The UAW’s Quiet Transformation

On August 18, 2026, a coalition filed suit in the U.S. District Court for the District of Massachusetts to block a DHS rule ending “duration of status” for F-1 students and J-1 exchange visitors. The plaintiffs are the National Association of Foreign Student Advisers (NAFSA), the Presidents’ Alliance on Higher Education and Immigration, the Association of Independent Colleges and Universities in Massachusetts, the American Federation of Teachers (AFT), a Brown University grad-worker local (GLO, an AFT affiliate), the UAW and its Massachusetts Local 2322, and The NewsGuild-CWA.

According to NAFSA’s own press release, the UAW now represents “more than 120,000 higher education workers at over 50 public and private institutions.” That‘s approaching a third of the UAW’s total membership of 383,000. That’s a union whose name still conjures images of assembly lines in Flint and Detroit that is now deriving a meaningful share of its institutional weight as well as dues base from academia. Its financial interest is increasingly tied not to the wages of American factory workers, but to the continued, uninterrupted flow of international students, postdocs, and researchers into U.S. universities. When an organization’s membership rolls depend on a policy outcome, it stops being a neutral advocate for “workers” in the abstract and starts being a stakeholder in a specific immigration pipeline.

Organized labor increasingly looks less like a movement organized around the interests of American workers and more like toadies for globalists who are not quite done gutting the American middle class. NAFSA’s presence in the complaint can be written off to protecting their own special interests. But the UAW’s presence on this complaint is mind-boggling, as they are fighting a rule that has nothing to do with auto plants, wages, or offshoring in manufacturing.

What the Rule Actually Does

The rhetoric surrounding this rule is misleading. DHS is not ordering legitimate international students to leave after four years. The rule replaces indefinite “duration of status” with a fixed admission period of up to four years, and a student whose legitimate program requires more time, a PhD candidate, for instance, can apply for multiple extensions.

The rule exists because of a documented pattern of abuse: DHS cites visa mills, pay-to-stay schemes, schools requiring minimal in-person attendance, and concerns tied to Day-1 CPT. For instance, the agency found nearly 77,000 F-1 students who had spent more than ten years in student status, some repeatedly re-enrolling at the same academic level. This isn’t a crackdown on scholarship. It’s a response to a cottage industry that has turned our colleges into de facto immigration centers. The coalition’s claim that the rule threatens the flow of “global talent” is a convenient way to defend a grift these institutions have a financial interest in preserving.

The NewsGuild-CWA’s Hypocrisy

Here’s what should really raise eyebrows: The Communications Workers of America has a documented history of fighting the exact kind of visa-enabled labor substitution this rule is designed to curb when it happened to American IT workers.

In 2016, the University of California system moved to lay off 89 IT employees and replace them with workers brought in on H-1B visas through the India-based outsourcing firm HCL. Members of University Professional & Technical Employees, CWA Local 9119, mobilized against it for months, building public pressure and drawing bipartisan congressional criticism that spurred Senator Chuck Grassley (R-IA) as well as several California Democrats into action. UC employees were being asked to train their own H-1B replacements before losing their jobs. CWA did the right thing defending the good-paying, taxpayer-funded jobs of American workers being handed to a foreign visa pipeline at the direct expense of American technology workers.

This is the same union, the CWA union family, now suing to preserve looser oversight of a visa category DHS says is being used as a backdoor path to indefinite U.S. residency and work authorization. CWA fails to see the possibility that F-1 abuse and H-1B abuse are two sides of the same coin, as employers and institutions are treating “temporary” visa categories, like the F-1 for foreign students, as permanent labor supply channels.

The NewsGuild’s stated rationale here is about press freedom, as the rule also shortens I-visa terms for foreign journalists, and NewsGuild president Jon Schless framed this as a First Amendment concern. Well, knock me over with a feather. . . .

That may be a defensible position to stake out on its own. But signing onto a broader complaint that also defends the F-1 status quo is, in my opinion, difficult to square. It suggests that “protecting workers” is a principle CWA applies selectively. Sure, they will apply it rigorously when the visa category displaces a union local’s own dues-paying members but much more loosely when the visa category benefits allied institutions that are part of the union’s broader political coalition. It’s a “you scratch my back, I’ll scratch your back” play.

The Pattern

None of this is really about the specifics of F-1 visa policy, let alone how it is being abused. It’s about what happens to a labor movement when institutional featherbedding, membership rolls, coalition politics, and dues income from academic locals start to matter more than the interests of the American workers these unions were founded to represent. The UAW fighting for looser visa oversight because a third of its membership now sits in academia and the CWA opposing H-1B-enabled offshoring in one breath and defending F-1 abuse in the next are examples of the rot that is afflicting many of the nation’s institutions.

Ten years on, I invited Dmitry back onto the IfSPP podcast to “reminisce” about his time writing the book and how he thinks the world performed in the ensuing years. Spoiler alert, it’s looking like the “technosphere,” his term for the self-interested, self-propagating network of technology that increasingly sets its own agenda is now bigger, faster, and more fractured than ever.

It was a far reaching conversation and we covered a lot of areas to include:

• Defining the technosphere vs. the biosphere

• AI: useful tool or dangerous crutch?

• Autonomous weapons and the loss of conscience

• The offshoring parallel: America’s disposable-engineer culture

• DeepSeek and the case for investing in domestic talent

• Vocational education, mentorship, and the “just be a welder” contradiction

• The Amish/Hutterite model of controlled technology

• Orlov’s own de-teching: Linux, no Microsoft, a Russian Army van

• A world splitting into separate technospheres

Dmitry now publishes weekly on Boosty and he can be found at: https://boosty.to/cluborlov

Earlier this month we published a report on the impact of STEM OPT workers on the Dallas-Fort Worth metro area. This week we published another short, easy-to-read report on STEM OPT workers, only this time for the San Francisco Metro Area.

And what better place to take a look at than the Bay Area/Silicon Valley, the great 20th-century geographic incubator that propelled the United States to dominance in the information age. For instance, in 1939, fresh out of Stanford, Bill Hewlett and Dave Packard started HP with $538 of seed capital. In 1957, eight engineers (aka, the Traitorous Eight) that included Robert Noyce and Gordon Moore (Moore’s Law) walked out of William Shockley’s semiconductor lab and founded Fairchild Semiconductor in Mountain View. In 1976, Steve Jobs and Steve Wozniak, working out of the Jobs family garage in Los Altos, built the Apple I computer. And in 1998, Larry Page and Sergey Brin, while Ph.D. students at Stanford, founded Google.

Given its history, one would think that top American talent would find a veritable Garden of Eden in the Bay Area. Sadly that’s not the case. The entrepreneurial, innovative, and inventive dynamism that was the heart and soul of Silicon Valley has been ripped out and substituted with corporate efficiency and a conveyor belt of cheaper, more compliant foreign workers. These days, the garage door isn’t open to the next Wozniak or Brin; it’s open to whichever visa or work authorization is cheapest to sponsor.

Our deep dive into OPT data in the San Francisco metro area between 2012 and 2022 tells a story that should trouble anyone who still believes programs like STEM OPT exist to fill unfillable gaps.

Jobs Grew Modestly. Guest Workers Exploded.

Between 2012 and 2022, engineering jobs in the San Francisco metro grew by roughly 25%. Computer and math jobs grew by about 60%, which is genuinely strong growth that reflects a decade of Bay Area tech expansion.

But look at what happened to the guest worker population riding alongside that growth. OPT engineers grew 300% over the same period, and OPT computer and math workers grew 570%.

Whatever is driving the number of foreign workers, it isn’t simply that the industry is booming. Let’s say a sector grows by 60% in a given period of time. You might expect the workers channelled into that sector to grow by 60% and not 1,000. That is not proportional growth, not by any stretch.

By 2022, OPT engineers made up 17.7% of the entire San Francisco metro engineering workforce, or nearly one in five engineers in the region. By 2022, OPT computer workers made up 5.5% of the entire San Francisco metro computer workforce.

Not PhDs. Not Scarce Specialists.

The usual defense of these numbers is that they represent talent that simply can’t be found here in the United States. As a result, employers are forced to look abroad for workers with rare talent and credentials. San Francisco’s data doesn’t support that defense.

Only 23% of OPT engineers in the metro pursued a U.S. PhD. That means 5,115 of the 6,619 OPT engineering jobs in 2022, or 77% of them, went to bachelor’s and master’s degree holders. Among computer and math workers, the picture is even more lopsided: just 7% held or were pursuing a PhD, meaning 8,407 OPT computer jobs, or 93% of the total, went to workers holding the exact same undergraduate and master’s credentials that tens of thousands of American computer science graduates hold every single year.

These aren’t hard-to-fill research positions. They’re the same entry-level and early-career roles that American new graduates are trying to break into. This is the first rung most people need to start up the career ladder. Moreover, over 90% of the OPT computer and math jobs identified in this report are entry or early-mid-level positions, meaning the real damage to American graduates in the entry-level job market is even larger than the raw percentages capture.

Why Employers Prefer the Guest Worker

This isn’t a mystery once you follow the incentives.

STEM OPT workers are exempt from Social Security and Medicare payroll taxes, and that is a 15% government-subsidized discount that employers get for hiring a foreign graduate instead of an American for the identical job. There is no wage floor tied to local market rates, no requirement that a company first attempt to recruit an American, and minimal oversight of whether the “training” component of these jobs is real or a formality on paper. In addition, there is no numerical cap for OPT, and universities devote as much time to recruiting foreign students as they do to matching them with an employer after they graduate. An administrator at the University of Maryland stated that the school has letters of agreement with over 200 companies.

Put a recent American graduate and an OPT-eligible international graduate side by side; with identical resumes, identical degrees, and the employer’s own payroll math, they can tell which one will be the most efficient hire. That’s not a talent decision. That’s a subsidy decision, and taxpayers are the ones footing it.

The National Trend Behind the Local Numbers

The Bay Area is at the epicenter of the American STEM joblessness crisis. It’s indicative of where the whole country is headed. A new Peterson Institute for International Economics analysis of government data puts the national scale of this shift in stark terms: the number of foreign STEM workers brought in and trained and educated at U.S. colleges and universities nearly tripled in twenty years, from 532,311 in 2003 to 1.464 million in 2023.

And here too, the PhD defense collapses under its own numbers. Of that 1.464 million, only 217,226, or about 15%, came up through a U.S. STEM PhD program. The other 85% hold bachelor’s or master’s credentials, the same tier of degree held by the American graduates competing for the same jobs.

The scale of the shift shows up in the growth numbers too as well. That same recent report found that foreign-born workers accounted for 36% of all US STEM job growth between 2000 and 2023. Not 36% of new PhD hires. Thirty-six percent of the growth across the entire STEM workforce. This has led to widespread displacement of American workers up and down the career ladder.

That is the injury. The insult is these foreign workers are not more talented; in fact, there is a body of evidence to suggest they are less talented than their American counterparts. And that is the insult.

Put the local and national data side by side, and the pattern is unmistakable. No one is going to argue against protecting a handful of irreplaceable foreign PhD researchers. The issue with Optional Practical Training is it has fueled a two-decade, nationwide substitution of the American STEM graduate’s first job.

What This Means for San Francisco’s Own Graduates

One might think the Bay Area would be the place where an American who earned a computer science or engineering degree from a California university, where a Hewlett, a Wozniak, and a Brin once did, would be able to gain a toehold in the industry that degree was built for. Instead, American graduates either graduating from or willing to relocate to the Bay Area for that first job are competing against a guest worker channel that grew many times faster than the jobs themselves.

Charlie Munger had a line that explains what our research is showing us. He said, “Show me the incentive, and I will show you the outcome.” The incentive here isn’t scarcity; it’s a 15% tax discount. The shortage story doesn’t hold up in Dallas. It doesn’t hold up in San Francisco either. What is being held up is a government labor subsidy to businesses that promotes not hiring American white-collar professional workers. It is an ever-increasing pipeline of foreign labor whose only purpose is to goose the supply of labor that will result in a generation of American graduates standing outside the garage door, wondering why it’s foreigners and they are not getting hired.

I Will Be In The Bay Area in August

I will actually be in the Bay Area on August 11 to speak on how American college students are being negatively impacted by legal guest worker visa and work permit programs that Congress designed with American citizen interests not in mind. You can attend with the registration link here or click on the image below. It would be great to meet you!

Also, join us in signing and sharing our petition to end STEM OPT. Find your state, click on it, sign the petition, and share it with your fellow students, neighbors, colleagues, parents, and alumni. Later this year we will deliver these petitions to each state’s congressional delegation.

If you do not see an active petition for your state, reach out to me at klynn@instituteforsoundpublicpolicy.org and we’ll discuss how you can start one in your state.

I however have another take on the basket, and it is none too rosy. As Venus Uninterrupted wrote earlier this month, a “basket can hold a golden age, sure. It can also arrive empty.”

I know few if any who read my substacks have any experience with astrology and I’m not writing this to litigate whether the planets effect individuals, society, politics and economics. I’m writing it because when you strip the “basket” down to its bare mechanics which is a decades-long index built by a French mundane astrologer, correlated against a century of political and financial history and set it next to three other forecasting systems that have nothing to do with astrology and nothing to do with each other, they all land on the same several years and point to the same direction. That convergence is worth taking seriously regardless of what one might think about the zodiac. And it argues for something more sobering, and perhaps more useful in the long run than a golden age.

Four systems, no coordination

1. The Fourth Turning: Neil Howe and the late William Strauss studied repeating cycles they tied to a generational theory. Neil Howe’s The Fourth Turning Is Here dates the onset of America’s current Crisis era to the 2008 financial collapse, and projects its climax and resolution by the early 2030s one or possibly three likely stressors that include another severe financial crash, civil conflict, and/or a great-power war. This isn’t astrology. It’s a sociological model built on several generations of recorded Anglo-American history, and it converges on the same window from a completely different discipline.

2. Demographics as economic destiny: Another forecasting system is Harry Dent’s demographic economics. Dent’s thesis that dates to before 2008, held that the mass retirement of the Baby Boom generation would trigger a deflationary depression as the largest cohort in American history stopped spending and started drawing down on their financial resources individually and in aggregate. He repeatedly got the timing wrong being forced to move the goal posts across a decade of books. But the mechanism was sound, and the reason he was wrong on timing is itself instructive. I believe the unprecedented monetary and fiscal intervention at the time (TARP, QE(s), ZIRP, deficit spending at a scale with no peacetime precedent) suspended the reset he predicted rather than preventing it. It didn’t disappear. It came out the other side butinverted as inflation instead of deflation seen most visibly between 2020–2022.

3. Barbault’s Cyclic Index: André Barbault, a French astrologer who began his professional career in the years right after WWII and who became a founding member of France’s Centre International d’Astrologie in 1946 built a running index by summing the angular distances between every possible pair among the five outer planets, from Jupiter and Saturn out through Uranus, Neptune, and Pluto.

   1. With five planets, that works out to ten distinct pairs (Jupiter-Saturn, Jupiter-Uranus, Jupiter-Neptune, Jupiter-Pluto, Saturn-Uranus, and so on through all the combinations). For each pair, you measure the shorter arc between them around the 360-degree zodiac but never more than 180 degrees, since you'd just measure the other way around past that point. You then add all ten of those numbers together into a single figure.

   2. A simplified illustration: if all five planets happened to sit within a tight 20-degree band of the sky, most of those ten pair-distances would be small, and the total would be low. If instead they were spread roughly evenly around the full circle, most pair-distances would run closer to the 180-degree maximum, and the total would be high.

   The premise, and the only part of it that matters here, is mechanical rather than mystical: when slow-moving planets bunch together, the index falls; when they spread apart, it rises. Barbault spent decades correlating the low points against the historical record and found major wars, revolutions, and systemic breaks clustering around the troughs. Whatever one thinks of the interpretive layer, the index itself is arithmetic applied to publicly available orbital mechanics and reproducible by anyone with an ephemeris. His index bottomed out at its lowest point of the 21st century in 2020–2021 which was the depths of the pandemic and has been climbing since, with July 2026’s Basket sitting on that ascent, not at a new trough. That ascent is the entire basis for the “golden age” reading circulating among the astrological community: the index is rising, therefore the crisis is supposedly ending. The 1922 comparison that I will discuss more about is the reason I think that inference moves too fast.

4. The federal government’s own actuarial math. This one isn’t a forecast at all — it’s an accounting fact, published by the government itself.

Suspending deflation, and what it actually bought

The throughline connecting Howe and Dent is that 2008 wasn’t resolved so much as postponed. Sixteen years of extraordinary monetary accommodation kept the system running past the point where Dent’s model said it should have broken. That’s not a conspiracy theory it’s the honest appraisal of what injecting capital directly into banks, QE, zero rates, and sustained deficit spending were for. What it bought was time, not resolution, and the postponed pressure surfaced anyway as the inflation shock of 2020–2022, which happens to be the exact window where Barbault’s own index recorded its century low. Two independent systems, one demographic-monetary, one astrological, flag the same years as the moment pressure deferred broke through.

Where the calendar actually stacks

Set the astrology aside for a moment and look at dates that don’t require the musings of a corporate mystic:

• 2026: the Social Security and Medicare Trustees released their 2026 report on June 9. The Old-Age and Survivors Insurance (OASI) trust fund which is the fund that pays retirement benefits to more than 60 million Americans, is now projected to be depleted in the fourth quarter of 2032, one quarter earlier than last year’s projection. At that point, absent Congressional action, incoming payroll tax revenue would cover 78% of scheduled benefits. That amounts to an automatic 22% cut. The Medicare Hospital Insurance trust fund is projected to hit the same wall in the second quarter of 2033. The combined Social Security funds (OASI and DI together) run to 2034, at 83% payable.

• The dates were moved up for because the Congressional Budget Office revised it forward from 2033 to 2032 in February 2026 because it now expects hotter inflation going forward. Higher inflation drives larger cost-of-living increases. The same inflationary dynamic that emerged when the 2008 postponement mechanism inverted Dent’s deflation into inflation is now the thing accelerating the entitlement funding crisis. It’s the same causal chain, showing up twice, twelve years apart.

• 2028: a presidential election, which will be at the point where transiting Pluto’s integration into Aquarius closes out its most active phase relative to the natal U.S. chart.

• 2030: Howe’s own projected climax of the Fourth Turning.

• 2032–2034: the entitlement cliff will be arriving right in the middle of whoever’s term follows the 2028 election, and just past Howe’s own climax date.

Howe’s climax sits almost exactly at the midpoint between the political marker and the fiscal one. You don’t need an astrological chart to find that arresting.

Why the “golden age” undersells its own best analogy

Here’s where I think the popular astrological reading gets the story backwards on its own terms. The last time Uranus and Pluto formed this trine (a trine is a 120-degree angle between two planets, generally read in astrology as the "easy" aspect, where energies flow together rather than clash, in contrast to the friction of a 90-degree square.) was 1922 which was 105 years ago, and the astrological press knows it because it’s the headline comparison in nearly every piece written about this summer’s Basket. But look at what 1922 actually was. It was the opening of the Roaring Twenties; a time in which radio and mass media exploding, the Harlem Renaissance, insulin discovered, King Tut’s tomb found. It was a genuine feel-good boom.

It was also the year Mussolini marched on Rome, Stalin became General Secretary of the Communist Party, the USSR was formally founded, and German hyperinflation began its run toward the Weimar Republic’s collapse. The same year produced both the party and the pathology. What followed the party, seven years later, needs no citation.

If 2026 rhymes with 1922 the way the astrological consensus itself argues, the honest reading isn’t “utopia arrives.” It’s “expect a genuine boom, and don’t mistake it for the destination.” That’s a materially different, more useful, and frankly more astrologically consistent claim than the one dominating the coverage and worth calling out.

Outlets built around consumer astrology content have a structural incentive toward uplift, since that’s what keeps readers subscribed. That’s not an accusation of bad faith. It’s a disclosure worth making before asking anyone to take the comparison seriously.

The 2026 conjunction of Saturn and Neptune, which perfected on February 20, tells a related story from a different angle. A "conjunction" simply means two planets appearing to occupy the same point in the sky from Earth's vantage point, the closest possible alignment two bodies can form. This particular conjunction’s three most recent occurrences which are 1917, 1953, 1989 don’t track with financial crashes. They track with the dissolution of political and institutional orders such as the Russian Revolution, the death of Stalin, the end of the Korean War, and the fall of the Berlin Wall.

That’s the more precise signature for this moment than “depression”. It won’t be an economic collapse specifically, but an old order coming apart at the seams, with whatever replaces it still unwritten.

The real economy is already telling us which reading is right

If the “golden age” framing were correct and this were genuinely a smooth transition into abundance it would be logical to expect capital to move first. Companies with real information advantages would be front-loading investment ahead of the payoff.

However, that is not what the data shows. Kearney’s 2026 Reshoring Index, the 13th edition of its annual gauge, found the index still in negative territory despite record announced investment and sweeping tariff changes: U.S. manufacturing imports hit a four-year high in the same period manufacturing capacity grew barely 1.5%. Domestic manufacturing employment has actually declined slightly since the “Liberation Day” tariffs took effect, even as investment announcements tripled. Executives are hedging They are not committing, at least at the moment and they are citing policy uncertainty, tariff whiplash, and labor availability as the reasons announced capital isn’t converting into poured concrete.

Corporations mirror in their own way what is in their selfish best interests and as far as they are concerned, nobody wants to be the first to take the pain of a real commitment. So, everyone extends and waits, and the extending itself becomes the risk. Companies watching for clarity before deploying capital are doing, at the balance-sheet level, exactly what Washington has done at the fiscal level since 2008.

But it would be a mistake to read this as capital sitting entirely on the sidelines as it isn’t. In a narrow way it’s being selective in a manner that is didactic. While general reshoring stalls, utility and infrastructure capex tied to AI data center demand is entering what analysts are calling a genuine super-cycle, with roughly $1.3–1.4 trillion in electricity infrastructure investment projected for 2025–2030 which is double the prior decade’s pace.

That’s real, front-loaded, high-conviction capital, moving fast but, into a single narrow category. The bifurcation is the tell on this one. This isn’t a broad-based industrial renaissance underwriting the reshoring narrative. It’s a concentrated bet on one technology, landing in whichever communities happen to have the grid capacity, water access, and zoning patience to absorb it.

What prudence looks like inside a soft mold

None of this is an argument for disengagement, and it isn’t a doom forecast dressed up in respectable footnotes. Howe’s own model argues the opposite and that is CRISIS eras are when the mold is softest, when a small number of credible, specific, locally grounded voices can actually shape what replaces the old consensus because the old consensus has already failed. That’s the case for doing careful, unglamorous policy work now rather than later. But it still has standing, and still usable, and especially so when the reckoning that four independent systems are all pointing toward actually arrives.

The astrology doesn’t predict the entitlement math, and the entitlement math doesn’t need the astrology to be true. But when a French mundane astrologer’s sixty-year-old index, a generational sociologist’s model, a demographic economist’s spending-wave thesis, and the U.S. government’s own actuaries all land on the same handful of years, that’s not a coincidence worth dismissing just because one of the four inputs is uncomfortable to cite in polite company. It’s in my humble opinion, worth taking exactly as seriously as the other three — no more, no less.

This is 1922. The party is probably real. So is what came seven years after it.

That standing-room-only crowd is indicative of what an emotionally charged issue AI data centers have become, as well as larger concerns over the consequences of a world succumbing to AI and all its generative iterations. My community’s journey into the world of data centers began officially in November 2025, with the signing of a Community Benefits Agreement (CBA) between the City of Lancaster and the data center developer known to us as Chirisa Technology Parks. As CBAs go, it wasn’t half bad, as it extracted environmental and operational performance commitments from the developer, and the city actually received payments of $20 million in total, delivered as two $10 million installments. This is opposed to places like Austin, Texas, that doled out incentives in the form of tax credits to AI data centers in the past. Moreover, as I pointed out in an earlier Substack, the two sites are former industrial properties that were sitting dormant and will now both be a source of jobs and tax revenues again.

Local Righteous Anger

But something went awry early on that is now fueling a portion of the local righteous anger against the project: the lack of public involvement and comment in the year or two leading up to the signing of the CBA in 2025. At the outset of Tuesday’s meeting, I had the good fortune to sit next to Arthur (Art) Morris, a man who served as Lancaster’s mayor from 1980 to 1990. He remarked to me that there will always be a price to pay when you don’t involve the public early on in issues such as this. In the end, this was a failure of process.

Lancaster’s two existing data centers were approved under a zoning determination that classified them as electronic data processing warehouses, which is the same broad category as wholesale trade and storage. This was rightly lampooned by local activist Tony Dastra. Nobody asked what a data center actually draws power in, or what its cooling system pulls from local water, because the zoning code didn’t ask those questions either. That is the reason Tuesday’s room was full of people who didn’t trust the process, and they weren’t wrong to feel that way. When government approves something without examining what it is, distrust is the correct response.

Fortunately, the data center zoning amendment is genuinely strong and addresses almost all of the shortcomings. Moreover, it gives our state and the country, for that matter, a model to follow that answers the deeper question of qui bono, who benefits. In our case, both the City of Lancaster and future developers will benefit. For instance, the amendment states:

Data centers will be pulled entirely out of Central Manufacturing zones and will only be permitted by special exception in Suburban Manufacturing zones. As Councilman Amos Stoltzfus stated, that leaves only sixteen parcels for data center development in the whole city.

There will be mandatory water utilization reporting with cooling system disclosure, thermal impact mitigation plans, and decommissioning bonds. Setbacks will be increased to two hundred and three hundred feet. Additional provisions strengthen heat mitigation and decommissioning standards and bar data centers from externalizing costs onto the grid or building within floodplains.

Although more needs to be done on the sound mitigation front, this is not a rubber stamp with better PR. It is a real, substantive constraint, arrived at through exactly the kind of public process that should make people trust it more, not less.

AI As An Overarching Issue Versus AI Data Centers

At this point, it may not be possible to separate these large AI training data centers from the overall debate on AI, since these large language models require enormous processing power, resting on the latest in central processing (CPU) and graphics processing (GPU) unit technology. And make no mistake about it — these processors are power-hungry, heat-shedding animals.

Years ago, the largest data center company in America was a state and local tax client of mine. When we spoke of cooling needs, we spoke about cooling floors and aisles. Today’s AI data centers cool individual chips in a process known as direct-to-chip liquid cooling. The layout logic then was distributed and modular, whereas today it is centralized and tightly clustered. Power requirements are orders of magnitude higher than in the past.

And there is a lot of fear and dread being spread about AI in general, from its role in creating a surveillance state to job destruction to the dumbing down of the citizenry. Some of it is warranted and worthy of spirited public debate, and some of it is hyperbole and outright fearmongering. Regardless of how these views were shaped, they are real and have consequences.

A recent Quinnipiac poll found that 76% of Pennsylvania voters follow news about AI data centers either very closely (26%) or somewhat closely (50%). Additionally, 74% said they would oppose an AI data center in their community, while 16% said they would support one and 10% offered no opinion.

Narrowing the conversation just to AI data centers, I fear we may be throwing the baby out with the bathwater just as Central Pennsylvania is showing the country how to do this right. In our case, we took a former industrial printing plant and converted it into a multibillion-dollar AI campus on already industrial land, touching no farmland, all the while drawing on a deregulated power market with nuclear generation coming back online nearby. Compare that to what’s been proposed elsewhere: tens of thousands of acres of open land in water-stressed states, betting that kind of risk can be exported onto a community’s water table. Lancaster doesn’t have to make that bet. We are providing a model for data center development that maximizes the public’s interest. Moreover, we have the water, we have the power, and we have the workforce.

But this may be lost in the overall AI debate.

Unrighteous Anger and Outside Provocateurs

And it is at this intersection of the debate that nefarious individuals and groups are inserting themselves, and this was on display last Tuesday evening in the city’s council chambers.

A group known locally as Lancaster Stands Up had worked the crowd beforehand, and they weren’t alone as a good chunk of the local left-of-center activist infrastructure showed up to voice their opposition to data centers. All night we heard from citizens calling for no data centers. Not fewer. Not better-regulated. None.

Lancaster Stands Up is part of a larger umbrella organization, Pennsylvania Stands Up, and it is not a small operation. According to their 2024 Form 990 (EIN 83-2880678), the organization took in $1,832,453 in revenue and held $2,899,875 in net assets. The grassroots are, at minimum, well-funded.

In his book I, Candidate for Governor: And How I Got Licked, Upton Sinclair wrote that it is difficult to get a person to understand something when their livelihood depends on not understanding it.

The Tell Is In The Ask

I’ve spent a long time in advocacy work, and I’ve learned to read a movement by what would actually satisfy it and by when it declares victory and calls it a day.

In this instance, a group asking for setbacks stops showing up once it gets setbacks. A group asking for water disclosure stops showing up once it gets water disclosure. Lancaster Stands Up and the coalition around it got both, plus a hard parcel cap, plus decommissioning bonds, plus a thermal mitigation requirement most municipalities in this country haven’t thought to ask for. If the ask was ever really about any of those things, Tuesday’s vote should read as a win worth claiming credit for.

I don’t think it will be, because I don’t think the ask was ever about the provisions or about cutting a good deal everyone could live with. Their ask is that AI data centers, as a category, are illegitimate; that they are an extraction economy dressed up as infrastructure, in which a handful of jobs are traded for someone else’s water and someone else’s grid strain; and that the benefits flow to shareholders in San Francisco and Wall Street sovereign wealth vehicles rather than the people of Lancaster.

That’s a coherent worldview. It is not, however, a policy position, because it has no answer to the question of what the city is supposed to do instead. Pennsylvania law requires every municipality to zone somewhere for every legal use, and “somewhere” was always going to be an argument about which sixteen parcels and not whether there would be any parcels at all.

This is a pattern I’ve watched across a lot of adjacent fights, from labor issues to immigration. The pattern goes like this: a coalition organizes around a maximalist demand, wins substantial concessions through a legitimate democratic process, and treats the concessions as a waypoint rather than a resolution because the actual goal was never negotiable in the first place.

You have to respect the organizing capacity of Lancaster Stands Up. They got a genuinely packed room on a Tuesday night, which is something. But you also have to be clear-eyed that a movement built around an unfalsifiable demand isn’t in the business of being satisfied by policy. It’s in the business of satisfying its donors.

Conclusion

None of this required a categorical answer, and that may be the real lesson of Tuesday night. Art Morris was right: the first two data centers went in without anyone asking the public what it thought, and the city is still paying that bill in distrust. But watching this ordinance pass, watching sixteen parcels replace the entire industrial map of the city as eligible ground, and watching a water utilization report become mandatory instead of optional, I don’t think silence was ever the fix. Rules were made. Real ones, with teeth, arrived at the way good rules are supposed to arrive: messily, publicly, over the loud objection of people convinced the whole exercise was theater, though it wasn’t.

There is a much larger, much harder argument to be had about what AI is doing to work, to attention, and to the shape of a society, and it deserves every bit of the seriousness people want to bring to it. But that argument was not on the agenda Tuesday night, and pretending it was is how a zoning vote about sixteen parcels in Suburban Manufacturing turns into a referendum on the future of the species. Lancaster now has language on the books that most cities sitting in far more precarious positions with regard to land, water, and power do not. Whether that satisfies the room that showed up Tuesday night is a different question from whether it works. I suspect it will keep doing the latter long after it’s stopped being asked to do the former.

Our latest report, which you should check out, does a deep dive into data regarding the Dallas-Fort Worth metroplex and how many jobs are being taken by foreign engineering students and graduates on OPT. Meanwhile, recent American engineering graduates are suffering from unemployment or underemployment. I’ll outline our key findings below.

Flat Jobs, But Double The OPT Engineers

Between 2012 and 2022, the total number of engineering jobs in Dallas-Fort Worth had essentially plateaued, hovering around 40,000 positions for that entire decade, even as the metro’s population surged past 7.9 million people.

But while the pie stayed the same size, the guest worker slice doubled. OPT engineers in Dallas went from 1,998 workers in 2012 to 3,913 in 2022. That is an increase from 5% to nearly 10% of the entire engineering workforce in the Metroplex. At the program’s peak in 2018, OPT engineers made up close to 19% of the sector, a fourfold increase in just the four years between 2014 and 2018. The COVID border slowdowns decreased the guest workforce between 2019 and 2022.

The rapid expansion in the 2010s tracks almost perfectly with the expansion of the program by the Obama administration. To rewind, the Bush administration first set the STEM OPT extension at 29 months. The Obama administration then expanded it further, to 36 months, in 2016. Understand that the creation of STEM OPT during the Bush administration and the extension that took it from 29 to 36 months were not authorized by Congress. Rather, they were created through agency deference.

In addition to never being voted on by Congress, there is no numerical cap on the number of work authorizations and no requirement that employers first attempt to hire an American. The year after Obama’s expansion took effect, the OPT share of Dallas engineers jumped from 12.5% to 17% between 2016 and 2017. Work authorizations like STEM OPT, OPT, and Curricular Practical Training work as magnets that attract foreign students to U.S. universities and to American employers, who want cheaper labor.

Who Actually Fills These Jobs

The defenders of these programs like to invoke the image of the genius researcher creating breakthroughs in the lab, leading teams of American engineers. However, the data doesn’t support that framing, at least not in Dallas, where only 11% of OPT engineers in the metroplex were pursuing or held a U.S. PhD.

The other 89%, or 3,487 foreign workers, had bachelor’s or master’s degrees, the exact same credential level held by tens of thousands of young American engineering graduates competing for entry-level roles in the same city.

We are not talking about scarce specialists filling gaps nobody else can fill. They are, credential for credential, direct competitors to American graduates for the same jobs—jobs that, again, did not grow in number over the decade in question.

Moreover, in 2022, for example, 61% of these workers who were on STEM OPT came from India, and a striking 21% of them graduated from a single school, the University of Texas at Arlington. Add the University of Texas at Dallas, which ranked as the second-largest STEM OPT-authorized campus in the country in 2018, and you begin to see that this is not the workings of a free and fair labor market. Rather, what we are seeing is an institutionalized pipeline funneling low-wage foreign workers, with universities as major players.

The Subsidy Nobody Talks About

Here’s the part that should trouble anyone who cares about fair treatment for American workers: OPT and STEM OPT workers are exempt from Social Security and Medicare payroll taxes. It’s a 15% built-in discount for every employer who hires a guest worker over an American graduate, a discount financed by the absence of oversight, the absence of a wage floor tied to displacement protections, and the absence of any requirement to look to hire an American first. It is, functionally, a government-subsidized underbidding mechanism that bars American graduates from that first job out of college.

What This Means for American Graduates

Every year, engineering programs across Texas and the country produce graduates who are told, correctly, that engineering is a stable and in-demand field. What they are not told is employers not only can but also prefer to hire foreign graduates rather than them.

In Dallas-Fort Worth, roughly one in ten entry-to-mid-level engineering jobs is now filled by a guest worker whose employer pays less in payroll taxes to hire them than to hire the American sitting in the next seat at graduation. The OPT program is not just an insult to Americans who are more than qualified, studied hard, and most likely took on student debt; it is a tragedy with broad implications for America’s future. Bypass enough of those graduates enough times, and you don’t just lose a job; you lose the first rung of a career ladder that never gets climbed up for an entire generation.

The Dallas numbers matter because Dallas is one of the largest, fastest-growing metros in the country. The Dallas-Fort Worth Metro area is home to a deep bench of engineering programs but a business community that insists on using guest worker programs when American talent is there. What’s missing is a labor market that gives it a fair shot before reaching for the subsidized alternative.

Our Call to Action

Frederick Douglass aptly stated, “Power concedes nothing without a demand. It never did and it never will.” STEM OPT and the litany of employment visa and work authorization programs used to displace American workers have strong support in Washington, DC, and in governor mansions across the country. These programs aren’t going away simply because they are bad news for Americans. If we are to end them, we must fight back, and we have a plan to do just that.

Join us in signing and sharing our petition to end STEM OPT. Find your state, click on it, sign the petition, and share it with your fellow students, neighbors, colleagues, parents, and alumni. Later this year we will deliver these petitions to each state’s congressional delegation.

If you do not see an active petition for your state, reach out to me at klynn@instituteforsoundpublicpolicy.org and we’ll discuss how you can start one in your state.

“The provision is, that ‘all persons born in the [U.S], and subject to the jurisdiction thereof, are citizens. . . What do we mean by [it]? Not owing allegiance to anybody else. That is what it means.”

The intention of the Fourteenth Amendment was to ensure those freed from bondage following our bloody civil war would not be returned to servitude; that they were and would remain fully enfranchised citizens of this country.

The above sentiments were echoed in Justice Thomas’ dissenting opinion:

“Congress implemented the principle that citizenship follows birth and domicile in the Civil Rights Act of 1866 and then in the Citizenship Clause of the Fourteenth Amendment. The Civil Rights Act guaranteed citizenship to the persons who were both “born in the United States” and, as relevant here, “not subject to any foreign power.”

Oddly, the majority opinion rested on a flawed interpretation of what constituted a broad interpretation of citizenship going back over 125 years to the 1898 SCOTUS ruling in United States v. Wong Kim Ark.

Chief Justice Roberts, a Bush appointee who wrote the majority opinion, even cited English common law stating, “children “born within [sovereign’s] dominions” owed a “natural allegiance” to the sovereign who protected them at birth.” This would be laughable if it weren’t so serious. Really Justice Roberts, we fought the revolutionary war 250 years ago to free us from English common law!

Notably, Justice Barrett, a Trump appointee, joined the majority opinion in full. Justice Kavanaugh concurred in part and dissented in part, believing it did not violate the Constitution but did violate the Immigration and Nationality Act. Thomas, Alito, and Gorsuch dissented.

This is a significant loss, and I won’t pretend otherwise. But understanding exactly what kind of loss it is matters for what we do next.

There won’t be a quick fix. Sure, the prestige of the Court has diminished greatly over the years, and it can be argued what needs to be done is change the composition of the court. But can this be counted on to right the ship?

The ruling was grounded directly in the Constitution itself, not just statute. That forecloses the easiest workaround, which would be Congress passing a law to override a constitutional holding. No statute, however clarifying its intent, can undo what the Court just said the Fourteenth Amendment requires.

But ways forward exist.

As Ryan Girdusky wrote in a post on X yesterday, “In the same way that the right started a legal movement to overturn Roe, we need a legal movement to overturn this terrible decision of birthright citizenship. The child of a Chinese spy born in the states is not an American.”

In the long run we must be strategic. The surest path that directly changes the rule is a constitutional amendment: two-thirds of the House and Senate, then ratification by 38 state legislatures.

This is an extraordinarily high bar and should be treated as a generational project, not a single election cycle. Winning more seats in Congress is necessary but is nowhere close to sufficient on its own, as the real bottleneck is the states. Thirteen holdout state legislatures can block ratification no matter what happens in Washington. Any serious long-term strategy has to build power in statehouses, not just on Capitol Hill.

In the short and mid-term we must react with viable, implementable policy changes. President Trump must exercise his authority using the Immigration and Nationality Act and proclaim that women of child-bearing years who aren’t legal immigrants are largely barred from the United States. As @USC12 stated in an X post, no admittance and removal of those here on F-2 and J-2 visas and H-4 EADs (essentially spouses.) We can also deal a severe blow to the birth tourism industry by restricting women who are pregnant from entering the country.

Congressman Chip Roy introduced late last year the PAUSE Act that would freeze all immigration to the United States with few exceptions until certain conditions are met to ensure our immigration system is working on behalf of the American People.

We need to explore potential legal and legislative openings to answer questions the Court didn’t address, such as emerging reproductive arrangements, jurisdictional questions in certain territories, and other statutory documentation standards distinct from the citizenship question itself. These options are narrower but represent opportunities where careful litigation can still do real work without inviting another defeat on the core question.

Pass the SAVE Act! Voter ID is something the vast majority of the American people want, President Trump wants, and the House of Representatives passed. Now it is time for the Senate to act. Eliminating election fraud is the first step in creating super majorities in the Congress and state legislatures.

In closing, we must be as stalwart as we are creative. We may appear to be hemmed in. But there is always a way forward. Remember the words of Sun Tzu in his Art of War:

“There are not more than five primary colors (blue, yellow, red, white, and black), yet in combination they produce more hues than can ever been seen.”

“There are not more than five cardinal tastes (sour, acrid, salt, sweet, bitter), yet combinations of them yield more flavors than can ever be tasted.”

The findings were significant. The press coverage was not.

I analyzed the research and built this presentation to explain what the data actually shows — in plain language, with full sourcing, and without the ideological filter that typically distorts this conversation.

What this video covers:

• 00:00 — Introduction

• 03:04 - The unprecedented scale of the 2021–2024 surge — 7 million net arrivals in three years

• 04:32 - The causal methodology — why this study is different from prior research

• 09:05 - Employment effects — the 0.96 multiplier and what it conceals

• 11:32 - Wage suppression — the -0.87% finding and why it matters

• 14:38 - The supply myth — what the permit data actually shows

• 21:49 - The fiscal picture — transfers, welfare enrollment, and the incomplete ledger

• 27:50 - What the mainstream narrative gets wrong

• 29:49 - Policy implications — and what enforcement data proves

Key findings from Wilson & Zhou (2026), Federal Reserve Bank of Dallas WP-2607:

→ Home prices: +2.2% per 1% unauthorized immigrant worker flow increase
→ Rents: +1.4% per 1% unauthorized immigrant worker flow increase
→ Wages: -0.87% per 1% unauthorized immigrant worker flow increase
→ New housing supply response: statistically zero
→ UIWF explains ~30% of home price growth and ~20% of rent growth at the average metro

Sources:

• Federal Reserve Bank of Dallas Working Paper 2607 — Wilson & Zhou, March 2026
  https://doi.org/10.24149/wp2607

• CBO Demographic Outlook 2026–2056

• Zillow Home Value Index & Observed Rent Index

• BLS Quarterly Census of Employment and Wages

• BEA Regional Economic Accounts

• Camarota & Zeigler (2026), Center for Immigration Studies

About the Institute for Sound Public Policy:
The Institute for Sound Public Policy (IfSPP) produces research, analysis, and advocacy on employment visa programs, offshoring, and labor market policy affecting American workers.

Surveying The Ground In Geneva

Place has always been important to me, and how I view a place has been informed by a number of lived experiences. Growing up on a farm inculcated in me a sense of my relationship with the soil, the flora, and the fauna of Bucks County, Pennsylvania. Later as an intelligence officer in the army, I learned to view the strategic value of a geography as I applied overlays of concern that included corridors of approach, demographics, politics, culture, history, the built environment, the economy, the weather, and much more. Later in life I studied feng shui under Steven Post, Edgar Sung, and Barry Gordon and even trained over a long weekend with the late Professor Lin Yun.

Geneva, and specifically its twentieth-century modern brutalist architecture, had a lot to tell me. The concrete edifices that are the city’s United Nations (Palais de Nations), ILO, and convention center (CICG) buildings represent something that can not only be seen but also felt by even the most casual observer.

The ILO headquarters sits on a hill above Lac Léman, a massive brutalist structure that was completed in 1974 at a time the developed world thought it could restructure global trade and commerce through institutional leverage. The building announces itself with the confidence of that moment: raw concrete, geometric mass of hard angles, and the unadorned honesty of a structure that believes its own importance so completely it requires no ornament to assert it.

Le Corbusier, the French-Swiss architect whose vision haunts everything modernist Geneva built in the postwar decades, understood concrete differently. His béton brut was emancipatory in intention. The exposed material, the board marks left deliberately visible in the Unité d’Habitation in Marseille, and the pilotis lifting buildings off the ground to return the earth to human movement were all oriented toward liberation. Light. Air. Dignity expressed in built form. He had gone to Greece as a young man, stood before the Parthenon, and understood geometry as a moral statement. Pure form as honest claim.

What Geneva built with his vocabulary was something else entirely. The ILO headquarters, the Palais des Nations, just down the road, are not emancipatory structures. They are administrative fortresses. They took the formal language of liberation and used it to house the opposite impulse. Not the freeing of human beings from industrial darkness but the management of human representation within acceptable procedural limits. In short, the béton brut aesthetic without the animating social vision.

Inside and outside those structures, as well as the CICG, where a lot of the workers’ meetings were held, the human is made to feel small, insignificant even. None of these modern constructions were made to a human scale. And that was purposeful.

Standing inside the ILO building for the first time, feeling the weight of the concrete above and around me as I crossed from the north to the south wings via the long corridor referred to as Pas-Perdus, a phrase came to mind—concrete and void. For

Le Corbusier, the “void” is not an empty absence but the dynamic, continuous flow of space that defines how humans experience architecture. The forces at work may not be visible to the naked eye, but they are there. Again, all of this was lost on the architects who interpreted “void” as void of humanity.

The ILO building has its own invisible activity, though, not architectural in nature. The technical meetings where resolution language gets softened before it reaches the floor. The quiet coordination between captured labor federations and compliant governments represents the ghost operations running beneath the ceremonial surface of tripartite consensus. The void inside this building is not empty either. It is full, full of obsolete and failed visions of institutions that no longer serve humanity.

Into The Void

On my day away from the conference I took the Metro to CERN, the home of European advanced particle physics research.

I could not book a tour of the superconducting tunnel housing the Large Hadron Collider, 27 kilometers of engineered precision buried beneath the Franco-Swiss border, the instrument that found the Higgs boson and confirmed our understanding of how matter acquires substance. But the exhibits were substantial, and the timing was fortunate: I caught the final day of the Ciné Global 2026 film festival on the CERN campus, where a program of science documentaries was screening.

One film gripped me, Invisible Energy, made in 1984, documented the first successful run of what was then CERN’s supercollider, a machine roughly 7 kilometers in circumference, modest by current standards, operating at the frontier of what was then possible. The film was dry by contemporary standards. It had the unglamorous texture of a workmanlike news documentary of its era. What it showed was people,

the engineers, physicists, and technicians doing extraordinarily complex work with extraordinarily modest tools. DOS-based computers with what we would now consider laughable processing capacity. Terminals generating data streams that the human eye had to interpret in real time. The complexity and scale of the undertaking were staggering. The ceremony surrounding it was essentially nonexistent.

1984. The same year the word entered the language as shorthand for institutional language carefully engineered to mean its opposite.

But what struck me most was simpler than that. The people in that film were trying to find something true. They had a genuine question about the invisible structure of reality, about what matter is made of, and about the forces that hold the world together, and they were applying every available instrument to answer it. The work was unglamorous. The computers were primitive. The question was real.

CERN makes the invisible visible. It exists to force the hidden structure of reality to reveal itself, to drag the unseeable into the observable through the application of sufficient energy and sufficient precision. Everything it does is oriented toward disclosure.

Two kilometers away, the institution I had spent the week inside operates on the opposite principle. Its invisible activity, the technical meetings, the softened language, and the managed consensus are oriented not toward disclosure but toward concealment. Not toward finding something true but toward producing something acceptable. The ILO’s invisible power is the power to make the visible invisible: to take the material reality of worker displacement, platform exploitation, and institutional capture and render it as resolution language that every delegation can sign and the majority of workers in developed nations will little appreciate nor feel.

Back to Reality

As I mentioned, the ILO was founded in 1919 as part of the Versailles settlement. Its foundational document, refined and restated in the 1944 Philadelphia Declaration, opens with a claim so radical it still startles: labour is not a commodity.

That sentence was meant to be the load-bearing wall of an entire international order. If labor is not a commodity, then the human being who sells their time and effort cannot be treated as an input to be priced, substituted, and discarded by market mechanism alone. It implies standards. It implies enforcement. It implies that somewhere above the transaction between employer and worker there exists a set of obligations that neither party can contract away and no government can legitimately abrogate.

Walk through the ILO’s current output on the platform economy, titled “Decent Work In A Platform Economy,” that deals with gig workers, the app-dispatched drivers, and the algorithmically managed piece-rate laborers of the 21st century, and you will find something remarkable. The diagnostic work is often excellent. The ILO’s researchers understand the exploitation clearly. They can describe with precision how the classification of workers as independent contractors strips them of the protections the Philadelphia Declaration was written to guarantee. They know what is happening. (With the possible exception of what they choose to overlook, such as the impact of immigration and offshoring they are having on workers in.)

The prescriptive work is another matter entirely. Resolution language carefully constructed to satisfy every delegation. Recommendations calibrated to offend no government, no employer federation, no captured labor organization. For instance, does the agreed-upon convention mention the impact of immigration and offshoring on the displacement of workers—NOT A CHANCE! However, it would have included lots of distracting diversity and gender topics had the US government delegation not aggressively intervened.

The gap between the accurate diagnosis and the toothless prescription is not an accident of drafting. It is the institution’s actual product, which equates to the form of governance without the force of it in any meaningful way.

“Labor is not a commodity,” a phrase first coined by Samuel Gompers, then founding president of the American Federation of Labor and later written into the founding statement of the ILO, has become a ceremonial statement. It is read aloud and nodded at, and then the conference proceeds to manage, with great procedural sophistication, the conditions under which labor is treated as exactly that.

In conclusion, the concrete edifices will endure. However, the intention they were built to contain has been replaced by its opposite. That substitution is the invisible power this institution actually runs on.

The same week, at the University of Central Florida, a commencement speaker told a gymnasium full of arts and humanities graduates that artificial intelligence is “the next Industrial Revolution.” They booed her off the stage. At another ceremony, former Google CEO Eric Schmidt told graduates “you will help shape artificial intelligence” and was met with the same chorus of jeers.

The air is coming out of the tires. The AI data center boom celebrated in boardrooms and congressional testimony alike is generating a mounting bipartisan backlash from Tucker Carlson’s audience to graduating students to Utah farm owners who don’t want a 40,000-acre industrial campus consuming twice their state’s entire electricity supply on their doorstep. The question worth asking is not whether to build the infrastructure America’s digital future requires. It is whether developers have the wisdom to build it in the right places.

Most don’t. And the consequences of that failure are beginning to compound.

The Wrong Places

The data center industry has a fundamentals problem disguised as a land problem. Developers have chased cheap acreage, generous tax subsidies, and permissive rural zoning and in doing so have systematically ignored the two resources that actually determine whether a data center succeeds over a decade or two: power and water.

The results are visible in the data. According to a comprehensive analysis using the World Resources Institute’s Aqueduct Water Risk Atlas, 65 percent of America’s 4,629 tracked data center facilities sit in areas classified as medium-high water stress or above. One hundred and forty-three operate in zones where water demand already exceeds renewable supply. Nevada leads the stress rankings with an average score of 4.8 out of 5. Arizona scores 4.1. Texas, home to 554 data centers scores 3.1, with 537 of those facilities in water-stressed areas.

The Ogallala Aquifer, which underlies much of the Great Plains from the Texas Panhandle to Nebraska, is being drawn down at rates that make the word “abundant” a reckless fiction. Water drawn from deep aquifers in parts of Iowa and the Texas Panhandle will not be replenished on any human timescale. O’Leary’s proposed Utah campus would consume up to 9 gigawatts of electricity, more than double Utah’s current total electricity usage and this is in a state that depends on an already over-allocated Colorado River and a rapidly depleting groundwater table.

This is not a business plan. It is a liability waiting to be recognized.

The Right Places

There is a better model hiding in plain sight, and it is already being built but it is not in the Utah desert. It is in Lancaster, Pennsylvania, on the footprint of two abandoned printing plants that had sat idle for years.

In February 2025, developers Chirisa Technology Parks and Machine Investment Group announced plans to convert 144 acres of former industrial land that was the old R.R. Donnelley and LSC Communications printing facilities at 216 Greenfield Road and 1375 Harrisburg Pike into a major AI data center campus leased to CoreWeave. By August, Blue Owl Capital had closed a $4 billion joint venture to fund the project, with CoreWeave committing up to $6 billion total. The campus will deliver an initial 100 megawatts of capacity, scalable to 300 megawatts, with $200 million in grid infrastructure improvements funded by the developers themselves.

No farmland was touched. No open space was converted. The city’s zoning officer confirmed the sites’ existing Suburban Manufacturing designation was fully compatible with data center use. Lancaster’s farmland preservation advocates — including the active watchdog group Respect Farmland didn’t lift a finger to oppose it. There was nothing to oppose. The land was already industrial, already served by heavy electrical infrastructure, already connected to municipal water and sewer. The community got construction jobs, permanent employment, and a tax base replacement for buildings that had generated nothing for years.

What makes Central Pennsylvania exceptional is not just the availability of brownfield land. It is the convergence of fundamentals that sophisticated data center capital should be chasing everywhere. The region sits within PJM Interconnection which is one of the nation’s largest deregulated electricity markets, where large customers can negotiate directly for power rather than depending on a regulated utility’s rate-setting process. The Susquehanna River watershed provides abundant, rechargeable freshwater that stands in stark contrast to the fossil water being mined beneath the Great Plains. And within forty miles of Lancaster, two nuclear facilities provide the firm, dispatchable baseload power that GPU clusters running 24 hours a day, 365 days a year actually require. The Crane Clean Energy Center, formerly Three Mile Island Unit 1, already powering Microsoft data centers — is back online. Unit 2 is on track for restart by late 2027, backed by a $1 billion federal loan. The Susquehanna Steam Electric Station in Berwick anchors northeastern Pennsylvania’s power profile.

A Distributed Model That Scales

Here is the argument that has not yet been made in the national data center conversation: you do not need 40,000 acres in a desert to build gigawatt-scale AI infrastructure. You need the discipline to think differently about scale.

A single 1-gigawatt hyperscale campus requires 500 to 800 acres with much of it land-banked for future phases that may never be built. But twelve 100-megawatt data centers, distributed across brownfield industrial sites throughout Central Pennsylvania’s Lancaster, Lebanon, Dauphin, York, Cumberland, and Berks counties, would together deliver 1.2 gigawatts on perhaps 1,000 to 1,200 total acres of already-industrial land, spread across a grid that can absorb the load without strain, fed by nuclear baseload, cooled by watershed water that renews itself with every rain.

The distributed model delivers additional advantages that the megacampus cannot. Smaller facilities in existing industrial zones move through permitting faster and with less community friction. Distributed grid draws are less disruptive to regional electricity markets than a single massive interconnection request. For the inference workloads that increasingly dominate AI computing demand, distributed regional facilities actually perform better at lower latency, closer to population centers, more resilient to single-point failures.

Pennsylvania Department of Environmental Protection’s land recycling program has cleaned up more than 6,000 sites since 1995. EPA brownfield assessments have identified thousands of acres of former industrial land across the Commonwealth awaiting productive reuse. The American Farmland Trust’s Cost of Community Services research has consistently found that agricultural land generates more in tax revenue than it consumes in public services while commercial and industrial development on existing infrastructure imposes zero marginal road, utility, or stormwater costs on local taxpayers. Every brownfield data center is a double victory: a derelict site returned to productive use, and a farm saved from a warehouse or a data center that had no business being there.

A Message to Developers

The communities that will welcome you already exist. The infrastructure is already there. The water is already flowing from a watershed that recharges with every storm. The power is already deregulated, already backstopped by nuclear generation that runs when the sun doesn’t shine and the wind doesn’t blow. The industrial land is already zoned, already assessed, already waiting.

You don’t need 40,000 acres in a Utah desert. You don’t need to fight farmland preservation groups, outrun water utility capacity, or ask three county commissioners to override a community’s objections before the cameras arrive. You need the wisdom to recognize what the developers in Lancaster, Pennsylvania already figured out.

It may seem improbable that the national model for responsible AI infrastructure would emerge from a county where horse-drawn buggies share the road with semitrucks. But perhaps it shouldn’t surprise us at all. The people who live among the Amish never forgot that land is not a commodity to be liquidated for the next speculative boom and that water is not an accounting abstraction but a finite gift, that a community’s consent is not an obstacle to be routed around but a privilege to be earned. They preserved their farmland precisely because they understood that prudence is not the enemy of progress. It is its prerequisite.

The fundamentals always win. In Lancaster County, the fundamentals are already here. The rest of the country would do well to take notes.

Both companies sell property management software. They compete in the same market, serve many of the same customers, and offer products that on the surface look remarkably similar. But they represent entirely different philosophies of what a company is for, who it serves, and what obligations it carries to the people who built it. One of them is a case study in what American capitalism can produce at its best. The other is a cautionary tale about what it has become.

The first company is Yardi Systems. The second is MRI Software. And the distance between them is the distance between a country that builds things and a country that financializes them.

The Builder

Anant Yardi founded Yardi Systems in 1984 in Santa Barbara, California. He has run it ever since. He has never taken outside investment. He has never sold to a private equity (PE) firm. He has never taken the company public. For more than four decades, he has simply built, year after year, product after product, customer after customer, a property management software platform so deeply embedded in its clients’ operations that switching costs are nearly prohibitive.

The results are extraordinary by any measure. By 2024, Yardi had grown to approximately $1.6 billion in annual revenue with more than 20,000 customers worldwide. Every dollar of that revenue was earned organically. There is no leveraged buyout debt on the balance sheet. There are no PE partners marking the company to market every quarter. There is no Goldman Sachs mandate to prepare for a dual-track sale process. There is a founder who built something, kept building it, and trusted his American workforce to grow it with him.

That workforce matters. Yardi’s competitive advantage is not, at its core, a codebase. Codebases can be copied, reverse-engineered, and eventually commoditized. Yardi’s advantage is the institutional knowledge, customer relationships, and organizational continuity that its employees have accumulated over decades of working inside one of the most complex and relationship-dependent industries in the American economy—real estate. You do not build that in Hyderabad. You do not acquire it through a roll-up. You cultivate it over years, and you keep it by treating the people who carry it as assets rather than line items.

This is what patient capital looks like. This is what a founder-led, long-horizon business produces when it is allowed to operate outside the quarterly earnings cycle and the PE return model. It is not glamorous. It does not generate carried interest. It will never be the subject of a breathless TechCrunch piece about a unicorn exit. It is simply excellent, and it endures.

Yardi is, in the truest sense, an aspirational American success story. A founder with a vision, a workforce that believed in it, four decades of compounding, and a company that today serves millions of people who live and work in properties managed on its platform. This is the kind of business that used to be the norm. It is increasingly the exception.

The Extractors

MRI Software has a different story. Founded in 1971, it was for decades a solid if unremarkable competitor in the same property tech space. It was a company with good products, loyal customers, and a reputation built on decades of reliable service. Then came the private equity era.

Today MRI is owned by a consortium—TA Associates, Harvest Partners, and GI Partners—who have grown it primarily through acquisition, layering product upon product in the classic roll-up model. The platform is wide. The integration is, by many accounts, something less than seamless. And the ownership group, having engaged Goldman Sachs to advise on strategic options, is now targeting an exit that could value the company at up to $10 billion.

That is the context for what was announced this week: MRI is laying off between 350 and 400 American workers — roughly one third of its domestic workforce — while simultaneously accelerating its offshoring operations to India. There will be no press conference. There will be no congressional hearing. There will be, in all likelihood, a tastefully worded press release about “operational optimization” and “global resource alignment.”

The stirrings of the current business model were first noticed by employees when MRI’s CEO Pat Ghilani posited something so banal it deserves to be framed: “Other companies are doing this.”

He is not wrong. They are. And that is precisely the problem.

The Mechanism Nobody Names

Before we go further, it is worth pausing on something, because MRI’s story contains a detail that is more important than the headline number of layoffs.

This is not the Disney scandal of 2015, where American IT workers were marched into rooms and required to train their H-1B visa replacements with the humiliation so visible, so documented, and so appalling that it generated genuine congressional outrage and front-page coverage in the New York Times. This is not the Vanguard or CSX model, where American workers were formally “rebadged” to Indian IT consulting firms in a transaction legible enough to litigate. Those cases had a smoking gun. A named consulting firm. A paper trail. A moment of identifiable wrongdoing.

What MRI has done—what many American technology companies are now doing—is something more patient. More methodical. More deniable.

An MRI employee described to me how the process unfolds in practice. It begins gently. American workers are asked to collaborate with a new team of developers in India. The framing is collegial: can you help these guys get up to speed? Over the following months, the offshore team begins sitting in on meetings. Then leading sections of meetings. Then owning sections of the codebase. Then owning entire products. The timeline varies with the complexity of the software involved. In most cases, the American worker is fully replaceable within one to two years. In one case this employee described to me, the software was so architecturally complex that it took three full years before management felt confident enough to act.

For those three years, the American worker showed up every day, performed their job with professional diligence, helped their overseas colleagues learn it, and waited.

“For a long time,” this employee told me, “it felt like we were just sitting in prison, waiting to be called before the firing squad.”

This is the mechanism that has no name in our political vocabulary. It is not outsourcing in the traditional sense. It is not a plant closure you can photograph. It is not a rebadging. You can’t make a Freedom of Information Act (FOIA) request about it. It is a managed, multi-year knowledge transfer, disguised as collaboration and executed as displacement, that ends with an American worker handing over the keys to their own livelihood and being thanked for their service on the way out.

It is the most insidious form of American worker displacement because it recruits the worker as an instrument of their own replacement and then fires them for having done the job too well.

Other Companies Are Doing This

Pat Ghilani’s justification, that other companies are doing this, deserves a fuller examination than he perhaps intended because he is describing not a defense but an epidemic.

Marc Benioff, CEO of Salesforce and one of the most celebrated technology executives in America, announced last year that his company would not hire a single new engineer in 2025. His support workforce had already been cut nearly in half, from 9,000 to approximately 5,000 employees. Meanwhile, Salesforce has been expanding its innovation centers in Hyderabad and Bangalore, aggressively increasing engineering, sales, and consulting hiring across India. On an earnings call, Benioff declared: “My message to CEOs right now is that we are the last generation to manage only humans.”

The last generation to manage only humans. There is a vision of the American workforce embedded in that sentence that should alarm every salaried professional in this country, regardless of industry. Benioff is not an outlier; he is simply the most candid.

The pattern repeats across enterprise software, financial services, healthcare administration, and insurance. American companies that once competed on the quality of their domestic workforce now compete on how efficiently they can transfer that workforce’s knowledge offshore and then eliminate it. The workers who built the products, cultivated the customer relationships, and generated the institutional knowledge that made these companies valuable are being systematically converted from assets into liabilities, and replaced by cheaper labor abroad before anyone notices.

The Schumpeterian Irony

Here is what makes the MRI story particularly instructive and particularly bitter.

Yardi’s success, $1.6 billion in organic revenue, forty-plus years of compounding growth, and a customer base so loyal it constitutes a genuine competitive advantage, is precisely what MRI’s private equity partners aspire to replicate. They want the valuation that comes with that kind of entrenched market position. They want the recurring revenue, the switching costs, and the enterprise relationships that justify a $10 billion price tag.

But they are pursuing it by systematically destroying the very things that produced it.

Yardi’s advantage did not come from its codebase. It came from the people who built it, maintained it, and sat across the table from property managers in Cincinnati and commercial landlords in Dallas and affordable housing operators in Phoenix and learned, over years, exactly what those customers needed. That knowledge lives in people. It is not transferable to a spreadsheet. It does not survive a layoff notice.

When you cut a third of your domestic workforce and offshore the institutional knowledge that justified your valuation, you are not optimizing an asset. What your actually doing is liquidating it: just slowly enough that the buyer may not discover the damage until after the deal closes. The PE partners will be long gone with their carried interest by the time the customer attrition begins showing up in the renewal numbers. The incentive horizon is catastrophically misaligned with the long-term health of the business. And the people who built that business will be the ones who absorb the consequences.

There is a word for this in economic theory. Joseph Schumpeter called it “creative destruction“—the process by which capitalism tears down the old to build the new. But what is happening at MRI, and at Salesforce, and across the American technology sector is not creative destruction. There is nothing being created. There is only extraction: of margin, of institutional knowledge, and of human capital accumulated over careers, in service of a transaction that enriches a small number of partners and leaves the enterprise weaker than they found it.

The Political Failure

Now consider the political moment in which all of this is occurring.

Donald Trump won the presidency in significant part on the promise that the economic displacement of the American worker — the dislocation that politicians of both parties had acknowledged and ignored for thirty years — would finally be addressed. On manufacturing, on tariffs, on trade policy, there is at least a coherent argument being made and a set of policies being attempted, however imperfectly and however disruptively.

But on this, on the slow, deniable, spreadsheet-mediated offshoring of the white-collar professional and salaried class, the Trump administration has no leverage, no policy mechanism, and no answer so far.

The H-1B visa program that provides the legal infrastructure for much of this offshore pipeline has some powerful champions both inside and outside the administration. The fiduciary and legal structures that compel PE-owned companies to optimize margins ahead of an exit are entirely untouched by any policy being contemplated in Washington. The regulatory framework that might give the government standing to identify, document, and intervene in a managed knowledge transfer disguised as professional collaboration simply does not exist. And the political constituency most affected, the white-collar professional class, the people whose displacement arrives as a calendar invite from HR rather than a plant closure announcement, does not yet fully understand what is being done to it or how to demand accountability for it.

This is a story of a thirty-year failure of both parties to reckon with the financialization of American business and its consequences for American workers at every level of the income distribution. The factory workers who lost their jobs to offshoring in the 1990s and 2000s were told they needed to get more education and move into the knowledge economy. The knowledge economy workers who are losing their jobs to offshoring now are being told what, exactly? To learn to code? The coders are going too.

What It Means

Anant Yardi built a $1.6 billion company by doing something unfashionable: he treated his American workforce as the source of his competitive advantage rather than the primary target for cost reduction. He was right, and his company’s four decades of organic growth prove it.

The private equity owners of MRI Software looked at a company with a genuine market position, real customer loyalty, and a talented American workforce and saw something different: a margin improvement opportunity and an exit multiple. They may get their $10 billion. But they will leave behind a company with a hollowed-out workforce, degraded institutional knowledge, and customers who are quietly beginning to wonder whether Yardi might be worth the switching cost after all.

Pat Ghilani told his workers, not long ago, that they were the people who made MRI’s success happen. That it was all because of them. He meant it, I believe. And now he is letting private equity fire them, one managed knowledge transfer at a time, while the workers who remain sit at their desks feeling, in the words of one of their colleagues, like prisoners waiting to be called before the firing squad.

This is what American capitalism looks like in 2026 for the white-collar professional class. Not a dramatic collapse. Not a visible injustice that photographs well and moves senators. Just a quiet, methodical, deniable process of extraction, disguised as collaboration, executed as displacement, that will not appear in any political speech and will not be addressed by any policy currently under consideration in Washington.

Winter is coming for the salaried class. The question is when will they begin to fight back?

That exchange reflects a broader frustration that has become difficult to ignore. For many recent graduates, especially in fields like computer science, the entry-level job market has tightened noticeably. Hiring has slowed, competition has intensified, and roles that once served as a reliable starting point now feel far less accessible.

In Washington, officials with power point to Congress, cite gridlock, and call for new legislation. Yet in this case, that framing obscures more than it clarifies. The executive branch already exercises substantial control over several of the programs shaping this part of the labor market. The question is not whether the government can act, but whether it is willing to use the authority it already has.

The debate over H-1B visas goes to the heart of how companies use guest worker visa programs to legally displace American workers. At the same time, focusing exclusively on H-1B risks overlooking a broader and more consequential shift. Over the past two decades, a set of programs, many created or expanded through regulation rather than statute, has come to play an increasingly central role in entry-level hiring.

The OPT Loophole

The Optional Practical Training program, known as OPT, offers the clearest example. Originally designed as a limited training period for foreign students closely tied to their field of study, OPT allows up to 12 months of work after graduation. Think of it as an internship, meant to let foreign graduates gain hands-on experience before returning home or transitioning to a different visa.

That baseline is what makes the subsequent expansion so significant. In 2008, the George W. Bush administration introduced a STEM extension through regulation alone, allowing graduates in science, technology, engineering, and math fields to remain in the workforce beyond that initial year. The Obama administration later expanded that extension further, ultimately allowing up to three years of work authorization.

What began as a one-year training benefit now functions as a multi-year work visa in all but name.

By 2024, more than 289,938 foreign graduates were granted permission to find work through OPT and its STEM extension. Since 2015, annual approvals under OPT, especially STEM OPT, have continued to accelerate without drawing nearly all the political attention. Even a substantial overhaul of H-1B would leave this parallel system largely untouched.

What makes this especially significant is not just scale but origin. Because the STEM OPT extension was created through regulation, not by Congress, it can be modified the same way: by the executive branch, without a single vote on Capitol Hill. Unlike H-1B, OPT is not subject to an annual cap. Employers can hire from a large pool of foreign graduates outside the numerical limits Congress set for temporary work visas. That makes it squarely an executive problem, but also an executive opportunity.

If this administration is serious about the effect of guest worker programs on entry-level employment, it is difficult to explain why attention remains fixed on Congress while the largest of these programs goes largely unaddressed.

The Spouse Loophole

A similar pattern appears in the work authorization granted to certain spouses of H-1B visa holders, known as H-4 EAD (Employment Authorization Documents) issued to H-4 dependent visa holders. In 2025, the Institute for Sound Public Policy estimates 150,000 H-4 EADs working in the country. The Obama administration created this benefit through regulation in response to severe green card backlogs, particularly for Indian nationals, who face the longest wait times in the employment-based immigration system.

In practice, the program has grown well beyond its original rationale. The benefit is overwhelmingly concentrated among Indian nationals, who account for roughly 90 percent of recipients. Unlike H-1B workers, these spouses are not tied to a specific employer and are not subject to comparable wage or labor protections. Congress never authorized this open-ended labor market access. It was created by regulation and can be rescinded by regulation.

The program has also begun to shape behavior within the green card process itself. Because spousal work authorization kicks in at certain stages of the sponsorship process, it can create incentives to pursue employer sponsorship even where the underlying job relationship is weak or largely transactional — pursued not for the job itself, but for the path it opens toward permanent residence. As more applicants enter the pipeline under these conditions, pressure builds on an already backlogged system and makes it harder to distinguish genuine employment relationships from those driven primarily by immigration incentives.

A System More Permissive Than It Appears

Taken together, H-1B, OPT, STEM OPT, H-4 spousal work authorization, and the Curricular Practical Training (CPT) program, which allows foreign students to work off-campus as part of their studies—these programs form a system that is far more expansive than most Americans realize. They are also riddled with fraud.

The good news is that the administration already has the authority to act on all of them.

On H-1B, the Department of Homeland Security (DHS) can tighten the definition of a “specialty occupation.” In plain terms: if a job does not truly require a specific degree in a specific field, it should not qualify. DHS can also require employers to demonstrate that they — not a staffing intermediary or outside client — actually control the worker’s duties, supervision, and worksite.

This approach has precedent. In 2010, U.S. Citizenship and Immigration Services issued what became known as the Neufeld Memo, which required employers to prove a real employer-employee relationship, particularly when placing workers at third-party client sites. The target was a widespread staffing model in which firms sponsor H-1B workers on paper, then lease them out to end clients who direct their work, set their hours, and control their assignments. That arrangement allows companies to access H-1B labor without taking on the legal responsibilities of being the employer of record and makes it nearly impossible to enforce wage and working condition protections. The Neufeld Memo made that model far harder to sustain. Legal challenges largely failed. Industry pressure eventually got it watered down and then shelved. Reinstating it through formal rulemaking would strike at the heart of how the H-1B program is most commonly abused.

On OPT and STEM OPT, DHS can return the program to its original purpose. The most defensible step would be to eliminate the STEM extension, the expansion from one year to three, which rests on a thinner regulatory foundation and would be harder to challenge in court. Restoring OPT to its original 12-month limit would bring the program back in line with its stated training purpose while significantly reducing its labor market footprint.

On H-4 EAD, DHS can rescind the regulation that created spousal work authorization. H-4 spouses could continue living lawfully in the United States as dependents, but without open-ended access to the labor market. A prospective phaseout, letting existing permits expire while ending renewals, would reduce disruption and strengthen the rule’s legal standing.

On CPT, DHS can clarify that employment must be a genuine component of an academic program, not a workaround for extending work authorization indefinitely. It can also increase oversight of schools that appear to use enrollment primarily as a vehicle for work permits rather than education.

Across all of these programs, DHS can strengthen enforcement through more frequent site visits, better data sharing with the Department of Labor, and closer scrutiny of high-volume employers and universities.

The Lesson From the First Term

None of these ideas requires new legislation. But it does require disciplined rulemaking.

Both the Obama and Biden administrations showed how to use the administrative process to advance immigration policy while building a legal record capable of withstanding judicial review. The Trump administration’s first-term effort to end Deferred Action for Childhood Arrivals (DACA), for example, was blocked by the courts because the agency’s reasoning was found to be arbitrary and capricious under the Administrative Procedure Act (APA), which is the law that governs how federal agencies make rules.

If this administration intends to continue pursuing a more restrictive approach to work authorization, it would be well served to follow that model: build the evidence, follow the procedure, and make the decision defensible.

For many young Americans entering the workforce today, the stakes are already clear. The tools exist. The question is whether they will be used.

If that makes this piece a lightning rod, then so be it.

The Problem Beneath the Problem

We tend to discuss the absence of critical thinking as an intellectual failure. We believe people don’t know how to evaluate evidence being presented. They defer too readily to credentialed experts. They mistake the complicated problems that have solutions with the complex problems that require ongoing navigation. All of that is true as far as it goes.

But it doesn’t go far enough.

The deeper failure is emotional. And until we reckon with that, the intellectual remedies such as teaching logic, distrusting authority, and asking better questions will continue to bounce off the problem without penetrating it.

Human beings are tribal animals at a level that precedes cognition. The terror of being cast out of the group, of being marked as the person who doesn’t accept what everyone else accepts, is not a rational calculation. It’s closer to a survival instinct. For most of human history, exile from the group was a death sentence. That ancient wiring does not disappear because we live in suburbs and consume our consensus through smartphones.

In his book, The Coming of Neo-Feudalism: A Warning to the Global Middle Class, Joel Kotkin coined the term Clerisy to describe the credentialed class of academics, journalists, nonprofit administrators, and credentialed commentators who legitimatize elite consensus. In the vein of the notion that “modern technology creates modern monsters,” the Clerisy has become extraordinarily sophisticated at weaponizing their fear of shaming. Social shaming is driven less by intellectual judgment than by emotional impulse. The labels, the credential signaling, the manufactured consensus are designed to trigger ancient fear before the mind has a chance to engage.

Consider how the argument for NAFTA was actually delivered to ordinary Americans in the 1990s. The intellectual case was it would enable us to maximize comparative advantage, efficiency gains, aggregate growth was almost secondary. Whereas the primary message was tribal: sophisticated, educated, cosmopolitan people support free trade. If you oppose it, you are revealing yourself to be unsophisticated, parochial, probably fearful of change. You are not one of us. Those of us who opposed NAFTA and supported Ross Perot were mocked as rubes who didn’t understand how trade worked. However, we were correct and we were correct about the substance decades before the credentialed consensus admitted it. The tragedy being we paid a social cost for being correct while the people who mocked us paid nothing.

That asymmetry where being right costs you and being wrong costs the credentialed class nothing drives the problem. Economists call this a perverse incentive structure and our grandfathers called this a racket.

Critical thinking, genuinely practiced, therefore requires not just intellectual tools but emotional ones. Specifically, it requires a cultivated indifference to the approval of people whose approval we have been trained to seek. That indifference, however, is not natural. It is not comfortable. It has to be built, usually through experience that makes the cost of conformity viscerally and personally clear.

Mental acuity must be teamed with mental toughness. Some people just get smart. Some people just get tough. Not enough get both.

What We Lost

The generations who built this country operated in a brutal but effective epistemological environment. Bad judgment produced immediate and concrete consequences. The farmer sizing up a land deal, the union man reading the company’s wage offer, the Coxey’s Army marcher asking who actually benefited from the gold standard while he went hungry — these were people for whom signaling was a luxury they could not afford. They needed to think because the cost of not thinking was immediate and personal.

What replaced that environment was modernity’s freedom from consequences caused in part by social safety nets and credential insulation. We became a very wealthy country off the backs of past generations that created a store of virtue that today is being drained faster than it is being replenished.

The mass credentialing of the twentieth century promised to substitute hands on expertise for certified competence. What it delivered, at scale, was a class of people whose livelihoods depend on believing the right things — and who are therefore structurally protected from the consequences of being wrong.

Nassim Taleb named this figure the Intellectual Yet Idiot (IYI.) Credentialed, consensus-affirming, fluent in the correct vocabulary, yet catastrophically wrong about anything requiring genuine skin in the game — trade policy, financial risk, public health, labor economics. The IYI’s defining characteristic is not stupidity but insulation. He is never in the room where the consequences land. Someone else absorbs those.

Jane Jacobs, in her underappreciated work on civilizational decline, identified credentialization itself as a symptom of stagnation. Healthy, dynamic societies transmit practical knowledge through doing — through apprenticeship, craft, trial and error, and the feedback loops that punish bad judgment swiftly. When credentials begin substituting for demonstrated competence, it signals that genuine knowledge transfer has already broken down. The certificate fills the void left by the collapse of the real thing.

We are living in that void. And the people most insulated from its consequences are the ones most confidently telling us there is nothing to worry about.

The Left Case Study: What the SPLC Taught Us

The Southern Poverty Law Center is the cleanest available case study in what happens when signaling substitutes for thinking at scale — and what it costs the people who trusted in that institution.

Founded in 1971 with a genuine civil rights mission, the SPLC discovered somewhere along the way that fear is more reliable than justice as a fundraising instrument. The formula was elegant in its cynicism. Identify threats. Amplify them. Monetize the anxiety of people who wanted to believe they were fighting hatred from the safety of a checkbook. Muckracker journalists were raising alarms by the late 1990s. Former employees described internal cultures that contradicted the organization’s stated mission. Financial disclosures showed hundreds of millions accumulating while the urgent direct mail kept arriving.

Thousands of decent, educated, genuinely motivated Americans kept writing checks.

They were not stupid. What they lacked was the habit of asking cui bono — not from the hatred the SPLC claimed to fight, but from the fear of it remaining permanently present. Not from solutions, but from the problem staying perpetually, profitably unsolved.

The signal was clean and compelling: this organization fights hatred, therefore supporting it makes you a person who fights hatred. That felt true. It felt good. It substituted for the harder question — is this organization actually fighting hatred effectively, or is it farming your moral anxiety for institutional profit?

The recent revelations have confirmed what the skeptics argued for decades. The gap between the signal and the reality was vast. And the people who paid for that gap were the donors who trusted, not the executives who collected.

Signaling is not thinking. And someone always pays.

The Right Case Study: The Capture You’re Not Supposed to Notice

The same machinery operates on the right. It wears different clothes. It speaks a different language. But the underlying mechanism — emotional capture through tribal signaling, followed by the extraction of trust, energy, and money — is identical.

The America First movement has achieved something genuinely significant. It cracked a credentialing consensus that had gone unchallenged for decades. It made trade skepticism, immigration restriction as labor policy, and hostility to endless foreign military adventures not merely sayable but politically dominant. It gave millions of ordinary Americans permission to name what had been done to them, in language the credentialed class had worked hard to make unspeakable. That is a real achievement against enormous institutional resistance, and it deserves honest acknowledgment.

But genuine movements are always targets for capture. The institutional right has decades of experience converting populist energy into votes while delivering for the donor class. The Tea Party is the recent cautionary tale: a genuine grassroots insurgency substantially absorbed into conventional Republican donor politics within a few election cycles, its energy harvested and its structural demands quietly shelved.

The capture of America First energy by Silicon Valley money represents a more sophisticated version of the same operation. Peter Thiel understood before most that the old donor class such as the Chamber of Commerce establishment and the defense contractor network was losing legitimacy with the populist base. Rather than defend that establishment, he moved early to position his network as the insurgency while pursuing fundamentally similar objectives that included access, contracts, favorable regulation, and a labor market kept artificially loose by visa pipelines that suppress wages for American workers.

Palantir’s deliberate embedding into the new conservative movement is not America First. What it is is surveillance capitalism wearing a nationalist costume. Their core business is data integration and intelligence infrastructure for governments. They don’t want a more limited government — they simply want to be the ones running the pipes. The H-1B policy dilution under donor pressure early in the current administration signaled whose interests are actually being protected.

Ask cui bono. Not what does this person say. Not what does this person perform. What structures and interests are actually being protected by their actions over time.

Steve Bannon could articulate the critique of the managerial class more clearly than almost anyone in American political life. He had read Samuel Francis. He understood James Burnham. He could name the disease with precision. He was also, we now know, quietly working to rehabilitate Jeffrey Epstein’s public image while publicly performing populist outrage. The gap between the signal and the reality is not a left-wing phenomenon. It is a human one.

This is not a reason for despair. It is not a reason to conclude that the movement is corrupt and nothing matters. It is a reason to apply the same question your grandparents would have asked immediately, without instruction, as a matter of survival: who benefits from my enthusiasm, my trust, my donated energy, my suspended judgment?

The Standard

The answer to capture can not and should not be paralysis. It is not the cynical conclusion that all movements are equally compromised and the honest citizen should simply withdraw. That conclusion serves the people doing the capturing. Disengaging is basically surrender with better aesthetics.

The answer should be having a standard, applied consistently, without tribal exception.

Ed Dowd, the former BlackRock fund manager whose actuarial analysis of excess mortality and disability data has made him one of the more significant heterodox voices of the past several years, has demonstrated what that standard looks like in practice. On the occasions when his analysis of financial topics was wrong, he said so publicly, explained his reasoning, and accepted the reputational cost of the admission. That is rarer than it should be, and it matters more than it might appear. A person who can admit error publicly has demonstrated that their relationship to truth is more important than their relationship to their own image. That is a meaningful signal.

The Whitefield principle is useful here. When Benjamin Franklin confronted the Reverend George Whitefield about the contradictions between his preaching and his conduct that included his use of slave labor to build an orphanage in Georgia, Whitefield responded with something approaching genuine humility: do not judge the teachings of Christ by a failed human like myself. It was an honest answer. It points toward the only workable standard available to us: judge the ideas on their merits, judge the person on their actions over time, and hold both assessments simultaneously without letting either collapse into the other.

A movement that can ask hard questions of itself is more trustworthy than one that cannot. The willingness to be a lightning rod which is to say true things that cost you something among people whose approval you value should not be taken as disloyalty to the cause. It is the highest form of investment in what the cause claims to be fighting for. Movements that cannot tolerate internal scrutiny do not get reformed. They get captured, and then they get harvested.

Apply cui bono before you donate. Before you march. Before you repeat the talking point. Before you suppress the doubt that flickered for a moment before the tribal instinct closed it down. Apply it to the people asking for your trust on the left. Apply it equally to the people asking for your trust on the right.

The Leash

The conformity instinct is a leash. It is not worn around the neck where it can be seen and removed. It is worn around the imagination. It operates before the thought is fully formed, suppressing heterodoxy not through argument but through the ancient fear of exile — the whispered warning that thinking this, saying this, will mark you as the kind of person who thinks and says things like that. In short, not a good fella.

Most people never consciously decide to wear it. It is fitted early, tightened gradually, and eventually becomes indistinguishable from the shape of one’s own thinking. The manufactured consensus feels like common sense. The approved narrative feels like reality. The boundaries of acceptable thought feel like the boundaries of thought itself.

Slipping that leash requires an act of will. It’s forged through experience and through the personal discovery that conformity’s promised protection is a lie; that the approval you performed for does not actually keep you safe and that the cost of being wrong alongside everyone else is paid in the same currency as the cost of being right alone.

Freedom from consequences made forgetting possible. The frontier generated feedback. Bad judgment had a price that was immediate, personal, and non-negotiable. That price was brutal, but it was honest, and it produced generations of people who asked cui bono not as a rhetorical flourish but as a survival reflex.

We have insulated ourselves from that feedback so thoroughly that the reflex has atrophied in most people. The credentialed class promises safety. The approved narrative promises belonging. The signal promises identity without the labor of thought.

The critical thinker is not the person with the highest credentials or the most sophisticated vocabulary. It is the person who has consciously decided, at some cost, that being right matters more than being approved of. Who has looked at the leash, recognized it for what it is, and chosen not to wear it.

That person is rarer than they should be. And more necessary than they have ever been.

I have been waiting a long time to write this piece.

Not because I anticipated a criminal indictment, I did not. But because the harm the SPLC did to our work, to our donors, and to the broader cause of honest immigration policy debate has been real, documented, and largely ignored by the mainstream press. The indictment changes that.

A Sixteen-Year Campaign

In July 2010, just two years after our founding, the SPLC published a formal report titled “Greenwash: Nativists, Environmentalism and the Hypocrisy of Hate.” Written by Heidi Beirich, then the SPLC’s chief architect of its hate group designation machinery, the report named our organization, then called Progressives for Immigration Reform (PFIR) as the centerpiece of what it called a coordinated “greenwashing” campaign: a cynical effort by nativist activists to seduce environmentalists to join their cause for purely strategic reasons.”

The report was comprehensive. With the exception of yours truly who was the Treasurer, it named our board members, documented our funding sources, questioned our progressive credentials, and placed us in a glossary alongside organizations the SPLC had formally designated as hate groups. It created a false foundational narrative claiming we are a nativist front group, a narrative that every subsequent journalist, advocacy organization, and foundation ran with and amplified over the next decade and a half.

Looking back on things, what is notable is the timing. Beirich published her report in 2010. The SPLC’s alleged secret payments to KKK-affiliated individuals began, according to the federal indictment, in 2014. For at least nine of the sixteen years during which the SPLC’s “greenwash” narrative was being used to suppress our work and stigmatize our donors, the SPLC was allegedly funneling donated money to the leadership of actual hate groups.

The article represented a poison pill that seeped its way into every article, every donor pressure campaign, and every grant screening decision that relied on SPLC characterizations of our organization must now be understood.

What the SPLC Did to Us

In January 2020, the SPLC’s Hatewatch unit published an article describing U.S. Tech Workers as “a project of the anti-immigrant group Progressives for Immigration Reform.” The article named me by name, linked my organizations to a network of “anti-immigrant hate groups,” and included an organizational chart connecting our work to VDARE and FAIR.

The article was not a good-faith analysis of our policy positions. It was a targeting document. And it worked as one.

Within months, that SPLC characterization had been picked up by The Daily Beast, Media Matters, and a roster of advocacy organizations. Each amplified the label. Each treated the SPLC’s designation as authoritative. None investigated whether it was accurate.

Our actual work that included advocating for American tech workers harmed by employment visas such as H-1B, drawing attention to the plight of unmatched American medical graduates, and pushing for evidence-based immigration policy — was rendered invisible. What the press saw was an SPLC-certified “anti-immigrant” group. That’s all they needed to see. If I had a nickle for number of times I have had to defend myself from these baseless accusations.

What They Did to Our Donors

The SPLC’s designation machinery was not only used against organizations. It was used against individuals who gave money to those organizations.

Case in point is that of Fred Stanback, a North Carolina conservationist, Warren Buffett’s best man, and one of his state’s most respected environmental philanthropists. Fred donated to our organization as part of a broader commitment to population and environmental sustainability. His help was instrumental in our creating the first ever impact study on the impact of immigration on the environment. His donations were routed through donor-advised funds at the Foundation for the Carolinas.

When advocacy groups identified that connection, they launched a campaign against the Foundation for the Carolinas, pressuring it to stop passing Stanback’s funds to us and related organizations. Some community foundations began screening grants against the SPLC’s hate group list as a matter of policy. Stanback’s Duke internship program, which had placed students at dozens of legitimate environmental organizations for decades was forced to remove us and associated groups after SPLC-linked reporting created pressure on the university.

The SPLC’s designation was a financial weapon. It was used to cut off legitimate philanthropy and to stigmatize donors who held views the SPLC opposed. Oddly, all of this, it now appears, was done by an organization that was simultaneously paying the leadership of the Ku Klux Klan.

The Mechanism of Reputational Destruction

It is worth understanding how this works, because the pattern did not originate with us and will not end with us.

The SPLC publishes a “hate map” and issues annual reports designating organizations as hate groups or “anti-immigrant” groups. Journalists treat these designations as authoritative. Advocacy organizations cite them. Foundations use them as screening tools. Social media platforms have used them to justify deplatforming decisions.

Once an organization is designated, the designation takes on a life of its own. It is cited in news articles, which are cited in subsequent news articles. It migrates into Wikipedia entries. It appears in grant application screenings. It becomes the lens through which every future action by the designated organization is interpreted.

The designated organization has no meaningful recourse. There is no appeal process. No evidentiary standard. No requirement that the SPLC demonstrate actual hate rather than policy disagreement. The label simply gloms on and the press amplifies.

We experienced this firsthand. Even though our first Executive Director was and an American descendant of slaves (ADOS) and at one point our board was 90% ADOS, our work advocating for displaced American workers was repackaged as white nationalism. Our donors were targeted. Our media reach was suppressed. And the organization responsible for that campaign was, we now know, engaged in conduct that included payments to actual hate groups.

What the Indictment Means

The federal indictment does not automatically vindicate every organization the SPLC has ever designated. The legal case will take time to resolve, and the SPLC has pledged to fight the charges.

But the indictment does several things that matter enormously to organizations in our position.

It establishes that the SPLC had a documented pattern of concealment and deception regarding how it operated its flagship mission of “fighting hate.” It demonstrates that the SPLC’s opinion as to which organizations constitute hate groups is no longer authoritative as it was built on an institutional foundation that was secretly in business with the groups it claimed to oppose.

Every journalist who cited the SPLC’s characterization of us as “anti-immigrant” was relying on an institution now accused by the United States Department of Justice of systematic fraud. Every foundation that used the SPLC list as a donor screening tool was using a document produced by an organization allegedly engaged in money laundering.

That context does not disappear because the SPLC says the payments were for informants. It raises, rather than resolves, fundamental questions about the credibility of an institution that positioned itself as the national arbiter as to who is and is not a hate group.

Reflections of Our Chairman

Our Chairman, Steven Lamb had the following to add regarding the challenges the SPLC created for our organization and why regardless of their assaults, our authenticity continues to shine through:

“When I joined the Board of Progressives for Immigration Reform, the entire Board except Ben Zuckerman was African American. They were all well known activists. In fact when I met the Board I was deeply honored to meet both it’s then Chair Dr. Frank Morris, whose work as the economist for the Black Caucus Foundation I had been aware of for thirty years, and Yvette Carnell whose work I had just discovered and who was noting that Black Americans descended from Slaves have a different experience and treatment in the American economy than Black immigrants.

At my interview, Dr. Morris informed me that the SPLC had labeled our organization a “Hate Group” and asked if that would concern me. I told Dr. Morris that I had once been a donor to the SPLC, but that something was wrong with that organization. People who I knew on the right and the left were getting the “Hate Group” designation who had strong policy views that were not based in race and some of them actually were totally race ignorant. We had a good laugh.

I’ve been on our board for a decade. I’ve become the Chair. I can tell you being on the SPLC’s anti-immigrant list hurts us both in fund raising and recruitment, especially since we are a group of anti racists. The SPLC has done incredible harm to American discourse.”

What We Are Doing

We are compiling a full record of every citation, article, and institutional decision that relied on the SPLC’s characterization of our organizations. We are documenting the concrete harm — to our funding, to our media presence, and to our donors that resulted from a designation made by an organization now under federal indictment.

We will be reaching out to journalists who wrote about us in the past. Some may now be willing to revisit their reporting. We welcome that conversation.

We are also consulting with legal counsel about the options available to organizations and individuals who were damaged by the SPLC’s designation machinery. We will share what we learn.

And we will keep doing our work. Because the displacement of American workers by employment visa policy is real, the crisis facing unmatched American medical graduates is real, the negative consequences of unbridled immigration are real and the need for honest, evidence-based immigration policy debate is more urgent than ever.

American medicine has its own breakage number. It is 6.5%.

Each year, roughly 6.5% of American MD seniors who enter the National Resident Matching Program — The Match — fail to obtain a residency position. This year it was 6.5%. The year before, 6.5%. The stability of that figure is not a coincidence of nature. It is the signature of a system operating exactly as its managers intend.

These are not marginal students. They are men and women who completed four years of medical school, passed licensing examinations, accumulated $200,000 or more in debt, and sacrificed the prime earning years of their twenties in preparation for a career in medicine. When the Match results are released on Match Day, they discover that the system has no place for them. The algorithm, we are told, has spoken.

Except the algorithm didn’t make this decision. The people who built and maintain the system did.

What the Algorithm Hides

The Match is administered by the National Resident Matching Program and uses a Nobel Prize-winning algorithm to pair medical graduates with residency programs. Its mathematical pedigree confers an aura of objectivity that serves a very specific institutional purpose: it makes policy outcomes look like natural results.

But the algorithm only processes what program directors submit on their rank order lists. The real decisions — who gets interviewed, who gets ranked, who gets a position — happen entirely upstream of any computation. By the time the algorithm runs, the outcome is largely determined by human choices made inside teaching hospitals and academic medical centers - choices that are subject to no meaningful public scrutiny.

What those choices reveal, when examined at the program level, is striking. Across social media, physicians and medical students have begun documenting cohort compositions at individual residency programs — and the patterns are not what a neutral, merit-based process would produce. Programs where the overwhelming majority of residents — eight, nine, ten out of ten — share the same country of origin, whether Nepal, Pakistan, or India. The statistical probability of such clustering emerging from an unbiased selection process approaches zero. Clustering at that level is not coincidence. It is policy.

This is the tribal dimension of the Match that its institutional defenders prefer not to discuss. Program directors who trained abroad, or who maintain active networks within foreign medical communities, systematically favor applicants from those networks. In some cases the mechanisms may be subtler — cultural affinity, shared language, established pipelines with particular overseas institutions. In others, the relationships between program directors and foreign recruitment consultants warrant harder scrutiny. Either way, American-trained physicians are being displaced not by superior merit but by insider preference operating behind the shield of algorithmic complexity.

The Managerial Class That Runs American Medicine

To understand why this persists, it helps to understand who benefits from the status quo.

The institutions that govern American medicine — the AMA, the AAMC, the ACGME, the major academic medical centers — are not elected, not meaningfully accountable to patients or to the public, and have constructed a self-reinforcing system of control over the physician workforce pipeline. They set the standards for medical education. They accredit residency programs. They control the number of residency slots, which grow only incrementally regardless of workforce need. And they administer the Match itself.

These institutions share an interest in physician scarcity — it maintains prestige, controls costs through monopsony labor market dynamics, and ensures that the institutions themselves remain the irreplaceable gatekeepers of medical careers. A 6.5% unmatched rate among MD seniors is not a crisis to this managerial class. It is a feature. It demonstrates selectivity. It reinforces the message that medicine is not for everyone, even for those who completed its most demanding preparatory requirements.

Medical schools, notably, collect full tuition regardless of match outcome. The student who fails to match after four years and $250,000 in debt represents no financial loss to the institution that trained them. The incentive to ensure match success is limited to reputation, and reputation is managed through selective reporting and statistical presentation.

The Hospitalist System and the Disposable Physician

The transformation of American hospital medicine through the hospitalist model is rarely discussed in the context of workforce policy, but it should be. The shift away from attending physicians who follow their own patients through hospitalization toward shift-based hospitalists — physicians who hand off patients at the end of each shift — was sold as an efficiency and quality improvement. It has also proven remarkably convenient for institutions managing a rotating international workforce.

Longitudinal care — the physician who knows a patient’s history, their family, their preferences, the nuances of their chronic conditions — requires continuity. It requires language fluency not merely in English but in the cultural idioms through which patients describe pain, fear, and uncertainty. It requires trust built over time between physician and patient. The hospitalist model systematically eliminates these requirements. Any physician, from any training background, who can manage a shift is interchangeable with any other.

This is not incidental to the workforce dynamics described above. It is complementary to them. A system designed around interchangeable shift workers is a system designed to accommodate — and conceal — the mass substitution of American-trained physicians with lower-cost foreign labor.

The Silence of the Unmatched

Perhaps the most effective protection the system enjoys is the silence of its victims.

Medical training selects relentlessly for individual perseverance and self-reliance. It rewards those who internalize failure as personal inadequacy and respond with harder work. When a physician fails to match, everything in their training and culture tells them the failure is theirs — their scores weren’t high enough, their interviews weren’t strong enough, their rank list wasn’t strategic enough. The system is meritocratic. The algorithm is objective. Therefore the outcome must reflect personal deficiency.

This internalized shame is not accidental. A system that looks meritocratic — standardized tests, objective criteria, a Nobel Prize-winning algorithm — creates the conditions for victims to blame themselves rather than the institutions that failed them. Shame silences. People who are ashamed do not organize, do not testify, do not become plaintiffs. The psychological architecture of the Match is doing institutional protective work.

Those of us who have followed the unmatched physician community know that reframing this experience — from personal failure to systemic harm — is one of the most important and difficult tasks in this space.

What Accountability Would Look Like

The good news is that the accountability mechanisms that could address this do not require new legal authority or radical policy innovation. They require transparency and enforcement of obligations that already exist.

First, consider what is actually at stake financially. Medicare alone paid an estimated $21.2 billion in fiscal year 2023 to teaching hospitals for Graduate Medical Education. When Medicaid, the VA, HRSA, and the Department of Defense are added, total federal GME spending reaches approximately $20 to $25 billion annually — and the Congressional Budget Office projects it will grow at 7 percent per year through 2034, roughly five times the rate of general inflation. This is one of the largest and least scrutinized public subsidies in American healthcare.

What does the public get in return? Remarkably little that is enforceable. Physicians who train in Medicare or Medicaid-supported residencies are under no obligation to accept Medicare or Medicaid patients when they enter practice. There is no requirement to serve the communities whose tax dollars funded their training. And there is no transparency requirement about who fills those federally subsidized slots — no public reporting on cohort composition, no disclosure of program director conflicts of interest, no audit mechanism to determine whether American-trained graduates are being systematically passed over in favor of foreign recruits.

Teaching hospitals receiving Medicare Graduate Medical Education funding are federally funded institutions. That funding comes with conditions — or it should. Expanding existing cost reporting requirements to include program-level cohort composition data — country of medical training for each resident — is a modest regulatory ask that would immediately illuminate the patterns currently hidden in aggregate statistics.

Requiring conflict of interest disclosure from program directors — undisclosed financial or personal relationships with foreign recruitment consultants or overseas medical institutions — would expose the corruption currently operating in plain sight.

Audit requirements tied to Medicare GME funding, focused on whether programs are meeting their obligations to consider American-trained graduates, would create accountability where none currently exists.

None of these proposals preference foreign physicians or domestic physicians on the basis of national origin. They simply demand that a federally funded system be transparent about what it is doing with public money. That is an argument that should be unassailable across the political spectrum — from open border advocates on the left to America First proponents on the right to the Make America Healthy Again movement that has made confronting captured medical institutions a central concern.

American medicine has a breakage number. It is 6.5%, it is stable, and it is a choice.

The unmatched physicians who bear that cost in debt, in lost years, in abandoned vocations deserve to know that the system did not fail them randomly. It failed them by DESIGN. And the institutions responsible for that design should be required to answer for it in public, with their funding on the line.

There’s a book most of the people running Washington, the universities, and the corporate boardrooms would prefer you never read. It was written in 1995 by a political analyst named Samuel T. Francis, suppressed in his lifetime, and only published after his death. It’s called Leviathan and Its Enemies and it explains, with cold precision, exactly what has been done to this country and to the people who built it.

Francis wasn’t a pundit. He was a scholar, and he did something scholars almost never do: he told the truth about power. Not the polite truth, not the approved truth, but the structural truth and that is the truth about who runs America, how they took control, and why every movement that has tried to take it back has so far failed.

His argument is this: beginning in the early twentieth century, a new ruling class quietly displaced the old America of small businesses, local communities, family farms, and self-governing citizens. This class didn’t seize power through revolution. It did something more insidious. It built and then captured the massive institutions of modern life: the federal bureaucracy, the corporate HR departments, the universities, the media, the nonprofit foundations, the regulatory agencies. Francis called them the managerial elite, and he showed that what unites them isn’t party affiliation — it’s their shared interest in keeping those institutions large, powerful, and beyond the reach of ordinary Americans.

Here’s what makes Francis essential reading right now: he understood that this class rules not primarily through force but through ideology. The language of diversity, equity, inclusion, expert consensus, and settled science - all of it serves a function. It’s not primarily about what it claims to be about. It’s about delegitimizing anyone who challenges the institutions these managers control. Call someone a bigot, a conspiracy theorist, or an anti-science extremist, and you don’t have to answer their actual argument. The ideology is a weapon, and it has been wielded with extraordinary effectiveness against exactly the kinds of Americans who voted for Donald Trump and who are now fighting for medical freedom under the MAHA banner.

Given his book was written in 1995, Francis witnessed the populist third party run for president of Texas billionaire Ross Perot and served as an advisor to presidential candidate Pat Buchannan during which time he no doubt saw the first glimmers of a movement like the one we’re living through. He called it the post-bourgeois resistance: working people, forgotten communities, the men and women who build things and fix things and raise families, who had been systematically stripped of political representation, cultural dignity, and economic security by a ruling class that viewed them with contempt. He said that movement would eventually find a voice. It did.

But he also issued a warning that every MAGA and MAHA activist needs to hear. The managerial class doesn’t need to defeat its opponents at the ballot box. It just needs to outlast them, co-opt them, and absorb their energy back into the system. It has done this before. It knows how. Winning an election is not the same as breaking the machine.

The essay that follows distills Francis’s major arguments from a five-hundred-page masterwork into a form that can be read in a single sitting. It is not light reading. Francis demands something from his reader but it is worth every page. Because you cannot fight what you don’t understand. And what Francis understood, more clearly than almost anyone else of his generation, is that the fight we are in is not about left versus right, Democrat versus Republican, or even freedom versus socialism.

It is about who controls the institutions that control everything else.

Read it. Share it. And ask yourself: what would it actually take to win?

Click here to be taken to the essay.

Wisconsin’s demographic picture is in line with the United States. Fertility rates sit at 1.6 children per woman against a current replacement level of 2.1, which is the same as the United States in 2020. No realistic immigration scenario can fully compensate for that. Meanwhile, the economic conditions that have made family formation increasingly unaffordable and unattractive for young Wisconsinites have not been addressed. As a result, the state has hemorrhaged roughly 10,000 net domestic Americans per year since 2010. Its labor force shrank by 50,000 to 60,000 workers in a single recent year due to Boomer retirements. It should be noted the Boomer generation was a byproduct of an all-encompassing World War and resulting unparalleled U.S. dominance, which is not replicable. Furthermore, Boomers were an exception because birthrates spiked after generations of decline from 7 children in 1835 to 2.1 children in 1935.

But Professor Burkham’s prescription, directing more immigration flows toward states like Wisconsin through a new visa architecture, treats a symptom while sidestepping the disease entirely. His proposal doesn’t seriously grapple with why Wisconsin has been losing its own people for decades, or with the uncomfortable reality that the guest worker programs already on the books have largely served corporate interests rather than the workers and communities they were sold as helping. Burkham himself does acknowledge the limits of a high-skill-only approach, telling the Daily Mail he worries about proposals that “prioritize high-skilled immigrants so much that I wonder who’s going to be doing the truck driving and other things like that.” Which is an uninformed statement to make to the millions of hardworking American truckers out there.

The Wound America Made in Wisconsin

Wisconsin’s demographic decline did not arrive from nowhere. It is, in substantial part, the downstream consequence of forty years of deindustrialization that hollowed out the economic foundation on which young Wisconsinites built households, formed families, and stayed.

In 1985, Wisconsin had somewhere between 630,000 and 660,000 manufacturing jobs, and manufacturing represented roughly 25 to 28 percent of total employment in the state. Today, after a 21.6 percent decline since 2000 alone, Wisconsin employs closer to 470,000 manufacturing workers. The losses were not evenly distributed across time. The decade from 2000 to 2010, when China entered the World Trade Organization and the Great Recession struck, was catastrophic, erasing the bulk of manufacturing employment that had survived the gradual erosion of the 1980s and 1990s.

These were not abstract economic statistics. They were the jobs that allowed an 18-year-old in Janesville or Kenosha or Oshkosh to walk off a high school graduation stage and into a wage sufficient to rent an apartment, form a household, and begin a life, without a college degree, without debt, and without waiting until their late twenties to achieve the economic stability that previous generations reached in their early twenties. The General Motors plant in Janesville, which closed in 2008, had employed workers who did exactly this for generations. When it closed, those workers and their children faced a choice between staying in a community with diminished economic prospects or leaving.

Many left. Wisconsin’s shift from a net in-migration state to a net out-migration state corresponds almost exactly with the acceleration of manufacturing job losses after 2000. The population decline Burkham seeks to solve with new visas is, at its root, a story about what happened when the middle-class economic pathway was systematically removed from a generation of Wisconsin workers through trade policy, corporate decision-making, guest workers, and automation.

This matters because it reframes the policy question entirely. Wisconsin does not have a population problem that immigration can fix. Wisconsin has an economic opportunity problem that drove its own people away, and that problem remains structurally unaddressed.

The Programs We Already Have — and Who They Actually Serve

Before designing new visa categories, it is worth examining in some detail how the existing high-skilled immigration programs, H-1B visas and the STEM Optional Practical Training program, actually function in practice. The gap between their stated purpose and their real-world operation is instructive.

The H-1B Program: Design vs. Reality

The H-1B visa was established by Congress in 1990 to allow U.S. employers to hire foreign workers in “specialty occupations” that Americans were working in. The annual statutory cap was set at 65,000, with 20,000 additional visas for workers holding advanced degrees from U.S. institutions.

What have been the results of the program? Consider the following pattern, documented by the Economic Policy Institute: the top 30 H-1B employers hired 34,000 new H-1B workers in 2022 while simultaneously laying off at least 85,000 workers in 2022 and the first quarter of 2023. Amazon, the single largest H-1B employer, hired 6,400 new H-1B workers in 2022 while laying off or planning to lay off 27,150 employees, more than twice the number of H-1B workers it had hired in the two preceding years combined. Meta, Google, Intel, Qualcomm…the pattern repeats across the technology sector with striking consistency.

A Bloomberg investigation revealed that from 2020 to 2023, the H-1B lottery itself was systematically gamed: staffing and outsourcing firms submitted multiple registrations for the same individuals through affiliated shell companies, dramatically increasing their selection odds. Nearly half of all H-1B visas during this period went to firms offering lower wages and less-skilled roles. The program was not selecting the best and brightest through a fair process and was instead being captured by the firms with the most sophisticated understanding of its regulatory loopholes.

Thirteen of the top 30 H-1B employers are IT outsourcing firms, companies whose entire business model involves placing foreign workers at American client sites, often replacing existing American employees in the process. A Cognizant class-action judgment in October 2024 found the company liable for intentionally discriminating against more than 2,000 non-Indian employees between 2013 and 2022, replacing them with lower-paid workers obtained through the H-1B program. What more needs to be said about the negative impacts of the H-1B visa program?

STEM OPT: The Program Congress Never Authorized
If H-1B represents a congressional program captured by corporate interests, STEM Optional Practical Training represents something more troubling: a program that was designed by corporate interests from its inception, created not by Congress but by executive-branch regulation at the explicit request of industry lobbyists, particularly Microsoft.

Here is what the program does in practice: F-1 student visa holders who complete STEM degrees at U.S. universities may work in the United States for up to three years (12 months of standard OPT plus a 24-month STEM extension) without being subject to the H-1B annual cap. There are no enforceable wage standards. Employers are not required to advertise positions to American workers before hiring OPT participants. And crucially, both the OPT worker and their employer are exempt from paying Social Security and Medicare payroll taxes, a 15 percent discount on total labor costs. It is not because the OPT worker is more productive or more skilled, but because a regulatory exemption makes them artificially cheaper. As a result, the program is an artificial government subsidy impacting entry-level American STEM positions.

According to the Institute for Progress, approximately 220,000 students participated in STEM OPT in 2022, yet the program was created without a single congressional vote. It was designed in 2007 by the Department of Homeland Security working in secret with industry lobbyists at Microsoft’s request, explicitly to circumvent the H-1B cap that Congress refused to expand. The program was challenged in court, vacated by a federal district court for procedural violations, and then reissued by the Obama administration in 2016 in its current expanded form.

The median annual wage in IT occupations rose just 3 percent between 2014 and 2019, while employment in those fields jumped 18.7 percent. In a genuine labor shortage, wages rise sharply. Instead, we have wage depression for American workers.

What This Means for Wisconsin Specifically

Burkham envisions immigrants filling the jobs Wisconsin employers can’t staff: healthcare workers, agricultural laborers, manufacturers, and yes, truck drivers. The Daily Mail piece lists these shortages as self-evident justification for his proposal. However, part of the reason these positions have gone understaffed is because the employment of guest workers has made them insufficiently attractive to the domestic workforce that already lives there.

Wisconsin’s nursing homes and hospitals face staffing shortages not because there are no Americans willing to do the work, but because pay and working conditions in those sectors have not kept pace with what workers can earn elsewhere. Wisconsin’s farms have held agricultural wages down by using H-2A visas in agriculture. Adding a new place-based visa category would add to the undercutting of American labor.

The proposal also does not confront the structural problems with the existing programs it would presumably complement or expand. If the H-1B program can be used by Amazon to hire 6,400 workers while laying off 27,000 others, a place-based variant could be used by Wisconsin employers in precisely the same manner, filling positions with lower-cost foreign workers while American workers in those same communities are displaced or priced out.

The departure of Wisconsin’s own people is not a demographic problem that can be solved with imported labor. Examining how to bring Americans back to Wisconsin is the better solution.

What Should Actually Be on the Table
Why are we designing new immigration pathways while existing programs remain vehicles for wage suppression and worker displacement? What would it actually take to make Wisconsin hospitable to the people it has already lost? Advanced manufacturing, the skilled trades, and healthcare all represent genuine economic opportunities, but they require wages that make staying in Wisconsin competitive with leaving. Wisconsin’s technical colleges and apprenticeship programs are underutilized assets. The state’s relative housing affordability is a genuine advantage, but it matters only if there are jobs worth staying for.

And finally, whose interests does this immigration debate actually serve? The bipartisan support Burkham cites, from mayors, from business associations, and from the 2025 Conference of Mayors, reflects, in the main, the interests of employers seeking labor cost relief. Their interests are not the same interests as the Wisconsin workers who left Janesville when GM closed in 2008, or the young Wisconsinites who can’t afford to form households on what the state’s labor market currently offers them.

A state that could in the past support working adults and families was ravaged by deindustrialization from policies set in Washington, DC. Immigration policy cannot repair the damage done from industrial policy failure, and only by investing in Wisconsinites can the state improve itself.

PS: If you haven't already, consider subscribing — and if you're able, a paid subscription directly funds our ability to expand our work.

The proposal aims to better align the prevailing wages paid to H-1B workers with those paid to U.S. workers in similar roles and locations. The stated goal is simple enough. Make sure the use of nonimmigrant labor does not undercut the wages and working conditions of American workers.

In practice, though, this looks more like DOL trying to make the best of a bad situation.

At a structural level, the system is still largely blind to actual worker qualifications. There is only so much DOL can do within the framework Congress built into the Immigration and Nationality Act (INA). The Department is still working within a four-tier prevailing wage system based on Occupational Employment and Wage Statistics (OEWS) data. The proposal raises those wage levels, but it does not fundamentally change how they are set.

Most notably, the rule moves the lowest prevailing wage level from the 17th percentile to the 34th percentile. That is being presented as a meaningful reform. But it still leaves a built-in discount relative to the broader labor market. Employers can continue to pay well below the 50th percentile, the median wage, for a given occupation. So the system does not eliminate the discounted wage tier. It just raises it.

That leads to a more basic question. If the H-1B program is supposed to help employers hire highly skilled workers they cannot find in the domestic labor market, why is it structured to allow hiring at below-median wages in the first place? If these workers are truly exceptional, you would expect wages to cluster at or above the 50th percentile, not below it. Why is the program being used to hire what are effectively entry-level workers at sub-median wages? That is not how the program is sold to the public.

More fundamentally, the system still runs on what is essentially a “trust me” model. Employers define the job, choose the occupational classification (SOC code), set the work location, and assign the wage level. The government largely takes those inputs at face value.

That happens through the Labor Condition Application, or LCA, which is the form employers must file with DOL before sponsoring an H-1B worker. In theory, the LCA is supposed to ensure that the employer will pay the required wage and that hiring the foreign worker will not adversely affect similarly employed U.S. workers. In practice, it is an attestation system with very limited substantive review, constrained by timelines that prioritize speed over verification. It is, for all practical purposes, a rubber-stamp process where the Department is largely checking whether the form has been filled out correctly, not whether the underlying wage determination or job classification is actually accurate.

By law, DOL must certify these LCAs within seven business days of receiving them, and the review is limited to form completeness and internal consistency rather than any substantive validation of the inputs.

Put simply, the prevailing wage is only as reliable as the inputs behind it.

And those inputs are flexible.

If the system becomes more precise on education and experience, the pressure will likely move to SOC code selection, occupational framing, and geographic assignment.

For example, a role that could reasonably be classified as a “Software Developer” can be filed instead as a “Computer Systems Analyst” or “Business Analyst,” which may carry lower wage benchmarks in certain markets. Job descriptions can also be written to emphasize narrower or less specialized duties so they fit into a lower-paying category, even when the underlying work is more complex.

Location creates another opening. With remote and hybrid work, a role tied to a high-cost labor market can be assigned to a lower-cost metropolitan area for prevailing wage purposes, even if the actual work and reporting structure remain anchored elsewhere.

These are not rare edge cases. This is exactly how the H-1B program is being gamed today.

To its credit, DOL does acknowledge the problem and gestures toward a more substantive reform through what it calls “Experience Benchmarking.” The idea is to tie wages more closely to workers with similar education, experience, occupation, and geography, with wage levels anchored closer to the 50th percentile and above. Conceptually, that is a more defensible approach. It at least attempts a true apples-to-apples comparison.

But DOL has not adopted that framework. Instead, it is only seeking comment on it, which allows the Department to signal interest in a more rigorous approach without committing to the legal and administrative risks that would come with formally proposing it.

That is not surprising. Moving in that direction would require a much more complex system, likely incorporating survey-based data such as the American Community Survey (ACS) to estimate education and experience. That brings its own problems. The data is lagged, the estimates are modeled, and it is not clear that those tools are well-suited for binding wage determinations.

Candidly, there is a real risk that this kind of complexity becomes a poison pill. It lets the Department signal that it is exploring a more rigorous approach, while also making the rule more vulnerable to legal challenge or harder to implement in practice.

There is also a simpler question that goes unanswered. Why not just anchor the lowest wage level at the 50th percentile using existing OEWS data? That would raise the floor in a meaningful way without introducing a new layer of modeling and legal risk.

Another weak point is the continued use of private wage surveys. The proposal talks about more scrutiny and better methodology, but it does not eliminate them. In practice, that does not change much. The LCA process still does not allow for meaningful front-end validation of wage inputs, and any added scrutiny remains largely theoretical without stronger enforcement. Private surveys will continue to offer a pathway to lower wage benchmarks.

All of this points to a broader reality. There is no real way to fix the H-1B program through incremental regulatory tweaks.

The system was built in a way that gives employers significant control over the inputs that determine wages, while limiting the government’s ability to verify those inputs. That is not an accident. It is how the program was designed. And no amount of fine-tuning can fully undo that.

That said, it is still worth acknowledging when the administration tries to raise wage floors or tighten standards. Those efforts can reduce some of the harm at the margins. But they do not change the underlying structure.

In the end, what was legislated into existence has to be legislated out of existence.

Agencies can only go so far. Real reform will have to come from Congress. But Congress has shown no meaningful movement toward seriously reforming the H-1B program, so credit is due to the administration for trying to address the problem on its own when Congress won’t.

The rule is now open for public comment, and this is where it actually matters. If you care about this issue, take the time to submit a comment. Be clear about the problem with anchoring Level I at the 34th percentile and make the case for using the 50th percentile as the baseline prevailing wage. If the goal is to prevent undercutting, the floor should start at the median, not below it.

Call out the continued use of private wage surveys and how easily they can be structured to produce lower benchmarks. And don’t ignore enforcement. If DOL is serious about addressing these issues, it needs to make audits, worksite enforcement, and post-certification enforcement a real priority, not an afterthought.

Most importantly, write your comment like it will be read by someone who has to respond to it—because it will. Keep it formal, structured, and focused on the substance. Avoid off-tangent or purely rhetorical remarks. A well-argued comment carries far more weight than noise.

And contrary to a popular myth being circulated these days, AI is NOT ripping out the rungs. Torsten Slok, chief economist at Apollo, stated on March 23^rd, “We are monitoring the unemployment rate in the Philippines and India for any signs that AI is reducing the need for outsourced workers in corporate America. So far, there are no signs of AI replacing offshore workers. By way of example, on January 7, 2024, IBM listed just 173 open positions in India, but by early 2025, that figure had grown to 3,866 open roles—while U.S. job listings remained well below 400 during the same period. Now if roughly 1/3 of IBM’s global workforce, a hundred thousand workers, are in India, why are the only job reductions happening in the United States and Europe? Again, it is not AI.

In addition to corporations becoming addicted to cheaper and more compliant workers, hiring in many places becomes tribal as leadership chains become predominantly Indian. Some may question if it is correlation versus causation. But it is a fact that Indian CEOs notably expanded their India operations under their tenures.

Moreover, here in the United States, there is unusually strong evidence by the standards of workplace discrimination that includes federal jury verdicts, EEOC findings, DOJ lawsuits, and detailed internal company documents that show a pattern of US citizens being discriminated against for hiring, promotion, and pay. For instance, the U.S. Department of Labor (DOL) sued Oracle for wage violations and hiring bias, alleging the company favored Asian applicants (90% of recent college graduates hired were Asian) and discriminated against qualified White, Hispanic, and African-American applicants. The DOL attributed the discrimination to strategies like targeted recruitment and referral bonuses that encouraged its heavily Asian workforce to recruit other Asians.

When I speak at campuses, I come armed with STEM OPT data for that particular school, which is available at optobservatory.org, produced by the Institute for Progress, and a state End STEM OPT petition, which you can sign if you’re from Massachusetts and can be found here. For instance, Boston University graduated 1,770 U.S. citizens and 1,552 foreign nationals on F-1 visas with master’s degrees in engineering between 2008 and 2022. 1,069, or 69%, of the international students obtained an engineering job right after graduation via STEM OPT. In addition, Boston University graduated 4,087 U.S. citizens and 2,615 foreign nationals on F-1 visas with master’s degrees in computer science between 2008 and 2022. 1,638, or 63%, of the international students obtained a computer job right after graduation via STEM OPT.

Painting with a broader brush, between the Class of 2008 and the Class of 2022, Massachusetts colleges added 55,042 foreign STEM OPT workers in those fourteen years. 19,049 foreign students were trained for engineering jobs, and 16,563 foreign students were trained for computer science and mathematics jobs in Massachusetts. You might say that’s not a lot. However, there were only 66,590 employed engineers in Massachusetts in 2023. Meanwhile, the employed Massachusetts OPT engineering workforce was 5,420 in 2022, meaning OPT workers actually made up 8.1% of the state’s total engineering workforce.

Now let’s zoom our analysis in even further, to the master’s computer science program at the University of Massachusetts Amherst. The University of Massachusetts Amherst graduate program ranks among the top 10 computer science public programs in the United States and is the number one public program in the Northeast. Unlike most colleges, the computer science program was created early in 1964 and is housed in its own separate school, the Robert and Donna Manning College of Information and Computer Sciences (CICS).

Interestingly, Robert Manning graduated from the UMass Lowell campus in 1984 with a degree in information systems management and rose to become president and CEO of MFS Investment Management. Donna Manning earned a nursing degree in 1985 and an MBA in 1991 and had a 35-year career as an oncology nurse at Boston Medical Center. In 2021, Robert and Donna Manning gifted $50 million to the UMass system (the largest in the school’s history), with $18 million earmarked to the computer science college that now bears their name.

Robert and Donna’s philanthropy philosophy is deeply rooted in the understanding their “parents made a lot of sacrifices to send them to school.” I wonder how they would feel if they understood how federal employment visa programs and the UMass system were undermining the career opportunities of today’s graduates, graduates who, like them, grew up in places like Methuen, MA, and whose parents invested and sacrificed in and for them so they could have a shot at the American dream?

Read our latest data research brief here in full, or I can just give you the short version below:

We have concrete data on job outcomes for the Class of 2022 six months after graduation, specifically at the Robert and Donna Manning CICS master’s program. Our data came from the First Destination Survey, which is done across the country by hundreds of public and private colleges. Moreover, UMass Amherst’s career department pulled in data on 98% of 2022 master’s graduates from sources such as “the Careers Survey, Graduation Celebration Registrations, LinkedIn Profiles, and Staff Knowledge.” Other data included already compiled Optional Practical Training data by the Institute for Progress and classroom citizenship data from the Department of Education’s National Center for Education Statistics. All that data allowed the Institute for Sound Public Policy to show how employers are prioritizing foreign graduates over Americans for jobs before and after graduation from the same academic program.

Sadly, the picture we were able to paint was not a pleasant one. Our key take aways were:

1. Only 60% of American citizens who graduated with a master’s degree from UMass Amherst’s computer science college were working in any capacity after graduation, while the remainder were looking for work, continuing their education, or had an unknown situation.

2. Meanwhile, 86% of foreign student graduates got jobs related to the computer science major through Science, Technology, Engineering, and Mathematics (STEM) Optional Practical Training (OPT).

3. Class of 2022 unemployment rate: 12% overall, 23.5% for American citizens, and 6.7% for foreign graduates.

The Mannings are a rare and admirable example of people who never forgot where they came from. But where they came from no longer exists in the way they remember it. Today, a student from Methuen who works as hard as Rob did, sacrifices as Donna did, and graduates from the very college that bears their names, faces a 23.5% unemployment rate — not because they failed, but because federal visa policy made them an afterthought in their own country’s job market. Robert and Donna Manning deserve to know that. So do their fellow Americans.

PS: If you found this research valuable, please like and share it. If you haven't already, consider subscribing — and if you're able, a paid subscription directly funds our ability to expand this research to more recent data sets and additional institutions.