version of this piece first appeared in Eureka Street on August 14.

Share

It is hard to believe, but Popular by Design turned one this month. When I started, this newsletter had only three subscribers. If you ask me how it all came about, it was… pretty random. On a whim, I applied for fellowship, and one of the required forms asked for my Substack account handle (which I did not have at the time, so I had to create one). When they accepted me, they made a public announcement basically implying that I write a newsletter.1 In any case, folks started signing up and DMing me about my new endeavor (which, to be clear, didn’t exist yet). So, to avoid total embarrassment, I started writing, launching my first piece on why skilled immigration is popular in August 2025. And it seemed like at least some of you actually liked it, which was a complete surprise.

Subscribe now

Today there are more than 2,600 of you. Apart from a few viral pieces, what actually drove the growth is mostly you: quiet recommendations from readers and fellow writers did more than any spike, and I am grateful for every one of them. Popular by Design is far from being a household name (probably for the better), but it has now been mentioned or cited in several major publications, including The Atlantic, the Washington Post, the Wall Street Journal, and CNBC. The newsletter also stopped being just me: we have now run guest essays, and the scope has quietly widened from immigration to AI and population issues and credible policy design more broadly.

In case you missed it: In August I published How to Use AI for Public Engagement, the practical write-up of my Machine Collaborators talk on what agentic AI can do for researchers who want to reach the public. And we published the fourth guest essay of year one: , a Spanish economist at Money & Macro, on making sense of the Spanish exception, or why the country running Europe’s biggest immigration boom has so little backlash so far. If you want to write the next one, the guidelines are here.

Since it is anniversary time, here are the ten most-read posts of year one:

• Academics Need to Wake Up on AI, Part II, Part III

• The Uncomfortable Truths About Immigration with Reflections

• LinkedIn Is Doing What Bluesky Was Supposed to Do

• Reality May Have a Liberal Bias, but the Social Sciences Still Need Conservatives

• Western Countries Do Not “Need” Immigration

• Why Japan Is So Uncanny… Uncannily Normal

• Public Engagement Is Good for Your Research

The top two alone were read nearly 100,000 times, and six of the ten are not really about immigration policy. I have made my peace with that, partly because the detours turned out to be intellectually in line with my pragmatic immigration writing. I started this newsletter because of all the gaslighting and bullshit I kept seeing on my pro-immigration side. Watching many academics respond to AI made me think about it the same way: people were simply not thinking clearly (and they still aren’t on AI writing). From there it was a short step to writing about academia itself as an institution, which is losing credibility with each passing day, with no end in sight. That is, unless we change something.

Speaking of which: ever since my piece on the need for more conservatives in the social sciences came out last month, I’ve been trying to write a follow-up reflecting on the most common objections. But I keep rewriting it, because reality keeps giving us more bangers. When the Jason Arday scandal consumed Cambridge, I wrote that while many folks lament the loss of trust in higher education, the bigger issue is clearly the loss of our trustworthiness.2 One common response (from left-of-center academic) was basically, why bother restoring conservatives’ trust if they don’t care about the truth anyway? Well, unfortunately, at this point the situation at places like Cambridge and the BBC is so bad that they need to worry about restoring the trust of rank-and-file moderates, and even of some other left-of-center academics themselves, like me. For instance, I’m not sure I’d still feel comfortable talking to a BBC journalist or trusting they would report my research findings objectively on any sensitive issue like immigration or AI.

I also signed my first open letter, opposing Ghent University’s suspension of Nathan Cofnas, the philosopher who first documented Arday’s misconduct for the public. It has been extremely depressing to watch so many mainstream institutions, academics, and smart people on the left still excuse Arday’s data fabrications and his attacks on journalists, or prefer to treat the whole thing as a single aberration. Many also seem unable to grasp the simple idea that defending someone’s academic freedom does not imply endorsing their views, whatever they may be. As you can imagine given my previous experiences, for about a week, publicly signing the open letter made me about the worst person on the internet, according to Bluesky.

I mention the pile-ons because saying unpopular things in public costs something, and what makes it sustainable is knowing that readers actually value this kind of principled independence, wherever it lands. The newly introduced paid subscriptions are perhaps the clearest such signal I get. I turned them on in July alongside my essay on liberal nationalism, which set out this newsletter’s broader ambitions. Every essay stays free and will stay free, so upgrading is patronage in the old sense: it keeps the writing open for everyone and tells me to keep going where the arguments lead. And there is more coming this fall, from new guest essays to our first reader meetups.

Yep, there will also be at least two chances to meet in person this fall, if you are in or near Washington, DC. On Thursday, October 1, the Making Immigration Better workshop I am organizing at Notre Dame’s DC office opens with a public reception and keynote from Johannes Haushofer, the Cornell development economist who founded Malengo—one of the most impressive immigration startups out there. Drinks start at 5pm, the talk at 5:40, and I am sharing the registration form exclusively with readers here.

Then on Saturday morning, October 3, from 10 to noon before I fly out, I am hosting the first ever Popular by Design reader meetup over coffee in Dupont Circle: a much more informal setting, going beyond immigration to talk about all the things, from AI and policy design to viewpoint diversity. RSVP here and I will send the exact location to everyone who does. Space is limited for both events, so sign up early if you want to come. If this proves successful, we can also come up with another meetup in London later in October.

OK, here are the August links (linking does not imply endorsement):

• The biggest immigration news of the month: reports that DHS has proposed a $103,265 fee merely to file an H-1B petition, stacking on the existing $100,000 proclamation charge. Petitions subject to the first fee already fell about 87%, and Bier calculates the government lost $28 million in net revenue in its first five months.

• won a ruling that the two statutes Secretary Rubio has used to revoke visas and deport legal immigrants over pro-Palestinian speech, as in the Khalil and Öztürk cases, are unconstitutional under the First Amendment. According to the judge: “In the United States, freedom of speech belongs to the people. It is not the government’s to take.”

• Mireille Paquet, Andrea Lawlor, and Colin Scott have a new open-access survey experiment with over 5,500 Canadians: immigrant candidates with prior Canadian work or study experience are clearly preferred, and the effect is larger among respondents with more negative views of immigration. A visible design feature that recruits its own constituency among skeptics is pretty much the thesis of my book, so I will be citing this one for a while.

• offers an insightful engagement with Justin Gest’s new book Democratic Drain: 88% of prospective migrants say there is limited freedom of expression at home, and in a conjoint survey experiment they were 27% more likely to pick a destination for being a democracy, a stronger pull than its economic growth. Politano pushes back on Gest’s worry that emigration drains democratization movements at the origin.

• treats the Immigration Act of 1990 as the last successful expansion of legal immigration and asks what made it possible: it passed the Senate 89-8 while nearly half the public wanted immigration reduced, four years after the 1986 amnesty had been framed as closing the back door. It is pretty crazy to think that nothing comparable on legal immigration has passed in the 36 years since.

• David J. Bier also reports that Senators John Curtis (R-UT) and Mark Kelly (D-AZ) introduced a state-sponsored visa bill: renewable three-year visas under a 500,000 cap, with allocations that expand or shrink based on states’ compliance records. Letting Utah and Arizona want different things is one of the few live ideas for breaking the congressional logjam, and it now has a bill number and a senator from each party attached.

• Cris Ramón at the wrote a manifesto for employment-based immigration design: immigration systems need explicit mission statements with measurable outcomes and the data infrastructure to adjust course. It is encouraging to see more people treating immigration policy as a design problem here on Substack.

• Alan Gamlen at the ANU Migration Hub asks whether temporary migration can sustain labor markets and communities over the long term, or whether churn-based guestworker models undermine exactly the sectors that need stable workforces, like care and construction. He treats the guestworker question more seriously than the usual Gulf-model discourse does, which makes the piece worth your time whatever your prior.

• Mihnea Cuibus at Oxford’s Migration Observatory walks through why UK small boat arrivals fell 44% in the first seven months of 2026 and lands on a straight answer: with EU-wide unauthorized crossings down and Syrian arrivals down 67% after Assad’s fall, it is genuinely hard to isolate what, if anything, UK policy did.

• and note the glaring patriotism gap: Democratic pride in being American has fallen from 88% in 2001 to 27% today, a record low; Republican pride sits at 93%. Whatever else you take from it, the constituency for the liberal nationalism I wrote about in July seems to currently live almost entirely in one party, and that should worry everyone.

• makes the most careful recent case for why even committed cosmopolitans and effective altruists should price in immigration’s institutional effects, drawing on and evidence that roughly 40% of home-country trust attitudes persist across immigrant generations. He is explicit that none of this justifies current restrictive policy.

• offers a mechanism for why sociology politicized more than its neighbors: weaker epistemic guardrails, from contested measurement to underdeveloped causal identification, left more room for politics when academia moved left. Meanwhile, argues in Persuasion that sociology must be saved from itself, recalling his Oxford advisor’s warning against squeezing the world into your own worldview.

• points out that below-replacement fertility is not new: under the mortality of 1950, replacement was roughly three children per woman, so Japan and Bulgaria fell below replacement while their fertility rates still exceeded 2.2. That deflates the idea that we are in unprecedented territory without denying the trend.

• makes a good point that the most consequential AI questions are effectively about people and institutions, and need expertise from all sorts of fields, from philosophy to social science. I should say that after my AI posts blew up, I was hesitant for a long time to claim any expertise here, but seeing how far behind most folks around me are has made me more OK with it, at least when it comes to opining on how researchers and public intellectuals can use AI. The fact that most smart people still don’t use agents is insane to me.

As before, if you want me to write more about one of these or other related topics, let me know! If year one taught me anything, it is that I should have started years earlier. The audience for pragmatic policy writing was always out there. So, thank you for being here, and see you in year two with more announcements and essays coming soon to an inbox near you.

Share

The images from Ceuta were everywhere this summer, and the obvious question is how long Spain's quiet immigration politics can last. Today's guest essay by , a Spanish economist who writes and hosts at Money & Macro, explains the system behind that eerie calm and why the country running Europe's biggest immigration boom has produced so little backlash, at least so far. Enjoy!

Few people realize this, but over the past five years, Spain has absorbed more than five million immigrants, adding the equivalent of 2% to its total population every single year. Across OECD countries, only Germany came close to these numbers, and even there, annual inflows seem to have cooled off. Yet the political reaction has been mostly absent: the country running Europe’s largest immigration experiment has one of its weakest far-right parties, and until this year most Spaniards barely ranked immigration as a national concern. This essay is my attempt to explain that exception, and to ask whether it can survive. I argue that Spain’s relative calm owes much to who its immigrants are and to a political system in which other conflicts have kept immigration from becoming the main divide.

Subscribe now

While the Spanish government continues to reaffirm its commitment to a migration policy grounded in “solvency, humanity, and efficiency,” a closer look at the landscape raises more than a few eyebrows:

• Most immigrants in Spain are relatively poor, earning around 29% less than the average local worker- one of the biggest gaps in the OECD;

• 40% lack a high school diploma and only one in four migrants have some tertiary education, one of the lowest shares in the EU;1

• In the last seven years, the country recorded over 150,000 irregular boat crossings to the Canary Islands and experienced 2 major border surges in its North African enclave, Ceuta, with the last one overwhelming the city with around 80,000 entries, almost the size of the city’s population;

• In light of the recent growth in the undocumented population, the government has recently implemented a general regularization that received over 1 million applications of foreigners that were residing in Spain without proper documentation.

On paper, this is hardly a policy model you would pitch at an international conference. When voters feel their government has lost control over who enters and who stays—or question whether the system serves the national interest—they can get pretty anxious. This is precisely the environment where right-wing populism thrives, leveraging concerns around crime, fiscal strain, and deservingness.

With far-right parties like the AfD in Germany now getting over 40% support in some polls, Germans seem to be quietly learning the lesson that while humanitarian goals matter, voters ultimately want migration policy to align with national interests. Yet in Spain’s 2023 general election—held amidst ongoing relocations of asylum seekers from the islands to the mainland—Pedro Sánchez’s left-wing coalition retained power, while the populist-right party VOX failed to surpass its 15% vote share from 2019, landing at 12.4%. By the end of that year, only 8.6% of Spaniards named immigration as a top national concern. Now that number has risen to 23%, making immigration the second highest concern for Spaniards according to July CIS Barometer surveys, though still well below the housing affordability crisis. Meanwhile, VOX’s polling momentum remains capped at around 17%, far below the heights reached by its European peers.

So, why is Spain different? Why so little backlash? To answer these questions properly, I think it’s important to know a little about the history of immigration in Spain.

The rise, the fall and the resurgence of immigration

The history of Spanish immigration is a story of two massive booms separated by a deep economic recession. Unlike other European powers that built guest-worker programs after WWII, Spain had virtually no foreign-born population a generation ago; my parents didn’t know a single immigrant growing up.

In the first boom, Spain’s foreign-born population skyrocketed from 2% in 1997 to over 10% in just ten years. Today, in its second boom, Spain is once again leading Europe in total and relative inflows. While distinct in certain ways, both booms were driven by a recurring formula, a specific mix of three ingredients that serves as the blueprint for the Spanish model.

A) Labor demand. Immigrant inflows in Spain closely track the unemployment rate. Spanish economic growth relies heavily on construction, hospitality, and personal care. The companies operating in these sectors are typically small, labor-intensive and quick to hire and fire, and are often willing to work informally. This creates an ideal structure for absorbing large numbers of workers rapidly, even if they lack high formal education or valid work permits.

The first boom was dominated by construction, alongside significant numbers in agriculture, hospitality, and a fast-growing domestic care sector. By the peak of that boom in 2007, the top countries of origin were Romania, Morocco, and Ecuador. The second boom has been driven primarily by hospitality and services, with Latin Americans making up the majority of new arrivals.

B) Visa policy. The steady demand for low-skilled labor is a feature of the Spanish business cycle, but it could never have produced over eight million foreign-born residents without accessible pathways into the country. Unlike the post-war guest worker programs enacted in other European countries, the Spanish immigration booms were not the result of an intentional recruitment program aimed at addressing any particular problem with the Spanish economy, but a combination of lax visa policies and the expectation of access to legal status after an irregular arrival.

During the first boom, citizens of several Latin American countries could travel to Spain with just a passport, thanks to long-standing bilateral visa exemptions that had been folded into Schengen regulations. Romanians gained similar visa-free travel access starting in 2002, and later became the largest immigrant community in the country. We know visa waivers fueled this surge because when Spain selectively got rid of them—for Colombia in 2001, Ecuador in 2003, and Bolivia in 2006—inflows plummeted almost overnight.

Overstaying short-term visas remained the primary route into the country in the early 2000s, causing the undocumented population to build up rapidly. To manage this reality, governments of both left and right ran four extraordinary regularizations, granting work and residence permits to over one million immigrants between 2000 and 2005. While these amnesties allowed migrants to contribute to the welfare state and move into better jobs, relying on periodic amnesties as a primary policy tool was widely criticized as arbitrary and politically volatile. Seeking a more durable mechanism, the socialist government introduced arraigo in 2005: an ongoing, individual pathway that allows migrants to apply for legal status after 1 to 3 years of continuous residency, provided they have no criminal record. Shortly thereafter, however, the 2008 financial crisis brought inflows to a sudden halt.

The second boom echoes many of these policy choices. It was kicked off by Schengen visa exemptions for Colombia and Peru in 2015 and 2016, which more than doubled arrivals from those nations over the following decade.

Once again, a new generation of overstayers turned to arraigo as their main gateway to formal work permits. Yet the sheer volume of recent arrivals has proven so big that even a flexible, uncapped mechanism like arraigo has been overwhelmed, driving a rapid rise in the undocumented population. To break the gridlock, this year the government stepped in with a controversial general amnesty that has registered over one million applications. Unsurprisingly, Colombian nationals represent by far the largest demographic applying for relief, a direct footprint of the 2015 visa exemptions coming full circle.

C) Push factors: When jobs are readily available and entry is straightforward, the economic incentive to move is set. But sometimes sharp domestic crisis at origin can push thousands to take the leap who might otherwise have stayed put or migrated elsewhere.

During the first boom, key external triggers included the severe 1990s Ecuadorian economic crisis and post-9/11 border tightening in the United States, which redirected migrant flows toward Spain. During the second boom, the defining shock was the catastrophic economic collapse of Venezuela. This exodus has driven Venezuelans to become the third-largest foreign community in Spain, behind only Colombians and Moroccans. This second Spanish immigration boom carries a distinctly more Latin American flavor, yet it retains all the core hallmarks of the first: it is largely unplanned, relies heavily on low-wage labor, absorbs a high proportion of undocumented workers, and frequently misallocates human talent.

As regular readers of this blog know, rapid influx of low skilled and undocumented foreigners are hardly ever popular among voters. Which brings us to the next obvious question.

Why so little backlash?

Spain was a latecomer to Europe’s populist-right wave, a fact that has sparked some speculation. One prominent argument centers around the post-Franco political culture that emerged in the late 1970s and 1980s: coming out of the long shadow of a 36-year dictatorship, the generation of my parents was raised to idealize democratic, egalitarian, and universalist values, creating a deep social stigma around rhetoric that echoed 20th-century authoritarianism. Under this view, this historical legacy imposed higher cultural barriers for parties like VOX to overcome. Tellingly, neighboring Portugal (which also emerged from a military dictatorship in the mid-1970s) was also a latecomer to Europe’s populist-right wave.

Beyond cultural taboos, geography and party politics likely played crucial roles. Neither Spain nor Portugal stood on the front lines of the 2015 Syrian refugee crisis. In other European countries, that crisis catapulted immigration to the center of the political stage, handing the far right a potent weapon against the ruling establishment. This mechanism simply failed to trigger in Spain and Portugal, where inflows had dried up after the 2008 crisis, and over 20% unemployment seemed like a bigger thing to worry about at the time. Additionally, in the midst of Spain’s economic crisis, public anger was initially captured by new challenger parties on the left and center. Rather than targeting migrants, these movements successfully channeled voter discontent toward political elites, potentially delaying the space for a right-wing populist breakthrough.

This Spanish exceptionalism appeared to dissolve in 2019 when VOX surged from obscurity to secure 15% of the national vote. Were Spaniards finally waking up to the immigrant threat their European peers had been shouting about for years? Not really. Granular voter analysis reveals that immigration attitudes were a surprisingly weak predictor of early VOX support compared to opposition to the Catalan independence movement.

Since then, VOX’s discourse has turned markedly more anti-immigrant, coinciding with the peak of the second immigration boom and record numbers of irregular boat arrivals in the Canary Islands. At this point, many would have predicted that immigration would finally become a politically defining issue. Instead, two things seem to go in the opposite direction:

• The public salience of immigration has decoupled from irregular arrivals. Historically, voters react strongly to maritime crossings, as images of unmonitored boat arrivals generate a perception of chaos and loss of control. During Spain’s 2006 Cayucos crisis (when around 40,000 migrants reached the Canary Islands) immigration briefly soared to become the public’s single greatest concern. Over the past decade, however, despite a larger, more prolonged humanitarian crisis and relentless anti-immigrant messaging on social media, few Spaniards consistently rank immigration among their top three concerns.2

• Negative attitudes toward immigration have declined rather than escalated. As VOX normalized far-right rhetoric, one would expect the reputational costs of expressing anti-immigrant views would decline. Yet, negative attitudes toward immigration have remained surprisingly contained. According to the latest wave of the European Social Survey (ESS), roughly one-quarter of Spaniards believe that immigration makes their country a worse place to live. That share is not only significantly lower than in countries like Italy or Germany, but it has somewhat trended downward alongside the rise of VOX and the start of the second immigration boom.

How do we make sense of all this? Whatever some of my international progressive friends like to imagine after hearing Pedro Sánchez challenge Trump’s approach to immigration, I don’t think my fellow Spaniards are more cosmopolitan or better informed than other Europeans.

Instead, I think there might be two distinct features about the Spanish immigration experience that could prevent immigration from becoming the top concern despite the high shares of undocumented and low-skilled.

• New immigrants mainly come from Latin America, a region that shares a language and deep cultural ties with Spain. If you live in any major Spanish city, Latin Americans are a seamless part of daily life, whether as friends, coworkers, or service providers. Unsurprisingly, anti-immigrant animosity toward this group remains extremely low. It is very hard to think that Spain could have accepted such large influx had it come primarily from majority-Muslim countries.

• Immigrants in Spain are relatively young. Immigration is a relatively recent phenomenon in Spain, and immigrants from the first boom are still for the most part of working age. Even though most immigrants earn below-average wages, the Spanish welfare state primarily redistributes through its pension system (72% of post-tax redistribution), with comparatively little provision for poor families and almost no public housing. There is simply less to fight over than in the northern countries, where a more generous welfare state makes newcomers look like rivals for benefits, family support, and social housing. As the Spanish government likes to remind, immigrants currently generate around 10% of Social Security revenue while drawing just 1% of its spending.

This makeup of Spain’s immigrant population pushes Spanish politics in two different directions. For the ruling Socialist (PSOE) government, foreign workers provide nice GDP and job creation figures and the tax revenue needed to delay unpopular reforms to the pension system. At the same time, Prime Minister Pedro Sánchez has mastered the art of weaponizing global issues like immigration, the war in Gaza and military spending to pick fights with polarizing figures like Donald Trump or Elon Musk. By linking domestic right-wing parties to Trumpist rhetoric, Sánchez transforms immigration into a high-stakes moral battle, framing support for his migration policy as a direct test of democratic values.

With VOX bringing anti-immigrant rhetoric into mainstream politics, left-wing voters could have reacted by adopting even more liberal views on migration out of the fear of sounding like them. As Sebastian Rinken notes, the recent decline in negative attitudes towards immigrants is driven almost entirely by left-wing respondents.3

You might expect VOX to make left-wing voters even more hesitant to speak out against immigration, while making right-wing voters feel freer to do so. Yet the share of right-leaning Spaniards who think immigration harms the country has held steady at around 35% for 15 years.

This highlights the second way the demographic profile of the immigrant population influences party politics. Because most recent arrivals come from Latin America, right-wing parties find it difficult to make broad anti-immigrant arguments. Certain Latin American groups, such as Venezuelans, hold pretty negative views on socialist policies, and have actually become reliable supporters of the Spanish right. As a result, VOX has to narrowly target its anti-immigrant rhetoric against African arrivals, a group that makes up a shrinking share of the overall migrant population.

So far, this unique alignment of party competition and immigrant demographics may have functioned as a political buffer, allowing Spain to defy the conventional wisdom seen across Europe and absorb significant numbers of low-skilled and undocumented immigrants without triggering a broad anti-immigrant backlash. Yet as immigration climbs higher on the list of voter concerns, that political immunity may be wearing thin.

Can the Ceuta surge be a breaking point?

Images of thousands of young Moroccans milling along the beaches of Ceuta have traveled the world—a vivid snapshot of border policy breakdown, accompanied by mounting anger and a sense of abandonment from local residents.

The surge occurred just three weeks after a landmark Supreme Court ruling restricted police from performing immediate pushbacks of migrants trying to reach Spain by sea. The legal saga began when an Algerian citizen—who had attempted to swim around Tarajal beach—sued the Spanish government after being sent back to Moroccan soil without the right of a court hearing. He argued that the 2015 Citizen Security Law authorizing “summary returns” (rechazos en frontera) applied strictly to land borders, not maritime arrivals.

Following a series of appeals, the Supreme Court ultimately ruled in his favor. The justices highlighted that the text of the law explicitly restricted fast removals to migrants trying to “cross border control elements”, reasoning that open waters do not qualify as a border fence or barrier.

Most people never hear about a ruling like that, but these things can spread fast in social media groups, and usually in a distorted version. Unfortunately, it is possible that a distorted version of a ruling is enough to convince thousands of disenfranchised young Moroccans that the border is open and mobilize them to coordinate a crossing attempt.

Group attempts at the border are nothing new, and Spain depends on coordination with Moroccan police and intelligence to break them up before they build. Given the scale of this attempt, it is hard to believe Moroccan security missed it entirely. Suspicion of an intentional “blind eye” is perfectly justified. After all, in 2021, Moroccan guards famously opened the gates to over 5,000 migrants during a diplomatic spat. Still, the geopolitical returns for Rabat aren’t obvious this time around, at least to me, so I’d hesitate to draw firm conclusions just yet.

Whatever the truth of the matter, the past few weeks have brought intense commentary on the fragility of Spain’s border and on what the relationship with Morocco ought to look like. But once the last crossers are returned and Ceuta settles back into normality, the question is how long before the media cycle moves on, before attention snaps back to partisan bickering and corruption scandals, and border policy fades into the background again.

The beginning of the end of the Spanish exception?

The high degree of cultural proximity among Spain’s immigrant population may have acted as a political buffer, blunting the sharpest edges of anti-immigrant sentiment. However, repeated humanitarian crises on maritime and land borders are slowly making immigration escalate in the list of concerns. While VOX vote intention seems to remain capped at a relatively low ceiling, events like Ceuta risk deepening the loss of legitimacy of the Spanish model, calling for more ambitious diplomatic efforts to curb irregular crossings and preserve a secure border.

The structural challenges remain formidable. On one hand, managing land borders demands constant, uneasy choreography with Morocco, a nation well-versed in weaponizing its border gates for geopolitical leverage. On the other hand, Spain faces thousands of kilometers of West African coastline spanning half a dozen sovereign nations, where closing one departure point may simply shift boats down the coast. Agreements with Morocco redirected flows to Mauritania; subsequent accords pushed them to Senegal; and Senegalese enforcement diverted launch points to The Gambia and Guinea-Conakry, transforming a voyage of days into a dangerous two-weeks journey at sea.

So far, the bilateral frameworks pursued by Pedro Sánchez—combining seasonal migration labor quotas with regional police cooperation—have yielded short-term gains. In Mauritania, local enforcement intercepted over 30,000 migrants, contributing to a drop in Canary Island arrivals. Yet disrupting departures is fundamentally different from deterring them. Without a credible mechanism for rapid repatriations, enforcement acts as a temporary tax on the crossing rather than a reason to abandon it, prompting migrants to simply re-attempt the journey from unmonitored shores.

Australia halted maritime arrivals by enforcing a policy of returning boats to Indonesian waters, sending an unambiguous signal that effectively eliminated irregular entries. Spain’s reality, however, is constrained by geography: reaching return agreements with all departure countries would require immense diplomatic capital, and turning back young migrants who have gambled everything on an ocean crossing poses a moral dilemma many Spaniards are unwilling to endorse.

As upcoming general elections approach—with the center-right People’s Party (PP) leading the polls alongside rising public attention to immigration—we may be witnessing the beginning of the end of the “Spanish exception.” Yet how a future PP-VOX coalition would actually dismantle or reshape this model remains an open question. Can a conservative government really risk strangling the low-skilled labor pipeline that keeps key industries afloat and feeds Social Security revenues? And how do they intend to manage the fragile, high-stakes diplomatic tightrope with Morocco?

I don’t think opposition parties have a coherent answer to this. Still, the rest of Europe and the world should watch the situation closely. Spain is the last big country on the continent running large-scale immigration with relatively quiet politics. The next election will show whether that combination survives the moment voters start paying attention.

Share

How to use AI for research and data analysis is by now reasonably well-discussed, and the issue of AI writing is still controversial but at least has a lively debate around it. Using AI to reach the public or assist your workflow as a public writer, however, is almost untouched. After a year of experimenting with those uses, my sense is that the constraint is imagination rather than capability. In May, I gave a talk about my experience for Machine Collaborators, a workshop series on how researchers work with AI. What follows is the written version, which is a more practical kind of post than I usually publish here, and it’s intentionally designed to be useful even to those folks who don’t want to outsource any of their precious writing to the machine at all.

Subscribe now

I spent the spring arguing in three separate posts that academics need to wake up on AI, changed my mind about detectors when Pangram turned out to work, and then doubled down on the idea that AI writing can be legitimate in many contexts. As I recounted to The Chronicle of Higher Education, some folks were calling to fire me for those blasphemous thoughts, while others kept privately messaging me to ask about my setup (sometimes the very same people). But even the sharpest critics I engaged with, at least outside of Bluesky, usually concede that AI has been good for doing research—just not writing about it yet. So I want to set the writing question aside and start from a friendlier premise.

Let’s say you have done your research, written it up, and had it published. It is now sitting in a journal that few people read. You would like more people—especially those of the influential kind—to read about your work and findings. With a custom style guide and proper verification, an AI agent can carry much of that promotional load by creating accessible write-ups, data companions, and interview prep packets. This post is a tour of how to start.

My academic readers may remember that I made the fuller case for scientific public engagement, especially vis-à-vis “scholar activism,” in an earlier post. In short, writing for necessarily diverse, non-academic audiences stress-tests your ideas and keeps good research from staying invisible. But I do not want to pretend the current incentive structure in academia encourages this kind of engagement outside the field. Unless you’re in a policy school, public engagement normally counts for little in promotion files. Skipping it has been a defensible use of your scarce time, especially if you don’t enjoy it. At least, that was true until AI agents came around.

Writing a public-facing piece built on research I had already done used to take me about a week. With some agentic assistance, it now takes me less than a day without sacrificing any quality. The bar for “worth doing” drops accordingly, and a whole set of projects that never made sense before—like a translated website or an automated idea tracker—suddenly do. The limits are again mostly about your creativity.

The basic setup you need for now (2026 Q3)

One reason it took me some time to publish this piece specifically is that the shelf life of any concrete AI tip is quite short given the rapid development of these tools. The only durable tip I’ve learned is that you can just talk to your AI agent and ask it to do things for you the way you would ask your human (research) assistant. Relatedly, I’ve come to believe in the general principle that, if there is something that’s ethical to outsource to your human assistant, it should be similarly ethical to outsource it to your AI assistant. With your own verification, this should be true regardless of whether we’re talking about creating outlines, summarizing sources, gathering data, and even writing formulaic email responses to administrative requests. The assistant analogy is only a floor, though.1 Today’s agents are also competent programmers and decent designers, so once the routine tasks are covered, the same setup can take on far more ambitious projects, like the interactive research companions I describe at the end of this post.

Reasonable people still disagree on how much actual writing you should be able to outsource to AI, but it is clarifying to think of it as a continuum rather than a binary. Here’s one good example by , building on my own case against Pangram from earlier.2 Archie and I don’t quite agree on where to draw the line (his yellow/orange is my green), but even for those of you who are even more conservative than Archie, there is plenty of exciting stuff you can do to improve your research’s public engagement.

Here are a few specific things all quantitative social scientists may find useful. This is very basic stuff, but I’ve seen folks who are still confused about the difference between Codex and ChatGPT, so it makes sense to be explicit:

• Pay for at least one premium model. The serious agentic tools (I use Claude Code and OpenAI Codex) sit behind subscriptions of about $100-200 a month, and the free tiers will mislead you about what is possible.3

• Use an agentic system, not a chatbot. That is, don’t use Claude or ChatGPT in your browser, even when you have a project folder. You need an “agentic tool” that gets access to your computer to work directly in your files until the task is done, while making changes in the real world, like updating your website.

• Get a GitHub account if you do not have one. Agents work best in folders with version history, and you will want the ability to undo. Here is a good guide for beginners.

• Talk to your agent when you do not understand something, preferably using dictation. The setup questions that used to require a programmer friend can now be answered, patiently and specifically, by the tool itself.

• Keep and periodically update a common instruction file such as CLAUDE.md or AGENTS.md. Every correction you make twice goes in there once, so the agent stops repeating the mistake.

• Always verify output, especially anything for public consumption, both with AI (yes) and personally. More on this below.

Why you still need an agent and why a chatbot with a project folder is still not that

I should say a few more words on this, because some of the folks who reviewed this piece were still not sold on using an agent. They already talk to Claude or ChatGPT in the browser every day, they keep project folders with their papers uploaded, and they reasonably ask what else a researcher could need.

The difference is what happens after the model finishes thinking. Everything you do within a chat window is just information: you still have to carry the answer somewhere and act on it yourself. Things you do with an agent are actionable in terms of impacting the real world around you, like updating and publishing your website, or fixing and uploading your code, or sending an email invite.

Using a project folder, useful as it is, is not enough. And it’s not just about having to copy and paste less from your chatbot window. Uploading your papers and relevant materials to the project folder changes what the chatbot knows; an agent changes what AI can practically do with the files you share. Since the agent also has access to your computer and the terminal, you can do more sophisticated things like the writing skill that I describe below, which lives in files the agent reads and updates every time it writes. The same goes for your browser: an agent working through it, with your logins and permissions, can reach far more than the AI’s own built-in browsing. My suggestion is, again, to just try it on one concrete task (e.g., writing interview prep) and compare it to the output from a chatbot for yourself (e.g., Claude Fable 5 vs Claude Code Fable 5). But the whole endeavor of building and updating your own website below is the perfect candidate to see the difference for yourself.

Step zero is always building an AI-assisted personal website

Everything that follows assumes you have a home on the web you control, which, for most academics, means a personal site. If you do not have one as a researcher, get one ASAP, and this is the first thing you can ask an agent for. I created my website by hand using Jekyll with the AcademicPages template about seven years ago. I’m stuck with it now, at least emotionally, but there are more good options available now to start from scratch.

Your website can be hosted for free on GitHub Pages or on a personal domain (e.g. on Squarespace), which costs about $10 a year. Back in the day, as a non-coder, it took me forever to set up my site the way I wanted, even with the help of various guides. With agentic AI, the whole thing can take as little time as one evening. From then on, the site effectively edits itself. Small upgrades cost an afternoon each. You can describe a change in plain English, like adding tags to your publication list, and the agent adds that feature by itself. The only thing you need to do, apart from any local verification to see whether the new features truly work, is to commit and publish the changes.

All the other possible add-ons you can think of, like publication highlights on the landing page or a sidebar of upcoming events, are possible—the limit is, as I keep saying again and again here, just your creativity. None of these would have justified hiring a web developer (especially on our meager academic budgets), but together, they make the site considerably more useful to a visiting journalist or colleague.

But my favorite new post-agentic addition is automatic website translation: alexanderkustov.org now exists in English and 11 other languages. It is not perfect, and it took me some time to set it all up by adjudicating based on translation into languages I speak myself, but it does the SEO job well enough. If a reporter in Germany, Mexico, or Japan googles something on immigration public opinion in their language, they may come across my research and reach out (which has already happened several times).

No grant or human translation agency was involved—my AI agent translated the site, and I now run an automated weekly check to compare the English pages with the 11 translations and list anything that has drifted out of sync. Even if I had initially hired certified human translators to do this job, I couldn’t afford to have them update their translations regularly (and I don’t ask people to work for free).4

What to do with research papers and ideas you already have

Once your research papers are published online, the AI agent can track what happens to them. We all know about Google Scholar—it tells you who cites your academic papers, but it can’t tell you who links to your public writing. So I made my own substitute out of Google Alerts on my name and major titles, a backlinks tool (e.g., with Ubersuggest), plus a scheduled scan that runs every Sunday morning and compiles new mentions and backlinks into one report I read with coffee.

A somewhat similar setup, it turns out, also helps with recording your research ideas and all the relevant writing across various platforms to help you develop those ideas. I have a whole backlog of things I come up with in the shower or during my conversations with colleagues, with one tracker for every idea and a scoring formula that includes subjective merit, expected involvement, and timeliness. When I stall, the agent tells me what to write next and suggests things to review on the topic from recent Substacks and academic journals I follow. You don’t have to do it exactly the same way, but I personally use Obsidian to organize and read all those markdown notes in one place while syncing it with my agent when necessary.

The same repository also lets me test ideas cheaply before committing to them. When I am unsure about a framing, I usually try it first as a social media post, and if it lands, I ask the agent to browse the responses and summarize the objections, which frequently become the counterarguments section of the eventual piece. Only after that does a pitch go out to an outlet, by which point I already know how the argument plays with a live audience.

Creating your own personal style guide and a writing skill

Once you have your website with all your ungated papers saved as PDFs and neatly available in one place, you should ask your AI agent to convert them to Markdown, extract each figure and table as its own file, and post all of it on the website next to the PDF. Believe it or not, the first reader of your paper is now often a machine; you want your work to be readable by AI so that AI can recommend that work to the human who prompted the relevant search.

You can also use the same clean .md (markdown) versions of your papers to create your “writing bank,” an authoritative truth about your thoughts for an agent to follow, which is very useful when you want to draft your new slides or a publicly accessible research summary. You can put everything you have ever written in this bank, saved as plain text (.md) files in one folder the agent can read: essays, op-eds, published papers, book chapters, even talk transcripts, with your voice already in that corpus.

You will also need a style guide of your own, because LLM writing is generic by default. Mine started as a single CLAUDE.md file and grew by accretion. I “trained” it on my own published pieces, corrected the first drafts it produced, then had it write the corrections down as rules. AI writing does have recognizable patterns, and by now my guide bans the tics everyone recognizes, including the “It’s not X, it’s Y” construction and artificially short paragraphs. The bans persist because these rules are stored in a file that the agent reads every time.

Over time, that one file grew into what Claude calls a skill, a small folder of instructions the agent loads whenever it writes as me; the diagram below shows the full setup. The writing bank feeds a voice profile (what I say and what I never say, down to sentence rhythm), while the style guide holds the rules. Before I see any draft, two checks run: a mechanical scan that flags banned constructions, and a fresh read by a second agent that sees only the final text plus the rulebook, because the agent that wrote a draft is blind to its own tics. My corrections then go back into the guide as new rules, which is the loop that makes the whole system improve.

The same agent can also serve as your reviewer, which may be the most underrated part of the setup. Before anything goes out, I ask for a critique of the draft from perspectives that are decidedly not mine: a restrictionist who thinks I am naive about enforcement, say, or a general reader who has never heard of “thermostatic opinion.” You can name specific writers whose judgment you trust, or simply describe ideological directions, and the notes that come back, while uneven, will reliably catch the objection I forgot to answer. For public writing, where no referee report is coming to save you, this kind of adversarial read is the closest substitute I have found.

With those two pieces in place, you can ask for whatever you need from one finished paper: an interview prep document, an op-ed pitch, a social media post, a slide deck, a Substack article outline, or even, if you dare, a draft of a full-fledged policy brief. For my Machine Collaborations talk, I ran all of this on one of the most sadly under-read papers of mine, a 2020 study showing that majorities or pluralities in most of the 30 countries surveyed want both less immigration to and less emigration from their own country, and the run produced, among other things, three genuinely different framings for a future essay that I had not considered.

Or let’s say a journalist contacts you about a paper you wrote years ago to ask about its implications for current events. With such a setup, you can produce a good draft of what you can plausibly say, which is almost certainly going to be better and more accurate than what you could come up with on the spot.

What to do about slop and hallucinations

Whenever I describe this workflow to colleagues, the first worry I hear is that it will just produce more slop. Don’t get me wrong, plenty of one-shot AI-generated text deserves it (just like much of human writing). But everything in the setup described here starts from the presumption that you already have legitimate peer-reviewed research and edited popular writing. The agent is here merely to repackage or contextualize your own claims in a new light for your own human shitty first draft. Just like in the case of relying on overly enthusiastic human assistants, the responsibility for what goes out under your name remains with you. Whether the result reads as slop comes down to the verification, editing, and all the extra hard work you do before hitting publish.

That verification is not optional, because the models still make things up, and many now argue hallucination is the biggest bottleneck in AI-assisted research. I have learned this the hard way. My workflow once renamed my co-author, James Dennison, as “Mick Dennison,” a first name it invented outright even though the correct one was sitting in the very files it was working from. On other occasions, it attributed a parody account (@GovAmyKlobuchar) to the actual senator and handed me a LinkedIn URL guessed from the pattern of my name.

So what do I actually do? I have the model triple-check every name, date, reference, and link; I make it paste the output of its verification so I can see what it fetched; and then I verify the things that matter myself. Although this may be overkill, I also routinely ask Codex to verify what Claude Code did, and the other way around.

On the hallucination side, regarding references and their attributions, has written the best practical treatment I know of. Denney’s fix is what you would build for a fast but sloppy human research assistant. Reusable instructions check every citation against Crossref and the DOI, and a project folder keeps the original PDFs alongside the machine-readable versions so claims get checked against all original sources on hand with minimal possibility of error.

I should admit that, even with all these checks and all the newest models (I’m using Fable 5 and GPT 5.6), the system is not perfect. My Claude agent seems to love the rule of three so much, for instance, that it keeps adding triads to my drafts even though my own style guide explicitly bans them, and it kept doing it in this very piece after I asked it to stop, over and over again.

What does the future hold for scientific public engagement?

If all this leaves you wanting to try it, you can start by taking one of your finished papers and asking an agent to create an artifact, such as an interview prep, in your own style. You tweak the result until it sounds like you, save the instructions that got it there, so the next one is better and faster.

You may also notice that I deliberately did not include my own formal writing skill here. That is because it encodes my voice, and you, my dear reader, should develop yours. The simplest way to do that is to point your agent at this very post and ask it to implement the setup I describe, with whatever variations fit what you are trying to do.

What I’m describing will soon be the floor, rather than the ceiling, of agentic capability. For instance, political scientist at Stanford revealed this week that he plans to publish much of his original research directly on Substack. This rapid release will allow him to discuss and promote his latest research not only with other academics but also policymakers and journalists, allowing him to skip the “often arbitrary” comments from journal reviewers. He is upfront about the costs of skipping peer review, and it is clear that this option is easier for tenured folks and others with secure positions. For timely work, though, the calculation increasingly favors speed, and I already see more researchers—myself included—making the same choice.

The frontier I find most exciting now is interactivity. Until recently, an interactive data feature was something only a newsroom with the resources of The New York Times could afford to build, while academic journals remain too outdated for anything more advanced than a static PNG. With an agent, any researcher can now publish an interactive companion to a published paper. Modelslant.com, the live companion that Sean Westwood, Justin Grimmer, and Andy Hall built for their working paper on the perceived political slant of AI models, lets readers explore the rankings themselves, and Hall’s lab now maintains a whole shelf of similar prototypes. Nothing about this, however, necessarily requires the resources of a Stanford lab anymore, and your own under-read paper is exactly the kind of candidate such a companion was made for.

The kicker of my first AI post still holds: lock yourself in a room with Claude Code and one tedious public engagement or research task you never had time to do, and see what happens. The specific tools in this guide will possibly look dated within a year, but the underlying shift is not going anywhere. The marginal cost of serious public writing has collapsed, and researchers who stay silent now cede the conversation to writers with worse information. You may not like AI or AI-assisted writing yet, but if you have something worth saying, staying silent while not using the best tools available is an active choice you’re making.

Share

Huge thanks to , , , and, of course, Claude and Codex for their helpful suggestions on the draft.

Last week I presented my book on how to make immigration popular at the Board of Latino Legislative Leaders summit in Niagara Falls, in front of state legislators and leaders from across the country. It was refreshing to talk to people who actually care about what works. I noticed the same pragmatism I found in the Vatican last fall: most of these folks really just want to do things that will help their constituencies, and they know migration can be a huge part of that, whether they are more liberal or conservative.

You may have also noticed that this section is now called Policy, But Better. The truth is that half of what I end up linking to here these days is about population, housing, AI, academia, or the general craft of policy design. This is all in line with the broader ambition set up by my July 4 essay, though immigration will still remain the spine of this newsletter.

Popular by Design also has a chat now, and, since we are building a community there, come introduce yourself. Last time we started with your biggest immigration bugbears. This time I am opening a new thread to ask for book recommendations: what do you think folks in our pragmatic-oriented community have overlooked?

In case you missed it: Starting with the big things I promised for the Fourth of July: Why US Nationalism is Essential after 250 Years, on how I learned to love my country and where Popular by Design is heading next. Then came Even In a Liberal Democracy, Some People Need to Be Deported, on why due process lacks legitimacy when losing never has consequences. And just this week I published Reality May Have a Liberal Bias, but the Social Sciences Still Need Conservatives, on why I was wrong about viewpoint diversity. Just so that you don’t think I am becoming some kind of a reactionary, this month we also ran a somewhat more programmatic and positive guest essay by Aubrey Westfall of Wheaton College on whether local governments should pick their own immigrants.

If you also want to write for Popular by Design, there are now actual guidelines for that, so please do reach out. And don’t forget that paid subscriptions are now also on: if you have been finding this newsletter useful, and you want to keep it free for everyone, upgrading is the best way to support it.

Subscribe now

Finally, as a sort of experiment, from now on, in each new monthly newsletter, I’ll be highlighting a relevant book I will recommend (or NOT) to my readers among the stuff I read recently. Let’s start with some of the positives:

👍 ’s new Kakistocracy: Why Populism Ends in Disaster argues that the real divide in politics now runs between populists and institutionalists, and that populism reliably ends in government by the least competent and virtuous people available. Given the somewhat unusual perspective Richard is coming from (right-wing populist to libertarian academic to conservative anti-populist), I thought the book was extremely engaging and his framing could be educational not just to regular readers and moderates, but also to many of my left-of-center colleagues who study populism. My sense is that these folks usually just assume populism is bad because it tends to be against minorities in the West or “liberal democracy” the way they envision it without thinking much about why it appeals so much to conservatives and why they may not like it either.

This is not necessarily detrimental to the main argument in the book, but my biggest quibble is that I am not sure populism is really a stable trait as the book portrays it. Given my own research showing that the same exact politicians like Barack Obama can decide to use it in some cases but not others, it seems to me more like a risky strategy of losing politicians. But, all in all, the book has many insightful points, including a genuinely novel conceptualization of who the elites really are and case studies of when populism can actually do some good in developing countries (which you wouldn’t expect to see given the title).

Here are the July links (linking does not imply endorsement):

• The biggest immigration news as this goes out: the Wall Street Journal reports that the administration is weighing attaching a $100,000 fee to Optional Practical Training, the program that lets international graduates work on their student visas. points out that the immigrant founders of many billion-dollar companies first came to the US as students: they would qualify for an O-1 visa today, after building their companies, but not back when they were getting started. There seems to be a very tough competition for sure, but this is the single worst thing the Miller administration could do on immigration and to our country, all for no good reason, to date. It certainly does not seem to be helping the administration’s own push for a new golden age of American science.

• This is July, which means now is the right time for your annual Gallup data update and immigration thermostat appreciation note. Only 29% of Americans want immigration decreased, down from 55% just two years ago, while 31% want it increased, almost back to the all-time high of 2020. Repeat after me: in this house, we believe that thermostatic public opinion is real. That is, Americans have not changed their hearts and minds on immigration much, but they have changed their concerns in response to the actual policy reality on the ground.

• at The Argument argues the left is wrong about assimilation. You cannot invite people into a culture you keep describing as rotten, and the share of Democrats calling it important that immigrants share American customs fell from 79% to 59% between 2016 and 2020. Assimilation talk, just like the stuff on patriotism and nationalism, makes many progressives uneasy, but giving up on it just hands the language of belonging to restrictionists.

  • Meanwhile, and show that America used to be exceptionally patriotic but is now below average. In the first World Values Survey wave in the 1980s, 95% of Americans said they were proud of their nationality; the latest wave puts it at 78%, below the European average.

  • also made the case that the Founders lived through, and loved, high immigration rates, with the immigrant share of the population at roughly a fifth in 1776 by his estimates.

• Martin Schröder and colleagues have a new PLOS ONE paper on AfD support using ten years of German panel data: concern about immigration alone accounts for roughly 75% of the explicable variation in AfD support, and about 90% for vote choice, with economic deprivation adding almost nothing. The “left behind” story keeps losing to the boring answer that far-right voting across Europe is mostly about immigration.

• Tiphaine Le Corre, Jane Green, and Alexander Yeandle have a big new review of whether political elites shape immigration attitudes in EJPR, synthesizing over 200 studies: elites barely move how extreme people’s views are, but they strongly shape whether those views get expressed, at the ballot box and in everyday life.

• My friend and co-author James Dennison asks whether the standard Western findings on the demographic predictors of immigration attitudes hold up globally in Public Opinion Quarterly, using World Values Survey data from 66 countries. Mostly yes, actually (his words), with instructive exceptions: education and left-right ideology predict attitudes far less outside the West.

• notes that Thailand’s population has already peaked, with fertility near one child per woman and the crossing below replacement made at a fraction of the income level of its neighbors. Getting old before getting rich is no longer a hypothetical about China.

• at looks at what a high-fertility subculture actually does differently: Yiddish-speaking women in the US expect 6.6 children against a national average of 1.6, and the difference looks structural (early marriage, communal childcare) more than attitudinal.

• says is partnering with Pangram now to fight what they call “Claudefishing.” The integration actually seems pretty reasonable to me: they acknowledge that not all AI writing is slop, and the scan is basically an optional feature for those who care. The good scenario is that all this will just nudge writers to be more careful about their AI use (like it just did with this post which is now apparently 100% human-written 😉). The bad scenario, of course, is that it will create yet another excuse for public witch hunts as I argued about earlier. Knowing the internet, I am not holding my breath.

As before, if you want me to write more about one of these or other related topics, let me know!

As always, if you would like to support the mission of Popular by Design and keep these essays free for everyone, you can do so with a premium subscription.

Earlier this month, The Chronicle of Higher Education asked Carolyn Rouse, the Princeton anthropologist who currently leads the American Anthropological Association: Is her field hopelessly politicized? A new report commissioned by the chancellors of Vanderbilt University and Washington University in St. Louis had just named anthropology the most extreme case among the six fields it reviewed.

I highly recommend the full interview to everyone, academics and non-academics alike, because her answers read like a very enlightening demonstration of the charge. In the course of one interview, she dismissed the report as chatbot output, waved away a survey of her own subfield because she does not “believe in opinion research,” and compared the teaching of binary sex to astrology.1 In short, the president of a major social science discipline’s flagship association does not seem to know what the public or even her own field believes, and does not treat the disagreement over many contested ideas as legitimate.

In public, there was very little defense of the interview’s substance, mostly just the reflexive assumption that any such critique must be a bad-faith right-wing attack. But this reflex reaches well beyond anthropology. And watching other social scientists—including my friends—defend those assumptions all week, in my own replies, forced me to finally reckon with an argument about the need for more viewpoint diversity I had previously mostly dodged. So, what I’ll try to convince you here is that, even if we grant that the reality famously has a liberal bias, we still need more researchers of conservative persuasion to understand that reality.2

Subscribe now

Some academics deny the discrimination is happening; others now defend it

The discussion started innocently enough. Two weeks ago, the left-wing Atlantic writer Tyler Austin Harper asked whether anyone seriously believes that a geographer critical of immigration could get hired by an elite American university. In response, a history professor called him a “right wing grifter” who had supposedly left his tenure-track job because none of his colleagues would eat lunch with him, then deleted everything when challenged.3

Then, many academic and left-wing commentators moved on to denying that any discrimination against people with conservative viewpoints is happening at all.4 And, to be frank, I’m increasingly losing patience with folks on this straight-up highbrow misinformation. We have years of anecdotal accounts, we have deans on record running demographic checks on hiring pools, and we have surveys of academics in which many openly said they would discriminate against conservative colleagues in decisions ranging from paper reviews to hiring. I have also seen it myself. I have sat in rooms where search committees explicitly discussed candidates’ ideology. And I have experienced the softer end of the same machinery many times, even though I am not even a conservative.5

The most clarifying development of the week, though, was the number of academics who skipped the denial and moved straight to defense. One widely shared post put it plainly: “academics with noxious right wing views are less likely to get hired,” and “personally I just don’t give a shit.” I appreciate someone finally saying it out loud. But that is not what I signed up for when I entered grad school, and it is not a good way to run a society where people disagree.

Why it would be OK to hire a climate skeptic in the environmental science department

Whenever someone makes the case of viewpoint diversity, the same gotcha arrives on schedule: so would you allow a climate skeptic in your environmental science department? A migration skeptic in your geography department? If you think about it for a moment, this is a much less reasonable gotcha than many assume. But I haven’t seen a single person in this debate is proposing to hire vaccine or evolution deniers in immunology or biology departments.

The assumption underneath the gotcha is that the missing conservative positions are “village idiot” views, ancient prejudices that could never generate scholarship. But if you cannot state the serious conservative case on immigration restrictions or biological sex, that is exactly an argument for hiring colleagues who can. A field that only ever meets the village idiot version of the other side never engages the serious one, and it arrives unprepared for a policy world made of exactly this messiness and disagreement. Indeed, much of the pushback was also a constant masterclass in scoring zero on Bryan Caplan’s ideological Turing test of trying to understand conservatives and their beliefs, not to mention liberals like me with some idiosyncratic opinions about the world.

The gotcha also works only by conflating two different people, the one who denies verifiable facts and the one who weighs trade-offs differently. Unlike flat denial of the facts, climate and migration skepticism is often a disagreement about the trade-offs involved. Most academics rarely, if ever, hear the argument for why migration should be restricted or carbon emissions allowed from inside their own field. That is why we slide into groupthink, give bad policy advice, and get distrusted in the first place.6

Some sociologists were still skeptical and asked which viewpoints are actually missing in their discipline. My serious answer, which clearly resonated with many, though sadly not with the sociologists themselves, was that the discipline would benefit from practicing colleagues who hold and can voice some genuine small-c conservative beliefs as their starting assumptions. I gave a few examples, not meant to be comprehensive: that inequality is not necessarily a problem, that migration restrictions can be justified, or that social order is important. The inequality claim, for some reason, proved particularly controversial and occasionally came back through the replies as the noted conservative Notre Dame professor believes inequality is good actually, which I thought made the case for viewpoint diversity in sociology better than I ever could myself. Since then, the list has kept growing, with nominations arriving from colleagues across the political spectrum.

Are any of those views completely absent in sociology or other social sciences? Probably not. But they are certainly vastly underrepresented and discouraged. To give you a sense of how bad things really are in sociology, the London School of Economics once produced a promotional film asking, in its own title, “Are There Any Right-Wing Sociologists?” The faculty’s answers were fascinating. One guessed that such colleagues probably exist somewhere, “perhaps there in criminology, perhaps they’re doing migration studies” (narrator: they are not). Another offered the reassurance that the department is “certainly not homogeneous,” containing as it does “hugely varied debates within a broadly left concern.”

I wish I could say this is not representative of the field, but much of the senior reaction was honestly almost as disappointing as the Rouse interview, with prominent sociologists seriously denying any underrepresentation of conservative views and wondering how anyone complaining about anti-conservative discrimination could be tenured in the first place. By that standard, of course, no social group with any employed members has ever faced discrimination anywhere. And this is exactly why we need divergent viewpoints in sociology: double standards like this get called out right away only when somebody in the room does not share them.

The taxpayer is always at the door, whether you like it or not

There is an important reason this all cannot just stay an internal academic conversation. Just days ago, The Wall Street Journal ran a widely shared essay by Pamela Paul asking whether it is time to “liquidate” the study of sociology. Back in March, I wrote that the standard explanations for academia’s drift—left-wing bias and self-censorship—always seemed incomplete to me. A bigger factor was simply forgetting about the taxpayer who funds us, directly at public universities and through grants, student aid, and tax exemptions everywhere else. As Tyler Austin Harper rightly pointed out, many of the angry responses to his viewpoint diversity points tacitly assume the public owes the Ivory Tower trust and money while deserving no say in what goes on there. Academic freedom is a privilege and it makes no sense for us to incessantly antagonize the mainstream society whose good graces we depend on.

Indeed, nobody is obligated or wants to hire vaccine deniers or worse, however such views happen to poll. But taxpayers’ views are not negligible, especially on normative issues that are not ultimately resolvable by science. Where a field’s disputes are mostly about values and priorities, and the college is financed by taxpayers who overwhelmingly believe those views are fake, full autonomy is simply not a stable arrangement, and academics should not be surprised when the public’s trust and funds disappear. As strikingly notes in his recent piece, “defunding police and the abolition of biological sex categories are two of the least possible ideas in politics today, and both are products of the critical humanities. If those ideas had stayed in unread academic journals, we might not be in this mess.”

So, which viewpoint diversity are we talking about here?

Let me be clear. I can see how the demographic diversification of the social sciences over the last decades has been, on the whole, genuinely good for knowledge. In my own field, scholars from migrant-sending countries and minority communities pushed migration studies beyond its old preoccupation with movement into rich Western countries to South-South migration, and migrants’ own decision-making. Diverse researchers asked questions the old professoriate had not thought to ask, which arguably improved our knowledge overall.

For some reason, folks in academia find it harder to accept the same logic when the underrepresented dimension is ideological, and yet the imbalance there is far larger. When ran the numbers for Heterodox Academy, comparing each group’s share of the professoriate to its share of the population, the lack of ideological diversity in social research fields came out “vastly more pronounced” than underrepresentation by gender, race, or sexuality. Now, you may object, as some academics did, that conservative arguments are hardly marginalized in our society, so why should universities of all places carry them? Fair enough, but I have yet to meet anyone who thinks conservatives must be represented in academia in proportion to the population. That standard is somehow reserved for demographic diversity, where we routinely demand proportionality despite everything we know about self-selection.

What people like me are actually asking for is much more modest: for these and other non-leftist views to be openly and institutionally represented at all without any drama. There are entire social science departments without, as far as anyone can tell, a single open conservative in them. And, again, we are often not even talking about actual conservatives, just people with a few heterodox beliefs who remain well to the left of the median voter. put it best: the faculty ratio is valuable as a proxy, and misleading if mistaken for the problem itself. What academia needs is people who pursue neglected questions and challenge orthodoxies, and, given the current makeup of the academy, such people now tend to be conservatives.

If you think about it, demographic diversity was often defended with the same proxy logic, the category standing in for the perspectives that usually come with it. Viewpoint diversity simply goes after those perspectives directly. A room that is diverse on every demographic dimension but homogeneous in its viewpoints—a common situation these days—has plenty of the diversity that matters least for truth-seeking and none of the kind that matters most.

So, it would be helpful to be more precise about what is currently missing. My sense is that the most absent sensibility is small-c social conservatism, the disposition describes as grounded in human flawedness, with its preference for order and its distrust of utopian schemes. The sociologist , whose essays over the past year have shaped my thinking here more than anyone else’s, calls these orienting assumptions, paradigms that shape which questions a scholar asks without predetermining any answers. Nobody, of course, needs colleagues who reject findings whenever the findings are inconvenient; that failure is already amply represented in academia. But the dispositional core—skepticism of rapid change and attention to what maintains social order—is a set of research priors, and these priors generate fruitful research questions. Here’s my attempt to group those succinctly:

• A degree of social hierarchy may be natural and even useful, and inequality is not necessarily a problem.

• Social order is hard to build and worth maintaining, and social change often happens too quickly.

• Many well-meaning government attempts to improve society will fail.

• Crime deserves punishment, and communities need protection from the small share of people who commit most of it.

• Religion is, on net, a social good. Traditional families and higher fertility deserve active support.

The questions immigration research keeps not asking

Let me come back to my own field of migration for a moment. My sense is that these are the questions that the actual majority of voters keep asking or, at least, wondering about: What are the long-term consequences of permanent low-skilled migration, including for migrants’ descendants? What are the local consequences of rapid demographic change? Under what conditions do temporary worker programs actually protect native workers? Which design features make legitimate law enforcement effective?

Every one of these can be studied with the field’s standard tools, and none of them is untouched: there are careful fiscal accounting exercises, evaluations of guest worker programs, scattered work on local reactions to demographic change. What the field lacks is an organized research program built around these questions at anything like the scale their public salience warrants, or anyone synthesizing the scattered answers for the conditional middle that keeps asking.

Relative to how much these questions move voters, my sense is that the attention they get is rather thin. And I do not think it is a coincidence that each of these questions is a small-c conservative one. Instead, a huge chunk of immigration research focuses on various aspects of aiding humanitarian immigration, which matters too, but is not where most voters’ concerns are.7

George Borjas has long been the most prominent economist pressing the restrictionist case, but we need more. One famous dissenter is not enough, and without serious competition, neither his arguments nor ours get properly stress-tested. , herself arguing from the left, says a lack of conservative pushback simply makes liberals lazy. I wholeheartedly agree. When there are more colleagues around you who have genuinely different priors about the world, the harder it is to publish weak work that flatters any shared assumptions. And a field that will not answer the questions voters keep asking should also not be surprised when voters take those questions to populist entrepreneurs who at least pretend to care.

What would actually help

So what to do? I want to be honest about my uncertainty, because the serious people in this debate genuinely disagree. Demsas bites the bullet and says that universities should adopt straightforward DEI for conservative academics and accept that some hires will be less qualified, alongside expanding elite universities and reforming admissions so the conservative pipeline exists at all. counters that nobody should be hiring Republicans for the sake of balance, the way DEI hires for their self-professed identity. Rather, the point is to get scholars who ask the questions the field neglects and push back on its orthodoxies. He has since made this case at full academic length and sharpened it further just this week. Meanwhile, other folks, including those who are conservative themselves, strongly disagree. , for instance, doubts increasing viewpoint diversity can be a good reform or even metric on the merits and would rather fix incentives for scholarly excellence directly.

I personally don’t have a strong stance on whether affirmative action for conservatives is a good idea. I do know that creating a metric of how many conservatives work in each department would not be healthy. Metrics, as always, would get gamed, and we would have self-declared heterodoxy and quota-filling long before we got the thing we want. And here I agree with Jesse Smith that we should want more people who push back on unreasonable priors before those priors harden into findings or are discovered by regular voters through the Fox News medium.

I also want to be plain about the political moment, because a version of this argument is currently being most actively made by an administration coercing universities with funding threats and ideological conditions. Nothing in this essay supports that campaign. Government-imposed viewpoint quotas would just replace one political test with another. That is precisely why I think universities should fix this themselves, precisely so that nobody else gets to (and they are currently failing to do so).

There are gentler levers, and they do not require anyone to count party registrations. Adversarial collaborations, including with the think tanks where many conservative researchers actually work. Post-publication review with teeth. Letting outsiders, funders included, put questions on the field’s agenda. It is also worth noticing what a decade of that movement has actually won. The concrete victories are rare and procedural, like the wave of institutional neutrality statements. So anyone promising a quick fix here is overselling you something.

In the end, intellectual disagreement should be welcomed and encouraged in academia. Right now, in too many spaces, it is quietly punished instead. I spent years assuming the people who said so were exaggerating for political reasons, but the past week showed me close colleagues saying the punishment is deserved. Whether the answer involves anything as blunt as reverse discrimination for conservatives, I honestly still do not know. But at the very least, making the environment friendlier to these ideas institutionally by design—because we need them to understand the world better—would be a good start.

Share

Huge thanks to , , and Codex for their helpful suggestions on the draft.

Last updated October 14, 2026.

Over the past months, Popular by Design (PbD) has published several guest essays, and their success convinced me that this should be a larger part of the project. The newsletter is also turning into the public face of a research lab I'm starting at Notre Dame with new postdoctoral and research associates. PbD keeps growing, and we're now open to guest posts on policies that work. So I'm making it official: if you have something that fits, I'd like to consider it. This post is going to be the one link I give to everyone for the guidelines, so if you're reading it months from now, it still applies.

We are primarily looking for pieces that are too evidence-heavy for an op-ed and too timely for an academic paper. Immigration and public opinion will still remain the spine of this newsletter, but we especially welcome posts that complement our expertise on adjacent topics like population change, housing, energy, and AI. The overarching theme for all of them is the same: policies that actually work and bring progress, made popular by design.

We primarily publish 2,000-4,000-word pieces of three kinds: a case study of an overlooked issue, an evidence review that takes a novel position, or an original empirical finding. Co-authored pieces are welcome, especially when they pair a researcher with a practitioner. The main test is whether a thoughtful reader of any persuasion would be glad they spent ten minutes on it.

Contribution formats

1. A case study looking at one overlooked policy, program, election, event, or live debate, here or abroad, and what we should learn from it. These pieces are especially valuable when they come from people close to the subject, whether they run a program or analyze the field. I saw Aubrey Westfall’s presentation at CES in Dublin, for instance, was blown away by it, and wanted to feature it in Popular by Design. Her piece on how Canada lets its provinces select their own immigrants was built on her own field interviews.

2. An evidence review explaining what the literature actually shows on a contested question and taking a position, no matter where it lands politically. Hanno Hilbig’s piece on accommodating the far right is a good example. Accommodation can work, which is why parties try it. The problem is that it can also become a lose-lose: some voters gained, others lost, and the far right legitimized.

3. An original empirical finding presenting a novel result from your own administrative data, survey, or sharpened re-analysis that hasn’t been published anywhere else. If you could share some new quality data on a question packaged in an accessible format (and explain why it’s better than existing data), that’d be great. The data should be open or shareable so the result is replicable. We can provide a template, and you can just email me your CSV, since we will redo any charts in our house style.

What we don’t run are vibes-only hot takes that add no new evidence or argument, jargon-heavy theory, or anything already published or under review elsewhere. If a smart non-academic policy person cannot follow the argument, this is the wrong outlet for it.

House style

• Plain and accessible. Don’t use jargon that is not necessary to advance your argument.

• Pragmatic and cross-ideological. Take the other side’s best case seriously.

• References as hyperlinks, always to the primary source.

• Every factual claim has to survive a fact-check before publishing.

• The mindful use of AI tools is encouraged for research, editing, and writing (!) as long as the argument is yours, you verified it, and you’re willing to sign onto it.

Submission and editing process

If you have something to share that fits Popular by Design, send your pitch to submissions@popularbydesign.org with a longer paragraph on what you’d write and a shorter paragraph on why you’re the person to write it. Our readership has some of the best policy folks in the US and other countries, so if you have something to say, it’s going to be a good, targeted way of reaching those audiences.

When your piece is solicited or your pitch is accepted, please share your Google Doc draft with me as an editor. I’ll then add a number of suggestions alongside some fact-checking and copyediting, and you can update everything right there without needing to track changes. Accepted pieces get one light developmental edit and one copyedit, with every change yours to approve, and we commission a custom cover image and write a short note introducing you and your work. Once it’s ready, I’ll send you an invite to be the guest author on Substack. You keep the copyright and can republish anywhere with a credit line—we only ask that the piece run here first.

Share

Is it morally justifiable to deport people from a liberal democracy? If you hesitated just now even for a moment, or if your answer was anything short of a resounding yes, this essay is for you.

Without any need for throat-clearing, my own answer is absolutely. Moreover, it’s possible liberal democracy even requires the real possibility of deportation. My sense is that a large number of pro-immigration advocates and normie liberals, including many of my fellow migration scholars, simply carry an implicit assumption that a just immigration system is one where no one—not a single person regardless of anything—ever gets deported. What I’ll try to convince you about here is that this assumption is wrong, and I think it has become one of the obstacles to building the immigration policy most of us actually want.

Subscribe now

This spring, I argued in The Atlantic that the old norm of opposing “only” illegal immigration, insincere as it often was, served as the political shield protecting legal immigration, and that its collapse should worry everyone who wants more of the latter. This essay takes that argument one step further, to the part hardly anyone on my side wants to defend, and argues for a narrow but essential proposition: in a liberal democracy, some people who lack a legal right to remain need to be removed, including by force, and to third countries, when nothing else works.

That sentence sounds harsher than it is because immigration debates have trained us to hear “deportation” as either unnecessary cruelty or restrictionist toughness. But the better way to think about deportation is to consider it, just like incarceration or even capital punishment, as one of the legitimate policy instruments our democratic governments have to enforce its laws and to maintain order. If our asylum procedures are meaningful, some claims will fail. And, if court review matters, some appeals will be exhausted. In the end, if we want to ask citizens to trust a generous immigration system, their government must be able to carry out the system’s own decisions.

I am not going to propose a detailed enforcement framework here, and the essay is deliberately agnostic about how many people should be removed and how. My goal here is simply to establish a baseline that everyone can stand on, so that we can have a productive conversation about enforcement design instead of the current bickering between “deport them all” and “deportations are illegitimate.”

The best case against deportation

Since I’m trying to provide the liberal case in favor of deportations here, I should steelman the opposite view first. Don’t get me wrong—deportations are pretty bad, to say the least.

When government agents deport migrants, they commit coercive state violence against people who are, in the vast majority of cases, guilty of nothing beyond wanting a better life. Deporting people costs serious money that could be spent on almost anything more constructive.1 It inflicts direct, lasting harm on the deported, their employers, friends, and families, including many citizens. It can violate international law when it deports people to persecution, while migration adjudication decisions are made under genuine uncertainty, with trauma, bad translation, missing documents, and uneven access to lawyers, all raising the odds of error. On top of all that, sometimes there is no functioning state to deport anyone to, or no willing state to accept anyone. In the end, deportation is clearly among the most severe things a modern liberal democratic state can do to a person outside its criminal law.

Some scholars and advocates take these premises to a radical conclusion, so what follows is no strawman. In the UCLA Law Review, legal scholar Angélica Cházaro argues that deportation is an indefensible act of violence whose abolition should be the horizon of pro-immigrant advocacy. Others scholars have mapped out what lawyering toward that horizon looks like.

I have no doubts this position is coherent and sincerely held. It’s also likely more influential than its explicit adherents suggest: for every scholar who writes “abolish deportation,” there are many more advocates and academics who would never sign the slogan but treat every actual removal as morally suspect. And a still probably larger group, including many of my more moderate and libertarian colleagues, simply keeps deportation out of sight and out of mind. For them, the topic feels vaguely discrediting, so their papers and policy proposals mostly route around it. I should admit I have been guilty of this myself. My own book barely discusses the need for deportations and enforcement at all, even though everything it argues about trust and order plainly depends on them. Writing this essay is my attempt to correct that.

I saw this worldview on display at the Council for European Studies conference in Dublin this June (a rare big conference which I actually enjoyed), where panels on migration kept running in two different directions, with some very heated disputes. Part of the room insisted that liberal democracies cannot survive without clear membership and public order, while many other researchers and advocates talked about removal as something morally contaminated, something decent states just do not do anymore. What struck me most was how rarely that second camp would say what should actually happen when a person’s asylum claim has failed. The question usually just hung there until the panel moved on to the next hard case.

The structure of this evasion is familiar from another debate. Prison abolitionists have built a serious intellectual movement on the observation that prisons are brutal, expensive, and unequally applied. Those observations are true as far as they go. But the movement has always struggled with the question of what happens to people who commit serious harm, and the most common answer is that in a just society, with poverty and desperation addressed, people would rarely commit such harm in the first place. Just last month, days before winning her New York congressional primary, democratic socialist and prison abolitionist Darializa Avila Chevalier was asked repeatedly in an interview what should happen to someone who kills another person; she never directly answered, saying she was talking about “the distance between the world we want to see and the world that we’re at.”

Many pro-immigration advocates hold the mirror-image view: in a just immigration system, with barriers low or gone, there would be few immigration laws left to break, so no one would need to be deported. Both moves relocate the hard case to a utopia where it no longer arises. Both leave the actual institution, the one operating in the world we have, without a principled account of its hardest and most defining task.

Of course, illegal immigration is literally not a murder. My friend Bryan Caplan made a version of the argument just last month: everyone who has ever driven has broken the speed limit, and “if you think it’s OK to break a law against driving 56 mph in the desert, you should think that it’s OK to break a law against mowing grass for money.” Unauthorized presence, in this telling, is a victimless regulatory violation, which makes deportation a grotesquely disproportionate punishment.

I hear Bryan and other libertarians who make this argument. But we live in a democracy, for better or worse, and that’s just not how most voters who have a say in who gets elected next think. And “victimless” is doing a lot of work in that sentence. In the United States the description is at least half right, since unauthorized workers pay federal taxes toward benefits they can never claim, but the costs of their presence land close to home: the schooling and local services their families use are funded by states and towns, a mismatch the National Academies documented in its landmark fiscal study. In Europe’s far more generous welfare states, where even people ordered to leave remain entitled to accommodation and benefits, the fiscal ledger is harder still to wave away.2 Besides, the current state of the law in the United States and other developed democracies is that illegal aliens, once ordered removed, must by statute actually be deported, while speeding—which is enforced sometimes even when it’s not absolutely necessary—will normally get you nothing worse than a ticket.

Political scientists Matthew Wright, Morris Levy, and Jack Citrin have also shown that Americans evaluating illegal immigration switch from weighing individual immigrants’ characteristics to categorical moral judgments rooted in the rule of law, and Levy and Wright’s book argues that civic fairness is the dominant frame through which Americans judge the issue. Voters treat unauthorized entry as a breach of the social contract, and a movement that keeps comparing it to a traffic ticket mostly convinces them that it has no plans to enforce anything.

So even the best case against deportation still falls short of the conclusion many advocates draw from it. Yes, the high economic and moral costs of deportation are an overwhelming argument for a fair and just deportation process: full hearings, competent interpretation, access to counsel (which, by the way, taxpayers would also have to pay for), and aggressive error correction to ensure no false positives. They make a strong, if less convincing, argument for making removal as rare as possible, prioritizing the worst offenders, and subjecting decisions to review. But the best case against deportation does not support the conclusion that no one should ever be deported, just as the best case against the horrifying conditions across America’s domestic carceral systems fails to prove that no one should ever be separated from society.

My test for anyone who hesitates is this single hypothetical case. Let’s say someone applies for asylum and receives everything the process can offer: proper notice, an interpreter, a lawyer, a full hearing, and an appeal. They lose at every stage; their country of origin is safe for them and willing to take them back, no legal protection applies, and the government even offers to pay for the flight home. They still refuse to leave. If you’re saying that the liberal state cannot remove even this person, then losing an immigration case has no consequence at all,3 and all those hearings were an expensive ritual of pretend law. Even Joseph Carens, the philosopher who did more than anyone to make the ethical case for open borders, writes plainly that nothing in his argument “denies a government’s moral and legal right to prevent entry in the first place and to deport those who settle without authorization, so long as these expulsions take place at a relatively early stage of residence.”

Germany: what happens when no one is deported

Germany is worth extended attention because it is a wealthy liberal democracy with elaborate legal protections and visible enforcement failure, and because Americans, who currently associate deportation with televised raids and daily arrest quotas, tend to find its situation genuinely unbelievable when they learn about it. On my research visit there last year, I talked to politicians and activists across the spectrum, and the mainstream center-right (CDU) kept telling me the same exact surprising thing: It is nearly impossible to deport anyone from Germany. Many people whose asylum applications were rejected years ago, they complained, still live in accommodation financed by taxpayers, and the authorities cannot do much about it. I was skeptical, since complaining about weak enforcement is what right-wing politicians do everywhere. Then I checked the numbers, and they were mostly right.

To an outside observer, this may all sound like a bureaucratic joke. At the end of 2025, about 232,000 people in Germany were legally required to leave the country, and about 82 percent of them held a “Duldung,” an official certificate of toleration: the state has ordered you to leave and simultaneously certifies that it will not, for now, make you. Over the whole of 2025, Germany deported fewer than 23,000 people, while more than 34,000 planned deportations were called off or otherwise never carried out, usually because the person could not be produced on the day. More planned removals fell through than succeeded, and just as the CDU claims, many of those ordered to leave remain entitled to state-financed accommodation and benefits while they stay, sometimes at reduced levels. Note that such an arrangement is no kindness to the “tolerated” either: years of insecure status and shifting work restrictions, with no way to plan a life.

Of course, if you talk to left-wing pro-immigration activists, they would tell you a different interpretation of these exact numbers. German deportations have actually risen for five consecutive years, voluntary departures exceed forced removals, and most people with a Duldung cannot lawfully be removed right now, for reasons that range from missing travel documents and unresolved identity to family and medical circumstances. Germany also keeps converting the stock into legal residents through a recent “opportunity to stay” law for the long-settled.

But even a state that deports more every year has built a durable category holding a quarter million people at a time, suspended, often for years, between a legal decision and any genuine resolution, whether departure or secure status. Carrying out removals is “fundamentally the responsibility of the Länder,” the sixteen states, whose practices diverge so much that the government’s own Expert Council on Integration and Migration diagnosed a “manifest coherence problem”: identical cases end differently depending on which local office holds the file. The local official who must put a family on a plane absorbs the newspaper story, while the costs of never enforcing anything land on no one in particular.

This is all no longer hypothetical. After an Afghan man who was legally required to leave the country killed a two-year-old and an adult in Aschaffenburg in January 2025, Friedrich Merz built his five-point enforcement plan around this enforcement gap and pushed a nonbinding motion through the Bundestag with AfD votes, shattering a postwar taboo. Meanwhile, the removals that dominate headlines remain almost ceremonial: two heavily publicized flights of convicted offenders to Afghanistan and one man returned to Syria, around a hundred people in total, against a stock of 232,000. When mainstream government looks simultaneously cruel and ineffective, the parties that promise to be merely cruel do not stay at 20 percent forever. The AfD now polls ahead of the CDU, pushing 30 percent.

None of this paralysis is the price a liberal democracy must pay for being humane. Canada, everyone’s favorite immigration success story, closed about 23,000 removal cases in 2025, up from about 15,000 in 2023, more than four in five of them involving failed refugee claimants. Most of these people left unescorted after a final decision, with force reserved for a small minority, and the government openly funds its border agency to sustain 20,000 removals a year to protect the integrity of the asylum system. A wealthy democracy can clearly run one of the world’s most generous immigration systems and still make sure that a final no is usually followed by an actual departure.

USA: large-scale enforcement without credibility

The United States has the opposite pathology, and I watched it up close. In November 2025, I was still living in Charlotte, North Carolina, when Border Patrol commander Gregory Bovino brought his agents to town for what the government called Operation Charlotte’s Web. In five days, agents made over 250 arrests, of which, according to an internal DHS document obtained by CBS News, fewer than a third involved people classified as “criminal aliens.” On the Monday after the operation began, more than 30,000 students, a fifth of the district, did not show up to Charlotte-Mecklenburg schools. Many businesses had to close for weeks. Even Republicans flinched: former North Carolina Gov. Pat McCrory told The Daily Beast that his party “had the upper hand on immigration, as long as they were going after the criminals and the gangs,” and was losing it “because of the apparent disjointed implementation of arrest.”4

Unlike Germany, America has been removing people at scale: by December 2025, the administration claimed more than 605,000 deportations in under a year, plus far larger numbers it counts as self-deportations.5 But I should be honest that the two countries have more in common than the cable news versions suggest. America has its own Duldung-shaped stock: roughly 1.5 million people sit on ICE’s docket with final removal orders that have never been executed, a heterogeneous pile that ranges from absconders to people the law itself currently protects, stalled by many of the same things that stall Germany, from countries that refuse to take their nationals back to missing travel documents and limited detention and flight capacity. And the process behind those orders manages to be slow and thin at the same time: the immigration court backlog stands at about 3.2 million pending cases, asylum cases take four years or more to decide, most people in those proceedings have no lawyer, and in fiscal 2025 roughly 63 percent of removal orders in newly initiated cases were issued in absentia, to people who were not in the courtroom. In the end, in both Germany and the United States, the visible lesson for citizens is pretty much the same: the immigration system’s formal decisions do not describe what actually happens.

Our American discourse has its own version of the no-deportation assumption, and it hides inside an apparently moderate position: the idea that enforcement is legitimate only against “criminals.” When Janet Murguía of the National Council of La Raza famously dubbed Barack Obama “deporter-in-chief” in 2014, his administration had just set the modern record of more than 438,000 removals in a single year. The label stuck precisely because most of those removed were not dangerous.

Today’s critics of the Trump administration lean on the same premise from the other direction, arguing the crackdown is illegitimate because agents arrest gardeners instead of gang members. Prioritizing serious offenders is correct triage, and I share the revulsion at what replaced it. But notice what the premise implies if you take it as a principle rather than a priority: that violating immigration law as such may never, even after a full process, carry the consequence the law prescribes. Every legal system runs on discretion, but a categorical bar on enforcement essentially repeals the law it claims to soften.

Obama’s own enforcement leadership understood the distinction; in his final year in office, over 90 percent of interior removals involved people with serious criminal convictions, yet the system retained the stated capacity to remove others with final orders. Losing that capacity does not produce a more humane equilibrium; it produces the German scenario, in which a quarter of a million people live suspended between a legal decision and its consequence. Many of the tolerated would surely rank limbo above deportation, even if what they actually want is secure status. But a liberal democracy is still a democracy, and its immigration rules are ultimately up to the citizens and their representatives.

Deportations do work

OK, now you may say: fine, deportations may be legitimate in principle, but they do not work, so all this cruelty buys nothing. Or to be precise, deportations do not help deter people from coming illegally.

This reasoning usually comes in two steps. Step one says that restricting legal immigration cannot work because it just displaces the same people into illegal channels. Step two says that enforcement and deportations cannot work either, because push factors decide everything. People who risk their lives to cross the border are pushed out by the conditions of their home countries, and nothing the U.S. government does will influence their decisions.

Neither step survives contact with the evidence. Hein de Haas, Mathias Czaika, and their colleagues, who have studied migration policy effects as systematically as anyone, conclude that these policies are “generally effective”: restrictions genuinely shrink the flows they target, and the documented deflection into irregular channels is only partial. In a related European study, Czaika and Mogens Hobolth found that a 10 percent increase in asylum and visa rejections raised irregular migration by only 2–4 and 4–7 percent respectively, far from one-for-one displacement. Immigration is hardly the only domain where smart people make this mistake—the fixed-quantity way of thinking goes well beyond, into debates over prohibitions of almost anything.

So, whether you like it or not, immigration policy does change who comes, how many, and through which door, and the last two years demonstrated it at scale. Border Patrol encounters at the southern border fell from over 2 million in fiscal 2023 to under 240,000 in fiscal 2025, the lowest annual level in more than 50 years, with 2025’s monthly totals among the lowest ever recorded. Economists at Brookings and AEI project that net migration turned negative in 2025 for the first time in roughly half a century, with people leaving in response to the enforcement climate possibly outnumbering the people formally removed. Yes, the decline began in January 2024 under Biden and Mexico’s stepped-up enforcement did much of the early work, but no one can look at these numbers and still claim with a straight face that enforcement has no deterrent effect on unauthorized migration.

And deterrence is only a part of the case anyway. Criminologists have long distinguished between deterring crime and simply incapacitating the person committing it, and deportation runs through the same channels. Like incarceration, it works even when it scares no one else: it ends a stay the courts have already ruled unlawful and stops the local costs I mentioned above. Even if the deterrent effect were somehow close to zero, removals would still be worth carrying out, just like any other final judgment.

People respond to credible rules, which is exactly why the rules themselves need to be worth enforcing. We absolutely should be mindful that every deterred crossing or a deportee is a person who never gets what would probably have been the largest income gain of their life, and for people fleeing persecution the loss can be safety itself. What justifies paying that price is what it buys, a system trusted enough to keep admitting people, ideally more of them, through the front door.6

Deportations are what democracy owes citizens

The deepest reason someone needs to be deportable in a liberal democracy is about what democratic government owes the people who authorize it. A liberal democracy is a self-governing political community: citizens bear the local costs of policy, vote governments out when policies fail, and (indirectly) make binding rules for a bounded territory, including rules about who may enter and stay. A rule that can never be enforced against anyone is an aspiration rather than a law, and voters can tell the difference. So no one in particular has to be deported on any given day, but deportation has to remain possible in principle, with the actual number settled by good process rather than by paralysis or spectacle.

In my own recent research, I argued that protection for humanitarian migrants stays politically durable only on a foundation of trust in the broader system, including its enforcement. How much removal performance itself feeds that trust is exactly the kind of question my field should be testing instead of avoiding. A government that asks citizens to accept immigration while visibly declining to enforce its own final decisions is asking for trust it has not earned, and the immigration thermostat eventually kicks in: when voters sense the system is out of control, they push policy back in the opposite direction, usually harder than anyone intended.

None of this dictates where people go, and the destination is where enforcement earns or forfeits its legitimacy: in the normal case, the country of origin or a third country under an agreement that upholds basic human rights, and never a maximum-security prison run by a third authoritarian country. The legal baseline is worth stating plainly here, because pro-immigration advocates tend to keep blurring it, and annoyingly so. The Universal Declaration of Human Rights guarantees the right to leave any country and the right to seek asylum from persecution; it contains no right to enter the country of your choosing, and no right against removal once a lawful process has said no. To the extent any of this is enforceable at all, international law’s hard limits, above all the ban on returning anyone to persecution or torture, govern where and how a removal may happen; they do not make removal itself illegitimate.

The obligation runs to noncitizens too, and this is where my argument departs from the restrictionist one. A liberal democracy owes every person in removal proceedings a serious hearing, competent interpretation, accurate information about their country of origin, protection from return to persecution, and treatment that never becomes degrading. The people most betrayed by a collapsed distinction between lawful removal and arbitrary force are the noncitizens with strong claims, whose cases drown in a system nobody trusts.

What follows for policy deserves its own essay, and I have already sketched the general formula in Reason: administrative capacity, the binding constraint in Germany as much as in America, that delivers faster protection for those who qualify and faster removal for clearly ineligible cases, and enforcement predictable enough that the rules feel real in everyday life for regular citizens and noncitizens alike. The goal of this all should be a deportation system that is credible, bounded, and boring.

Serious pro-immigration advocates already know much of this, whatever they think of my framing. The American Immigration Council’s blueprint for rebuilding the asylum system concedes that even with every one of its reforms implemented, people without meritorious claims “will still be deemed ineligible for relief and ordered deported.” And as Dara Lind, whose chronicling of American enforcement failures I trust more than almost anyone’s, put it at that blueprint’s launch: “The problem isn’t that too many people are seeking asylum. The problem is that the U.S. hasn’t done the work to handle those who come. That’s what creates backlash.”

Whoever governs after Trump, and whichever mainstream coalitions survive in Europe, will inherit the same choice: own enforcement and make it lawful, or keep ceding it to people who enjoy it. The backlash to this administration’s methods will tempt Democrats to define themselves by the opposite of everything ICE does, just as humanitarian pressure tempts German officials to leave every file untouched. Both instincts feel humane, and both keep producing the politics we currently have.

So my ask to my skeptical readers—assuming you all are still here—is a rather small one. What I want from pro-immigration folks is a plain concession that in a liberal democracy with any immigration rules worth having, the number of justified deportations is not zero. I don’t want you to cheer for masked agents or endorse anyone’s removal targets to grant it. But once we say that deportations are legitimate and needed out loud, we can finally argue about the questions that matter: how many, after what process, and what to do when removal would be unlawful. A liberal democracy that cannot say no will not be believed when it says yes.

Share

Huge thanks to Mike Riggs and for their helpful suggestions on the draft.

OK, folks, real talk. Just so that we’re all on the same page, I’m an American nationalist, too. Not just a patriot in the sense of being proud of America, but a nationalist. And today I’m finally doing something about it.

By nationalist, I mean that we, American citizens, are one unique people, that the people who govern and the people being governed should belong to the same community, and that the American government should serve the common interests of this community. When necessary, it should serve those interests over both narrower domestic interests and foreign interests. Yes, especially as an immigrant, I also think America is superior1 to other countries in the ways that matter most for human progress. And, yes, I can believe all of that while still valuing human beings everywhere, regardless of origin.

Subscribe now

But long before I was a nationalist and started writing about immigration politics, I was writing about globalization and cosmopolitanism. In my undergraduate thesis, supervised by the late Ronald Inglehart, known for his ideas on postmaterialist values, I tried to measure how many people saw themselves as citizens of the world and why supranational identities varied across countries. I was young and naive enough, and living in Europe at the time, to think the future might be cosmopolitan (and even European). I was also eager enough to turn that hope into a dataset to prove globalization is good and inevitable.

That position still makes some moral sense to me. All humans are part of the same family, or at least close enough that any decent person, regardless of their religious beliefs, should feel the force of that idea. Borders are morally arbitrary in many ways, and the accident of birthplace remains the largest determinant of life chances in the world, more so than race, gender, or anything else. Much of my work on immigration still begins with that discomfort and with the fact that immigration empowers people to escape this unfortunate reality and contribute to this world.

Yet the more I studied public opinion and democratic politics, the harder it became to believe that cosmopolitanism (just like socialism) could do what its admirers wanted it to do. True, identifying with humanity as a whole can and probably should guide moral reflection for some people in positions of power. But no one has yet used it to build a durable democratic community, a welfare state, a public school system, or a legitimate and capable government that ordinary people feel belongs to them. In the end, most politics still runs through countries, whether we like it or not.

This is why I have gradually become more sympathetic to the idea of nationalism, or at least to American nationalism of the classical-liberal variety, certainly more than my younger self would have expected. Of course, we also know that nationalism, just like anything in large doses, can be dangerous and easily abused. But nationalism is also the political form through which modern democracy and equal citizenship have actually been built. In fact, liberal nationalism is arguably the most inclusive political identity that works at scale right now.

As for the something I’m finally doing about it: it comes at the end of the essay, where this newsletter becomes a bigger project and you get an invitation to become a part of it. But my argument on the virtues of US nationalism comes first, because the invitation depends on it.

What nationalism is and can do

Many folks I respect will still reject the label or even the idea itself. and , for instance, have made a forceful case against nationalism, and they have good reasons for doing so. If nationalism means ethnic hierarchy, coercive cultural uniformity, trade protectionism, or a state empowered to decide who counts as a “real” American, I oppose it too.

If you wish, you can call the good parts liberalism, patriotism, or equal citizenship. But I care much less about the label than about the ideas underneath it. Alex has also argued that American identity is based on a broad creed, not religion or ancestry. That is pretty much what I and many others mean by American liberal nationalism, too, whether you like the label or not.

So, forget the meager defenses of nationalism by self-proclaimed white nationalists, right-wing populists, or self-styled dual national philosophers with obvious foreign policy agendas. My personal favorite conceptual and empirical account of nationalism comes from Andreas Wimmer because it starts with nationalism as a principle of political legitimacy. The nation-state provided the ideological basis for democracy and the provision of public goods because these institutions could be justified in the name of a people of equal citizens, held together by a shared purpose and mutual obligation.

Yael Tamir, whose work is among the best-known liberal defenses of nationalism, makes a related point in a more personal register. People need meaning and continuity alongside liberty and opportunity. Liberalism supplies rights and freedoms, while nationalism supplies belonging and obligation. A good, successful society needs both.

That is also exactly the part of nationalism that many of my cosmopolitan and libertarian friends underrate. The welfare state, or if you’re skeptical, state capacity more generally, does not run on universal benevolence or market mechanisms alone. Public schools and other services command support because people see other citizens as part of a shared project. This kind of solidarity is imperfect and morally partial, but it is also how large-scale generosity usually becomes politically durable.

My own empirical research pushed me in the same direction from a different angle. In my PhD dissertation and, later in In Our Interest, I show that most voters are closer to what I have called altruistic nationalists than to pure selfish homo economicus or principled humanitarians. They genuinely care about others, but they weigh the well-being of compatriots first when judging policy.

We know from most other research that people are pretty groupish. But even I was still surprised by how stark this was in the data. In both UK and US surveys I ran with an incentivized charity choice game, only 10 percent of respondents decided to donate money to global charities over domestic charities or their personal pocketbook. Importantly, this held even among left-wing voters often wrongly accused of being cosmopolitan.

That sounds limiting if you want universal solidarity, but it helps explain why democratic politics works at all. International cooperation is indispensable, but no global organization can command democratic legitimacy or popular loyalty. The nation-state remains the primary site where democratic accountability and public trust converge. Of course, I should be frank that this is a compromise for some of us: politics built on national solidarity will always do less for the world’s poorest than pure cosmopolitanism promises on paper. But a promise that cannot survive an election delivers even less.

The beauty of American nationalism

The American experiment is about to turn 250 years old. That is a long run for a constitutional republic built on the repeated expansion of citizenship across the continent. And this is all beautiful.

Some readers will cringe at that sentence. But the flag, the oath, the Declaration, and the stubborn idea that people from everywhere can become part of one political people are not empty words and rituals. They are part of what made my own American life and success possible. And I do not feel embarrassed to say that. And neither should you, dear reader.

Policies that ask people to trust one another, make sacrifices, and wait for results need a shared we behind them. recently argued that America needs liberal nationalism back. I think he is right because the more realistic choice is between better and worse forms of nationalism: a liberal nationalism that treats newcomers and minorities as potential co-authors of the national project, or a narrower nationalism that treats them as permanent threats. A country with no shared national story will usually become more fragmented and vulnerable to even narrower forms of identity politics and conflict.

The American version specifically has an unusual promise. The United States is not the only civic nation, and it doesn’t always live up to its own principles. Still, as an immigrant who has lived in several countries, I find it hard to avoid the emotional conclusion that there is no better country for human life and progress.

Yes, American institutions sometimes run badly. Our immigration system needs serious reform, housing policy fails in many places, and the country regularly makes its own advantages harder to use. These shortcomings are worth fixing precisely because they keep Americans from the opportunity the country already offers. But never underestimate how much is working: America still has the size, the markets, and, more unusually, a national identity that almost anyone can adopt. Those are all the right ingredients for renewal.

This is one reason American debates about progress have so much force. Americans expect the country to solve problems and get better. In many places I’ve lived, the failure to build enough housing or even ensure people are not dying from the heat can feel like ordinary decline. Here, it feels like an offense against the country’s promise. ’s distinction between progress and abundance is useful here: abundance starts with making it easier to build, while progress also depends on culture and ambition. A better liberal nationalism can acknowledge exclusion and broken promises while still insisting that the American project is worth fighting for. That better version is not automatic by any means, but it is still available if we choose it collectively.2

Immigration needs nationalism, and nationalism needs immigration

This brings me back to the argument that started Popular by Design. The original manifesto promised writing about immigration policies that work: policies that are evidence-based, demonstrably beneficial, and built to earn public support. A few years earlier, in one of the first public essays where I laid out the argument that became my book, I argued that immigration policy needs more enlightened nationalism. But now I see that the argument goes both ways.

Immigration needs nationalism because voters need to see a national benefit. Most voters will not support freer immigration because economists tell them it increases global welfare or because advocates tell them that borders are morally arbitrary. They support immigration when they can see how it serves the country they belong to: by filling real labor needs, strengthening universities, reuniting families under rules they trust, and making the country more dynamic.3 Pro-immigration politics that sound embarrassed by the national interest often fail because they ask voters to treat their attachment to compatriots as a moral defect. A more durable pro-immigration politics starts from the fact that voters care about their country and then shows how better immigration policy can serve their country.

But nationalism also needs immigration because a liberal nation that stops adding new members slowly becomes less liberal. Membership begins to look like ancestry. The more a country treats the national community as closed, the more its creed becomes decorative. America especially cannot defend a civic identity while treating newcomers as contaminants.

Days before the country's 250th birthday, the Supreme Court reaffirmed birthright citizenship, rejecting the idea that a child born on American soil could be denied membership because of their parents' immigration status. Indeed, a confident nation can afford to be generous because it expects to absorb newcomers and be improved by them, while an anxious one treats every arrival as a referendum on its survival. We absolutely can and should decide who can join us and under what conditions.4 But the phrase “nation of immigrants” is more than a sentimental slogan. It is a theory of national strength.

The same logic applies to other important issues

Immigration is the clearest case for me. The deeper principle is that good policy has to be built for imperfectly informed humans who care about their communities and judge policy by visible consequences. National interest is often the bridge between substantive merit and democratic legitimacy. People need to see how a policy makes their country more capable, more prosperous, more trustworthy, or more open to the people who can join and strengthen it.

Housing, not to mention newly controversial heating and cooling issues, is another useful example. For a long time, many supporters of new housing described opposition as the product of selfish homeowners or exclusionary racism. But housing politics also runs through people’s ideas about community, aesthetics, and whether change will make the place they live better or worse. and have been especially good in showing how local veto points turn ordinary family needs into a national housing problem. Recent work on the symbolic politics of housing and sociotropic aesthetic judgments points in this direction. If people can see housing growth as improving their communities, the politics can change.

Artificial intelligence is another case where the public argument is already too crude. I came to the AI debates almost by accident, after using the tools in my own research, and found a messier story than either panic or accelerationism allows. has made the interesting case that large language models may push public opinion toward more expert-aligned information, partly reversing the fragmentation of social media. ’s approach to AI policy is useful here because he takes transformative possibilities seriously while keeping a high bar for policy claims made under uncertainty. I am less sure than either of them, but good AI politics will require institutions and publics capable of deciding what should be accelerated or constrained, especially as we weigh the more consequential aspects of AI, such as job disruption, alignment, and catastrophic misuse.

Population decline is the third case, and probably one of the biggest issues facing humanity alongside AI now. Dean Spears and Michael Geruso argue in After the Spike that the future challenge may be too few people to sustain the progress America and the world need. Fewer people means fewer workers, fewer caregivers, fewer ideas, and fewer chances for the lucky accidents that drive progress.

Immigration cannot solve global fertility decline by itself. Despite my weakness for housing theories of everything, permitting reform cannot solve it on its own, either. AI could offset some labor scarcity and help people do more with less, but a country that treats children and national membership as afterthoughts should not expect robots to rescue it from demographic and institutional decline. All these questions point back to the same political fact: progress depends on people who trust one another enough to build, adapt, welcome, and take risks together.

What Popular by Design is for: A new chapter

When I launched Popular by Design, I described it as a newsletter about immigration policies that work. Almost a year in, that remains true. Immigration and public opinion will stay at the center because they are the topics I know best and because immigration is the best test of the broader idea.

But the broader idea was always there. Good policy has to work substantively and politically. It has to account for trade-offs, limited information, group attachment, institutional incentives, and the fact that voters often judge policies by whether the benefits are visible to people like them.

Over the next year, I want Popular by Design to keep making that argument in a wider set of cases without becoming a generic politics newsletter. The center will remain migration, public opinion, and policy design. Around that center, I want to write more about population change, AI, housing, and the other areas where progress depends on turning good ideas into policies that ordinary people can understand and support. Behind the scenes, this newsletter is also becoming the public face of a research lab I’m starting at Notre Dame. More on that soon.

If Popular by Design is doing its job right, readers should come away with a better map of politics: where voters are wiser than elites think, where voters are wrong, where experts are useful, and where policy design can change the incentives so good ideas survive contact with democratic consent. Liberal nationalism fits the project because it lets us be ambitious without pretending that politics can float above membership, trust, and country.

So today, I am turning on paid memberships for Popular by Design. As I promised, every essay and all other substantive content will always remain free. That is central to the project because the point is to influence the public conversation, not to hide the core argument from people who cannot pay.

By now, you should have a clearer sense of what I am trying to do here. Popular by Design will remain rooted in immigration, but the community around it is for people who think politics works better when ideals are attached to institutions, national interest is treated as a democratic constraint rather than an embarrassment, and policy claims are judged by their consequences. If that way of understanding politics is useful to you and you want to be part of our community, paid membership is the way to help make it sustainable.

Popular by Design is now almost a year old, and it grew during my sabbatical, when I had more time to write, edit, and build the audience. Next month, I am getting back into the normal life of teaching, research, and university service obligations. If the newsletter is going to stay serious, publish regularly, and host useful conversations, it needs a base of recurring support.

There are two tiers at launch. Supporters contribute $15/month or $120/year. They help keep the essays free and receive invitations to reader meet-ups online and in person, where thoughtful folks can trade ideas and build connections around immigration, public opinion, and policy design.

Founding Members contribute $1,000/year, recurring annually. They join a smaller circle that gets two hour-long conversations with me each membership year, helps shape topics, and gives the project an informal kitchen cabinet. If your institution would benefit from a deeper engagement with these ideas through a book talk, workshop, or consulting engagement, reach out directly.

The money funds continuity: editing, guest posts, occasional reader events, and the time required to keep doing this well, while I have a full-time job. I am asking readers who care about this pragmatic policy community to help keep it open and improve it. I’ll be honest, though: giving folks the version they’d rather not hear has come at a price, from Bluesky curses and public calls to fire me to a few genuine death threats, ugly enough to end up in the news. I have thick skin and a steady job, so I can take it, but knowing there are real people behind this work, and not only the loud ones telling me to shut up, is a big part of what keeps me doing it.

If you want a public conversation where immigration is treated as a policy problem instead of a tribal signal, where patriotism is open enough to welcome newcomers, where AI and population are discussed with curiosity and discipline, and where progress is judged by whether policies actually work, I hope you will become a paid member. If money is genuinely a constraint, email me. I want students, early-career researchers, practitioners outside the United States, and thoughtful readers who cannot pay to stay in the conversation.

The United States of America, the country I became part of, is still the best bet for human progress. But no country keeps that status by inertia. It needs better policies, better institutions, and, yes, a better nationalism: confident enough to love what America is, honest enough to improve it, and open enough to keep adding the people who will make it stronger.

Share

Huge thanks to all the folks who read and commented on this essay: , , , , , , and .

Update: Just as this went out, the Supreme Court upheld birthright citizenship in a 6-3 decision (Trump v. Barbara), striking down President Trump's executive order and holding that the Fourteenth Amendment guarantees citizenship to nearly everyone born on U.S. soil.

This month I went to Ireland for the first time, splitting two weeks between Dublin and Belfast for two conveniently timed political science conferences. To make sure this was not just a nice European vacation (before the heat wave), I took one of those “both-sides” walking tours of “the Troubles” through the peace walls and murals, standing in the streets where neighbors were killing each other just a few decades ago. I teach courses on war and conflict, and “the Troubles” is one thing I always assign to my students, to show how even a relatively wealthy democratic society slid into something close to civil war. It is all a useful corrective to the comfortable assumption that prosperity completely inoculates a country against political violence.

Now for the good news: not every academic conference is a slog. I had two on this trip, the Conference of Europeanists (CES) in Dublin and the inaugural European Political Science Society (EPSS) meeting in Belfast, and both were genuinely good. I came away impressed by the venues and the quality of the presentations, which is not something I say lightly. Regular readers know how little I enjoyed my last big conference, the 2026 ISA convention in Columbus, where I sat through (mostly absent) tenured scholars reading bad arguments off their slides. These were a different experience entirely, and a reminder that both human and AI slop is avoidable when people care. Another benefit of a good conference: in Dublin I caught a presentation by Aubrey Westfall on Canadian immigration federalism that genuinely blew me away, and I asked her on the spot to write it up for this newsletter, which she did.

Subscribe now

There was one EPSS panel that deserves its own paragraph. I sat in on a session where people shared how they use AI in their research workflows, which was practical and mostly civil. But you could feel the polarization in the room, and at a conference for quantitative political scientists of all places. A vocal share of the audience was deeply skeptical of the recent AI advances, citing the environmental footprint, equity, and European technological sovereignty, with a few proposing that Europe build its own large language models to rival OpenAI and Anthropic (bless their souls). I admire the ambition. But we, academics, are people who, collectively, have not yet figured out how to run a journal submission system that works, so I am not holding my breath.

In case you missed it: On June 1, I (or Codex?) published Pangram Policing is the New Grammar Nazism, my defense of the possibilities of AI writing against the rise of detector-policing. After that came How Canada (Un)made My Thinking on Immigration, on how a points-based system and broad elite consensus still produced the steepest immigration backlash its pollsters have recorded in half a century, and Why Evidence-based Policymaking Is Overrated.

That last one turned into the thread of my whole month. Most of my time online went to a running fight over degrowth and air-conditioning denial, which the Wall Street Journal picked up as an op-ed on Europe’s strange war on cooling. It is the same argument as the evidence-based piece: a comfortable expert class keeps moralizing against policies that are demonstrably beneficial and popular, whether that means asylum work bans or keeping people alive in a heat wave.

We also just ran the guest post I teased above: Aubrey Westfall of Wheaton College on whether local governments should pick their own immigrants, drawn from her own fieldwork interviewing the Local Immigration Partnership directors who run those programs. If you want to write something yourself, DM me.

Here are the June links (linking does not imply endorsement):

• The National Constitution Center has a calm, non-partisan rundown of the Supreme Court’s TPS decision and the companion asylum ruling. In Mullin v. Doe, the Court held 6-3 that it cannot review how DHS ended Temporary Protected Status for roughly 350,000 Haitians and 6,000 Syrians, who now stand to lose work authorization and protection from removal.

• has an honest ten-years-on retrospective on Brexit, conceding the economic case was oversold and that sovereignty over immigration was the actual driver. His line that “constituencies are created who are perpetually disappointed” is the salience problem in miniature, and a good companion to my running argument that you cannot wish immigration off the agenda.

• Hiroko Yoda, in a guest post for about what defines Japanese national identity, argues that belonging in Japan runs on shared culture and conduct, “how you read the air,” more than on creed or bloodline. I think this squares pretty well with my running claim that Japan is uncannily normal, and it complicates the lazy story that Japan is simply closed by ethnicity.

  • It pairs well with Ryoko Yamamoto‘s new JEMS article, “Fashioning homogeneity”, on how Japan keeps a self-image of homogeneity while quietly importing workers, a process she calls decoupling.

• has a sharp takedown of the claim that immigration researchers are ideologically biased. Borjas and Breznau found that researcher ideology predicts findings, but other researchers show the effect appears only in one preferred specification and vanishes across most reasonable alternatives. As usual, the study accusing the field of doing something shady turns out to be the one doing it. Note that this decidedly doesn’t imply immigration research is not biased!

• lays out the strongest version of the skills-and-culture case against betting on immigration, adapted from his UATX debate with . As someone who is probably somewhere in-between Jones and Caplan on the issue, I disagree with much of it, including the very deep cultural-persistence pessimism, but it is the cleanest steelman of selective immigration in one place, and worth reading precisely because he takes the tradeoffs seriously.

• Mihnea Cuibus at Oxford’s Migration Observatory has a clean evidence review on why asylum seekers come to the UK. The finding worth sitting with: countries that deported a larger share of refused applicants saw no subsequent drop in applications. Deterrence through enforcement is a weaker lever than its advocates assume, which is exactly the inconvenient kind of evidence my recent piece on honest judgment says we should take seriously.

• has a thoughtful essay on why right and left see two different worlds. His point is subtle: the two sides often accept the same verifiable facts but filter them through different interpretive frameworks, the “pseudo-environments” that decide which facts register and what they mean. Borrowing Walter Lippmann’s line that “we define first and then see,” he captures why misperceptions are so sticky, which fits my long-running point that they are downstream of identity and rarely move when you correct them.

• wrote a visual follow-up that turns my AI-writing spectrum into a two-dimensional chart. In the Pangram piece I put writing on a single provenance spectrum, from where provenance matters most to where only the content does. Archie rebuilds it on two different axes: how much of the work the AI did, from a one-shot prompt down to just fixing typos, and how mechanical or ideas-led the writing is, from admin emails to novels. It seems like Archie (and other folks I deeply respect like , , and ) strongly disagree on both the usefulness and morality of AI writing, but having a chart like that is a genuinely useful upgrade that helps us see what exactly we disagree about (Archie’s yellow edge cases are my clear green yeses). It also makes me genuinely happy that nobody is trying to defend the human authenticity of admin emails :)

One last thing before I go. Big things are coming on the Fourth of July, for the world and for Popular by Design, and I will have much more to say very soon. Stay tuned!

Share

I have a new op-ed (ungated) in the Wall Street Journal. (I didn’t pick the headline, but “Air Conditioning, Scourge of the French Left” is hard to beat, and Chad Crowe’s illustration of a café standoff over an unplugged window unit is pretty neat.)

Subscribe now

The short version: as Europe bakes, the boldest plans to cool people come from the populist right, while much of the left treats air conditioning as a selfish indulgence and an ecological menace. Cooling people in a deadly heat wave is about as demonstrably beneficial as policy gets, which makes this one of the rare debates where the far right simply has the better of the argument.

I did not plan to become an air-conditioning expert this month. What started as one cranky take turned into a whole week of near-total fixation, and it pairs unexpectedly well with my last piece on why evidence-based policymaking is overrated. Basically, the same voices who spent the pandemic insisting we “follow the science” are now reasoning from a narrow read of climate models to the conclusion that you should just sweat it out.

A few things the op-ed only gestures at. First, this is all bigger than spoiled American tourists missing their comfort. We Americans fail at plenty of things the rest of the rich world finds baffling, from public transit to gun violence, so the point is not that the US has it all figured out. It is that almost everywhere cooling is normal, from Japan to the Americas, people look at Europe with the same sincere puzzlement: how does a continent that gets so much else right decide that a machine invented in 1902 is the one thing it cannot bring itself to use?

Second, the death toll, grim as it is, is probably the smaller part of the story. keeps making the point that deaths averted are likely a minority of the well-being lost to a lack of cooling. Heat that wrecks your sleep, your concentration, and your ability to work or study is an enormous and largely invisible cost. When someone shrugs off a heat wave as “a week of discomfort,” what they are shrugging off is billions in lost productivity and health, every summer, borne by the people least able to escape it.

Third, the debate is also missing the people best placed to settle it. It has been left largely to sustainability advocates reasoning from their picture of climate change toward conclusions about how the rest of us should live, while the quantitative social scientists who could actually measure what cooling does to human welfare, and what the public wants, have mostly stayed out. I went looking for good survey data on how Europeans feel about air conditioning and found almost nothing. What little I have come across even hints that support for cooling may run stronger on the political right, which would be its own puzzle worth taking seriously. So I will probably end up fielding some of these questions myself.

And this is finally why I keep circling back to it. People ask why I am so allergic to degrowth, and Europe’s response to the heat is the clearest answer I have. Air conditioning is just one instance of a much larger instinct: that wanting to be comfortable is a little shameful, that the rich world should consume less and endure more, that even a cool room counts as a kind of vice. The trouble is that heat waves are happening, and they will keep happening, more often and more severely. You cannot wish that away with an appeal to sobriety, and no amount of environmental concern changes the fact that a cool room is one of the cheapest and most reliable ways we have to keep people alive and well. And, as I mention in the piece, you do not have to like Marine Le Pen, or agree with her or other populists on immigration, to admit she has this one right.

As always, I’d love to hear what you think I’m getting wrong, especially if you live somewhere this fight is live. And if anyone knows of good survey data on how people outside the US actually feel about cooling, I’d be grateful to see it.

Share

As more US states and cities demand a say over who settles within their borders, it helps to know that Canada has been doing a version of this for thirty years. Today's guest essay by Aubrey Westfall, a political scientist at Wheaton College, draws on her own interviews with the people who run Canada’s local immigration programs to show how provinces and towns select immigrants to fit local needs. She’s just as good on the harder part: how the federal government keeps that patchwork from splintering into conflict. I hope you all enjoy the read!

In 1996, the Winnipeg Jets left for Phoenix. For many Manitobans, losing their NHL team to a Sun Belt city was more than a sports heartbreak; it confirmed a quieter fear that their province was being left behind, its people and economic life drifting toward Toronto, Montreal, and Vancouver. Out of that anxiety came an unlikely experiment. Rather than wait for Ottawa to direct newcomers its way, Manitoba set out to choose its own immigrants, and pioneered a model that has since reshaped how the whole country decides who gets to come.

That model is immigration federalism: provinces, and increasingly local communities, selecting and admitting immigrants to fit their own labor needs instead of leaving the choice to the national government. Canada has spent three decades building it, and its experience speaks directly to a question now pressing on the United States and Europe: can you give regions meaningful control over immigration without fracturing the country or feeding the backlash now reshaping democratic politics? Canada’s answer is a qualified yes. The conditions that make it work, and the one that makes it fragile, are what the rest of us should be studying.

Subscribe now

Canada is the quintessential country of immigration. Its population is made up of over 450 ethnicities, and 23% of the population is foreign born (compared to 15% in the US). Canada’s robust economic immigration is facilitated by a points-based system that determines eligibility based on a prospective migrant’s human capital, and their potential to meet current and emerging market needs. Points are awarded for skills, education, language, work experience, age, and other characteristics. Prospective migrants exceeding a certain threshold are invited to apply for a visa. Such a system is intended to clearly match immigration levels to Canada’s demonstrable needs. It is also intended to be transparent, flexible, and responsive, as governments set the criteria for awarding points, and regularly review and update the criteria and the planned levels of migration. The system assumes an admitted individual responds to the domestic labor market to find appropriately skilled work, wherever the jobs may be.

And yet, at the turn of the millennium, the vast majority of immigrants in Canada disproportionately settled in just three major cities in Ontario, Quebec, and British Columbia, many in jobs mismatched to their skills. During the 1990s, the other provinces and territories began to recognize that the national immigration policy was not working for them or the newcomers, and that they were missing out on a critical influx of human capital.

Where immigration federalism began

For Manitoba, the message of losing out was driven home when their beloved hockey team moved due to financial shortfalls. Devastated by the loss, and driven by a desire to grow Manitoba’s economy, a group of businessmen formed the Business Council of Manitoba (BCM). The BCM worked with other stakeholders to build immigration capacity, and as a result, Manitoba became a forerunner in developing the first of Canada’s Provincial Nominee Programs (PNPs). Manitoba’s innovation worked—migration numbers started to rise shortly after the implementation of the PNP, and now it receives a proportionate share of Canada’s immigrants. Encouraged by the population and economic growth, the Winnipeg Jets returned to Manitoba in 2011.

More broadly, the PNPs are intended to facilitate the settlement of migrants outside of the three largest Canadian cities, and to help employers address their workforce needs in more context-sensitive ways. Through a PNP, provincial governments can nominate and select a potential economic immigrant, either from the candidate pool in the federal government’s system, or their own provincial selection systems, which are designed to reflect provincial priorities and labor shortages.

Historically, the vast majority of the candidates nominated by the provinces are invited to apply for a visa by Immigration, Refugees and Citizenship Canada (IRCC - the federal department responsible for immigration). Between 1998 and 2009, PNPs were introduced in every province (excluding Quebec, which had already made its own immigration arrangements based on its special status as a minority nation) and every territory (excluding Nunavut). By 2019, provincial nominees accounted for 35% of all new economic migrants to Canada and the PNPs became the single largest economic immigration stream into Canada.

More place-based policy innovations

The PNPs achieved their objective of getting more immigrants to land in different parts of Canada. They also demonstrated how delegating the management of migration to lower-level governments can better respond to diverse needs and capacities. Even better, according to a government report, around 89% of those who immigrate on PNP schemes tend to stay in their province for the first year, and 86% of those are still there five years later. However, there is significant variation across provinces. Those provinces that need immigration the most—especially the Atlantic provinces—suffer the lowest five-year retention rates (59% on average).

In response, the Atlantic Immigration Program (AIP), which started as a pilot in 2017 and became a permanent program in 2022, was developed as an employer-driven program. The Atlantic provincial governments designate employers to participate in the program, and the employer commits to intercultural competency training and enhanced settlement support for newcomers by connecting with service providers. The program then connects employers with qualified immigrants from all over the world. Once the employer makes a job offer, the province endorses the candidate to IRCC. By requiring local businesses to invest in the success of their newcomer employees, the AIP makes the immigration process a community enterprise, and as a result, improves long term retention of newcomers.

For many within Canada, the existing provincial and region-based programs still didn’t go far enough. Those in rural, northern, and francophone communities advocated for a distinct policy focus to meet their unique needs. In response, the IRCC issued Ministerial Instructions to quickly introduce a series of employer-driven pilot programs within selected communities, built on the example of the AIP. The pilots are centered on the community or municipality more than the province, as the community identifies their labor needs, designates employers for the program, and recommends the migrant for permanent residency through the program. These pilots are ongoing (and could be stopped at any time), but early assessments indicate success at attracting migrants to these less conventional destinations, as demand outstrips allocated spaces.

Avoiding the risks of immigration federalism

The layering of fragmented immigration programs and pilots at varied levels of government creates a very complex system. There are over 100 pathways of migration into Canada for skilled workers, business people, and families, all with potentially moving targets. Just like in the US, this complexity undermines the transparency of the points-based system and makes it challenging for immigrants to navigate. At the same time, the sophistication of the system allows it to better serve the purposes of the Canadian government and meet varied needs within the country.

Still, fragmented systems come with risk. Where the central government allows provincial governments to assert control over key policy areas, it effectively establishes the institutions and competencies of a “mini-state,” prepared to function more independently. This has the potential to undermine national unity and increase conflict between levels of government, even to the point of secession. This possibility became reality for Canada with the announcement of an Alberta referendum on October 19, 2026, focusing on ten questions relating to immigration, constitutional amendments, and independence. The Albertan premier partly blames provincial budget deficits on federal immigration policy, even as her government maintains an active “Alberta is Calling” program to attract skilled tradespeople.

Even with the threat of the Alberta referendum, and despite the fragmented nature of Canada’s immigration federalism, there is still remarkable national coherence in the general approach to immigration. Xhardez and Tanguay analyze the design, requirements, and distribution of provincial nominations within PNPs, and note policy variation within prevalent employment-based PNP streams, which require prearranged work. The emphasis on employability aligns with federal policy that requires skilled work experience, or targets selection for specific in-demand sectors and occupations.

Why it all doesn’t fall apart

I saw similar trans-provincial convergence in interviews with directors of Local Immigration Partnerships (LIPs), which are bodies intended to integrate the inclusion of newcomers in the general community planning agenda. I was told about distinct programs intended to serve the particular needs of their community. Examples include targeted programs to welcome Ukrainians in rural Ontario, collaborations with the police department in Medicine Hat (AB), music programs aligned with London’s (ON) designation as a UNESCO city of Music, even distributing vitamin D to newcomers in the north. The activities vary, but the ethos driving the activities and the vocabulary used to describe the work were the same. As one LIP director in Ontario put it, “The bottom line for all the local immigration partnerships is creating a more welcoming community, a more inclusive community. Whatever the political system they have, this is the goal…I haven’t seen many differences… every community is unique, every LIP is different, but I think the objective is the same.”

How is interprovincial agreement in the management of migration achieved, given the robust opportunities for immigration policy fragmentation? Analysis of my interviews offers three explanations.

Federal funding. Among those working directly in migration, such as the LIPs, service providers, and the IRCC, federal funding is perhaps the most powerful coordinating tool. The Canadian government provides extensive funding to support the many actors in the immigration sector. In July of 2025, the government announced an allocation of $3.2 billion over three years to more than 520 organizations intended to improve the integration of newcomers. Recipients include service providers and LIPs, both of which are almost entirely reliant on federal funds. The amount of funding is tied to the landing rates of immigrants, and is also determined by federal strategy. High reliance on federal funding allows the government to define the programmatic direction and eligibility, which manifests in parallelism and congruity in the migration sector across provincial or regional entities.

Formal and informal collaboration. Consistent collaboration, within and across provinces, and between the federal, provincial, and local actors, supports coherent shared governance. Much of the collaboration within the migration sector is informal, built on relationships and a desire to share and implement best practices. However, collaboration is also encouraged through formal federally-supported structures. The federal government supported the development of a LIP secretariat as a coordinated planning body for all LIPS. The federal government also sponsors annual professional conferences for those who work in the migration sector. There are ministerial, provincial-municipal, and municipal advisory “tables” (working groups) that promote dialogue and information sharing among stakeholders at all levels of government.

Low politicization of immigration. Immigration in Canada is bureaucratized, at least in part because the system is so complex. And, because immigration is managed as a public policy embedded within formal structures of multilevel governance, the immigration sector is relatively depoliticized. The federally-funded actors within the immigration sector are required to remain non-partisan, which is not extremely difficult, because immigration has not been a partisan issue for the last few decades. There is partisan consensus about the desirability of immigration, at least at the symbolic level, and even if a party is not ideologically committed to immigration, there are strong electoral reasons for parties to maintain pro-immigrant and pro-minority messaging. Until recently, the majority of Canadians supported the size of immigrant flows, which made it less politically salient. As a result, it has not been used to support divisive politics, notwithstanding the very recent case out of Alberta.

These three factors go a long way in explaining how immigration federalism has been implemented in Canada without succumbing to the serious risk of fragmentation.

What can other countries learn from Canada?

The Canadian case demonstrates that immigration federalism can work to serve place-based interests in countries with multi-level systems serving regions with distinctive needs. Insistence that a common border requires a centralized immigration policy that exclusively governs all migration into a country ignores the obvious reality of diverse economies, cultures, and polities within a country, and how they can benefit differently from immigration. There are risks associated with decentralizing powers over immigration, but if the federal government financially supports a multi-level migration sector, and assists in coordinating intergovernmental networks, it can create efficiencies and push back against forces of fragmentation.

The development of Canada’s immigration federalism is a feature of Canada’s willingness to experiment and fail. The LIPs began as a regional effort in southern Ontario, and have since spanned the nation. The directors at LIPs shared a willingness to discuss both successes and failures in an effort to better inform others about what to adopt and avoid. The stable financial support from the federal government allows them to be honest—they do not have to oversell their successes to private donors. PNPs were adopted gradually over a decade because the early adopters of the PNPs were watched by other provinces, who took cues from policy successes to tailor their own PNP agreements.

The pilot programs implemented with Ministerial Instructions allows for the design and implementation of small initiatives without time-consuming Parliamentary review. If the programs succeed, they can go through a process of regularization, with alterations to fix identified problems. Any multi-level system could embrace their ability to use lower-level systems as a type of “laboratory” to explore policy options, but systems of collaboration and review should be implemented to facilitate learning.

In my view, the most difficult element to replicate, and the most fragile one, is the political and social consensus around the desirability of immigration. In contexts where immigration is hotly contested, or where political distrust is high, experimentation and policy flexibility may not be democratically viable, at least not in the short term. International policy analysts like have identified a persistent representation gap, where there is a disconnect between public attitudes about immigration and the immigration policies passed by governments. The far right has built a base of support in Europe and the United States on this very issue. Mainstream parties have responded by moving right on immigration, with mixed results.

and others in this newsletter argue that another solution could be found in smart policy implementation that effectively demonstrates the benefits of immigration to the national population. Place-based migration policies are an underutilized way to build trust in the system by making the contribution of immigrants clear and directly applicable to the needs of individual communities.

For the US in particular, the Canadian example demonstrates the viability of a state-based visa program. American states are institutionally capable of administering such a system; however, the current American immigrant integration sector is typified by infrequent and ad hoc collaborations. In order for a state-based visa system to achieve policy coherence and operational efficiency without compromising national consensus, more formal structures of intra- and interstate collaboration must be developed.

Crucially, this structural coordination must extend to funding; the current federal system of competitive grant programs for NGOs—which remain heavily reliant on private donations—results in a piecemeal approach that does not serve the national interest. In order to promote equity and systematic unity, the federal government should develop a consistent federal budget line to provide high-level strategic direction while continuously supporting service provision and community inclusion at the local level. Instating a national or hybrid federal-state program modeled after Canada’s LIPs would help communities coordinate resources and develop coherent strategic plans.

In the absence of political will at the federal level, Manitoba’s experience reminds us of the power of community-level policy entrepreneurs. States can take the lead by funding and organizing regional initiatives. Ultimately, as the development of Canadian immigration federalism proves, good policy ideas are contagious.

Share

Three years ago, the United States declared the end of the pandemic emergency, and the country exhaled. You, dear reader, have probably filed the whole thing away by now. I’m sorry, but I just cannot let it go, and reading Macedo and Lee’s excellent “In Covid’s Wake” on my way back from Ireland last week reminded me why. The virus, of course, was not the fault of any of us in particular.1 But how we reasoned about it as a democratic society was absolutely our fault. The so-called experts spent years telling the rest of us to “follow the science,” then claimed a certainty the science never gave them, dressed value judgments up as technical ones, told plenty of noble lies for our own good, and waved away every cost that sat outside their own narrow expertise. We were drowning in data, but what we lacked was honest judgment.

This knee-jerk evidence-based mindset is not just a pandemic thing, and it shows up wherever we treat one kind of data as the only kind that counts. In 2018, a top medical journal published a randomized controlled trial (RCT) finding that parachutes did nothing to prevent death or injury when people jumped from an aircraft. The catch was that the planes were parked on the ground, and the mean jump altitude was about half a meter. The whole study, of course, was a joke. The authors were not against experiments as such, but they were mocking the common reflex among their colleagues that treats RCTs as the only respectable form of knowledge, even about a claim you could check by looking out the window of a plane.2

Subscribe now

I have spent much of my professional life contributing to and asking for better evidence in immigration debates, usually to the quiet exasperation of people on my own side, so I am not about to start sneering at data and acting on “vibes.” Good policymaking does need evidence, cost-benefit analysis, and careful counterfactual thinking. But it also needs humility and better judgment about what kind of evidence a given question can actually require, if any.

There are some policies that are so obviously good that we should not need a perfect study to try them, like legally allowing more housing where demand is high, keeping reliable low-carbon power online when the substitute is fossil fuel, or increasing visas for foreign top talent that everyone says they want. Reasonable people can argue about details and trade-offs. But on questions like these, the case for action does not depend on a perfect RCT, and the burden of proof should not be infinite.

There are also policies so obviously bad that we should not need a study to stop them, like asylum seeker work bans. Seriously, you do not need a randomized trial, or any hard evidence for that matter, to predict what happens when you forbid a willing adult in a legal limbo to work, pay to shelter them instead, and then point to their idleness as proof the system is broken. Most of the hardest calls in policy look more like this than some great mystery requiring a well-designed experiment.

What “evidence” actually means

One rung down from the parachute sits dental floss. I personally don’t enjoy flossing, but most of us can probably feel the difference after flossing out a big meal, and our dentists can see it at the next cleaning. Yet when some journalists seized on a review that found little high-quality randomized evidence for flossing, a run of headlines announced there was no evidence it works at all. As the historian of science Naomi Oreskes points out, that was a misreading. We should be “broad-minded about evidence,” she argues, counting professional experience and ordinary observation, especially where a clean long-term trial is impractical or will never be funded.

The same confusion, though with far higher stakes, ran through the COVID pandemic. In their recent (highly-recommended) book on the topic, Stephen Macedo and Frances Lee make a startling case: much of the relevant knowledge on how to deal with a respiratory pandemic was already there, but governments across the West largely set it aside. Before 2020, the dominant pandemic preparedness plans were skeptical of sweeping measures like lockdowns and prolonged school closures, warning that the evidence for them was weak and the human and economic costs high. In the panic of early 2020, those governments scrapped that guidance almost overnight and then projected a confidence that science never supported.

It helps to sort things by the kind of feedback they give, and to what end. A parachute gives the simplest kind: the benefit of staying alive is immediate, individual, and impossible to miss, so a trial would only confirm what anyone can already see. A new drug or vaccine gives the opposite: the benefits and costs may be genuine, but they are also much more diverse and often invisible. The infection that never arrives is easy to confuse with luck or the body healing on its own, which is exactly why a randomized trial with placebos is essential and why modern medicine runs on them.

Most government policy sits in between. You can usually tell whether a policy points toward gain or loss long before you can put a clean number on it, and the effects run through labor markets, prices, and politics all at once, so no single trial can isolate them, and often none can be run at all. But you cannot settle it just by looking, the way you can with a parachute, and you cannot settle it with one clean experiment, the way you can with a drug. What is left is judgment based on cumulative evidence, comparisons across places and times, and honest cost-benefit reasoning under uncertainty, all stated openly instead of disguised as settled science.

Another casualty of our actually existing knee-jerk evidence-based policy is the prior question of what even counts as relevant evidence, and whose expertise gets to decide. Here, Macedo and Lee are devastating. As they document, a niche cohort of public health and infectious disease experts was suddenly treated as the only legitimate authority on a crisis that touched every part of life, and their lens was narrow by design: fixed on minimizing infections, it pushed almost every other effect, economic and non-economic alike, off the table as someone else’s department.

Francis Collins, who ran the National Institutes of Health, admitted that the “public health mindset” he shared led him to “attach infinite value to stopping the disease” and “zero value to whether this actually totally disrupts people’s lives, ruins the economy, and has many kids kept out of school.” In Britain, the chief medical officer, Chris Whitty, told the official Covid inquiry that adding economic or social experts to the government’s advisory group would have made it too “unwieldy.” Stopping the virus became the only goal that counted as following the science, and the costs any honest policy has to weigh were ruled out of scope.

The economists already had this fight

Development economists have been arguing about this for 20 years. The “credibility revolution” taught social science to distrust sloppy causal claims and to prize identification, which was a genuine advance. But the economist argues that some of its champions then performed a strange trick: having demanded the most stringent possible evidence for a narrow estimate inside a paper, they would turn around and accept sweeping, system-level claims built on those estimates “with complete and total gullibility.” He calls it the credulity revolution. An experiment can be internally airtight and still tell you almost nothing about whether a program will work at a national scale in another country.

The complaint comes from the center of the field. Angus Deaton and Nancy Cartwright, hardly enemies of quantification, argue that randomized trials earn their keep only “as part of a cumulative program” alongside theory and mechanism. Demanding external validity from a single trial “expects too much of an RCT while undervaluing its contribution.”

Pritchett offers a blunter smell test for his own field of economic growth: if rich countries do not have more of some fashionable factor than poor ones, we should be suspicious of claims that it explains development. No country has ever run a randomized trial on its way to wealth. Poland did not climb out of communism and into prosperity because of a well-run cash-transfer experiment; it did so through messy, large-scale changes in markets, institutions, and politics that no trial could have tested in advance. The method should fit the question. When it does not, more rigor on the wrong question is just a costlier way to be confidently beside the point.

When the trial earns its keep

Don’t get me wrong—I like my RCTs and have carried out some myself. Randomized studies are essential when intuition gets ahead of knowledge. The best case for a trial is the mirror image of the parachute: sometimes the answer is not obvious at all, the intuition almost everyone shares turns out to be wrong, and the only way to find out is to run the experiment and pay its cost. When you cannot see the effect immediately just by looking, you absolutely have to measure it carefully and systematically.

Cash transfers in rich countries can be an interesting example here. In poor countries, the evidence that handing people money improves their lives is about as strong as social science gets, with randomized studies finding large gains in income, assets, food security, and even lower infant mortality. When economists first proposed simply handing poor people cash, the worry was that they would fritter it away on alcohol and other temptations, which is why aid so often came as food or as cash with strings attached. The trials found that worry was largely unfounded: across dozens of studies, cash did not raise spending on alcohol or tobacco and often reduced it, because people in poverty turn out to be good judges of what they need.

Now, since it’s getting increasingly obvious to more folks that cash transfers clearly work, you might assume the same logic carries over to a rich country: if you give a struggling American a few hundred dollars a month, their life gets measurably better. At least it did seem obvious to the researchers who believed in it. But then they ran the trials.

As details, a run of careful American studies, including an OpenResearch experiment that gave people $1,000 a month for three years, found no sustained improvement in health, employment, stress, or children’s outcomes. Piper, who went in expecting cash to help more, called the evidence “shocking.” The developing-country findings were sound, but they did not transfer cleanly to this context, which is exactly that external validity problem Angus Deaton was so worried about. Here the trial earned every dollar it cost, because the stakes were high, the intuition was strong, but it was ultimately wrong. The problem, of course, is telling a parachute from a medical trial before you decide whether the study is worth running.

Even then, the right tool is not always an experiment, and sometimes it is not available at all. Some of the most important things we know rest on theory and modeling more than on any single trial. No one has run, or could run, a randomized study on whether a country should open itself to free trade. The case here rests on the theory of comparative advantage, worked out two centuries ago and refined by mountains of non-experimental evidence since. No one randomized nations into democracy and dictatorship to learn which produces better lives. We reason about questions like these from theory and values, and we lean on thought experiments and formal models, the kind of computational and thought simulations social scientists build precisely because some questions can never be put to an experimental test.

When everyone is wrong at once

I’m not at all an expert on the pandemic. But it is a very important case that still shapes public trust for the worse. We saw the uncertainty in the evidence already. But the deeper failure was the refusal to admit it. Officials at every level kept repeating that “we know what works against Covid-19,” as Macedo and Lee document, “even as it became increasingly evident that policymakers were improvising and did not, in fact, know with certainty what worked.” Certainty was claimed where it did not exist. Science can tell you what a policy is likely to do. But it cannot tell you what you ought to value, and pretending otherwise spends the credibility you need the next time.

The critics often made the mirror-image mistake, treating the absence of a clean trial as proof that a measure was worthless, which is the flossing error scaled up to a national emergency. Macedo and Lee fault both reflexes at once: it was wrong, they write, “to mock and censor mask skeptics, as it was to insist with certainty that masks do not work.” And underneath the shouting sat the trade-off nobody wanted to name. The studies on masking measured “only one side of the equation,” they point out, and said “nothing about the costs of masking to children’s learning, communication, socialization, and psychological well-being,” while school closures “most hurt poor kids.” What both errors share is a refusal to say the honest thing: the evidence was incomplete, and the choices imposed heavy costs on people who never got a vote.

The “follow the science” failure took a disagreement about values, about how much weight to put on the schooling of the young against the safety of the old, on liberty against caution, and dressed it up as a technical dispute the data had already settled. Immigration runs on the same move constantly. A great deal of what looks like an argument about evidence is actually an argument about what a government owes its own citizens against what it owes foreigners. No experiment can tell you how much to weigh a citizen’s wages against a stranger’s safety, or whether a more diverse society is better than a more cohesive one.

These are all questions of value, and if people simply do not want a certain outcome such as a more diverse society, no clean estimate that a policy would deliver it will change their minds. I have made this case at length: the most useful thing evidence can do here is discipline a values argument by telling us what a given policy will cost and produce. It cannot make the values argument disappear, and recasting a value conflict as a scientific one mostly just hides what people are fighting about.

makes a closely related point in his new essay on how political tribes construct rival realities. Political disagreement often runs through rival systems of interpretation that decide which facts matter, what counts as representative, and who belongs in the story as victim, villain, or hero. That is why appeals to “the evidence” so often disappoint. The fight is partly over facts, but also over the frame that tells people what the facts mean.

Demonstrable benefits

No randomized trial will also ever tell you everything you want to know about admitting a productive, skilled worker who pays taxes, fills a documented shortage, starts a company, or treats patients in a town that cannot recruit a doctor. The feedback here runs through a variety of institutions, and you cannot hold the rest of the world fixed while you turn the dial. Yet the basic direction is not at all mysterious. A policymaker weighing whether to route more eligible scientists through a specialized visa or to clear a needless skilled-worker backlog will never have clean experimental evidence that settles the whole question, and waiting for it just means letting a bad default persist.

This is why skilled immigration is so popular: its benefits are intuitive and visible without anyone reading an econometrics paper. About 80 percent of American voters support high-skilled immigration across party lines, and policymakers who wanted to act on that could streamline the uncapped O-1A visa for extraordinary-ability workers tomorrow, without a single new law. This is what I mean by demonstrably beneficial: a policy whose contribution to the country ordinary people can see in practical terms, explicitly and straightforwardly serving the national interest. The persuasive framing is already baked into the policy itself, so you do not need a campaign to explain it.

But setting immigration policy, of course, is not the same thing as choosing parachutes. “A skilled worker who pays taxes and fills a documented shortage is an asset to the country” is a claim about as close to observational saturation as social facts get. “This particular visa reform will produce that benefit at this particular magnitude” is a genuine empirical question that does need identification, and it is the kind of thing my colleagues spend careers trying to estimate. Treating the second claim as if it were as self-evident as the first is the same error the parachute authors mocked, just aimed in a friendlier direction. As I have argued before, “immigration” in the abstract has no clean effect waiting to be discovered; specific policies admitting specific people under specific rules do.

Not every visible benefit is as fraught as immigration. Since we were already talking about flossing, consider… the idea of a Japanese toilet. You only have to use a heated washlet once or twice to know it beats what most regular bathrooms offer, and nobody needs a randomized trial measuring cleanliness or satisfaction to settle the matter, as just argued. I also bought one recently after spending some time in Japan and have opinions.3 I am fairly confident that installing washlets at scale, in hotels and airports, would be a good policy to have for a lot of American institutions, and just as sure it does not need to be evidence-based in any serious sense. The holdup here is not a shortage of evidence. It’s our cultural habits, building codes, and the electrical wiring most American bathrooms were never built for.

Demonstrable harms

The same logic runs in reverse, and it produces what may be the clearest self-defeating immigration policy in the rich world: the ban on letting asylum seekers work while their claims are being processed. In many countries, that exclusion lasts half a year or more, and in practice it can stretch much longer.

Before considering any of the politics, let’s think about what this does to a person. A government accepts an asylum claim for consideration, pays to shelter the applicant, then forbids the one activity that would let them support themselves, build a record, and start to belong, which is work. Months of forced idleness drain savings, erode skills, and corrode the habits and confidence that make someone employable, and the damage outlasts the ban by years. You do not need a clever instrument, or a randomized trial, to see that this is a bad trade.

Then there is the fraught politics of the issue. The ban manufactures the exact image that anti-immigration politicians point to as proof the system is broken: able-bodied newcomers idle in taxpayer-funded hotels, or selling fruit at stoplights because the formal economy is closed to them. That visible dependence can sour voters on immigration more broadly, including the skilled pathways voters otherwise like.

Whenever I argue that immigration policy has to be intuitively beneficial to stay popular, someone asks what a non-beneficial pro-immigration policy would even look like. This is what it looks like, and no amount of aggregate evidence about the long-run fiscal contribution of refugees will make a voter unsee the visible disorder of the asylum crisis in the streets. The honest fix is a policy that stops generating the visible failure in the first place.

Against evidence theater

It helps to see why the maximalist version of evidence-based policy keeps disappointing its champions. Pushing every chip onto “the science,” and deferring to whatever the experts currently agree on, still does not deliver a policy, because the hardest questions are not the ones science was built to answer. Macedo and Lee make the case for Covid: scientists, they write, “with their narrow bases of expertise, should not make policy,” because the choices turn on values and trade-offs no study can weigh. The same gap shows up on climate, where the economist argues that the loud framings, “no big deal” and “existential threat,” are both probably wrong, and that responsible policy has to weigh trade-offs and judge which scenarios are plausible, something no appeal to the consensus can settle. And it shows up on immigration, where “the evidence” gets invoked by both sides to settle what is at bottom a fight over what we owe one another.

So what does this look like in practice, for someone who wants to take evidence seriously without hiding behind it? The following questions help:

• What kind of claim is actually on the table: an empirical one that data can settle or a normative one no regression will ever resolve?

• What kind of feedback does the policy produce, and would a clean study even be valid and available before the decision has to be made?

• And what does it cost to gather the evidence, what does it cost to wait, and who pays for the delay?

When the mechanism is clear and strong, the downside is reversible, and the people who would benefit cannot afford to wait, the honest move is to act now and keep studying as you go. And when the disagreement is at bottom about values, experts should say so, instead of laundering it through “the science.”

From evidence-based policy to honest democratic judgment

Every study has a price, and so does every delay. The writer recently named a failure mode he calls the tyranny of the edge case in piece, where any conceivable harm to anyone, however rare, becomes a reason for nobody to act, and the demand for one more study is among its favorite instruments. That demand almost never falls evenly: the bar rises for the reform someone dislikes and disappears for the status quo, which usually rests on no trial at all.

An isolated demand for rigor, aimed only at the conclusion you would rather not reach, is one of the most effective ways to block action while sounding admirably careful about the evidence. Running a trial means spending money, burning time, and sometimes withholding a promising policy from the people in the control group, while every month spent waiting is a month the current rule stays in force. You can usually weigh this informally: nobody needs a randomized trial to see that paying people to sit idle, as the asylum bans do, and then resenting them for it is a bad trade.

The discipline that keeps all of this honest is a willingness to say what would change your mind. If I claim the asylum work ban is obviously indefensible, I owe you the conditions under which I would drop the claim: solid evidence that it deters fraudulent applications at a scale worth years of lost earnings and delayed integration for people who would otherwise work. I have not seen that evidence, but I would genuinely look at it. Naming the finding that would move you is what separates a considered judgment from a convenient one, and it is a test the loudest demanders of “evidence-based policy” too rarely apply to themselves.

Don’t get me wrong—none of this is a license for vibes. The parachute study is funny because obviously nobody needs a randomized trial at altitude, but most of the choices that matter look less like a parachute and more like our immigration system: the evidence is partial and the feedback is slow or invisible. Evidence is indispensable for those choices. But it still cannot make the hard judgment for you, and a good judgment will not survive long in a democracy unless the policy behind it can show its value in the world ordinary citizens of any political persuasion actually live in.

Share

A year ago, I published a book arguing that democracies can make immigration popular by making it demonstrably beneficial. In the months since, the United States ran something close to the opposite experiment. Mass deportations and a “border” enforcement surge reached even in decidedly non-border cities like Charlotte, NC where I was living at the time. It became a year-long case study in what happens when a policy many voters backed in the abstract becomes, in practice, hard to see as serving their interests. And it came right on the heels of the Biden years, when an administration that wanted to be humane lost control of the border and paid for it politically. In barely two years, the country had produced two opposite failures and a fast lesson in how quickly immigration politics can swing.

But the test I was most worried about did not come from Washington. It came from Canada. The book leans on Canada as its clearest case of immigration done right, and I had said, more than once, that if Canadian support for immigration ever collapsed, it would invalidate part of my thesis. Over the past two years, it came closer to that than I thought possible. So I want to spend this publication anniversary explaining what happened and why the backlash against immigration in Canada still fits the argument I made in In Our Interest, and recalibrate a bit. In short, Canada’s “backlash” has been much more about speed, housing, and temporary flows than about crime or ethnic identity, while the selective core of the system remained uncontested.

Subscribe now

Is there a popular backlash against immigration in Canada?

Of course, I never claimed in the book that Canadian opinion could never turn. But a swing this fast, in the country I held up as the model, cuts against the grain of my argument, and I owe readers an honest account before I move to the defense.

Drawing on Environics survey data going back to the 1970s, my book showed that the share of Canadians who think immigration is too high had fallen for decades, from around 70 percent in the early 1990s to under 30 percent by 2020. All consistent with the idea that Canada achieved a high-immigration system that most people actually like by running it as a selective, orderly program whose benefits were visible to ordinary people. This is what I meant by demonstrably beneficial: a policy whose payoff you can grasp without a degree in economics, not a claim that all immigration is automatically good. My last data point sat near the bottom of that long decline. It was, in hindsight, almost the worst possible place to stop drawing the line.

Because here is the picture now. Agreement that immigration is too high jumped 17 points in 2023, then another 14 points in 2024, reaching 58 percent, the first clear majority since the late 1990s and the steepest climb in the half-century where Environics has asked the question. The line I published as evidence that immigration can be made durable turned and shot upward almost the moment the book came out. I would be lying if I said that it did not sting.

After nearly every talk I give on the book, a hand goes up, and someone asks, “What is going on with Canada?” The vibes are bad, and they are bad in places that did not feel that way before. For instance, in October Alberta is set to hold a province-wide referendum carrying five explicit anti-immigration questions, a ballot move with no real precedent in modern Canadian politics.1 You could line the chart’s spike up with that and tell a clean story: Canada has finally joined the rest of the developed world in a right-wing immigration backlash, and the fuss I made about demonstrably beneficial policy turned out to be wishful thinking.

What actually did and did not happen

As I show in the book, this is usually what we see across the world: there are people who can be labelled “committed nativists” and “committed cosmopolitans,” each around 10 percent of the public, who rarely budge on immigration. What usually moves is the large middle. I call this majority of voters altruistic nationalists: they care about newcomers but weigh their own compatriots first, and they back immigration when they can see it working for the country.

Over the last few years, Canada’s middle majority have not seen immigration working as well as it used to. Canada scaled temporary residents and international students faster than housing and services could absorb them, the textbook condition under which the diffuse and often invisible benefits of immigration shrinks while its concentrated costs swell. Even so, a majority of Canadians still say immigration is good for the economy, and when Environics asks the people who want less why, they point overwhelmingly to housing and the cost of living, not to crime or culture.

The book actually flagged this risk in the Canada chapter, noting that the post-pandemic surge in international students, plus a housing shortage, had already begun to produce dissatisfaction. What I got wrong was the speed of the public confidence decline: I assumed a system this good would keep absorbing shocks without a serious political reaction.

But notice, too, what the most recent opposition actually targets. Pierre Poilievre and the Conservatives ran hard against temporary residents and students while leaving the skilled, points-based core of the system alone; even the country’s most restrictionist federal party, the People’s Party, has proposed capping permanent admissions at 100,000 to 150,000 a year, a sharp cut from today’s pace but a level that would still admit several times what most other rich democracies do. The argument in Canada has effectively been about how much and how fast, while the principle of selecting immigrants for national contribution goes essentially unchallenged.

Even at its angriest, the backlash never turned Canada into a restrictionist country. Gallup still ranks it the most migrant-accepting society in the world, and even after the recent cuts, Canada is on track for roughly 395,000 permanent admissions in 2025, near one percent of the population a year, more than double the United States’ rate per capita and above the OECD average. Even if every one of Alberta’s referendum questions passed, they could not override the federal totals. A Canadian who complains of “too many immigrants” is still open to an extraordinary number of immigrants. A Canadian who says there is too much immigration and, let’s say, a Japanese voter who says the same thing are not asking for remotely the same world.

The dial has also already started to move back. Faced with the backlash, Ottawa did the both responsive and responsible thing: it cut permanent-resident targets from 500,000 to 365,000 by 2027, imposed the first-ever caps on temporary residents and international students, and set out to shrink the temporary population to about 5 percent of the country’s population.

Opinion is already tracking the policy. Environics’ 2025 reading shows the “too many” number leveling off at 56 percent, the first year in three it stopped climbing, and the people who still hold that view increasingly describe it as a failure of government management, not as a problem with immigrants.

The refinements: thermostatic politics, credibility, and salience

The whole recent Canadian sequence is what thermostatic politics looks like: while deep attitudes barely move,2 the dial of “too much” or “too little” adjusts to whatever the government is actually doing in response to those attitudes. Overshoot what the public can absorb, and the dial swings; correct the policy, and the dial swings back. I gestured at this in the book. But I should have built a chapter on it, because it turns out to be the engine of the whole story.

The same mechanism just ran in the opposite direction south of the border. As enforcement escalated from the border into workplaces and neighborhoods, the share of Americans who told Gallup they wanted less immigration fell from 55 percent in 2024 to 30 percent in 2025, one of the sharpest reversals on the question in decades. Same thermostat, opposite setting. As I argued in a guest piece for ’s Substack, competence builds trust, and chaos invites dissatisfaction, whether the chaos is restrictionist or expansionist.

The other refinement is about the importance of credibility. Demonstrable benefit only holds support when voters find it credible, and credibility, as Canadian housing politics has shown, is much easier to lose than to build. Once a community sees an acute housing shortage that can plausibly be blamed on a surge in arrivals, even a system that had been working stops being seen as beneficial. The visible benefit and the credible attribution have to travel together.

The same point came through more sharply from the American side. Consider what happened to the Harris campaign’s late attempt to look tough on the border. In her September 2024 Douglas, Arizona speech, she promised to “bring back the border security bill that Donald Trump tanked” and to “do more to secure our border to reduce illegal border crossings,” language that would have read as a Republican script a few cycles earlier. It did not move the needle. Voters continued to trust Trump on the border by wide margins. No matter what Harris said by then, voters had little reason to treat the pivot as credible after the previous Democratic administration had presided over visible border disorder. Their skepticism made total sense.

This all now points us back to the actual policy. Demonstrable benefit and credibility are the same project seen from two angles: a policy earns credibility by being implemented and producing benefits people can see, which is why you cannot shortcut it with a pivot. A promise of orderly, beneficial immigration is worth very little until voters have watched you keep it. And it is the credibility of mainstream parties and politicians, specifically, that matters most: the way trust in the center-left or center-right to handle immigration well is built up over multiple terms and then squandered in one episode of total dysfunction.

The third refinement is issue salience. For a long time, Canada came unusually close to what I still think is the best political outcome for immigration: a mostly boring, technical policy area nobody cares about, closer to something like “international measurement standards” than to a culture-war battlefield. Voters cared about whether the system worked, but they did not organize their politics around it. That is much harder to sustain once immigration becomes one of the main things people blame for housing shortages, crowded services, or a general loss of control.

That is why Canada now worries me more than it did when I finished the book. Once a quiet technical issue becomes a top-line political issue, the margin for error shrinks. Every backlog, administrative failure, or local pressure point starts to stand in for the system as a whole. And because immigration opponents tend to care more intensely about reducing immigration than supporters care about expanding it, the politics can move quickly once the issue rises on the public agenda.

​​Where the project goes next

So where does my book and research go from here? Three directions, concretely.

First, the empirics of demonstrable benefit, and the credibility that follows from it. I want to know what voters actually have in mind when they say a policy is nationally beneficial: which features they register, which they ignore, and how long a government has to deliver before its promises start to seem credible. The book basically asserted the construct, and the next step is testing it rigorously, which is among the first questions I want to take up through the immigration lab I am starting at Notre Dame.

Second, issue salience and ownership. I want to understand when immigration becomes one of the issues voters use to judge whether a government is competent, and who they then trust to handle it. In work I am now doing with James Dennison, we’re trying to figure out how that dynamic travels beyond immigration as well: when an issue becomes politically available, who gets to own it, and how mainstream parties can keep hard policy problems from becoming permanent trust problems.

Third, humanitarian design, which is the hardest case for the argument and the part of the book I most want to extend. The way to protect humanitarian immigration is to build public trust first, through visible benefits and shared responsibility. That is the case I made in a recent policy brief, and the next step is testing which combinations of sponsorship, faster decisions, and credible limits actually hold up politically.

In Our Interest was an academic book by design (I had to finish my dissertation and get an academic job, folks!), and I am proud of the research behind it. But its argument deserves a more direct public life too, which is part of what this newsletter is for: the same case, in a register that meets readers where they actually are.

A year ago, the question under every other question, in rooms from an Oxford lecture hall to a retiree center in Charlotte, was whether any of this is realistic, or whether making immigration popular is a fantasy that collapses the moment real politics arrives. After the year we just had, the year that bent my favorite chart the wrong way and then began to bend it back, my answer is steadier than when I started. Immigration support becomes durable when people can see it is working, and the work of showing them is never finished. That is still, for everyone involved, in our interest.

Share

Many thanks to more than a thousand3 of great folks who read, reviewed, argued with, or hosted the book over the past year. If you haven’t had a chance to check it out yet, for the full argument and the evidence behind it, the book is still the place to go. For the next stage of the project, this newsletter is where I am taking it. And if you want a say what should come next, tell me in the comments. You can find the rest of the year’s talks, reviews, and podcast conversations on my media page.

The latest AI-writing scandal arrived, fittingly, with the Pope. On May 25, 2026, the Vatican published Pope Leo XIV’s first encyclical, Magnifica Humanitas, dated May 15, on safeguarding the human person in the age of artificial intelligence. Within days, people were feeding it to Pangram, the detector that has suddenly become the respectable instrument of literary and academic suspicion. Some portions were apparently flagged as AI-assisted, and the accusation was easy to understand: the Pope, or at least the Vatican, had used AI to write about AI.

My reaction was basically: OK, and then what? Encyclicals already pass through institutional drafting, staff work, consultation, revision, translation, and committee-like smoothing. If some Vatican official used Claude to turn Pope Leo’s ideas into prose, the relevant questions would still be whether the document is accurate, thoughtful, and worth reading. But I should admit I was wrong.

Subscribe now

In Part II of my AI series, I wrote that AI writing detectors were bad and would probably remain bad. Pangram changed my mind: it has third-party evidence behind it, claims a very low false-positive rate, and has become the detector people reach for when they suspect undisclosed AI writing. Kelsey Piper recently wrote about Pangram Labs’ claims that several prize-winning short stories were AI-generated or substantially AI-assisted, and The Atlantic’s Matteo Wong has now written about Pangram’s growing power in schools, publishing, journalism, and the AI-writing accusation economy.

A working detector makes the ethics more urgent because it tempts people to treat provenance as a verdict. That is why the joke about using Pangram to filter out human-written content is sharper than it first sounds: in plenty of settings, AI-assisted writing may be more readable and more useful than unaided human prose. If the data underneath are sound, the human sentence-level struggle contributes little.

I should also be upfront that I am an interested party here. I am that infamous AI professor who proudly writes with the help of AI. An essay arguing that AI detection can become status policing is conveniently a defense of my own practice, so readers should ask whether I am drawing the line in a way that flatters me.

The provenance spectrum

AI-writing ethics starts with the promise the writer made. A writer can make a promise to a teacher, editor, reader, institution, or recipient. The ethical question depends on that promise before it depends on a detector score.

Let’s start where the skeptics are right. A student assignment that explicitly bans AI is the obvious case. A creative writing contest that promises to recognize new human writers is another. A condolence note belongs in a different category from an exam, but if someone grieving expects words from you, outsourcing the emotional act to a machine feels like a betrayal.

Some decisions also require accountable human judgment. If I am deciding whether someone gets a scholarship or grant, provenance matters because the applicant is owed my judgment. AI can help organize evidence or check consistency, but the evaluative act has to remain mine.

Consequences matter too. The more influential the decision, and the more it depends on personal judgment, the stronger the case for knowing who or what made the call. Human discretion can also be worse than AI discretion: a committee can be biased or arbitrary, and a well-designed AI system may eventually make some decisions more consistent.

There is also a special rule for first-person claims. When I write “I think” or “I feel,” that conviction should actually be mine. AI can help me phrase it, pressure-test it, or make it less awkward. It cannot supply the conviction itself.

Research and journalism sit closer to the middle. A byline is a promise that the author stands behind the claims, evidence, and judgment calls. It has never meant that the author personally typed every sentence without help from search engines, copy editors, co-authors, translators, or now LLMs. If my name is on an argument, the argument has to be mine; the prose can be assisted.

Much technical writing belongs closer to the content-matters side. If I ask AI to describe a chart, write a methods paragraph, or translate a regression result into normal English, the important question is whether the output is correct. I still have to verify the numbers and own the final text. Accuracy and accountability carry the moral weight.

At the far end are administrative messages where almost nobody cares about the human act of writing. If a department asks you to send a polite note confirming a committee meeting, use AI freely. The relevant standard is whether the note is true and clear.

A single AI score cannot answer the ethical question. The same level of AI assistance can be harmless in an administrative email, useful in a technical report, questionable in a personal essay, and disqualifying in a no-AI classroom assignment. Context is the whole point, even when the underlying detection is accurate.

Detection has a spectrum too

The ethics of detection should follow the ethics of use. If a teacher has told students to write without AI on a particular assignment, a detector can be part of an academic-integrity process. A Pangram score should never be the only evidence, especially given the stakes for students.

Creative contests face a similar problem. Piper’s argument about the Commonwealth Short Story Prize should be taken seriously because fiction prizes are partly about human craft. If a prize is rewarding a human writer’s voice, a fully AI-generated submission violates the premise. The organizer can allow AI, ban AI, or create a separate category. Trust alone will not settle the problem.

Peer review is harder. Seth Lazar gave the strongest version of the pro-detection case in response to my earlier Pangram post: AI-generated research output can become a denial-of-service attack on peer review. The cost of producing plausible-looking papers collapses, while the obligation to read them remains expensive. In that context, a detector may help preserve scarce review capacity.

The peer-review case still depends on the goal. If the goal is to catch students violating an explicit rule, provenance is the target. If the goal is to protect reviewers from worthless submissions, provenance is only a proxy: a detector estimates the probability that a text is AI, never the probability that it is bad. The real target is bad work: hallucinated data, fake citations, nonexistent methods, and papers with no question worth answering. A detector might help triage some of that, but someone still has to check the actual claims.

My worry is that we will police the em dashes while ignoring the hallucinated data underneath them. That would be a very academic way to lose the plot: exquisite attention to the surface marker, little attention to whether the thing says anything true.

Why disclosure mostly fails

The obvious compromise is disclosure. Let people use AI, require them to say so, and let readers decide how much it matters. That sounds attractive because it treats AI assistance as information and lowers the moral temperature.

I argued in Part II of the AI series that disclosure norms collapse under the incentives they create. The detector half of that argument now needs revision because Pangram appears to work much better than I expected. The disclosure half still looks right to me.

The more ethically questionable the AI use is, the stronger the incentive to hide it. A student who used AI after promising to write unaided, a contest entrant who submitted machine-written fiction to a human-writing prize, or a researcher who used AI to paper over fake citations has every reason to stay silent.

The people most likely to disclose are the ones using AI in low-stakes ways: cleaning up a paragraph, translating a chart, or turning rough notes into readable prose they still own. Those are also the cases where disclosure matters least. The likely equilibrium is a world full of ritual acknowledgments about harmless AI assistance, while the genuinely deceptive cases remain hidden until someone investigates them.

Disclosure can still help when the disclosure itself explains the work, as it does here. Editors, teachers, prize committees, and people exercising institutional authority should also be clear about the rules they enforce. But if the whole system depends on honest confession, it will punish the conscientious and leave the strategic users alone.

The new grammar policing

I know the phrase “grammar Nazism” is abrasive, and I mean something specific by it. I was born in the Soviet Union, and Russian elite culture can be intensely sensitive to grammar, pronunciation, stress patterns, and the small status markers embedded in speech. In practice, grammar correction often doubled as social sorting: the wrong school, the wrong region, the wrong family background, or the wrong kind of education could leak through the way you spoke.

America has its own version of this. Academic English is full of status signals masquerading as standards. The right kind of fluency makes you sound smart before anyone checks whether you are right, and the wrong accent or idiom can mark you as unserious before your argument gets a hearing.

AI detection is turning this old habit into a new technical ritual. The same people who once policed grammar now police “AI tells”: em dashes, smooth transitions, generic metaphors, oddly balanced paragraphs, prose that seems a little too clean. Sometimes they are right. AI writing does have recognizable patterns, which is why I have a style guide full of them.

If someone reads a piece, learns something new, and then makes the conversation about one suspicious phrase, the Pangrammar instinct has wasted everyone’s time. A reader’s attention should go first to the claim, the evidence, and the payoff, with style policing saved for cases where the prose actually blocks understanding or signals deception.

This status dynamic is very familiar. The detector score gives a scientific-looking license to dismiss work without reading it carefully. The people who benefit are usually the incumbent writers and credentialed gatekeepers who can turn a judgment call into a score. The accusation becomes especially convenient against lower-status writers and people who do not write English well but can now use AI to translate, draft, and reach an English-language audience. Too polished looks fake. Too awkward looks low quality. Either way, the gatekeeper wins.

The moral contamination logic makes the problem worse. Once AI involvement is treated like impurity, any trace of assistance becomes enough to condemn the whole work. That is a strange standard for a world where human writing has always been socially produced by editors, reviewers, co-authors, translators, and the sentence you read yesterday.

The funniest possible equilibrium is already here. AI tools write prose that is too clear, detectors punish the clarity, and then new “humanizer” tools rewrite the prose to look more awkward. TIME recently described people inserting mistakes and oddities to avoid sounding AI-generated. This is Grammarly in reverse: make the writing worse so it looks more authentic.

What to do instead

I am arguing for detector modesty. Pangram should only be used where provenance is part of the agreement: exams with explicit no-AI rules, contests that promise human craft, or institutional settings where the source of the text is part of the job. The institutional rule should be written before the score is consulted: define what AI use would violate the promise and what appeal process follows a high score.

In many domains, the standard should be quite simple: if you put your name on the work, you own it. You own the facts, the claims, the errors, the taste, the structure, and the judgment. If AI helped you produce an accurate technical summary, good. If AI helped you produce nonsense faster, that is on you.

Because attention is scarce, people will still rely on shortcuts. They will trust names they know, journals they respect, editors with a track record, friends who have read the work, and institutions that have something to lose if they publish junk. That is imperfect and often unfair. Outsiders and newcomers pay a price when reputation becomes the filter. But at least reputation is accountable over time. If a journal, prize, professor, or writer keeps endorsing bad work, people can notice.

A Pangram score is different. It gives a quick guess about textual provenance and invites us to stop reading before we have asked what the text is doing. Pangram seems to work, so the question is no longer whether we can detect AI. The question is what we should do with that information. Use it when provenance is part of the bargain and the stakes justify an investigation. Treat it as a prompt for judgment, never as a substitute for judgment.

If the work is fake, wrong, plagiarized, emotionally fraudulent, or a violation of a clear rule, say that and act accordingly. If the work is accurate, useful, and owned by the person whose name is on it, the fact that Codex, Claude, or ChatGPT helped assemble the sentences is a weak basis for scandal. The scandal would be building a culture where everyone learns to make writing worse so it can pass as human.

One last disclosure, since the whole piece is about this question: the essay above was written entirely in Codex from several hours of my dictated thoughts, earlier posts, saved style instructions, and recent social media exchanges. The cover image and spectrum chart were also produced by Codex. This was not a one-shot prompt. We went through more than a dozen iterations, mostly refining the argument and the chart. Yes, I am using Codex more than Claude Code now. Yes, I read the draft before publishing but did not line-edit the prose at all. By my own chart, this essay sits on the content-matters side of the spectrum, and I stand behind it.

Share

Some good news first. I am opening Popular by Design to occasional guest contributors, with formal guidelines coming soon. The first two posts were a perfect reminder of why this is worth doing: asked whether intervention would cause a migration crisis in Cuba, and explained why mainstream parties seem to lose either way when they accommodate the far right. Gil’s piece asks what a Cuba intervention debate misses when it treats migration as an afterthought. Hanno’s piece asks what mainstream parties actually gain or lose when they borrow far-right immigration positions. Both pieces say something new and practical about the problems facing our politics.

Subscribe now

May was a lighter month for my own writing after the April sprint, which was probably healthy. I still managed to publish my Washington Post op-ed on why high-skilled immigration is a winning policy that needs a champion, plus a Popular by Design essay on why LinkedIn is doing what Bluesky was supposed to do. I also recently spent two days in DC at the Teaching Progress workshop, hosted by at the JHU Bloomberg Center, where faculty reworked syllabi around the question of how to teach about progress, especially now in the era of AI. I went in with my migration and conflict courses and came out with a clearer understanding of what I need to change. After a stretch of academic conferences that felt more draining than useful, this one felt like a breath of fresh air. I actually learned something new and met interesting people in a cool setting.

Here are the May links (linking does not imply endorsement):

• argues that Trump closed the representation gap on immigration. I agree with the broad thrust that Trump’s immigration politics are more popular than many smart people want to admit. My main criticism is that the usual Gallup increase/decrease item is a shaky way to diagnose representation: it mixes what people want from immigration, what politicians in power and opposition are doing, and thermostatic reactions to current events. I would also separate Trump’s popular moves on asylum and border security from his much less popular moves against legal immigration, green cards, travel, and high-skilled migration. Bundling all of that as “Trump’s immigration policy” hides too much of the real politics.

• at Cato reports on the DHS quitting granting green cards almost entirely. USCIS now says adjustment of status should be granted only in “extraordinary circumstances” and that most applicants should use consular processing abroad. It is still unclear if the memo will be implemented as written, including forcing many people with pending green-card applications to leave the United States and apply abroad. But crazy stuff nonetheless.

• Tara Watson, Matthew Wich, and Johnny Willing at Brookings give the broader talent-pipeline version of the same concern in a long report on how the Trump administration is eroding high-skilled immigration. They project a 29% decline in new F-1 student visa issuances in 2025 and estimate a green-card backlog of roughly 1.2 million. Their useful point is that the pipeline is sequential: international students become OPT workers, some become H-1B workers, and some eventually become permanent residents. Breaking one step does not stay confined to that step.

• The American Immigration Council released “Restoring Credibility and Humanity”, an enforcement framework built around compliance, safety, proportionality, and accountability. As I have been saying before, there is very little programmatic work on enforcement from the left, so this is exactly the kind of thing the pro-immigration side needs more of: a concrete alternative to the mass deportation agenda that doesn’t just brush off prevalent people’s concern about legality and order.

• uses recent CBO projections to show that aging is in the future for the United States too. In his zero-net-migration scenario, the 20-64 population falls by 20.5 million by 2056 while the 65-plus population rises by 21.9 million, and the ratio of working-age adults to people over 65 falls from 3.1 to 2.1. This is the immigration-demography link in its least sentimental form: even countries that avoid near-term depopulation still age, and aging changes what labor, care, and growth require.

• pushes back on a too-simple fertility narrative in “The (Fake) Long Decline of Fertility”. His target is the viral story that fertility has just been sliding down the same long trend line for centuries. Instead, he separates long-run forces like child mortality, norms, and the cost of children from the more recent post-2007 decline, where he emphasizes partnership, sex, contraception, and digital technology.

As before, if you want me to write more about one of these or other related topics (or now if you want to write something yourself!), let me know. And if you came across something I missed this month, send it my way.

Share