A quick note: A reader and friend asked me to consider writing a piece on the Lindsay Clancy trial. I normally avoid true crime cases for a lot of reasons, not least of which is a dislike of seeming to profit off other people’s tragedies, but given my past research on maternal health and my current research on institutional brittleness, this case seemed worth grappling with. This piece follows in some ways on my piece on Jason Arday from last week, so read that one too. Due to the nature of the case, this is a tough one to read about. Readers should proceed with caution, though I don’t discuss the deaths in detail here.

In the four months before January 24, 2023, Lindsay Clancy did nearly everything we tell struggling mothers to do. She reported her symptoms. She filled the prescriptions: thirteen psychiatric medications in four months, a rotating cast of antidepressants, benzodiazepines, sleep aids, mood stabilizers, and antipsychotics. She sat through fourteen telehealth appointments with her psychiatrist, the last one on January 23. She enrolled in a perinatal day program. She called a crisis line twice. On New Year’s Eve she admitted herself, voluntarily, to McLean Hospital, the most famous psychiatric hospital in America. And on January 24, about three weeks after she came home, she strangled her three children, Cora, Dawson, and Callan, in the basement of the family’s home in Duxbury, Massachusetts. Neither the prosecution nor the defense dispute these facts.

Her psychiatrist, Dr. Jennifer Tufts, testified this month that she never had the records from Clancy’s other treatment and never sought them. She trusted that Clancy would be candid. The perinatal nurse practitioner treating Clancy through a hospital program testified that she never communicated with Tufts at all. Across fourteen visits, thirteen drugs, and five settings not a single person held Clancy’s whole record.

The trial is at closing arguments in Plymouth County as I write. Coverage has already sorted Clancy into the two available binary bins: monster or victim. I don’t know what the jury will decide and for my purposes it doesn’t really matter. Whether Clancy was floridly psychotic, criminally responsible, or some combination the law handles badly, the institutional facts are the same. The convenient binary skips the institutional question altogether. Institutional brittleness, as it turns out, happens in 4D.

Brittleness Isn’t (Just) Scarcity

The standard narrative for a case like this is one of scarcity: not enough mental health funding, providers, or beds. But the structure of resources may matter more than total resources at the end of the day. It certainly mattered more in Clancy’s case.

Massachusetts itself makes the scarcity story hard to sustain. Clancy lived in one of the wealthiest towns in the state, within an hour of the densest concentration of psychiatric expertise on the planet. She was herself a labor and delivery nurse at Mass General. She had insurance, an engaged family, and access to the very institutions the scarcity story says other women lack. She got into McLean by choosing to go.

What her case reveals instead is brittleness. Brittleness is not the absence of resources. It is, in part, the absence of load-bearing connections between them. In the world of institutions, a pile of bricks is not a bridge. And a hundred fragmented providers don’t add up to quality care.

I’ve written before about brittle versus resilient institutions: brittle ones optimize for appearance and cave under pressure, while resilient ones treat friction as information and bend without breaking. Maternal mental health care in America is brittle in a particularly revealing way, and the problem shows up at three levels. Combine them and the result is catastrophic.

Individual Brittleness: Honesty Is Expensive

When I was doing work on my book on healthcare in January 2018, I ran across the case of a Sacramento mother named Jessica Porten. She went to her postpartum checkup and did what every pamphlet in every OB’s office begs new mothers to do: she described her symptoms, including intrusive thoughts of self-harm, and said she needed medication and therapy. She asked for the help everyone says you should ask for. She sent a signal to an institution that the institution itself asked for.

The nurse called the police. Officers escorted Porten, with her baby, to an emergency room, where she waited hours, never saw a psychiatrist, and was discharged around midnight with a pamphlet. Her candor bought her a police escort and a pile of paperwork, but it didn’t buy her treatment.

Disclosure carries serious weight for mothers. As a mother of three who struggled with mood disorders following the births of all three of my kids, I can say firsthand that there’s a qualitative difference between admitting that you’re struggling at, say, work, and admitting that you are struggling at what culture tells us should be instinctive: loving your kids.

Apart from those internal struggles, which are powerful in their own right, those disclosures can function as evidence. Every clinician who screens a new mother is a mandatory reporter, and mothers know it. Tell your providers about a drug problem, or simply take the addiction medication they prescribed, and the drug screen at delivery can convert candor into a case file: a Marshall Project investigation published this year found tens of thousands of mothers flagged to police or child welfare over childbirth drug tests, some flagged for poppy seeds, and others for the very medications their doctors prescribed (speaking of fragmentation…). The research on punitive prenatal substance policies finds what the incentive story predicts: women don't stop using when they’re punished for honesty; they stop showing up, skipping prenatal care and delivering across state lines. Tell a doctor your partner beats you, and "failure to protect" doctrine can land you, the victim, in the caseworker's file too. New York City removed children from battered mothers on that logic until a federal court stopped it, and survivors of abuse report avoiding help for exactly that reason. I still remember reading the fine print in my welcome folder while idly waiting in my OB’s office during my first pregnancy: the routine urine test given to pregnant women to screen for proteins that can indicate pre-eclampsia would also, they helpfully noted in tiny print, be screened for drug use as well. Reassuring! (I quit that practice shortly after, for semi-related reasons.)

Prosecutors have leaned hard on Clancy’s restlessness with treatment: she repeatedly pushed to change her medications, pressed for the drugs she thought she needed, and the state presents the pattern as calculation. Whatever was true of Clancy in particular (and the jury will unenviably have to make that call), one way to explain that pattern is that it’s what a rational patient might look like when full candor with any one provider is expensive. To put it another way, it’s what happens when an institution punishes the signal it asks for. If telling one clinician everything risks the psychiatric label that ends your nursing career, or the report that brings child welfare to your door, you distribute the information. A little insomnia here, a little anxiety there, and nothing anywhere that adds up to the sentence I am afraid of what I might do. The fragmented record that made her treatment incoherent could also be the record that made her disclosures safer, at least in the moment, until the snap.

The perverse result is that our entire screening apparatus - the Edinburgh screening forms and the “how are you feeling, mom?” scripts - assumes honest self-report from a population we have given good reason not to self-report. We built the resilience of a successful maternal health system on a signal we punish people for sending.

Medical Brittleness: A Disease With No Name

Postpartum psychosis, the condition Clancy’s defense says she had, does not exist as a diagnostic category. There is no code for it in the DSM and none in the ICD. A woman can be treated for it in 2026 and her chart won’t say what she actually has. Last October an international consortium of researchers published a consensus statement in Biological Psychiatry laying out the evidence that postpartum psychosis is a distinct disease and arguing for its formal classification. That the field’s leading experts still had to make that case, in 2025, about a condition described since Hippocrates, is the institutional failure in miniature.

The absence of a name has institutional consequences. No diagnostic category means no registration pathway for controlled trials. It no means treatment guidelines built on cohort studies and expert consensus. It means a condition barely present in training curricula, so the emergency physician doesn’t know what to look for or what the signals that do come through mean. It means no billing code, so no hospital service line, so no dataset, so no research base.

A diagnostic category does one institutional job above all others: it gives clinicians a schema to pattern-match against. Without it, thirteen medications in four months looks like titration. With it, the same record looks like what it apparently was: decompensation happening faster than anyone tracking a single slice of it could see.

But that diagnostic research hole sits inside an even larger one. The National Academies concluded in late 2024 that women’s health research received 8.8 percent of NIH funding over the previous decade, and, more damning, that conditions disproportionately affecting women are underfunded relative to their disease burden. The share isn’t the scandal; the burden-normalized gap is. Within that gap, the mechanisms most likely to explain postpartum psychosis, the hormonal cliff after delivery and the immune shifts around it, sit among the least studied questions in the field.

As a mother who struggled with postpartum mood disorders after my own kids’ births, though thankfully not anywhere near the extent Clancy did, I can testify firsthand to the “solutions” offered:

Option 1: a prescription for Prozac that left me with brain fog and flat affect, weight gain, zero libido, and not caring much about anything at all (an improvement, I guess?).

Option 2: pulling myself up by my bootstraps and going about the work of caring for infants and kids, maintaining a marriage, writing books, and trying to mask periodic despair and anger as I muddled through the hormonal whiplash that caused rage episodes and motivation and energy swings that made daily life feel like a roulette wheel. Will I be able to function today? Spin the hormonal wheel and I guess we’ll find out!

I chose option 1 until I couldn’t take the side effects and then slid into option 2 mostly by default.

Having a diagnostic category doesn’t do much work if we don’t have the research to know what to do about a problem when it manifests. In my case I had a diagnosis, but no one knew what to do about it other than throw anti-depressants at it to see what would stick. From the many moms I’ve talked to over the years that’s the standard response and it’s not a great one. For every mom I know whom anti-depressants worked well for (and they do work well for some people) many others I know ended up in a Clancy-esque rotation through medications and providers, trying to find something that works, while also trying to care for an infant, maybe other kids, themselves, their marriages and jobs, and whatever else all at the same time. The burden of provider visits alone is exhausting and many people give up for that reason alone.

And that leads us to the third level where brittleness hits and maybe hits hardest.

Institutional Brittleness: Nobody Sees the Whole

The United States has five inpatient perinatal psychiatric units: Chapel Hill, Little Rock, Baton Rouge, Glen Oaks, and Mountain View. Forty-five states have none. The UNC unit, the country’s first, opened in 2011 and has five beds. The number of American units where a mother is admitted together with her baby is zero. People who have not been new mothers cannot really understand what that kind of separation does to both mothers and babies, but suffice it to say that separation from your baby while struggling with mental health issues is unlikely to make the underlying issue better. Quite the contrary.

And this separation also, critically, loops back to the individual level brittleness. A mother whose only option to pursue psychiatric care requires total separation from her infant is much less likely to pursue precisely that care. Again, we’re creating care models that new mothers have very good reasons to avoid. The United Kingdom, with a fifth of our population, runs twenty-three mother and baby units built on the premise that separating a psychiatrically ill mother from her newborn is itself a clinical harm. Who knew?!

While this could be the scarcity story sneaking back in, we need to look at where those five perinatal psychiatric inpatient units are. They’re not in the states with the most to spend: Massachusetts, with more psychiatric expertise per capita than almost anywhere in the country, has none. Arkansas, the same state that is last to extend postpartum Medicaid, has one. The gap tracks the brittleness, and it does not track the resources. Those five units exist because of individual champions in those states who worked to make them happen: a researcher at Chapel Hill, an academic program in Little Rock, a psychiatric hospital on Long Island, a community hospital in Silicon Valley, and a women's specialty hospital in Baton Rouge. Capacity that depends on the luck of a champion is steward-dependent, and steward-dependence is its own marker of brittleness: a resilient system builds what the category requires and keeps building it after the founder moves on. Our existing inpatient perinatal psychiatric units weren’t conscious design choices on the part of medical systems or state public health offices. They were luck. And institutions that depend on luck eventually snap.

The good news (I guess) is that we did solve the part that was easy to legislate. Twelve-month postpartum Medicaid coverage is now nearly universal; when Wisconsin’s extension took effect this summer, Arkansas became the only state without it.

The coverage duration problem was solved while the care integration remained untouched.

For better or for worse, a woman can now be insured for twelve months postpartum while seeing a psychiatrist who has never spoken to her nurse practitioner, who has never seen her hospital discharge summary, and who does not know what the crisis line logged. We pay for the fragments. We just don’t connect them. Lindsay Clancy had excellent care, by all accounts. The quality of each piece itself wasn’t the problem.

Psychiatric care in this country is segmented by payer and by specialty in ways nobody would tolerate for any other organ (though we also see this fragmented care at the end of life; it’s much more common in areas that are more holistic and harder to isolate, like maternity care, end-of-life care, dementia care, and so on). If an endocrinologist prescribed insulin for four months without ever obtaining the records from the patient’s two other prescribers, we’d call it malpractice. In psychiatry we call it Tuesday. And the coordination failure is not for lack of federal awareness: the 2024 Task Force on Maternal Mental Health strategy names as its first pillar, building “a national infrastructure that prioritizes perinatal mental health and well-being.” That was two years ago. It’s fair to ask what has been built. Certainly not inpatient perinatal psychiatric units, that’s for sure.

Distress Signals: Expensive to Send, Cheap to Ignore

The three levels are the same failure in different guises. Maternal honesty about symptoms risks police and child welfare involvement, so patients ration the truth. The condition has no name, so the truth that gets through has no schema to land in. Nobody holds a unified record, so the truth lands in five places. At every level, a clear signal about a woman struggling to care for herself and her kids is expensive to send and cheap to ignore.

That’s pretty close to my working definition of institutional brittleness. A resilient institution is, among other things, one where reporting a problem is low cost to the reporter and high salience to the system. In academia, the colleague who flags your error is thanked, and the flag travels to whoever can act on it. Invert both properties - make reporting costly to the reporter and invisible to (or ignored by) the system - and you get brittleness that looks like adequacy: a bridge that holds its shape right up to the moment it snaps. Clancy’s record looked, to every individual provider, like a patient receiving abundant care. It looked that way because abundance was the only thing anyone could see from inside their own fragment.

In the taxonomy I’ve been developing in my current book, this is a Correction failure first: nowhere in that four-month record was there a mechanism for converting distress signals into a system-level response. It is compounded by a Containment failure: even when the signal finally got through and Clancy admitted herself to McLean, there was no specialized setting to escalate into, no unit built for what she may have had, and she was discharged back into the same disconnected outpatient scatter she came from. And under both sits an Exit and Voice failure of the kind Albert Hirschman would have recognized immediately: voice is punished, exit to another provider is free, so patients exit repeatedly, and every exit fragments the record further. The system trains its patients to behave in exactly the way that keeps the system blind.

The prosecutorial narrative is that Clancy’s repeated exits indicate intent. What’s much more likely is that her exits indicate the fear and confusion that arise when someone is ill inside a brittle institution and when their distress signals, signals the institution is created to capture, instead disappear into the void.

What Non-Brittle Would Look Like

A resilient, or at least a non-brittle, healthcare system that Clancy could have found herself in would look quite different. There are four shifts we can make and none of them are particularly radical. They operate in all four dimensions, asking us to align individual level risks, social systems and support, institutional tradeoffs, and time across a patient’s care.

The first is a name. Classify postpartum psychosis in the DSM and ICD, as the consensus statement urges, and let the trials, the training, the billing code, and the service lines follow the way they follow every other named disease.

The second is a safe harbor. Create reporting rules that decouple asking for help from losing your children, so that a mother describing intrusive thoughts (or drug abuse or domestic violence or any other risk-factor) to her OB is triaged as a patient rather than processed as a suspect. This should be obvious by now, but is somehow not.

The third is a record-holder. Make care coordination a specific person’s actual job, with the authority to pull records across providers. In end-of-life care we’ve created palliative care teams that (in the best of circumstances) do this kind of coordination, but we don’t have any equivalent in maternity care (in most cases). We could change that. We have the model.

Finally, a place to go. Mother-baby units, starting with the forty-five states that have nothing. If the UK has 23 surely we could have one per state. Or something close.

Notice, the order matters here (that’s the fourth dimension in action). The safe harbor has to come before the record-holder, because coordination without safe disclosure just builds a better surveillance system: a unified record accessed by a mandatory reporter creates even greater disincentives for honesty. Decouple disclosure from punishment first, then connect the records. None of this requires new science. All of it requires institutions willing to value the signal over the ease of ignoring it.

One Last Thing

Soon, twelve jurors will decide whether Lindsay Clancy was capable of criminal responsibility on January 24, 2023. I (fortunately) don’t have to answer that question, and my institutional verdict doesn’t depend on hers. If she was psychotic, the system missed a treatable emergency while medicating it thirteen ways. If she was not, the system spent four months prescribing to a woman and no one connected the dots. The failure is identical in both stories and that’s what brittleness means.

Somewhere in those four months the truth might have existed in one piece, but every institutional incentive Clancy faced told her to keep it in pieces. And no one was there to put it back together. We created a system that made telling the truth both dangerous and unproductive, creating brittleness just waiting to snap. The snap was the death of three children and the destruction of the lives of those who remain. The only person on trial is the one whose signals, had they been heard, might have saved them all.

Your Turn

Tell me: Where else do you see systems that punish the signal and are then surprised by the silence (and then the snap)? Let me know in the comments. If you know someone who works anywhere near maternal health, I would love it if you restack this piece; it was written to be shared. If institutional brittleness is a frame that works for you, please subscribe: this is part of an ongoing series on institutional repair, and radical moderation is where it starts.

Finally, if you liked this piece, you can read some of my other work on brittle institutions and what happens when they snap:

Jason Arday, Brittle Institutions, and False Choices
The Quiet Cruelty of Fragmented Systems
Universal Governance Design

Subscribe now

Share

Leave a comment

At a manuscript workshop for my second book, a colleague who takes sourcing very seriously flagged one of my citations. I had cited a study and generalized its findings to physicians as a whole. My colleague actually checked the footnotes and found the study had an n of five. Even worse, the five weren’t physicians, they were medical students. Hardly strong evidence for my claim. That colleague, James Stacey Taylor of the College of New Jersey, pulled me aside after the session and gently told me I should double check my citations as I finalized the book to avoid similar errors.

It was an important correction, it stung a bit at the time, and I should have been more careful. I was writing that book for a major university press with zero research assistance, no TAs, no grad students, relying entirely on my own organizational systems (thank goodness for Zotero) to keep close to 2,000 sources straight while teaching a full load while also parenting two young kids and pregnant with a third, before AI existed to check or help me. None of that excuses the error and I’ll come back to how all the relates in just a bit.

But that experience, and over twenty years of experiences like it in academia, give me a lot of thoughts about the death of Jason Arday late last week.

If you haven’t followed the case (and are either unaffiliated with academia and/or living under a rock): Arday rose from a PhD at Liverpool John Moores through Roehampton, Durham, and a full professorship at Glasgow to become, in 2023 at age 37, the youngest Black professor in Cambridge’s history. His backstory was irresistible: autistic, didn’t speak until age eleven, didn’t read or write until eighteen, and the son of Ghanaian immigrants in South London. The life story and the scholarship started to unravel around the same time. His 2015 dissertation contained extensive near-verbatim passages from a 2009 Brunel dissertation. A 2018 article lifted text from an epidemiologist’s study and included interview quotes that appear to have been fabricated. Cambridge was warned about “substantial plagiarism” in May 2023, within weeks of hiring him, and settled for a quiet correction. When Times Higher Education prepared an investigation in 2025, Arday hired the defamation firm Carter-Ruck and the story died. It took a Substack post this summer to force everything into the open. Arday resigned in early August. A week later he was found dead.

The culture war has already sorted this story into its two available bins. In one, Arday is what you get when DEI replaces merit: a fraud manufactured by identity politics, QED. In the other, he’s a Black scholar hounded to death by racists who were never going to let him keep what he’d earned.

Both bins are one-dimensional. The actual landscape exists in all four dimensions.

Arday committed extensive, deliberate academic fraud, and he was responsible for it. The institutions that hired and promoted him failed at the one task that justifies their existence. The incentive structure of modern academia manufactures misconduct at industrial scale, most of it far less dramatic than his. And the way we’ve turned representation into a checkbox harmed the very people it was meant to help, Arday included.

All of those statements can be true at the same time. Radical moderation is the practice of refusing to pick just one.

Not All Misconduct Is the Same

Academic misconduct takes a lot of different forms, and the extensiveness of the errors matters as much as the intent behind them.

There’s careless sourcing. There’s misremembering the outcome of a study. There’s the phrase you’ve stared at for five years and now believe is your own when it started as someone else’s. The human brain is fallible, research is complicated, and anyone who has tried to hold 1,500 sources in their head while throwing animal crackers at a toddler to try to keep her occupied so you can finish a paragraph knows the limits of human cognition. My n-of-five mistake lives in this category. So, I’m sure, do other sloppy citations scattered through things I’ve published.

None of this is okay. Errors should be caught and corrected, and researchers should be held accountable for those errors. But accountability is a range. Firing is one lever at the far end of it, but the range runs from the kind correction my colleague gave me through performance improvement plans, having your scholarship monitored by a faculty committee for a period of years up to losing your job. It shouldn’t involve global humiliation and it should not involve death.

Matching the response to the offense is what taking integrity seriously looks like. Treating every error as moral depravity creates its own perverse incentives by teaching everyone to hide their errors rather than fix them.

Of course, what Arday did was different in kind, not degree, from run-of-the-mill citation errors. His work was plagiarized from soup to nuts. The dissertation itself was compromised, meaning every credential built on top of it was compromised too. He should never have received the PhD, let alone the first academic job, let alone the Glasgow professorship, let alone a chair at one of the top universities in the world.

To be fair, my own dissertation was derivative. It leaned heavily on my advisor’s work, my advisor knew it, and it was never going to be my first book. But it was good enough to justify a PhD, and it was mine in that every word, however imperfect, came out of my brain and was typed with my own hands. My work got better over the years, (and it would have gotten better faster with more support). That’s the normal arc of a scholarly life. Arday’s arc skipped the part where you do the work.

As I’ve argued before (and people keep resisting), understanding is not excusing. Bad incentives can enable bad people and some very bad incentives can make otherwise good people do bad things they otherwise would not have. I don’t know which bucket of human being Arday falls into (a meaningless question, really) but academia falls into the “very bad incentives” bucket without qualification.

The System Asks Everything and Funds Nothing

The expectation for tenure and promotion at most universities, including universities with relatively heavy teaching loads, is that you are constantly publishing. It is very hard to do careful scholarship while teaching a lot. It is nearly impossible to do groundbreaking scholarship while teaching a lot, because ideas take time to generate and require resources to germinate. They need conferences, workshops, and people stress-testing your arguments. All of that takes time and money.

Pre-COVID, my entire yearly travel funding was $1200. Once you pay $400 for a professional membership, $300 for the registration, $300 for two nights in a hotel and $400 for a flight from a regional airport, that money covers (maybe) a single conference a year and you still have to put part of it on your credit card. That’s a single opportunity every year to get the kind of in-person intensive feedback academic excellence requires. Then you attend that conference and find only four people in the audience at your panel (one of whom is asleep), and a discussant who didn’t actually read your paper, and the entire project feels a bit farcical.

For most of my academic career I had no research assistance of any kind and I relied on external networks to support my conference travel. I was lucky to have them, given that I work in political theory, a not-terribly-fundable discipline. But even with that help, the reality of the fourth-dimension hit hard: When I was on a book deadline, my teaching suffered. When I was in a heavy teaching semester, my research slowly died on the vine. There are excellent teachers who shouldn’t be forced to publish and excellent scholars who shouldn’t be in a classroom, and the system demands that everyone be both, always, simultaneously. There are ways to align research and teaching, but most universities don’t bother so you get crappy outcomes in both directions.

Meanwhile the research-industrial complex comfortably hums along. I get a couple emails a week from predatory journals asking me to publish with them. I have never been desperate enough to consider paying thousands of dollars to publish an article simply to get the publication I need for tenure. But I can imagine how the incentive might look to someone who is. The academic job market is brutal. If you don’t get tenure, you may not get another academic job. If you lose an academic job, it’s very unlikely another will appear. If you’re lucky enough that another one does appear, you will have to uproot your entire life and family and move somewhere else, usually not in a place you actually want to live. The system we’ve built demands that we enter a Hunger Games-esque competition for far fewer jobs than there are applicants and once we’re lucky enough to get hired we then publish constantly, with minimal support, under threat of permanent exile, while peer reviewers, unpaid and overwhelmed, wave through work nobody has time to check and for-profit journal corporations rake in the profits. We should be asking what kinds of behavior that system encourages and we should stop being surprised by what that system produces.

Everyone Checked the Box

How did Arday get through? It wasn’t a single hiring committee or a single failure point. It was four institutions, ascending in prestige, each of which had more reason than the last to scrutinize the record.

But he was a perfect storm. Neurodiverse, a racial minority, with a story of struggle that academics absolutely love. He checked every box. And so nobody bothered to look at the actual scholarship, which was the one part that really mattered.

The problem was never that universities value diversity. The problem is what happens when any credential, diversity included, substitutes for reading the work.

When I got my current job, a faculty member came up to me in the hallway and said she was so glad they’d finally hired a woman. It was meant in solidarity, said by another female academic, and I remember thinking: what a weird thing to say to a young female academic. The fuller story is more complex than identity, as these things always are. I was absolutely not the best candidate in some abstract meritocratic ranking. I had a specific skill set, I came from the specific political theory tradition the department was looking for, I met the qualifications, and I happened to be a woman, which I know made me easier to hire than another white man, given that the department had just lost its only female member a year or so before. Did I get my job because I’m a woman? A little, I guess? It’s just not a meaningful thing to say, given that I also got it for a pile of other reasons. Most hires, viewed honestly, are some kind of nepotism: the right tradition, the right network, the right fit for what a department wants that year, the right pedigree, or the person who manages to make it through the deeply subjective and messy search process that is every faculty search.

But when institutions treat identity as the credential, they put every person who carries that identity under a cloud of suspicion they didn’t ask for, and they put the occasional fabulist on a rocket. Arday’s story of struggle functioned at Cambridge the way my gender functioned in that colleague’s mind: as the explanation, rather than one fact on a landscape of facts. One critical difference is that my work existed and was mine. Nobody checked whether his was, because the story was doing the work the scholarship was supposed to do.

That harmed everyone downstream. It harmed the scholars whose work he stole. It harmed every honest Black academic and every honest autistic academic who now works under a heavier cloud.

The Copycat Don

If you’re tempted to conclude that Cambridge protected Arday because of the story, we can look at what the same institution did for a white historian with no inspirational story at all.

William O’Reilly, an associate professor of history there, is in some ways a control case. He published work containing large chunks of an undergraduate student’s essays. An internal investigation concluded it wasn’t plagiarism. He stayed in post while further cases piled up on PubPeer, where he earned the nickname “the Copycat Don.”

The O’Reilly case is concerning on its own terms. An institution like Cambridge allowing any kind of academic misconduct to continue is a failure. He should have been on a performance improvement plan, with his scholarship monitored by an independent faculty committee. AI now makes plagiarism far easier to detect than it was even five years ago; there’s no excuse left for letting it slide.

But O’Reilly and Arday are still qualitatively different cases. One is a scholar with real credentials who took unethical shortcuts that should have triggered serious, proportionate consequences. The other built an entire career on work that wasn’t his. When your accountability system responds identically to both, or, as Cambridge managed, responds to neither, you’re no longer in the business of knowledge generation; you’re in the business of reputation management.

Brittle Institutions Protect Themselves

I’ve written before about the difference between brittle and resilient institutions. Brittle institutions optimize for the metric and the appearance: the Carnegie classification, the diversity statistic, the prestige hire, or the news cycle survived. Resilient institutions optimize for the thing itself, and treat friction, the colleague who flagged my n of five, as information rather than threat.

By that measure, academia is brittle all (or most of) the way down. The Arday case is just brittleness under pressure.

Most of us in academia have seen the friction disappear. My dissertation advisor was once asked to serve as an external reviewer on a dissertation in an education department. He told the committee he couldn’t support it because two entire chapters were lifted from a book; the most straightforward kind of plagiarism. The department’s response: we’ll let the student rewrite it. My advisor’s response was that this was absolutely inappropriate, because someone willing to plagiarize huge chunks of a dissertation will be willing to do much else, and they get worse rather than better. He was initially overruled before he threatened to go public, which is nuclear option that should not have been necessary. (Arday’s field, for what it’s worth, was sociology of education.)

I see firsthand where it goes now. I’m my college’s academic integrity officer, and the vast majority of student integrity violations never get reported, because faculty have learned that reporting is more work for them than letting it go. While each individual decision is understandable, collectively, they dissolve the infrastructure the whole enterprise stands on. My colleague James Stacey Taylor (of the workshop citation correction) has spent recent years chasing down serious plagiarism cases and doing amazing work, but I think he knows he’s tilting at windmills because the misconduct is that widespread.

The irony is that this is one place where the tools have outrun the will. I am a more careful scholar now than I was writing that second book, because AI functions as the research assistant I never had, checking my sources and my memory of them. Detection has never been easier. (Of course, it also facilitates slop and that slop overwhelms reviewers, so tradeoffs are real). But we have solutions, if we only care enough about our own mission to implement them: peer reviewers could be supported and compensated for real scrutiny; the for-profit journal giants could spread some of their remarkable margins in that direction. What’s missing isn’t capability. It’s an institution willing to value the finding of fraud over the embarrassment of having found it. Cambridge had the warning in hand for three years. A defamation firm was enough to keep the lid on for one more.

That brittleness undermines trust and it causes serious harm, as we saw last week.

It harmed Arday himself, in the most final of ways, as he was lifted, stage by stage, into positions his actual record could never survive, and then left standing at the center of a global culture war when the scaffolding came down. To be clear, he willingly climbed. He lied at every stage, and he knew better than anyone how much deceit the climb required. But if the system had done its job in 2015, or 2021, or in May 2023 when Cambridge was warned in writing, he would have faced consequences scaled to a much smaller life.

There’s a version of this story where he’s quietly stripped of a PhD a decade ago. Everyone shares blame for the brutal version we got instead.

Burn It Down, or Build Small

Academic institutions have become extraordinarily complicated. They are too big to fail in some ways and too big to succeed in others. They have accumulated layers of incentives and expectations that don’t necessarily reinforce one another.

We want excellent teaching. We want groundbreaking research. We want high publication counts. We want high rankings. We want research classifications. We want diverse faculties. We want external funding. We want student satisfaction (and student money). We want administrative efficiency.

And then we act as though all of these goals can simply be stacked on top of one another without tradeoffs.

They can’t.

My solution is (almost) never burn it all down, but this is the closest I get.

Academia in its current form cannot take seriously the misconduct that violates its own foundational project, because taking it seriously would implicate the rankings, the classifications, the hiring narratives, the publication counts, and the entire apparatus of prestige that has crowded out the thing the apparatus was built to certify. The core of the academic project is integrity. That is the only thing that makes any of it work. Knowledge creation is one of the strangest and most beautiful (and deeply four dimensional) things humans do together: thousands of strangers across generations building on each other’s work, each trusting that the last person told the truth about what they found. The emergence of knowledge from the messy process of scholarship only works if people can trust the footnotes; if the credentials actually mean what they claim to mean; if the ideas belong to the person responsible for claiming them (for good or bad). Unfortunately, integrity is now the value our academic institutions are least organized to protect.

Where do we go from here? Short of arson, I’d say build small. Create pockets where the standards are real. Universities, departments, presses, and journals where the output can be trusted because someone reads the footnotes, where accountability is proportionate and therefore credible, where a correction is a normal event rather than a scandal, and where a warning letter from a senior scholar gets an investigation instead of a quiet edit.

People are doing this work. Academics are creating journals with real rigor. Organizations like Retraction Watch are keeping an eye on when and how retractions happen and how to prevent the errors that caused them. Centers like the Adversarial Collaboration Project at Penn work to create resilient scientific institutions where integrity comes first and foremost. And brave scholars are choosing to retract their own research when it fails to replicate, treating correction as a normal part of scholarship rather than a scandal. There are increasing numbers of people developing models and ideas for making science not only trusted but trustworthy. That kind of rigor won’t always be aligned with elite institutions. Ironically, when it comes to integrity, the evidence suggests elite institutions arrive last.

The workshop colleague who caught my n of five did more for academic integrity that afternoon than Cambridge managed in three years of holding a written warning about substantial plagiarism. The funding for the workshop itself, ironically, came from the Institute for Humane Studies (where I’m currently a fellow), an organization working to make academia more resilient. It didn’t come internally. That academia increasingly relies on external funding to support academic rigor is exactly the problem.

Jason Arday was a flawed man and Cambridge is a flawed institution. Both failed their own claimed values at the moment they needed them most. But there’s a critical asymmetry: the institution gets to keep going. For Arday there are no second chances; there’s no kind correction, no dissertation reader catching him before the PhD, no improvement plan, and no quiet return to a smaller life. That’s what brittleness costs. A resilient institution catches failure while correction is still possible; a brittle one ignores failure until the only outcomes left are catastrophic. Cambridge was lax for three years; the world was merciless for one summer. Lax and merciless aren’t opposites. They’re the two halves of a brittle system: the long silence and the sudden snap. In that sense, at least, a brittle institution killed Jason Arday.

The kindest form of accountability turns out to be the most rigorous one: the correction that comes early, from a colleague who actually read the work.

[This post is dedicated to my friend and colleague James Stacey Taylor, who I hope continues tilting at windmills in defense of academic integrity for many years to come.]

Your Turn

These problems exist outside academia and not all my readers are academics. Where else have you seen brittle institutions? Where else do we focus on metrics instead of actual principles? Where else do we see misaligned incentives that distort the very mission of institutions? I would love to hear your thoughts and examples in the comments. And if the incentives question interests you, subscribe and share. I have a broader series on higher ed. You can search the publication to see the top posts or start here: Part I: The Higher Ed Death Spiral, Part II: What About The Wage Premium and Part III: Halting the Higher Ed Death Spiral along with Higher Ed Doesn’t (Just) Have a Trust Problem.

Leave a comment

Subscribe now

Share

I remember sitting in my grandparents’ house in rural Tennessee decades ago listening to my grandmother and father arguing about whether the moon landing was real. The argument was jovial, if animated, from what I remember. My grandmother and father, long antagonists, eventually reached a status of mutually respectful if occasionally tense debate partners, even while my grandmother loathed the heathenism my mother had married into.

If you think "trust the experts" is a straw man nobody actually deploys, it was a busy year. When the Trump administration dismissed the CDC's vaccine advisory panel last June, the governors of California, Oregon, and Washington responded that their states "depend on the best science and knowledge available." By September Newsom was signing a law to break from future federal vaccine guidance altogether, pledging to put "science, safety, and transparency first." In January he called the administration's revised childhood vaccine schedule a move that "flies in the face of decades of medical expertise." Meanwhile, at the Make America Healthy Again summit in November, Vice President Vance told RFK Jr. that "all the experts were wrong" and that the administration needed to be "comfortable challenging some of these old orthodoxies." Both sides are taking a one-dimensional view. One collapses all expertise into a single authority you either accept or reject. The other collapses all expert skepticism into ignorance. Neither is doing what my grandmother ultimately did, which was trust but verify in four dimensions.

Last time I asked readers to think about the phrase “I believe in science” with a little more complexity, because all knowledge is contextual and incomplete: a snapshot that will look partial from a hundred years out. This time I’m asking you to do the same thing for the related claim, “trust the experts.” That saying asks you to treat a snapshot as the whole landscape, and it strips out the humility that would make an expert worth trusting in the first place. It’s also incredibly damaging to our public discourse over the long-run.

Trust needs to be earned

I don’t remember precisely what my grandmother’s argument was, but knowing her my guess is that she was probably agnostic on the moon landing itself. Her point was likely a broader one based in distrust of government. If she refuted the photographic evidence it wasn’t because she truly believed it false, but that she “wouldn’t put it past them.”

My grandmother was the youngest of eight children whose father died when she was seven, plunging the family into poverty. She spent the next ten years suffering from malnutrition and subject to her mother’s ire. Despite a very difficult adolescence, she was a woman of striking intelligence and intellectual curiosity, the only one of her siblings to graduate high school. In the rural Tennessee town they eventually retired to she was called to give guest sermons at various basement evangelical churches based on her almost photographic memory of the Bible and her powerful analyses. Her financial acumen took her from lower-class to comfortably upper middle class across her lifetime, carefully investing my grandfather’s blue collar earnings in the stock market while her more skittish siblings continued to invest solely in government bonds.

My grandmother denied the moon landing not because she was ignorant or stupid but because in her experience of the world experts had done little to earn her trust. Almost everything in her life had been fought for, concretely and in the moment and by her. Academics, lawyers, government bureaucrats and physicians had done little for her and she knew it. If she vaccinated her kids it was because she had firsthand experience that vaccines worked; she was suspicious of the moon landing because she hadn’t seen it with her own eyes. She had, undoubtedly, benefited from invisible experts over her lifetime as public health and medical expertise grew and she herself trusted medical experts when it seemed prudent to do so.

But that qualification did a lot of work.

She grew up in a world where she had to figure things out on her own and she took verification very (very) seriously. God and the church showed up for her. Government and experts, not so much. And for whatever it’s worth, she subjected church leaders to the same verification model and found they often failed too. She was nothing if not consistent. It’s also true that as a woman of faith, experts had not treated her beliefs with much respect, let alone deference.

As my grandmother cynically understood, experts of all kinds are human. Experts respond to incentives, and not all incentives point toward your wellbeing. My grandmother was viciously critical of failures of church leaders, politicians and public health officials alike. In the end, she knew that she was ultimately responsible for the choices she made about her own life and she made those choices, taking advice from experts when it seemed to make sense and ignoring it when it didn’t.

She died of lymphoma, likely exacerbated by an avoidable medical error, when I was in grad school. Against medical advice (and the pleading of my grandfather), she chose to cease the radiation and chemotherapy that were ravaging her body, arguing that she knew she was going to meet Jesus and she might as well do it sooner rather than later. She knew what her doctors refused to say: she was dying, her body riddled with tumors, whether she liked it or not, and she saw no reason to beat around the bush. Two weeks before her death she was puttering around their farm making chicken soup from scratch for my grandfather and she died at home in her living room rather than in an ICU. Her decisions to buck the experts at the end of her life were, on most measures, the correct ones.

She also happened to be wrong about the moon landing. If her method was generally sound, her conclusion was false, which happens to everyone at least some of the time. But it was false not because she was ignorant or stupid. It was that verification is expensive and nobody had ever given her a reason to spend those scarce resources on NASA. Being right on concrete issues and wrong on the abstract is probably a pretty good pattern of error for the average person, for what it’s worth. But the gap between method and conclusion is most of what this post is about, so let’s dig in.

Experts are human

Most doctors are conscientious and care about their patients. But there are a range of incentives in medical practice that cut against patient wellbeing. My grandmother knew this and took it at face value. She trusted but verified.

At one extreme, there are bad faith actors. Every profession has its psychopaths, because that’s just how humans work. There are psychopathic doctors, psychopathic nurses, psychopathic professors. Wondery’s Dr. Death is a great deep dive into the failure of expertise, where every single sociopath had both top-notch medical credentials and standing. “He’s a doctor” was, for a long time, treated as a full character reference.

It isn’t one.

But you don’t need bad actors for the system to fail. Sometimes you just need people to trust expertise over evidence. For decades, bite-mark testimony helped convict people, often in sexual assaults and murders, delivered by a small number of credentialed experts the courts treated as authoritative. It took the field years to run the blind tests it should have run at the start, and when a board of bite-mark examiners finally ran one, they couldn’t reliably agree on whether the marks were even human. Thirty-eight board-certified forensic dentists looked at a hundred photographs, and on only fourteen of them could even eighty percent of the examiners agree on the basic questions. Bite-mark matching is about the junkiest junk science there is. Not fraud, necessarily, but confident expertise that no one had ever verified, with real human lives on the line. People went to prison on made-up evidence delivered by people with all the right letters after their names.

And bite marks aren’t the outlier. In 2009 Congress asked the National Academy of Sciences to evaluate American forensic science, and the answer came back that with the sole exception of nuclear DNA analysis, no forensic method had ever been rigorously shown to reliably connect evidence to a particular person. Not hair comparison, footwear, or even ballistics, which had claimed near-perfect accuracy for decades on the strength of exactly one study. When the FBI audited its own hair examiners, it found erroneous testimony in 96% of the trials it reviewed, from 26 of the 28 analysts involved. Thirty-five of those defendants had been sentenced to death. Nine have been executed.

Some dental forensics is perfectly sound: matching dental records to identify remains, for instance. The problem was never the whole field. It was a specific claim nobody had bothered to test that was protected by the aura of expertise. That’s the (in this case catastrophically dangerous) failure a blanket “trust the experts” produces.

Trust, but verify

Where I land is not “distrust everyone.”

Do I want to trust experts when they are trustworthy? Absolutely. The trouble is that I can never know for certain, in advance, who is trustworthy. So the best I can do is treat it as a probability. I can do some research. Extend trust based on what I find. And keep verifying as I go.

My family has a wonderful family-practice doctor. When we first joined her practice, I checked her credentials. I have five years of experience with her now. When she tells me one of my kids has an ear infection and writes a prescription, I do not launch an investigation. I go get the antibiotics (though if you don’t know the doctor you should ask questions, since 84% of ear infections resolve without antibiotics and more than 95% of children get prescribed them anyway). That is a low-stakes decision inside an iterative relationship where she has earned trust over time, and spending my skepticism there would be irrational. But if it were a high-stakes call, I would ask for a second opinion, and this is the critical part: my doctor would say yes immediately. Her response to a request for a second opinion would be, “Absolutely, let me refer you to a couple of specialists, and if they disagree with me I want to know, because it makes me a better doctor.” The willingness to be checked is itself one of the best signals of trustworthiness there is.

That’s the whole disposition. Trust, but verify, even (or maybe especially) yourself. It takes the radically moderate virtue of humility and applies it to everyone, equally. It’s how I treat elected officials, and I don’t see why the people diagnosing my kids or testifying at a murder trial should be exempt from a standard I apply to a city council member. For most of us that verification doesn’t mean re-running the science ourselves. It means checking the checkers: the credentials, the track record, and whether the claim has survived people trying to knock it down.

Verification as insult

There’s a longer-term danger to the “trust the experts” sloganeering that I heard during COVID and increasingly again now that various scientific fields are under politicized attack. The slogans do real damage because “trust the experts” has trained too many disciplines to treat any question as an attack. Push back on a diagnosis and you’re a difficult patient. Question a consensus and you’re a conspiracy theorist.

Surgery is a great example because the resistance there had nothing to do with politics. Checklists in medicine got fought for years on the grounds that they were insulting, that they were a nurse’s tool, and that a trained physician did not need a piece of paper to tell him his job. Then the WHO ran the numbers on a nineteen-item surgical safety checklist across eight hospitals and found that deaths fell from 1.5% to 0.8% and complications from 11% to 7%. Ironically, the very experts who insisted they didn’t need to use checklists wanted other experts to use them: 93.4% of the operating-room clinicians said they would want the checklist used if they were the one having the operation. We know verification works. We would just prefer to be on the other end of it.

The counter-example is eyewitness identification (and one where I’ve been wrong before and am glad to be corrected). For thirty years the research consensus was that eyewitnesses, and especially confident eyewitnesses, could not be trusted. Then researchers looked harder and found that the problem was mostly the procedures: non-blind administrators, suggestive lineups, confidence recorded after the witness had already been told she did well. Clean up the procedure and a confident identification turns out to be accurate around 97% of the time, in the field as well as the lab. And then last August another group ran the numbers and said no, that estimate is too generous and here’s why. And they’re still arguing about it.

That’s what a field looks like when verification is working and where the scientific method trumps bare credentialing.

The same dynamic shows up even worse in politically charged spaces. The New York Times spent a year reporting on the University of Michigan’s DEI program and found that after a decade and a quarter of a billion dollars, students and faculty reported being more frustrated and more divided, not less. Many DEI “experts” offered expensive expertise and trainings that were useless at best and made racial animosity worse rather than better at worst. (Musa al-Gharbi’s analysis here is still one of the most comprehensive I’ve seen).

Because experts are often granted status simply for being experts, they don’t enjoy being questioned. And because humans are primed to respect status, particularly within our tribe, we often don’t question expertise at all. We’ve built a culture where the reasonable act, “trust, but verify”, gets coded as hostility, and the expert is incentivized to resist exactly the verification that would make the whole system stronger.

I can think of two fixes, but both are hard to pull off.

The first is cultural and it’s on us. Retire “trust the experts.” Retire “I believe in science.” Replace them with something closer to how knowledge actually works: this claim has been vetted, this person has the chops they say they have, and here is how I checked, with the caveat that the information might change and all knowledge is incomplete.

The second is on the experts, and it’s structural. We have to build in the expectation that expertise should be asked to justify itself, and train experts to feel that as normal rather than as a personal wound. It’s not an insult to ask a doctor why she’s recommending this procedure (or even if they’re operating on the right body part). It’s not an insult to ask a forensic analyst to defend a method under blind conditions (in fact, justice demands it). Having your expertise questioned should be a standard part of becoming an expert, in medicine, in law, and in science. That means we need to build in more of the machinery that actually catches error: adversarial collaboration, rigorous blind testing before we let something count as evidence in a courtroom, and real verification mechanisms for the experts themselves (maybe more emphasis on standardized exams and less on prestige-based degrees). Blind review is what finally caught the bite-mark people. We shouldn’t have needed a generation of wrongful convictions to run it.

Shaming feeds error

“Trust the experts” has deeply damaging political effects too. When you treat people like idiots, they stop listening to you about everything else, too. I watched this happen in real time during COVID. I was not a denier of anything. My kids were masked, and they were among the first children vaccinated. But the way vaccine skeptics and people with ordinary, good-faith questions got treated - as morons, as enemies, and as a problem to be managed rather than people to be persuaded - did lasting damage to public trust and to the scientific community’s own ability to do its job. Treat people like idiots and they’ll ignore you the next time, even when you’re right.

The raw milk debate is maybe the most recent example. There’s solid evidence that pasteurization is a very good thing. My kids drink pasteurized milk, for whatever it’s worth. But you cannot get from “this person is wrong about raw milk” to “therefore they should shut up and defer to the medical establishment on everything forever.” That’s not how humans work. The more we mock and shame people for their suspicions about official scientific claims, the more we push them, not toward the scientific method, but straight into the arms of the actual conspiracy theorists, who are always happy to welcome anyone who’s been laughed out of the room.

It’s also worth noting that many of those who are the most shamed are those whom medicine has failed the most. Their skepticism of expertise is, if anything, well-earned. I see people dismiss middle-aged women’s interest in alternative medicine and supplements as some kind of hysterical lunacy, when until (very) recently the medical establishment had decided that a massive hormonal shift that affects everything from energy to cognition (not to mention heat tolerance, and ask me how I know) doesn’t exist or isn’t worth devoting research time to. Menopause is not a priority of a single one of the NIH’s twenty-seven institutes, which is a remarkable fact about a life stage half the population goes through. Women’s health got 8.8% of NIH grant spending between 2013 and 2023, and that share has been falling even as the agency’s budget grew. Menopause research got $56 million in 2023, out of roughly $47 billion. No wonder women are desperate enough to try crystals or whatever bizarre over-priced garbage Gwyneth Paltrow’s Goop sells. At least the latter pretends to care.

And sometimes, of course, the person laughed out of the room is right. Lady Mary Wortley Montagu survived smallpox in her twenties, encountered the practice of inoculation while living in the Ottoman empire, and brought it back to England, where she had her own children deliberately infected with a mild dose to protect them. To respectable European medicine at the time, this was insane. Deliberately give your children smallpox? It was probably the raw-milk-drinking of its day. She was also completely right, and inoculation was a direct ancestor of the vaccines that eventually erased the disease from the earth.

Leaving room in science for the fringe isn’t just important politically; it’s how science actually works. Shame the people with strange questions and you don’t just wound a few feelings. You slow down the whole enterprise, because some of the people getting laughed at are actually right or at least partly right or maybe just pointing in the right general direction. Take their belief seriously and test it honestly, and if it’s wrong the method shows why, and if it’s right you’ve captured it. Ridicule it and drive them out, and they build a parallel world where quacks and real dissenters are indistinguishable, and you’ve lost the signal along with the noise.

I’m much more deferential to expertise than my grandmother was, not because I am smarter or better educated than she was, but because I’ve lived in a comfortable world where, by and large, experts could be trusted. It’s not the world my grandmother grew up in. So before my comfortably middle class friends throw disdainful “trust the experts” signage at apparently ignorant sectors of the population, it’s worth considering how little expertise has done for those folks and how much the expertise that has worked for them is currently invisible. People really do occupy different parts of our shared landscape, and their location on that landscape can help explain why one person finds “trust the experts” reassuring and another finds it deeply disturbing. In reality, given all the evidence, everyone should find themselves much more in the “trust but verify” middle.

I mostly hear these slogans deployed from the left, in large part because I live in a left-y bubble, but I still think they’re one of the most harmful habits we have. Not because expertise doesn’t matter. It’s precisely because expertise and evidence matter enormously that we cannot afford to allow them to be polarized and politicized in the simplistic way they have been. A demand to “trust the experts” is how you get an expertise too brittle and too coddled to deserve the trust it demands. It’s also how you thrust people on the fence right into the arms of the conspiracy theorists you’re condemning.

“Trust the experts” is an inaccurate understanding of how expertise actually works. It corrodes the legitimacy of the very experts it’s meant to defend. And it pushes distrustful people further into the margins with every repetition.

Most humans dislike being patronized and looked down on and they will seek out the company of those who do not treat them with active disdain. Who knew?

So the next time you’re tempted to call out some conservative raw milk aficionado or vaccine skeptic as ignorant or dumb, it might make sense to ask the following question: does this person, across the sum of their life, have good reason to trust experts? And does calling them stupid or ignorant help convince them that experts are trustworthy, or does it just further undermine that trust in the long-term? Because it’s better to be effective than right, I’d say giving each other grace while trusting but verifying is the better bet. And it’s the best way to preserve a meaningful deliberative democracy in a time of deep division.

Your turn

One question I keep asking myself: what is the thing I currently trust because the experts told me to, that I’ve never once verified for myself? I’d love to hear your version of this. I’m not saying distrust it. I’m asking whether you’d know how to check. Or just tell me how I’m wrong or what I missed!

Tell me one in the comments. I read ‘em all. If you’re new here, this is the second half of a pair. Part 1, What’s Our Bloodletting?, is where the argument starts. Subscribe for the next one, and if this piece clarified something you’ve been thinking about, restack it so it finds the next person.

Leave a comment

Subscribe now

Share

As Anthony Fauci testifies before the Senate this week in a deeply partisan spectacle that bears no small resemblance to a modern witch hunt, it’s worth thinking about how science became partisan in the first place. Certainly, mistakes were made, as the saying goes, during COVID. But it’s also the case that people have always turned to science to justify various partisan claims. Far from being morally or politically neutral, science and the scientific process is and will always be embedded in social, historical, and political moments that challenge our claims about objectivity and pure standards of truth. Even individual preferences affect what we call a scientific truth in critical ways. Science, like everything else in this world, exists in all four dimensions. And until we recognize that reality, we will continue to find ourselves (and our opponents) using and mis-using scientific “truth” and “consensus” in all kinds of dubious ways.

The Ambiguity of Chemo

A few years ago a physician named Lisa Masson got a text from her daughter with a bar-game question: what medical treatments are the norm now that our grandchildren will consider barbaric the way we now consider bloodletting barbaric? Her answer was chemotherapy.

Bloodletting was the gold standard for confident, mainstream, credentialed medicine; that it killed more people than it helped (including President Washington) was a wrinkle that no one noticed, lacking control groups. It wasn’t fringe science (though we could ask whether it counted as science at all). It was the standard of care.

Questioning chemo on the other hand sounds like something a crank would say (and cranks, have, in fact questioned chemo). But even those cranks have a point: chemo is poison and that’s exactly why it works. We’re basically introducing something toxic enough to kill the cancer, and crossing our fingers that your healthy cells can outlast the poison long enough for it to finish the tumor first. When it works, it is one of the great achievements of modern medicine. Along the way it’s often brutal, causes extreme suffering and can even kill people sooner than the cancer would have. Both can be true at the same time.

Asking the Wrong Questions

Ask whether chemotherapy is a legitimate treatment for cancer is kind of pointless because the answer is: it depends. Which cancer? Which patient? How advanced? How many rounds have they already survived? How strong is the body you are asking to absorb the poison?

For most treatable cancers, this isn’t really a close call. The evidence for chemo is overwhelming. The contextual point is about the margins and the future, not the median case. But the margins are real (I keep trying to convince people of this fact, but they remain resistant). There is good evidence that for some patients approaching the end of life, moving to early palliative care and less aggressive treatment at the very end produces a higher quality of life, and in some cases a longer one, than pushing chemo to the last possible day. In a landmark trial of metastatic lung cancer patients, the group that got early palliative care alongside standard treatment, and less aggressive treatment at the end, actually lived meaningfully longer. So a person who says “I don’t want any more chemo” isn’t necessarily a nut job who has rejected science. They may be making a rigorously science-based decision that fits where they are standing on our shared 4D landscape, which is a different place on the landscape than a younger, healthier patient whose cancer hasn’t advanced as far.

When I say all medicine is contextual I don’t mean there’s no such thing as strong evidence or clear standards of care. But all medicine is relative to a specific patient at a specific time with a specific set of available alternatives. A hundred years ago, offered chemo against the surgical butchery that passed for cancer treatment, you would have grabbed it with both hands. A hundred years from now, when we may be able to build targeted cells that hunt tumors without poisoning the whole body, “just give the patient chemo” might sound exactly like “just bleed the patient.” The science didn’t change its mind because it was stupid before. Copernicus wasn’t an idiot. Brilliant people have held incomplete pictures of the world through no fault of their own, working with the tools they have at the time.

Context Isn’t Relativism

Does chemotherapy save lives? Yes, demonstrably, with mountains of evidence. Does it also cause enormous suffering, and is it the wrong call for some patients? Also yes. Is it the best tool we have right now for a lot of cancers? Yes again. All of those live together, and none of them cancels the others. Saying “the science is contextual” does not mean the science is fake or that every opinion is as good as every other. It means the opposite of a slogan. “The science” is not a single settled object you either believe in or don’t. Science is a method, a way of generating knowledge, applied inside specific disciplines, working from whatever evidence exists at a given moment, some of it more solid than the rest.

It’s this context that yard signs that proclaim “science is real” flattens, often in deeply partisan ways. The same person who vehemently defends the COVID response might be the same person expressing skepticism about some other mainstream medical practice. (And of course, not all mainstream medical practices are scientifically based, which I’ll touch on next week).

I believe in the scientific method. But “I believe science is real” treats a moving, contested, always-incomplete process as though it were a fixed creed you sign. It isn’t. The honest position isn’t declaring faith in the consensus to flout the other side. It’s understanding that any current consensus is a snapshot: it’s real, likely useful, the best we have, and certain to look partial from a hundred years out. Every expert, however good, is standing on a single point of a much larger and still-contested map, and there are things even the best of them don’t know they don’t know.

That’s partly why I can say that the current GOP attacks on Fauci are deeply disturbing and an abuse of power and I can also say that Fauci’s COVID response was imperfect and required a real scientifically grounded postmortem so we can learn from all our mistakes. There’s nothing inherently contradictory about those positions. But many people will see them as contradictory because - whether they’ve admitted it or not - they believe science supports their “side” when in fact science operates in 4D and doesn’t really takes sides at all.

In my post next week I’ll attack the phrase “trust the experts” and argue that it’s one reason (among many) why Fauci is in the position he’s in right now. That’s not to blame Fauci himself for what is a deeply unjust persecution, but instead to say that we’ve created incentives in public health for exactly this kind of partisan attack. Subscribe if you haven’t already so it shows up in your inbox, because those incentives are where this argument stops being philosophical and starts being about innocent people in prison.

Your Turn

As always, let me know what I missed! Why do you think science has become so polarized? Is it worse now than it was during, say, the Scopes Monkey Trial or has science (as I claim) always operated within complex social and political contexts? What does that mean for how scientists should think about their work and their messaging? I’d love to hear your thoughts in the comments and if you haven’t already, please subscribe and share. Reader shares are still how most people find my work in this crowded space.

Leave a comment

Subscribe now

Share

After my talk this weekend, a man asked if I was safe.

I blinked. I didn’t understand the question. He was the part-time mayor of a small rural borough, a veteran, and he had come up to talk once I’d finished speaking to the Pennsylvania Mayors’ Association. Somewhere in the conversation he stopped and asked, with real concern, whether a woman who writes and speaks publicly against mainstream ideas was safe from threats and harassment. It honestly hadn’t occurred to me. I get my share of vicious internet comments. I have never once felt unsafe.

He feels unsafe almost constantly.

He came to this country from somewhere else and he doesn’t look like what many of his neighbors pictured when they pictured a mayor. He runs a borough that is overwhelmingly white; a sizable minority of his neighbors won’t answer the door when he knocks. He was put in office by a silent majority who wanted better government but who mostly won’t say so out loud, because they’re afraid of what the neighbors would think. He lives with a steady drip of whisper campaigns: that he’s corrupt, that he’s a criminal, that he’s acting in bad faith. He goes door to door to serve people, some of whom have made it plain they’d rather he didn’t exist. He makes maybe a couple hundred dollars a year as mayor while juggling his own full-time job on the side. And he keeps doing it, because he thinks the work is worth doing and because he knows that if he doesn’t, no one else will.

I left that conversation inspired and also heartbroken. The heartbreak isn’t entirely about him. It’s about us. We have built a country that runs on people like him and hands them almost nothing to run on.

The room where it happens

There are close to a thousand mayors in Pennsylvania. Most of them you will never hear about. They aren’t running for anything higher. They spend their evenings on feral cats and a borough website and mental health crises and the problem of how you get a neighbor into drug treatment when the nearest open bed is two counties away. The ones I listened to were dealing with housing shortfalls, with immigrant families who needed a translator, and with the slow economic transitions that hollow out a main street and leave a town trying to figure out what it’s for now. A number of the younger ones told me they only ran because the previous leadership had been so bad that somebody had to, which is its own quiet kind of civic faith.

It was also a mixed room. Politically these were people who would disagree with each other, and with me, about nearly every national question you could name. And they were spending their lives on their neighbors’ worst days anyway.

That combination is exactly what the national story tells us is impossible.

Local works, and not because the people are better

When people look at Washington right now and conclude the country is coming apart, I get it. I write about it all the time! But I also think it’s a mistake, and that room full of mayors is why. The dysfunction in Washington tells us an enormous amount about our national institutions: their incentives, their veto points, the perverse machinery that turns every disagreement into a war. It tells us almost nothing about who we are as a people. Because who we are was in that room, getting things done across exactly the divides that paralyze Congress. The data backs my anecdotal evidence. Americans trust their local governments far more than Congress (67 percent to 32, in Gallup’s 2024 survey), and a survey of more than 1,400 local officials found the smallest communities the most insulated from partisan dysfunction.

So what’s different? Not the people. The same American who is a partisan combatant online is a pragmatist at the zoning meeting. (Though I’ll admit I do think our institutions create harmful selection effects at the national level, so in some sense the people really are different). What changes is the institution he’s standing inside, and local government happens to be built (mostly by accident, partly by design) in a way that pulls people off the one-dimensional battlefield and back into the actual world.

Every mechanism is the same move: it forces you off the flat left–right line and back onto the four-dimensional landscape where people actually live. I call this way of seeing 4D thinking and it’s the kind of thinking that keeps local communities running even while Congress can’t tie its own shoes.

In my talk, I focused on four primary mechanisms that make local governance resilient, but there are likely many others too.

To start, local government runs on concrete problems instead of abstract ideology. National conflict is fought over symbols (identity, values, what kind of country we are), and symbols are binary (#blacklivesmatter vs. #bluelivesmatter!) and non-negotiable by design. A pothole is not. Snow removal is not. Neighbors dying of overdoses are not. The moment the question becomes where does the money go you are forced to weigh real tradeoffs, real costs, real people who gain and real people who pay. You can’t grandstand your way through a budget. Complexity does the work that civility lectures never could.

Local government makes you govern alongside people you can’t reduce to a caricature. Affective polarization runs on a fantasy: that the other side is a monolithic, cartoonish enemy. That fantasy cannot survive the person across the table also being the person you see at the grocery store and the youth soccer game. Your opponent in November is your kid’s coach in the spring. The identities refuse to stack. Party never gets to become the whole of anyone, because everyone in a small town is visibly, stubbornly more than one thing.

And local government tends to keep the party label off the ballot, or at least turn its volume down. Strip out the partisan tag and you’ve removed the single strongest cue telling people to sort themselves into teams. What’s left is the work.

Put those together, and local government turns out to be a resilient institution. Not resilient the way we usually mean it, not tough, not stubborn, but resilient in a specific sense. It’s a structure that takes different people, with different values and interests, and metabolizes that difference into decisions a town can live with. It runs on disagreement rather than requiring everyone to agree first. That’s what the best of our institutions do; markets, courts, science, and federalism all work this way. They’re the opposite of the brittle national institutions that cave with every election and every executive pen.

We talk about “small town America” as though it names a single thing. Spend a day with the people who run these places and that dissolves. Some of these boroughs are booming and unprepared for it; others lose population every year and watch the tax base shrink. Some have large immigrant communities and a real need for translators; some are almost entirely homogeneous and aching as their young people leave. A fix that works in one borough will fail in the one twenty miles down the road, because the two only look alike from a distance.

This is deep pluralism, and small towns are one of the cleanest examples of it we have, which is also why local government works. Nobody in that room was waiting for a national template. They were adapting, constantly, to the specific place they’d been elected to serve. A thousand small laboratories, each learning something the others can borrow.

None of this makes local government bulletproof. National politics is starting to seep down, especially in bigger cities and in the culture-war fights that have colonized school boards, where the same identity battles that flatten Washington flatten the curriculum meeting too. And local government can fail - and frequently does - when it’s isolated, lacks support and accountability, and when corruption is allowed to fester. (One of the most common narratives I heard from younger mayors who stepped up was being appalled by the corruption of the previous administrations). The resilience is real, but it is not guaranteed. That’s an argument for shoring up the conditions now, while they mostly still hold, rather than waiting for the seep to become a flood.

Conditions, not character

Toxic polarization is not our fate, and it’s not a flaw unique to the American character. It emerges from specific conditions — and conditions can be changed. Change the structure of the interaction and you’d be surprised what becomes possible. That was the lesson I gave that room, and it is the hopeful half.

The mayor is the other half.

Because the same logic runs in reverse. If good behavior comes from good conditions, then we cannot keep pretending that bad conditions can be redeemed by good people. And right now, at the level of local government, that is precisely the bet we’ve made. We are running the most resilient institution we have on the personal virtue of individuals who receive almost nothing in return: no budget, no staff, no networks, no protection, and a standing invitation to be harassed. It works, when it works, because a veteran decides the harassment is worth eating so his neighbors get a functioning town.

That is not resilience. That’s a heroism tax, and any system that depends on heroes is by definition brittle, because heroes are rare and they burn out and the whole thing fails the moment they stop showing up. We’ve built the counter-example to national dysfunction and then starved it of everything that would let it last.

I’ve made a version of this argument before, about immigration of all things. My point there was that you can’t run a humane immigration system on enforcement and slogans alone; you need absorption infrastructure, the unglamorous capacity that lets a community take people in. Local democracy needs its own absorption infrastructure, and it’s the same list: capacity, resources, and above all networks, the connective tissue that means a mayor going door to door in a hostile town is not doing it completely alone. My mayor is almost entirely isolated. He shouldn’t be. There is no reason, other than our inattention, that a person doing this work should have to invent it from scratch and absorb every cost himself.

If we want the conditions that produce pragmatic, cross-partisan, problem-solving citizens, we have to fund the conditions. You do not get resilient institutions by admiring the people holding them up with their bare, bleeding hands.

Education in self-governance, if we can keep it

Tocqueville saw this coming almost two hundred years ago, and he said it more sharply than we let ourselves say it now. The township, he wrote, is to liberty what the primary school is to knowledge. It’s where a free people actually learns self-government — not the theory of it, the muscle of it — and where each generation relearns the habit from the one before. Take that away and the knowledge doesn’t survive at the national level on its own. You can’t lecture a country back into self-rule from the top. The habit is grown locally or it isn’t grown at all.

Which is why I don’t think the man who asked if I was safe is a feel-good story about one good person, and I’ve resisted writing him up as one. He’s evidence about an institution. He is what the school of self-government looks like when it’s still running and no one has bothered to keep the lights on. Every time he knocks on a door that won’t open, he is teaching his town something about what a citizen is willing to do for the people around him, whether they deserve it or not. That is the most important education happening in that borough, and it is being delivered by a man we pay two hundred dollars and leave to fend for himself.

The dysfunction in Washington is real, and I’m not asking anyone to pretend otherwise. But it’s a symptom of brittle institutions, not a verdict on the American people. The verdict on the American people was in that room in Pennsylvania: mixed, pragmatic, exhausted, and still showing up. On that evidence we are in far better shape than the headlines will ever let on. The question isn’t whether the people are up to it. They plainly are. The question is whether we’ll build the conditions worthy of them, or keep asking a handful of unpaid heroes to hold the republic together by hand.

What you can do

We can’t keep running local democracy on the personal heroism of a few people. So while the problem is bigger than any one of us, there are some things you can do to help build the conditions instead, on whatever time you have.

Today (five minutes, from your couch)

• Name them. Look up who your mayor, council or township supervisors, and school board members actually are. Most people cannot name a single one. You cannot support a person you cannot name.

• Send one thank-you. Message one local official and thank them for a specific thing. Not the office in the abstract, but the pothole they fixed, the four-hour meeting they chaired, or the budget they balanced.

• Fund the local signal. Find your town’s newsletter, beat reporter, or borough page and subscribe. If it takes money, pay. A town that can see itself governs itself better.

Tomorrow (this week’s errands)

• Read one agenda. Pull up the agenda for the next council or school board meeting and see what your town is deciding this month. It is almost never what cable news says it is.

• Introduce the person before the party. When a neighbor brings up the council member they cannot stand, name what else that person is. The coach. The veteran. The one who plows the road. Caricatures collapse on contact.

• Ask what they need. Go back to the official you thanked and ask them directly what would actually help. Then close your mouth and listen.

Next week

• Show up once, in person. Sit through one public meeting, including the boring parts. Presence is the cheapest support there is, and officials feel the difference between an empty room and a full one.

• Connect one wire. Isolation is what breaks local officials. Introduce yours to one other person who can help: another official two towns over, a reporter, a nonprofit, someone who already solved the problem they are stuck on. This is the network I keep insisting is missing, and you can lay one strand of it yourself.

• Say the quiet part out loud. If you are part of a silent majority that wants better governance but stays quiet for fear of the neighbors, stop being quiet. A yard sign, a comment on the record, a letter to the editor. Silence is what leaves good officials exposed.

Next year

• Fill a seat. A startling number of local offices run uncontested. Run for one, or find the person who should and back them the way the younger mayors in that room got backed.

• Turn heroism into infrastructure. Help start or fund the connective tissue: a county association, a training, a small stipend fund, a mentor paired with every new official. The goal is a system that no longer needs a hero to function.

• Change one condition. Pick a single structural fix and push it locally. Keep party off a race where it adds nothing. Add a neutrality standard to council rules. Fund a reporting beat. Emphasize conditions, not character, all the way down.

• Come back and ask again. Twelve months from now, ask the same official the same question: what do you need now? Sustained attention is the rarest resource local government ever gets, and it is the one entirely within your power to give.

Your Turn

Tell me about a hero in your town. You most likely have one: a supervisor, a council member, a school board volunteer; whoever is doing the invisible, unglamorous, load-bearing work of self-government for almost nothing. Pick one thing on that list and do it. Sometimes there won’t be much you can do. But at the very least they’ll know they aren’t doing it alone, and not being alone turns out to be most of the infrastructure we’ve forgotten to build.

If this put a face to something you already felt, send it to someone who serves, or to someone who should. That is usually how the people doing this work find out they aren’t the only ones. And please: If you like what you read, subscribe and share! Shares are how most people find my work.

Read next: America Is All Right. The Trump Administration Isn’t., the longer case for why the national dysfunction tells us about our institutions, not about who we are.

Leave a comment

Subscribe now

Share

[I’m returning from a much-needed vacation this week and pulling some oldies but goodies from the archives so they get a bit more love and attention. This one was originally posted Nov. 4, 2023 when I had a scant 120 subscribers, so it’s time to brush it off again. I’ve made some minor edits to bring it into 2026 in light of changes in my own thinking but the original is still available here. There’s also a follow up post, Landscapes, Binaries, and Scales, where I respond to some of the criticisms of the original, but read this one first!]

I just got back from giving a talk on my 4D framework for radical moderation and the political/social landscape metaphor that I use as the foundation of that work.

Here’s a very brief summary of 5+ years of work:

In general, talks are a great way to figure out what I don’t understand about my own arguments and what I understand but I’m not communicating well to other people.

One question from the audience that I didn’t answer well was “how is this landscape an alternative to the binary or the more complex frameworks that you’re complaining about?”

I started the presentation with the binary below, arguing that it’s not just empirically wrong, but it’s also morally wrong. It not only doesn’t reflect human political, social, and moral reality, but it leads to morally bad outcomes.

But, one of the audience members asked (or rather, the question implied), “that’s outdated, so what about more complex articulations of the political landscape, like the political compass?”

Great question! So what about the political compass above? Isn’t it more complex? Doesn’t it have three dimensions? Isn’t it more accurate? Even more specifically, the question was, “If this isn’t helpful, what’s your alternative? Where do values map onto your landscape?”

This question goes to the heart of the entire project, so I’m glad this person asked. It was clear that he thought I was presenting the complex landscape as a *different* way to locate or organize political beliefs. And that’s exactly what I’m *not* doing.

The radically moderate landscape isn’t an alternative to binaries or compasses at all, but a radical rejection of that way of thinking about politics. And that’s because social, political, and moral values don’t map neatly onto human life the way we think they do. There are no permanent locations of clusters of political ideas because that’s not how humans navigate their shared world. And, I’ll argue, as helpful as these tools are for intellectuals trying to map out the broad landscape of political ideas, they may be actively harmful in practice.

Thank you for reading The Radical Moderate’s Guide to Life. This post is public so feel free to share it.

Share

Are binaries/compasses accurate reflections of real human moral/social/political worlds?

Let’s start with the empirical case.

There is no stable location of any of these political values in the real human social, political, and moral world. We can distort things in order to shove events and people into these categories, but it’s always a distortion because it requires jettisoning so much complexity along the way.

The real social, moral and political landscape looks a lot more like the photo at the top of the page. There are lots of ways to navigate that landscape, lots of decisions that need to be made, and many idiosyncratic values and goals that might animate those decisions. Pluralism runs deep in that landscape.

Moreover, even the most complex of the political paradigms don’t have a fourth dimension, which means that they locate people and human problems at a particular point in time. But social, moral, and political problems morph in really important ways over time.

To be fair, the goal of the political compass isn’t really to think about the entire human social, political and moral landscape. It’s (usually) a way to think about the power of the state or how particular politicians or political thinkers think about the power of the state. And in that very narrow sense, it can be useful. But too often we expand the compass or the binary and stretch it into areas of human social life where it has no real business going.

There are a range of complex human social, moral, and political problems where the compass doesn’t really help us locate anything like solutions. Children and the family, the mentally ill or severely disabled, small local communities, and friendships are all areas where the compass fails to give us much in the way of guidance.

But, someone might reply, that’s only because you’re using it in a way that it’s not meant to be used! Sure, that’s true. But that’s how many people *do* use it. Precisely because the political spectrum or compass is a convenient way to think about the political world it becomes tempting to place oneself on it and once you do that it’s much easier for tribalism to kick in. Once that’s kicked in it’s even more tempting to try to align your beliefs to your particular location on the spectrum, without realizing that you’re distorting yourself in the process.

The political compass is a lot like the body mass index or BMI in that sense. BMI used to be a basic and somewhat helpful screening tool - it was meant as a way to invite someone to ask more questions or do some more digging. But precisely because it was so simple, it very quickly became a diagnostic tool, labeling someone as obese simply based on their height-to-weight ratio. This is obviously very dumb because a simple height-to-weight ratio obscures an enormous amount of important variation, including muscle mass and bone density. That variation is crucial to assessing any person’s actual health status. And that’s very much the case for the political spectrum or compass too. It might be helpful as a first-round screening tool, but when we accept it as a fully accurate depiction of our social, moral, and political landscape we lose all the complexity that makes that landscape meaningful.

Academics, pundits, and wonks think about the political world in terms of spectrums because it’s a convenient way of organizing our ideas. Ironically, that convenience ends up obscuring a lot of truth and a lot of complexity. And that obscuring can actually be harmful, particularly when translated into the way the average person thinks about our shared political life.

Do binaries/compasses help us make better decisions?

Elites are far more polarized than the rest of the population and that may be at least in part because our simple tool has lured us into thinking about ourselves on these kinds of spectrums. As another audience member pointed out, and which I argued in a somewhat different way not long ago, binaries and even the more complex political compass trick us into thinking that the people on the other side of the binary or the compass are wrong, bad or even our enemies. They animate our tribalism, which in turn gets us thinking about political problems and their solutions in polarized terms. Once you’ve identified a politician as authoritarian, you’re much less likely to accept any of her policies, even if some of them happen to not be authoritarian at all.

The important questions we want to ask about these binaries and compasses are: Do they make us better humans? Do they help us navigate our landscape better? Do they provide better outcomes for humans trying to navigate their worlds?

The answer is “probably not.” I can’t think of any examples of real-world situations where someone’s location on a political compass helped them make a more humane decision, particularly in the real world. That doesn’t mean it doesn’t happen, but it’s probably not common. If anything, locating politicians and voters and ourselves in a stagnant value framework means that we’ve pinned them and ourselves like preserved butterflies in a particular political context. We’re no longer moving through an environment and making decisions together about how to navigate that environment, but we’re stuck. And if we’re stuck we’re not thinking creatively about how to solve deeply important human problems and we’re not working with people who we see as located in an incompatible part of the political spectrum.

Unfortunately, the political compass is probably here to stay. Because it’s convenient and easy, we’ll still stick people on it and it will continue to animate tribalism and create political enemies when what we need most right now is creative four dimensional thinking. We need people building bridges and clearing away barriers and asking other people what they want out of our shared landscape, which we’re not going to do if we’re pointing at each other across invisible lines. So I’m not sure what the solution is, other than to keep pointing out the complexity and beauty and messiness and colorfulness of our shared moral, political and social landscape in the hopes that it can lead us to a better way.

Your Turn

There’s a lot more to say here and there were some other really helpful questions that I’ll tackle in subsequent posts, but I’ll leave this question for readers and hope that they respond in the comments: is the political spectrum or compass actually helpful? If so, when and in what contexts? And if not, how do we convince people to let it go?

And if you’re in need of more radically moderate summer reading, read the follow-up post to this one, Landscapes, Binaries, and Scales, and let me know what you think about that one too!

The Greeks had a hard little rule about happiness: don’t call any man happy until he’s dead. Solon says a version of it to Croesus. The chorus says it at the close of Oedipus Rex, watching a man who had been king, solver of riddles, and savior of the city, end up impoverished, guilty, and blind by his own hand:

Let every man in mankind’s frailty consider his last day; and let none presume on his good fortune until he find life, at his death, a memory without pain.

(Radically Moderate trivia: This was, incidentally, my senior yearbook quote, which tells you a lot about my moodiness level at age 17).

The Grecian point isn’t that life is misery. The point is that a life can be undone at the end. You can lose your fortune, your family, your sight, and your good name, right at the finish, and the losing rewrites everything that came before. You don’t get to be called happy at the start of the fifth act. You have to make it to the curtain first.

When I heard the news of Lindsey Graham’s death at age seventy-one, that line came immediately to mind because Graham is a nearly perfect case. Here was a man with a long career in public office, a military lawyer’s mind, a Bronze Star, more than two decades in the Senate, and a genuine reputation (for a while, at least) as one of the few Republicans who would work across the aisle and mean it. And yet I’d argue the memory he leaves behind is not a savior of the city but instead its most useful undertaker.

What Graham Used to Be

Graham was, for a long stretch, a recognizable and even admirable kind of conservative. Small government, with the exception of an outlandish defense budget. A lawyerly, institutional cast of mind. A willingness to cut deals on immigration and judges and climate that made his own base furious and made the Senate function slightly better than it otherwise would have. He was, whatever else, a legislator, someone who understood that the machine only works if somebody is willing to turn the crank.

He also saw Trump coming, which means he couldn’t plead ignorance. Graham called Trump (among other things) a “kook,” a “jackass,” and “unfit for office.” He said, on the record, that if the Republican Party nominated Trump the GOP would be destroyed, and that it would deserve it.

He was right. He knew. That’s the tragedy of him. Oedipus, while also forewarned, at least had the excuse of ignorance of who he actually killed. Graham read the prophecy, out loud, on cable television, and then walked straight into it with his eyes open. Hamartia, as the Greeks called it, is the fatal flaw that brings a great man down. It’s in many ways a critical but often overlooked trait or choice that takes an otherwise happy life and destroys it. For Oedipus it was over-confidence. For Graham it might have been hubris: that despite reading the danger better than those around him, he believed that reading it was the same as controlling it. It could have been other things though: pride, fear or simple miscalculation. Whatever the reason, Graham had a choice in life that will forever change the way his legacy is understood now that he’s dead.

The Hinge

In every tragedy, there’s a hinge: a moment where it could have gone the other way, and Graham’s was the winter of 2020 into 2021. I wrote about it at the time. I’ve gone back and reread what I said then, half hoping I’d been too harsh. Given what we know now, I’m pretty sure I wasn’t harsh enough.

At the time the puzzle was real. Graham was safe. Six more years in the Senate, no primary to fear, and no election that required him to grovel. The usual excuse for a politician’s cowardice, the next campaign, simply didn’t apply. And yet there he was, insisting the movement “needs to continue,” that “Trump-plus is the way back,” that a Trump daughter-in-law mind-bogglingly “represents the future of the Republican party.” A man with nothing to lose electorally chose to lose everything else instead.

I offered some theories back then, all in the spirit of QAnon, because at some point analysis fails and you may as well have fun. Was he a Trump-built robot? (Ruled out. The Trump Organization has never successfully innovated anything that wasn’t a tax dodge.) Was he being blackmailed? (I kept that one open.)

More seriously, I don’t know and none of us likely will. The reasons have never been clear, and I’ve come to think the murkiness is itself the verdict. When a man abandons everything he claimed to believe and can’t, or won’t, tell you why, the “why” stops mattering. The abandonment is the fact. Underneath might be the oldest political miscalculation there is: the belief that he could ride the movement without being ridden, that he could use a demagogue like Trump without being used. Every tragic hero is certain he is the exception to the rule that is about to destroy him.

Had Graham chosen differently in that critical 2021 window, he would have a different legacy today, and a far better one. He was positioned, better than almost anyone, to be the adult in the room: safe seat, military credibility, decades of relationships, and a documented record of having called the danger by its name. If a principled Republican revolt was going to have a face, it could have been his. Maybe it would have failed anyway. Maybe the base was already gone, and one senator from South Carolina was never going to call it back single-handed. But that was never the charge against him. The charge is that he didn’t try, and trying was the job. He had the standing to lead. He chose to kneel.

No Longer a Public Servant

We call it “public service,” and for most of Graham’s career the phrase fit, even if you disagreed with his politics. At the end it stopped fitting, and that’s not a mean-spirited rhetorical flourish. It’s closer to a definition. A public servant serves the public. It became clear, and clearer every year, that Graham was serving something else. Not his constituents, who had just returned him for six more years and asked for none of it. Not his party’s future, which he actively mortgaged. Not the republic, which he’d sworn to protect and then abandoned at the one moment protection was actually required of him. Something else. Proximity to power, maybe. The warmth of the fire. Whatever it was, it wasn’t us.

I said it in 2021 and I stand by it in 2026: Graham made a choice to become a different kind of man. He became a politician instead of a leader, a groveling yes-man where there had once been something like a maverick, a principled man who sold his soul for a stake in a game he had to know was rigged against him. What I didn’t fully see then was how much the sale would cost the rest of us, and how little he’d get for it.

The Real Damage

We talk about Graham as though the worst of what he did was done to the country, and there’s plenty of that to focus on. But the country is large and durable and full of institutions built precisely to survive men like him. The country will be, in the long run, fine.

The movement he claimed to represent is another story. The saddest and maybe most consequential part is what Graham’s final choices did to principled conservatism.

I’ve written a great deal over the years about why principled conservatism matters anyway, why a healthy polity needs it, and why its absence makes the rest of us poorer.

Conservative thought, at its best, reminds us that institutions take a long time to build and no time at all to burn.

That a stable and flawed human nature needs guardrails.

That individual virtue and choice matter, even when they’re constrained (they may most when they’re constrained).

That love of one’s own is where love of humanity starts.

Those aren’t political slogans; they’re a real intellectual inheritance. Losing them is a loss for progressives and libertarians and radical moderates too, whether we admit it or not.

Graham didn’t invent MAGA. But he was one of the men who could have stopped it inside his own party and chose to feed it instead. And the feeding was the betrayal, because the thing he fed devours precisely the tradition he spent his career claiming to defend: limited government, checks and balances, skepticism of executive power, a defense of virtue and character, rule of law and mercy for the vulnerable. Watch the movement he championed and count how many of those survive. He helped pull out the load-bearing walls of modern conservatism and then acted surprised when the roof came down.

Ironically, the America First movement Graham threw himself behind is turning on its own establishment, and one of the fault lines is foreign policy, the endless hawkishness abroad that Graham personified more than almost anyone. A movement that sold itself as “America First” is now eating the very interventionism he embodied. This is peripeteia: the tragic reversal, the hero consumed by the machinery he helped build. Graham spent his last political years serving a coalition that was always going to spit him out. He died coming home from Ukraine, of all places, still pressing the hawk’s case for a war his own movement’s base had soured on. He mistook the mob for a home.

Graham’s Last Day

I’m not writing this to erase whatever good he did throughout his life. He passed laws. He served in uniform. He was, by the accounts of people who knew him, funny and warm in a room. The Greeks weren’t saying those things don’t count. They were saying the deeds you’ve done and the character you’ve built don’t get the last word if you throw them away at the end. A memory without pain is earned across a whole life, including the final act, and the final act is the one Graham controlled and got wrong.

So here’s the question: If a man can see the flood coming, name it, warn everyone in earshot, and then spend his last years building the levee in the wrong direction, what do we owe his memory? I think we owe honesty, at least. And honesty says that Lindsey Graham had the chance to be remembered as one of the few who stood in the breach at one of the most important political moments in the history of our republic. He chose instead to be remembered as one of the men who widened it.

Tragedy is supposed to end in catharsis, the pity and fear that leaves the audience somehow cleansed. Today, as I mull over the tragic choices of Lindsey Graham’s final years, I don’t feel cleansed. I feel the smaller, colder and angrier thing you feel watching a man throw away what was his to keep.

Count no man happy until he’s dead.

Graham made it to the curtain.

I don’t think he made it home.

Your Turn!

What do you think? Too harsh? Not harsh enough? Theories on why Graham changed? Was he blackmailed, broken, or just afraid of the cold once he stepped away from the fire? Tell me in the comments. If I said something you think others should read (for good or ill), please share it. And subscribe if you haven’t; reader shares are still how most people find my work (and I appreciate each and every one).

Leave a comment

Subscribe now

Share

[This post originally appeared in Reason Papers vol. 46 on the 250th of the Declaration. I’m reposting it the week of July 4th partly because the timing is appropriate and also because I’m on family vacation this week! Enjoy and as always drop thoughts, feedback, questions, and anything else in the comments.]

The Declaration of Independence turns 250 this year, and everyone agrees it is worth celebrating. It is. But the celebration obscures something the founders themselves would have taken for granted: the Declaration’s rights language was not radical in the way we tend to imagine, as a theoretical leap, deduced from abstractions. It was instead deeply moderate: concrete, practical, and grounded in rights the colonists had already claimed for generations. Its very success has created a problem. We have taken rights forged as tools for resolving specific political conflicts and transformed them into weapons for escalating those conflicts. Edmund Burke spent his career warning about precisely this danger. Far from becoming less relevant with time, his warnings are more urgent now than they have been in a very long time.

We should start with the Declaration itself. Thomas Jefferson’s language about self-evident truths and unalienable rights is usually read as a statement of Enlightenment philosophical principle, rationalism delivered from on high. But that reading misses what the document actually was: a legal and political instrument designed to justify a specific act of separation to a specific audience. Jefferson himself described its purpose not as the invention of new ideas but as the expression of “the common sense of the subject.” The truths it declared self-evident were truths the colonists had already been living out in their local institutions, their common law traditions, and their practices of self-governance for more than a century. The rights language was experience-driven. It articulated principles that had already been tested in the messy reality of colonial political life. That grounding in experience is what made it work.

Madison and Jefferson, Side by Side

The distinction between experience-driven rights and abstract rights becomes clearer when we hold James Madison and Jefferson side by side. I’ve written about this contrast before, and I keep returning to it because it illuminates a tension at the heart of the American project itself. Madison, the architect of the U.S. Constitution, was deeply suspicious of abstract principle applied directly to political life. His genius lay in institutional design, in creating structures that channel human ambition and self-interest toward productive ends rather than destructive ones. In Federalist 51, he captured the moderate core of the whole liberal project: the goal of constitutionalism is to “first enable the government to control the governed; and in the next place oblige it to control itself.” That is the language of a statesman who understood that human nature is a mixed bag, and that our capacities for both freedom and bondage require institutions that support and channel the full complexity of human life.

Jefferson was more of an idealist, and I do not mean that entirely as a compliment. His Kentucky Resolutions of 1798 argued for nullification, the claim that states could declare federal laws unconstitutional and simply refuse to enforce them. Madison’s parallel Virginia Resolutions proposed the more cautious doctrine of interposition: states could step between the federal government and their citizens to slow things down, to create friction and space for democratic contestation and deliberation, without claiming the final word. The difference matters enormously. Madison refused to draw a firm line demarcating where federal power ends and state power begins. This was not indecision. The whole point of the proposed constitutional system was ongoing contestation rather than settled answers. The power struggle is the feature, not the bug. Jefferson wanted clear principles clearly applied, principles that would, usually, prioritize individual rights above much else. Madison understood that the application of principles to real political life requires prudential judgment, the kind that resists bright-line rules.

Where Burke Becomes Essential

This is where Burke becomes essential. Burke never wrote about the Declaration directly, but across his career he applied a single, consistent test to revolutionary rights-claims: were they grounded in inherited experience or deduced from abstraction? By that test he defended the American colonists in 1775, in his speeches urging conciliation, even as he would condemn the French revolutionaries of 1790. The Declaration sits on the side he defended. And, despite many claims to the contrary, he did not deny the existence of natural rights. He was explicit on this point. His concern was with what happens when rights are ripped from the social and emotional contexts that make them safe for practical application.

Burke distinguished between what he called the “real rights of men” and the “pretended rights” of the revolutionary theorists. The pretended rights were, in one of his most penetrating formulations, “metaphysically true” but “morally and politically false.” Their “abstract perfection” was “their practical defect.” By claiming a right to everything, they ended up wanting everything. Burke’s alternative was to insist that natural rights must be filtered through what he understood as the social affections: the emotional bonds of family, community, habit, and tradition that tie individuals to particular times and places. These are not irrational sentimentality. They are the passionate intermediaries through which abstract principles can be safely transported into political life.

Burke offered a precise analogy: metaphysical rights entering common life are like rays of light piercing a dense medium. They refract. They bend. In the complicated mass of human passions and concerns, they undergo so many refractions and reflections that it becomes absurd to talk of them as if they continued in the simplicity of their original direction. Without these intermediaries, rights become untethered from the duties and relationships that give them meaning. And here Burke makes a point critical for us today: abstract rights do not simply fail to help. They actively harden the heart. They corrupt the moral sentiments that bond members of a society together. The revolutionaries were so taken up with their theories about the rights of man, Burke wrote, that they had “totally forgot his nature” as an emotional as well as a rational being.

What the Declaration Got Right

The Declaration of Independence, at its best, avoided this trap. Its rights language was anchored in concrete grievances, rooted in actual experience of governance, and directed toward a specific political end. Jefferson appealed to necessity, noting that the colonists were “impelled” to separate, rather than to ideological enthusiasm. In this sense, the Declaration’s natural rights were closer to Burke’s “real rights of men” than to the abstract rights of the French Revolution that Burke would later excoriate. They were rights filtered through local affection and connection, through a century of self-governance, through the common law and inherited traditions of English liberty. This is what made them moderate. This is what made them work.

Many readers encounter the first few paragraphs of the Declaration and never arrive at the specific grievances at all. In some measure, then, the Declaration’s own success planted the seeds of our current difficulty. Because the language was so powerful, so elegantly universal, it invited abstraction from the concrete context that gave it meaning. Over 250 years, rights language has been progressively detached from the practical, experience-driven, compromise-oriented tradition in which it was born. Rights have migrated from tools for resolving political conflicts into partisan trump cards that foreclose deliberation. When everything is framed as a right, nothing remains to negotiate. Burke saw this coming. Madison did too. The abstract rights of the revolutionaries, as Burke observed, “admit no temperament, and no compromise; anything withheld from their full demand is so much of fraud and injustice.”

The Paradox of Our Moment

The paradox of our current moment is visible here. We invoke rights language more expansively than at any point in American history, and yet our political culture has never been less capable of the compromise and deliberation the founders understood as essential to self-governance. Left and right alike have weaponized rights, turning them from shared premises for political argument into conversation-ending assertions. Burke warned that purely rationalistic rights, severed from emotional attachment to a specific community, would “temper and harden the breast.” The inflation of rights language has done precisely this. In contrast, Burke argued, the British (and by extension the American) model required that “we give and take; we remit some rights, that we may enjoy others; and, we chuse rather to be happy citizens, than subtle disputants.”

Madison understood this trade-off clearly. His institutional design was premised on the recognition that rights alone cannot sustain a republic. Instead, we need structures that force compromise, that create friction, and that allow ambition to counteract ambition. And we need what Burke called the social emotions, attachments to community, to place, to the people and institutions we actually know, as the connective tissue that holds a rights-bearing political order together. Jefferson’s more ideological approach tended to treat rights as self-executing principles that needed only to be declared in order to be realized. We have inherited far too much of Jefferson on this score and not nearly enough of Madison.

How to Honor It

The best way to honor the Declaration’s 250th anniversary is to recover the moderate and grounded spirit in which it was written. Rights are powerful. They are among the most important tools we have for protecting individuals against the predations of power. But they need to be filtered through local affection and connection before they are safe for practical application. They need the grounding of community, the wisdom of experience, and the humility that comes from recognizing that political life involves balancing competing goods rather than declaring absolute principles. The Declaration’s real achievement was to deploy rights language in exactly this way: concretely, practically, and in the service of a specific political community’s struggle for self-governance.

We honor that achievement by remembering the choice the founders made. When Burke rose to urge conciliation with the colonies in 1775, he reminded Parliament that all government “is founded on compromise and barter.” A year before independence, he laid out the choice the founders could make, and they made it: they chose to be happy citizens rather than subtle disputants. That choice, not the abstract declaration of rights but the decision to treat them as instruments of a shared political life, is what made the American experiment worth undertaking.

Your Turn

If this is your first time here: I write The Radical Moderate’s Guide to Life, where the running argument is that the most radical move available in our political moment is to refuse the binaries everyone keeps trying to force on you. The framework behind this post lives in the pinned essay on radical moderation.

Subscribe to get these in your inbox each week. If you know someone who argues about rights the way half the internet does, share this with them. And tell me in the comments: when it comes to rights, are you more Jefferson or more Madison? I’d love to hear where you land.

Leave a comment

Subscribe now

Share

A new study in PNAS Nexus, published earlier this summer found that 37% of Americans have ended a relationship over politics. Most of the breakups were with friends. The number has been climbing since 2016, and the breakups produced by the 2024 election reached the 2016 election’s totals in half the time. The researchers, Mertcan Güngör and Peter Ditto at UC Irvine, surveyed almost four thousand people across four datasets and triangulated the answers against a national-election panel. They were measuring something most of us already know: we are losing each other.

This study also undermines a comfortable (and false) binary that some of us lean into, namely that the only options are fight harder or pull away entirely. Either you keep slugging it out with your brother-in-law or you stop returning his calls. Either you wade back into the comment section or you mute everyone who voted differently than you did. Either you keep going to the family dinners and grit your teeth through them or you stop going.

The new finding cuts against the option that feels safest. Güngör and Ditto found that the people who had experienced political breakups were more hostile to their opponents than the people who hadn’t, perceived the other side as more extreme, and were more likely to ascribe selfish motives to ordinary voters on the other side. Walking away does not relieve the polarization. It intensifies it. The phantom in your head gets bigger, not smaller, after you stop talking to the actual person it is supposed to represent.

To be fair, some breakups are right. If someone in your life is cruel to you, or treats your existence as a debate, the healthy move is distance, and no amount of technique fixes that. The study isn’t about those. It’s about the much larger pile of relationships we let go quiet over a vote or a yard sign, where the person was never actually the phantom we replaced them with.

I have been working through this for a talk this week, and the thing that has come into focus for me is that the binary is wrong because both halves of it make the same problem worse. We are not, mostly, fighting about policy. We are fighting with phantoms. When we walk away from the actual person, the phantom is the only one left in the room.

The Argument You Think You’re Having Isn’t the One You’re Having

The political science term for what has happened to us is affective polarization (and I’ve talked about it a lot). It is a precise and useful phrase. We do not disagree more than we used to. On the substantive policy questions that get coded as the Big Issues (abortion, immigration, healthcare, even guns), Americans cluster surprisingly close to the middle, and have for a long time. What has changed dramatically in the last twenty years is how much we hate each other.

We dislike the other side more. We trust them less. We are more willing to believe the worst about them and less willing to spend time with them. Marriage rates across party lines have collapsed. The share of Americans who say they’d be unhappy if their child married someone from the other party has climbed from around 5% in 1960 to roughly half today.

The uncomfortable part is that we are wrong about who they are. Not slightly but substantially. Surveys keep finding the same pattern. Republicans imagine that Democrats are far more progressive than they actually are. Democrats imagine that Republicans are far more reactionary than they actually are. Both sides imagine the other side hates them more than they actually do. More in Common’s 2019 work found that partisans imagine almost twice as many of their opponents hold “extreme” views as actually do. The “perception gap” is one of the most replicable findings in American political psychology right now, and the new Güngör and Ditto paper sharpens that finding to highlight the vicious cycle: people who break up over politics are the ones whose perceptions of the other side are the most distorted. The breakup and the distortion travel together.

(On a side note, Democrats were more likely to have experienced a breakup, and were usually the party who initiated it. The pattern may track something durable about moral foundations, or it may be a reaction to a specific cluster of figures and events since 2016. Güngör and Ditto don’t know yet, and neither do I, though I suspect it has something to do with how progressive-leaning folks talk about harm. I have some posts relating-ish to that here and here.)

This is bad news and good news at once. The bad news is that the country is genuinely sick, and the breakup data gives the sickness interpersonal teeth. The good news is that a meaningful portion of it is a perception problem, which means it has perception solutions. The other side, the actual one with a face and a name and an opinion about your kid’s soccer schedule, is not who you think. The elite voices in the media ecosystem (politicians, pundits, podcasters with monetized outrage cycles) are extreme. Most of the people in your life are not.

Steel-Manning for Curiosity

The most important move in this whole conversation is to make the best version of three positions that don’t usually get steel-manned in their own preferred terms. Pick the one you are least tempted by and read it slowly.

Libertarians. The caricature is that libertarians are selfish, anti-government, and indifferent to other people’s suffering. The good-faith reading is that libertarians see something the rest of us tend to miss. Coercion is real even when it is wrapped in good intentions. The person closest to a problem usually understands it better than the person who is going to write the regulation. Voluntary experiments at the local level often beat top-down plans designed for an averaged-out population that does not actually exist anywhere. The libertarian skepticism is about which institution is best equipped to solve a given problem, which is a different thing from indifference, although the two get conflated all the time.

Conservatives. The caricature is that conservatives are intolerant, stuck in the past, and afraid of change. The good-faith reading is that conservatives see something the rest of us tend to forget. Institutions are slow to build and easy to break. Moral knowledge accumulates in traditions and families and communities in ways that abstract reasoning cannot fully replace. “Let’s tear it all down and start over” is a sentence that historically precedes a great deal of suffering. The conservative position is a bet about how much load-bearing weight the inherited structures are doing, and a worry that we routinely underestimate it. That is a coherent worry. It is not the same as fearing the future.

Progressives. The caricature is that progressives are naïve utopians who want to throw away everything that works or condemn us all to an egalitarian dystopia. The good-faith reading is that progressives see something the rest of us tend to ignore. Power asymmetries are real. Structural causes shape lives in ways individuals cannot opt out of. “Everyone’s voice counts equally” is a sentence that conceals a lot when some people don’t have access to the room. The progressive insistence is that the system as currently arranged leaves real people on the outside, and that this matters. That is a coherent commitment. It is not the same as wanting to torch the building.

These are traditions, not parties, and they don’t map cleanly onto red and blue. Plenty of Democrats carry a conservative streak about institutions, and plenty of Republicans are libertarian about coercion. Turns out humans are complex and they can hold all three of these at once in different contexts and on different timescales. None of them requires you to abandon your own first commitments. What they require is the recognition that the other people in this conversation are working with a partial map of reality (so are you, so am I) and that the partial maps are not random. They track real features of the world that the other maps tend to flatten.

The caricature wins debates. The good-faith reading is what makes dialogue possible. It’s also what spurs curiosity and real problem-solving, instead of the partisan mud-slinging that has prevented policy progress over the past decade plus.

We’re Playing the Wrong Game

I have done the thing I am about to describe. I am a political theorist, I moderate difficult conversations for a living, and I will still find myself, twenty minutes into a discussion with someone whose politics I find genuinely unworkable, listening for the flaw in what they are saying so I can pounce on it. The default setting of the modern political conversation is listening to win, and the default is hard to override.

The shift that matters is the shift from debate to dialogue. They are different games. They have different goals. In a debate the goal is to prove the other side wrong. You listen for the flaw. You steel-man your own position. Politics is a contest in which someone has to lose. The outcome is harder lines and more heat.

In a dialogue the goal is to understand what the other side actually thinks and to surface what they see that you do not. You listen for what is true. You steel-man their position before you respond to it. Politics becomes a shared problem-solving exercise in which the people in the room are trying to build a more accurate picture of what is going on.

Most of us are stuck in debate mode because the media ecosystem we live in is structurally a debate. Cable news is a debate. Twitter (excuse me, X) is a debate. Most of the podcasts you listen to are a debate. We have been trained, every day, to listen for the flaw. Then we wonder why our family dinners feel like cable news.

The first move toward dialogue is the recognition that you can do it without giving up your positions. You can think the other side is wrong about almost everything and still ask them what they actually mean. Curiosity is not concession.

(I really like Mike Lee’s work at the Civility Initiative in this area. You can check out a couple podcast episodes with Mike and me here and here.)

Three Tools, One Shift

The research on depolarization interventions is sobering. Most work a little, then fade. A recent meta-analysis found about a five-point improvement on a hundred-point scale that washes out within two weeks. This is something. It is not a cure.

What works better, consistently, are approaches that involve personal story and genuine curiosity. The variable that shows up in nearly every successful intervention is the same: people ask questions instead of making arguments. That is the move. Everything below is a way to make the move easier.

Looping (Amanda Ripley)

Amanda Ripley, who wrote High Conflict, calls it the looping technique. It is the simplest of the three and the easiest to underestimate. The structure is listen, reflect, check.

You listen closely to what the other person actually says. You reflect it back to them in your own clear, plain language. Then you ask, “Did I get that right?” If you didn’t, you let them correct you. You do this before you respond to the substance.

It feels a little cringe (to quote my 13 year old) the first time you try it, like you are doing a workshop exercise in real life. It also reliably changes the temperature of a conversation. People who feel heard fight less. People who feel mischaracterized fight more. Looping interrupts the second loop before it starts.

Try it once this week, in any conversation, political or not. The technique transfers.

Complexity (Peter Coleman)

Peter Coleman runs the Difficult Conversations Lab at Columbia, and the finding from his lab that I think most surprised even him is this: the most effective way to reduce conflict is not to find common ground. It is to introduce more complexity.

Binary framing (us versus them, red versus blue, my side versus the bad guys) locks people into fight mode. Multiple perspectives and additional layers cool the tribal response. The brain has a harder time hating a position that has six dimensions than one that has two.

Coleman’s practical exercise, and the one I keep recommending to people, is this. Before your next political conversation, list three things you think the other side gets right. Just three. It is harder than it sounds. If you find yourself unable to come up with three, that is information about the size of your phantom.

The goal is not agreement. The goal is a more accurate picture of reality.

Charitable Interpretation (Pluralist Lab)

The Pluralist Lab runs a workshop format called UNDIVIDE US in which participants show their stance on a question with a colored paddle, and then have to make the strongest case for the other side’s view before they make their own. They call this “charitable interpretation, not conversion.” You are not required to change your mind. You are required to demonstrate that you understand the position you disagree with well enough to articulate it in terms its own holders would recognize.

The logic of the exercise is simple. It’s hard to throw stereotypes at someone who is sitting right in front of you, especially when they’ll have a chance to respond. Once you have made the best case for the other side, you cannot unmake it. The landscape becomes more complicated than you assumed, which is the point (see Coleman’s work above).

The portable version for your own life: before a hard conversation, ask yourself what a thoughtful, well-meaning person on the other side would actually say. Then ask the person you are talking to whether you got it right. Notice that this links back to looping. The tools are not separate. They are facets of the same shift.

Where the Action Is

What looping, complexity, and charitable interpretation share is that all three slow down the reactive brain and ask you to get genuinely curious. None of them requires you to change your values. None of them requires you to abandon your positions. You can do all of this and still disagree, vehemently, about the things that actually matter to you. That is the point. The goal is not a world without conflict. It is a world where conflict is productive rather than corrosive.

If you try one of these this week, the question I would ask you to sit with is not “did it work” (you cannot evaluate a single conversation that way) but “where did I feel the most resistance.” Resistance is information. Resistance is usually where the action is.

I am still working out a lot of this myself. I find the affective polarization frame hopeful and I find the perception-gap data hopeful and I also notice, on a regular basis, that I am still listening for the flaw but that I get better at listening each and every time. Hopeful is not the same as easy. The phantom is sticky. The Güngör and Ditto paper finding that’s worth remembering is that the people who walk away end up with bigger phantoms, not smaller ones. There are friendships in my own life that I have let go quiet, and the paper makes me wonder whether the quiet has been doing the work I told myself it was doing. The actual person across from me is, almost always, a person I can talk to.

Your Turn: Your Summer Survival Kit

It’s cookout season, which means a lot of us are about to be standing in a backyard, paper plate in hand, while someone’s uncle says The Thing. The Fourth is just a few days away! So consider this your survival kit for the season.

At the next BBQ, pick the tool that made you most uncomfortable: looping, complexity, or charitable interpretation. Use it once, on the actual person across the picnic table whose politics you find hard. Don’t try to win. Just get curious.

Then tell me what happened in the comments. I read all of them, and the backyard stories are where the rest of us learn what actually works.

If the post was useful, send it to the one person you’ve been quietly drifting from over politics, ideally before you see them this summer. Forward the email and, if the timing feels right, ask to reconnect.

And if you’re new here, this is what The Radical Moderate’s Guide to Life does every week. It refuses the binary and tries to map the real landscape instead of the phantom one. Subscribe and pull up a chair.

I’m curious where you land on this, and which of the three feels most doable in your life. The one that feels most uncomfortable is probably the one to try first.

Leave a comment

Subscribe now

Share

Resources (in no particular order, some more academic than others) if you want to go deeper:

• Mertcan Güngör and Peter H. Ditto, “Political breakups: Interpersonal consequences of polarization,” PNAS Nexus 5(5), May 2026

• Shanto Iyengar et al., “The Origins and Consequences of Affective Polarization in the United States,” Annual Review of Political Science (2019)

• Joseph Voelkel et al., “Why depolarization is hard,” PNAS (2025)

• Amanda Ripley, High Conflict (2021), and goodconflict.org

• Peter T. Coleman, The Way Out: How to Overcome Toxic Polarization (2021), and thewayoutofpolarization.com

• The Pluralist Lab and UNDIVIDE US, pluralistlab.org

• More in Common, “The Perception Gap” (2019)

More from me on this topic (some linked above already):

• My podcast interview with Mike Lee’s When We Disagree on not giving up on people

• Other Voters are Not the Enemy

• Fear is the Mind Killer

• Life in 4D: Surviving Until (and After) Election Day

“I’m supposed to work like I don’t have kids and supposed to parent like I don’t have a job.”

That line is from a mom in Pew’s new survey of working parents, out last week. The report is good; It documents what it calls the mental load, the running background process of holding a family’s needs in your head while a job demands your full attention, and it shows that the load lands hardest on moms and on lower-income parents. Seventy percent of full-time working parents handle parenting tasks while they are on the clock. A majority say balancing the two is hard, and far more moms say so than dads (62 percent of full-time working moms call it difficult, against 47 percent of dads). Lower-income parents are the least likely to have paid time off or paid leave, and the most likely to fear losing pay or their job if they have to miss work for a sick kid.

All of this is true, but the report measures the load as a tug-of-war between two things: what your job wants and what your family needs.

There is a third thing in our lives creating cognitive load that the survey never names, because we have all been trained not to see it as a separate category. It is the administrative work of running a life inside broken and fragmented systems.

For many people I know, the stress is not really about the parenting. And it’s not the job. It’s the paperwork, the phone calls, the prior authorizations, the portals, and the hold music, all of which takes time, energy and attention away from work and parenting. It’s the labor it takes to make brittle institutions actually deliver the thing they promised. Once you see that third thing, the Pew numbers read differently. A lot of what looks like “work-family conflict” is actually parents doing unpaid administrative labor for fragmented institutions, on time stolen from both their job and their kids.

Meet Anna.

I like focusing on the concrete, so we’ll start with a specific person, who is actually a composite of women I’ve known and worked with over the years.

Anna works in medical billing (but she could work anywhere). She is 34, makes $19 an hour at a regional health system in upstate New York, and spends her workday on the phone with insurance companies, untangling claims for strangers. She knows the codes, the appeal windows, the difference between an explanation of benefits and a bill, and the specific tone of voice that gets a denied claim re-reviewed. But none of that expertise spares her when it is her own family on the other end of the system. At 12:01 every afternoon she stops being the person who fixes this for a living and becomes the single mother of two who cannot get a human being on the line before her lunch break ends.

In what follows, I’m going to dig into a typical year in her life, focusing only on healthcare. It’s not a dramatic year. It’s actually the best possible year in the world she lives in. It’s a year in which nobody gets seriously sick, nobody breaks a bone, and nobody develops anything that requires a specialist. It’s the boring, best-case, everything-goes-fine version. It turns out that is the version that exposes the problem most clearly.

Since Anna is a composite, there’s no single mother whose year ran exactly like the one I am about to describe, because the details depend entirely on which state you live in, which employer you work for, which Medicaid managed-care plan got assigned to your kids, and which dentist happens to be in which network this year. Change the state and the dental coverage changes. Change the employer and the deductible changes. That variation is the point, not a flaw in it. I’m not trying to prove that this specific year happens to this specific woman. I’m trying to show the shape of a problem that takes a thousand different specific forms, all of which share the same underlying structure.

The Household

Anna has two kids: Diego, 10, and Sofia, 7. Her coverage is what a lot of working-poor and near-poor American families actually carry:

• The kids are on Medicaid (they may flip to CHIP in a month where her hours run high, which is its own small administrative event)

• Medicaid in her state covers kids’ medical but carves dental out to a separate managed dental plan with its own card and its own network

That carve-out is why parent and kids end up on different logics under the same roof. Every state covers comprehensive dental for children under 21 through a federal mandate called EPSDT: exams, cleanings, fillings, even orthodontics when it is medically necessary. Adult Medicaid dental is optional, and states treat it accordingly. As of 2025, only 11 states and DC meet the American Dental Association Health Policy Institute‘s strictest “extensive” tier, an annual cap of at least $1,000 covering the full range from cleanings to dentures, as of the end of 2024. At the other end, one state, Alabama, covers no adult dental at all outside pregnancy, and a handful offer emergency-only coverage, which means they will pay to pull a tooth but not to save it. Many states have cut or eliminated adult dental benefits during budget downturns, because when budgets tighten, the optional benefit goes first. So depending on where Anna lives, her own teeth might be fully covered, capped at $1,000, or covered only when something is already an emergency, while her kids’ dental is comprehensive and guaranteed in all fifty states. So we end up with one household with two different dental systems, determined by a line in a state budget.

• Anna is on her employer’s high-deductible PPO, $3,000 deductible, because it was the cheapest premium and she does the math on premiums, not deductibles, the way you do when cash flow is the binding constraint

• She has a separate employer dental plan

• She has a vision discount plan, which is not insurance, a fact she will rediscover at the optometrist’s front desk

That is five distinct coverage entities for three people. Five cards, five networks, five customer service numbers, five sets of rules about what is covered and who is in-network and what needs prior authorization. None of them can see the other four. The only place the whole picture exists is in Anna’s head.

The Appointments

Anna’s baseline year assumes thirteen medical appointments for basic preventive care.

That’s the number if the universe cooperates completely, but we all know it almost never does. Once we add the things that are not “something went wrong,” just “people have bodies” and the load gets much worse. Say one kid has a cavity at a cleaning: one filling visit. Anna’s annual physical turns up bloodwork that has to be drawn at a separate lab because the in-network lab for her PPO is not the one in her doctor’s building. Her women’s health visit flags something that needs a six-week recheck, the routine kind, the kind that turns out to be nothing. Now we are at sixteen or seventeen appointments.

It’s still a good year, in one sense. Nothing catastrophic happened. Anna should be grateful, right?

As any parent knows, no appointment is just an appointment. Each one is a chain, and the visit itself is the shortest link:

1. Schedule it. Confirm the provider is in-network for this specific plan, which differs by child and by service. Compare against disaggregated school and work calendars and car maintenance schedules and factor in driving time (because of course the pediatric dentist who takes Diego’s Medicaid dental plan is not the one near the house).

2. Verify coverage. Get any referral or prior auth. This often requires a separate call, usually to the insurer, frequently bounced back to the provider, occasionally bounced a third time.

3. Attend. This means time off. For an hourly worker, almost always unpaid or, in the best case, scarce PTO that conflicts with every other administrative need in Anna’s life. Almost every appointment has a dollar cost before anyone in a white coat even starts talking.

4. Decode the bill. The EOB arrives weeks later, says “This is not a bill,” and is followed by something that is a bill, and the two do not match. And while this is happening, scheduling for the next appointment is underway, using a different system and a different portal.

5. Fix the discrepancy. The dental plan and the medical plan each decide the other should have paid. A claim denies for a missing code. Anna, who fixes exactly this for strangers all day, now has to fix it for herself, on her own time, which she does not have.

Run sixteen appointments through that chain and assume only half generate a billing question. Assume each question takes one or two calls. Anna is looking at thirty to fifty discrete phone interactions a year, nearly all of which can only happen between noon and 12:30, which is also the only time she has to eat.

The noon window is the unintentionally cruelest part of the design. Provider offices and insurer call centers are busiest at midday, because that is when everyone like Anna is calling. Many of them close their phones for lunch in the same window. The one slot she has is the one slot the system has decided to make least available.

The Year, Month by Month

Here is what a good year looks like laid against a calendar. Watch how the billing tail of one appointment lands on top of the scheduling of the next. The cognitive load compounds and never really eases.

January. New plan year. Deductible resets to zero. Anna calls to re-verify that the kids’ Medicaid dental network still includes their dentist (it changed last year without notice). Schedules Sofia’s spring cleaning. Twenty minutes on hold; lunch ends mid-call.

February. Diego’s annual physical. School requires an updated form afterward; the pediatrician’s office charges nothing to fill it out but takes nine days to do it, and Anna calls twice to check. The physical also means a half-day of unpaid leave because the only pediatric slot is 2 p.m.

March. Sofia’s dental cleaning. Cavity found. A filling gets scheduled for April. The EOB from Diego’s February physical arrives, showing a $40 charge that should have been fully covered. Call one of two.

April. Sofia’s filling. The dental plan processes it as out-of-network because the dentist’s billing ID, an NPI, the provider’s national identifier number, doesn’t match what the plan has on file, a clerical error on the plan’s end. Anna spends two lunch breaks fixing this. The March physical billing error from Diego is still open.

May. Anna’s own annual physical, finally, because she scheduled everyone else first. Bloodwork ordered. The lab in her doctor’s building is out-of-network for her PPO; she has to find the in-network lab and go on a separate day because she knows from experience that the lab lines are out the door by 10am most days. That means another partial unpaid afternoon.

June. School’s out, which sounds like relief and is the opposite because: camp forms. Both kids need a physical-within-the-last-12-months on file and an immunization record. Diego’s February physical covers him. Sofia’s lapsed in March, so she needs one she wasn’t otherwise due for. The lab bill from May arrives, coded wrong.

July. Anna’s women’s health visit. Something flags for a routine six-week recheck. Her eye exam, which she books at the discount-plan optometrist, ends with a $180 bill she thought was covered, because the discount plan is not insurance. She decides to wait on updating her glasses, keeping her current outdated pair that give her headaches if she wears them for too long.

August. The kids’ eye exams. Diego needs glasses; the kids’ vision coverage under Medicaid has a frame allowance and a list of approved frames, which is a separate phone call to confirm before the appointment so she doesn’t get stuck with a bill. The July recheck gets scheduled.

Fall. The women’s health recheck comes back exactly as expected: nothing. Back-to-school brings a second round of whatever forms the district has decided it needs. Diego’s second dental cleaning surfaces the April out-of-network error all over again, because the plan never corrected the NPI on file, only that one claim, so Anna fixes the same error a second time. Sofia has her second cleaning. Open enrollment lands at Anna’s job, and deciding whether to keep the high-deductible PPO requires comparing plan documents she can only read after the kids are asleep. Somewhere in there Diego goes overdue for his annual eye recheck, because the glasses saga ate the reminder.

December. Year-end. Anna tallies what hit the PPO deductible and realizes a chunk of her own care got deferred to “next year” to spread the cost, which means January starts the whole thing again with a fresh $3,000 deductible and a body that has been waiting.

Across that calendar, the billing tail of nearly every appointment lands during the scheduling or attending of the next one. There is no clean month. There is no month where the administrative load is zero. The work is continuous, and it is all hers, because she is the only node in the system that can see all five plans, both school calendars, and both kids’ bodies at once. She is the integration layer. She is the API between five insurers, six providers, two school districts’ worth of forms, and one employer who docks her pay every time she leaves to make a call.

The Portals Were Supposed to Help

I haven’t even counted the logins (my own personal nemeses). The modern answer to all of this is “just use the app,” so now we can list the credentials Anna has to maintain: her PPO portal, her employer dental portal, the kids’ Medicaid state portal, the kids’ dental carve-out portal, the pediatric practice’s MyChart, her own primary care system’s MyChart (a different MyChart instance that does not talk to the kids’), the OB-GYN’s portal, the dental office’s scheduler, and the optometry chain’s system. Nine or more credential sets, each with its own password rules, its own timeout, several requiring a code texted to a phone she’s not supposed to check on the clock. The portals, intended to reduce the calls, end up being a second full administrative system layered on top of the first. (And that’s before you add in every single (different) portal or app used by schools, after school care, kids’ sports, and so on.)

But There’s More!

Everything above assumes the easy version. How much I had to assume away to keep it “easy” makes the argument even stronger.

I assumed neither kid has special needs. No IEP, no therapy authorizations, no durable medical equipment, no specialist on a recurring schedule, none of the prior-auth machinery that families with disabled kids live inside permanently. The CDC puts the share of US children with a developmental disability at around one in six. For those families, every number in this post multiplies.

I assumed Anna herself has no chronic condition. No diabetes, no hypertension, no asthma, no thyroid management, nothing requiring a maintenance medication and the quarterly refill-authorization dance and the periodic “your insurer no longer covers this drug, here is the formulary alternative, please get your doctor to submit a new prior auth” letter. About six in ten American adults have at least one chronic condition. The genuinely healthy single parent of two genuinely healthy kids is not the typical case. It is the lucky case.

I also assumed Anna is not caring for elderly relatives, which is also something of an assumption in this day and age. Pew estimates that about a quarter of U.S. adults are sandwiched between the needs of older parents and young kids, though not all provide direct care for the former.

And I barely touched school and camp health forms, not to mention school events, camp registrations, and any and all health forms for extracurricular activities for Diego and Sofia and all the coordination that goes in there (alongside the fact that almost zero summer camps align with parental working hours) which are their own seasonal administrative weather system. Physicals timed to the district’s calendar rather than the body’s, immunization records pulled from one portal and uploaded to another, the sports physical that is somehow a different form than the school physical, and the camp that wants its own version notarized.

Strip those three things out and you still get thirty to fifty phone calls a year crammed into a half-hour lunch, all held together in Anna’s head. Put any one of them back in and the system stops being merely exhausting and starts being something a person cannot actually do while holding down an hourly job.

Adding all this up and a healthy single mother in a no-disasters year spends about a work-week of unpaid time, plus a few hundred dollars in lost wages, just coordinating healthcare institutions that have decided coordination is her job.

And this is only healthcare. Every other institution in Anna's life runs on the same assumption, that somewhere in the household is a person with spare daytime hours to make it all line up. School wants forms, parent portals, permission slips, conferences, the lunch-account balance, and the fundraiser. The kids' sports and activities want registrations, waivers, season schedules, payment plans, and a carpool she has to arrange herself. Then there are taxes, the car (registration, inspection, the DMV line, the insurance renewal, the maintenance nobody sends a reminder for), the utilities and the phone bill, the bank's fraud alerts and overdraft warnings (and all those portals and passwords), the landlord and the lease, the childcare waitlist, the employer's own open-enrollment paperwork, and the annual benefits recertification that demands she re-prove she is still poor. Each system assumes a free coordinator. For most families that coordinator is the mother. Whoever is doing this work, it takes away from time spent with their kids, on their job, or just staring out the window after a hard day of work.

Millions of Annas

Because Anna is a composite, you could fairly ask whether I have built a worst case and called it typical. I don’t think I have and there’s a fair amount of data to back it up.

The most careful study of this is Michael Anne Kyle and Austin Frakt’s 2021 work in Health Services Research, built on a nationally representative survey of insured working-age adults. They looked at five tasks: scheduling, getting information, prior authorizations, resolving billing issues, and resolving premium problems. Seventy-three percent of people had done at least one in the past year. And about a quarter of everyone surveyed, 24 percent, reported delaying or going without care because of one of these administrative tasks. Paperwork keeps roughly as many Americans from care as cost does.

The regressivity shows up directly in their numbers, and it’s a little more interesting than “the poor do more paperwork.” Higher-income and college-educated people actually report doing more of these tasks, because they engage the system more, push harder, and appeal more. But higher income predicted fewer burdens, meaning fewer cases where the task turned into delayed or missed care. The wealthy do the administrative work and absorb it. The poor hit the same wall and it converts into care they do not get. Disability had the strongest association of all with both doing the tasks and being harmed by them. The people with the most medical need carry the most administrative weight, which is exactly backward from how a working system would distribute it.

The billing-specific piece deserves its own note, because the law recently changed in a way that matters for Anna. The dramatic surprise bill, the kind where you go to an in-network hospital and get a separate five-figure charge from an out-of-network anesthesiologist you never chose, was genuinely common. Before the federal fix, roughly one in five insured adults reported a surprise bill from an emergency visit or an in-network hospital stay. The No Surprises Act, which took effect in January 2022, banned most of that category. This is good! But like most of our healthcare policy “reforms,” it was written for the dramatic case. It does not touch the routine one. It does not cover ground ambulances, and it does not cover the situation where a patient picks an out-of-network provider in a non-emergency, where the plan’s normal out-of-network rules still apply and the balance bill is still legal.

Which is to say it does almost nothing for Anna. Her billing problems are not dramatic. They are the out-of-network lab her own doctor’s building happens to use, the dentist whose billing ID does not match what the plan has on file, the claim that bounces between the medical plan and the dental plan while each insists the other is primary. The law cleaned up the emergencies and left the grinding, routine, coordination-failure billing entirely intact, the exact category that eats her lunch breaks and keeps her up at night trying to remember which of her 32 spinning plates she’s going to drop that week.

This is, incidentally, pretty common. Our policy conversations tend to focus on financial cost of using a system. But they often ignore the costs of navigating the system itself: people’s mental health, their stress levels, their quality of life, their lost time with loved ones, or the lost productivity at work trying to handle the second part-time job of life admin.

What the Pew Report Misses

Let’s go back to that Pew survey. When Pew asks a parent whether work makes it harder to be a good parent, and a parent says yes, what is that parent actually answering? Some of it is genuinely the job. Sometimes it really is the late meeting, the travel, and the boss who emails at 9 p.m. But some unknown and uncounted share of it is Anna’s afternoon on hold with a dental plan, logged in her mind as “work got in the way of family” or “family stuff bled into my workday,” because there is no third box to check. The survey offers two categories, work and family, and the administrative labor gets sorted into whichever one it interrupted. It never gets counted as what it is.

This matters because the two framings point at completely different solutions. If the problem is work-family balance, the fixes are the ones the report gestures toward: more flexibility, better leave, and the option to work from home. But notice that Pew found something odd buried in its own data. Parents who regularly work from home were no more likely to say balancing job and family was easy than parents who did not. Flexibility helped them attend a recital. It did not lower the load. That should be a clue. If telework moves the deck chairs but the water keeps rising, maybe a meaningful part of the load was never about where you sit or when you work. Maybe it is the volume of administrative coordination the parent is absorbing on behalf of institutions that have quietly decided she will do that work for free.

You cannot flex your way out of thirty phone calls a year that can only happen between noon and 12:30. Working from home doesn’t shorten the hold time. A generous leave policy doesn’t make the dental plan and the medical plan agree on who is primary. The Pew report is measuring the symptom, the felt sense of never giving 100 percent anywhere, with real precision. The underlying cause it cannot name, because the cause does not live in the relationship between work and family at all. It lives in the relationship between families and the fragmented and increasingly brittle institutions that are supposed to serve them.

Institutional Failure is (Always?) Regressive

An easy binary is that this is either a story about a broken system or a story about a hard-working mom doing her best. It is both, and the relationship between the two is the whole point.

The load is regressive, and it is regressive along several axes at the same time.

Time is the tax, and Anna has the least of it. A salaried professional like me makes these same calls between meetings, or pays someone, or has an HSA-funded concierge service do it, or simply eats a surprise bill rather than spending a second lunch break fighting it. Anna pays in the one currency she has least of: uninterrupted daytime hours, which for an hourly worker convert directly into lost wages. The system charges the same administrative price to everyone and then collects it in a currency that is far more expensive for the poor.

Fragmentation scales with poverty. This is the part that bothers me the most. The wealthier a family is, the more likely everyone in it sits on a single integrated comprehensive plan. It is precisely the lower-income, mixed-eligibility household, kids on public coverage, parent on employer coverage, dental carved out, vision a discount card, that gets handed the maximum number of separate entities to coordinate. The system imposes the most complexity on the people with the least capacity to absorb it. It is means-tested in reverse.

The penalty for a mistake compounds downward. A missed prior auth, an out-of-network slip, a clerical NPI error on the plan’s end: a high-income family treats the resulting bill as an annoyance and a phone call. For Anna the same bill is a crisis with a tail, to collections, to credit damage, to the choice between the bill and something else. The identical administrative misstep carries radically different consequences depending on who makes it, and the people most exposed to the misstep are the ones least able to survive its consequences.

It’s tempting to look for a villain, but there isn’t one (other than maybe total political dysfunction). Medicaid is doing what Medicaid does. The dental carve-out is administering its network as designed. The employer chose a plan menu that is legal and even reasonable. The pediatrician’s office fills out the form for free. Every single institution in Anna’s year is behaving rationally and roughly as intended. The dysfunction is not inside any of them. It lives in the gaps between them, in the spaces no institution is responsible for, the coordination that each one assumes someone else, meaning Anna, will perform for free. This is what I call structural hobbling at the life admin level.

And that is what institutional failure actually looks like most of the time. It’s not a single broken rule. Instead it’s a system that has quietly externalized its own coordination costs onto the individual, and distributed that externalized cost regressively, so that it falls hardest on the household least equipped to carry it. There is no office you can call to fix the gaps, because the gaps are the product of every office working exactly as designed.

Part of the reason for writing this is that I’m exhausted trying to stay on top of life admin and that’s with a supportive partner, flexible job, and good insurance. I know a lot of other similarly situated parents who struggle too. If the system still sucks for the most privileged among us, it’s worth asking how much worse it is for those who can least afford to embrace the suck.

When that mom in the Pew survey says she is supposed to work like she does not have kids and parent like she does not have a job, I believe her, but I think she is describing something even larger than she was asked about. She is also doing a third job nobody named on the questionnaire: unpaid administrator for every institution that decided coordination was her problem to solve. That job has no office, no hours, and no pay. It just has her lunch break.

I’ve written before about the life admin tax, the cognitive overhead of running a life across too many systems at once. The tax sometimes comes out in money via late fees or people just paying errors instead of fighting them. But more often we pay that tax with our mental health, our patience with our kids, our attention at work, our cognitive load, and our leisure time. For most of us, it’s the far crueler way to pay it.

Your Turn

Since everyone loves a horror story, tell me about your own life admin horror story in the comments! I have a great one about spending hours trying to get a “free” breast pump after the birth of my first kid, but I know you have them too. What unpaid life admin job sucks the life out of you? More importantly maybe, what did I get wrong? What can’t I see or what did I miss? Tell me in the comments. And as always, if you like what you read please share and subscribe!

Leave a comment

Subscribe now

Share

For years there was a bridge in my town that I drove under several times a week. From underneath, stopped at the light, you could see exposed rebar and bare patches where chunks of the deck had spalled off and dropped onto the shoulder. Rust bled down the concrete in long streaks. At some point somebody strung netting under the span to catch the next piece before it landed on a windshield. The netting sagged.

Now imagine a public information campaign. The bridge is inspected. The bridge meets standards! The engineers have signed off, and the engineers know more about load-bearing structures than you do, which is true. Trust the bridge, Because Experts.

You can run that campaign forever. It might even work for a while. But at some point your eyes override the expert consensus, because the expert consensus is asking you to disbelieve the rebar. And the campaign gets one thing exactly backwards: your distrust of that bridge is not a misinformation problem. It is information. It may be the most rational data-processing you do all day.

I say this as a card-carrying member of the expert class in question. I’m a political theorist who studies institutions for a living and an associate dean who runs a few (or at least tries to keep them running). I have spent a good portion of my adult life telling students and neighbors some version of “trust the process.” I still believe in processes. I also drove under that bridge for years, and I noticed that I had stopped parking anywhere near it.

Trust in nearly every major American institution has been falling for decades: medicine, media, higher ed, Congress, the courts. The standard diagnosis treats this as a demand-side problem. Voters have been propagandized, or they’re cynical, or they don’t understand how the sausage gets made, and the fix is to educate them back into deference. Some of that is real. Conspiracy thinking exists, and some perfectly sound bridges get slandered.

But there is a supply side to trust, and we barely talk about it. Before you ask why people don’t trust institutions, you have to ask whether the institutions are trustworthy. Those are different questions. The first is about psychology. The second is about design.

“Because Experts”

The official answer to institutional distrust has mostly been reassurance: fact sheets, media literacy curricula, louder credentials, a fresh coat of paint over the rebar. Reassurance is what you offer when you believe the problem is in the audience. Repair is what you offer when you believe the problem is in the bridge.

Reassurance also carries a hidden cost. Every “the bridge is fine” that contradicts what a driver just experienced on the bridge does nothing to transfer trust to the structure. It drains trust from the speaker. People are quite good at integrating expert testimony with their own crossings, and when the two conflict repeatedly, they don’t stop believing their tires. They stop believing the experts. This is how institutions burn through the credibility of the entire professional class defending them. The pandemic-era whiplash on masks and school closures worked exactly this way: each confident pronouncement that got walked back cost more than silence would have.

Which means the question for anyone who cares about expertise, and I do, is what makes an institution worth trusting in the first place. Trustworthiness turns out to be something you can inspect for, the way you’d inspect a bridge.

Every Bridge Is Built for Somebody

Every institution is built for someone: a default user it quietly assumes.

With physical bridges, the assumption is visible. The freeway overpass with no sidewalk is a perfectly good bridge if you own a car; if you’re on foot, it isn’t infrastructure. It’s a wall with traffic on top. The on-ramp only locals know about works fine for locals. The posted detour assumes you can read the language it’s posted in.

Institutions make the same assumptions, often with less visibility. A school day that ends at 2:30 assumes a parent who is free at 2:30. A court system assumes you can either afford a lawyer or decode filings written for one; you learn the rules by being towed. A benefits program assumes income rises smoothly, so in Kentucky a $4,000 raise can trigger $10,000 in lost benefits for a family of three, a toll charged mid-span, after you’ve committed to the crossing and can’t back up (the Kentucky response is an interesting case study in institutional reform itself). An occupational licensing regime assumes the gatehouse fee is small and the gate is single.

Mariana, a Certified Professional Midwife, catches babies for a living and is good at it; whether she may do it legally depends on which side of a state line she is standing on and on a collaborating physician’s signature she cannot get. The toll that is a rounding error for a corporatized hospital system is, for her, a closed road.

The distributive math runs the same direction most of the time. It’s deeply regressive. The blown tire, the lost day in line, the compliance fee, the form that must be filed in person on a weekday: these costs are roughly flat, and they land on people with wildly unequal capacity to absorb them. The toll is flat but the damage isn’t. A bridge built for a narrow default user is regressive by construction, and nobody had to intend that for it to be true.

So who is the default user? The problem again comes back to visibility: we design institutions for the most visible people. These are often the most powerful; those in the room with us. Availability bias is real and we’re also simply limited by what we don’t know that we don’t know. Turns out that deep pluralism runs deep (thus the name!) and the people designing institutions almost always miss the actual range of humans trying to cross them: religious and secular, neurodivergent and neurotypical, rural and urban, the trades and the professions, and the four or five political tribes that don’t fit on a left-right line.

This is partly why I find the current diversity wars so exhausting. One camp counts demographic representation; the other counts ideological representation; both try to measure, mandate, and manage pluralism from the top down. It’s the same managerial premise on a different variable. The real diversity on the road is deeper than either checklist, and the real question is structural: which institutions can carry the range of travelers that already exists?

Run the Circuit in Reverse

Trust is built the way a bridge earns its reputation: crossing by crossing. You drive over it and nothing happens. You drive over it a thousand more times and nothing happens. “Nothing happens” sounds like faint praise, but it is the entire product. Legitimacy is a reservoir filled one uneventful crossing at a time. The authority that owns the span cannot issue its own reputation, however many press releases it prints.

Now run the circuit in reverse. A bridge built for a narrow default user starts shedding the users it wasn’t built for. First they route around it quietly: the informal childcare swap, the cash arrangement off the books. Exit, in the vocabulary of Hirschman’s Exit, Voice, and Loyalty, after voice failed or was never on offer. Black markets are the bridges people build when the official one won’t carry them.

The people with the most resources exit first, and they exit upward. Where the public crossing has failed or never existed, the crossing privatizes: the concierge physician, the private school, the $50,000 admissions consultant, or the gated community with its own security contract. The wealthy charter a helicopter over the river while everyone else stands at the wall, and the void becomes a sorting mechanism that sorts by wealth.

And once the people with the most political power have helicoptered out, watch what happens to the bridge. The constituency for maintenance evaporates. Repair budgets become raidable. The feedback that should have driven the redesign stops arriving, because the remaining users are precisely the people whose complaints carry the least weight. Quality falls further, which justifies more exit, which weakens the institution again. I’ve watched higher ed run this spiral from the inside. Felt illegitimacy concentrates exactly where buying power runs out, and the people still standing on the failing deck draw the reasonable conclusion that the bridge was never engineered for them.

That conclusion is the part the trust campaigns refuse to metabolize. The commuter who has stopped trusting the bridge is not failing civics. She has audited the institution with her own crossings and reached a defensible finding. Divestment on that scale calls for repair rather than reproach. You cannot message your way out of a vicious cycle. You can only build your way out.

How to Design a Trustworthy Bridge

My bridge was the easy case. Rebar is visible to anyone stopped at a light, and the case I can defend hardest is exactly that kind: the bridge you cross with your own body. The school that lets out at 2:30, the benefits cliff, the licensing gatehouse, the courthouse, the campus. There your eyes really do override the experts, because you are the instrument reading the load.

Most institutional failure isn’t visible from the car. You can’t eyeball the soundness of the FDA or the Federal Reserve, and most of our crossings of those arrive secondhand, filtered through feeds engineered to show us rebar whether or not the deck is cracked. Manufactured rebar is a real industry. So the perception gap cuts both ways: people keep trusting some institutions that are quietly failing them and distrust some that are sound, and distrust can overshoot into a taste for rubble. That is real, and I’m not pretending otherwise. But notice where it cuts hardest: precisely where you can’t inspect for yourself. That’s an argument for building institutions you can inspect, not against it. Posted limits, published inspections, predictable outcomes, and working feedback channels: these are what let an ordinary user tell a failing bridge from a slandered one. An institution that shows its work gives you something better than reassurance. It gives you a way to check.

Trustworthiness, in other words, is a design property, and you can write the spec. The spec is partial, and every line translates off an actual bridge.

Posted load limits, applied evenly. A trustworthy institution publishes its rules and its inspection reports before you commit to the crossing, and then applies them the same way for years, regardless of whose truck it is. Predictability is the product, and “nothing ever happens” is the highest praise a bridge can earn. The load rating cannot quietly change depending on connections or identity. The fastest way to destroy an institution’s trustworthiness is for users to learn the rules by being towed, or to discover that outcomes track who they are rather than what’s posted.

The pothole hotline. Trustworthy institutions hear about cracks before the collapse, through working feedback channels and scheduled inspections. In the brittle version, the first feedback anyone receives is the failure itself.

The retrofit. A good bridge can take a bike lane later. Institutions need the same property: adaptable to users the designers never anticipated, rather than facing every new traveler with a choice between as-is and abandonment.

More than one crossing. Redundancy keeps everyone honest. One bridge, one owner, one chokepoint is an invitation to capture, and a captured chokepoint always charges a regressive toll. Real alternatives let exit discipline the institution instead of merely escaping it.

Comment while the concrete is wet. Voice has to arrive while the design can still change. Hearings held after the blueprints are finalized are theater, and users can tell.

The engineer drives across it daily. The people who design institutions need skin in the game. The people certifying the bridge should be on it regularly. Regulators who never use what they approve, and administrators exempt from their own systems, are why so many institutional decks crack.

Underneath all of these sits one architectural choice. You can build for a default user and bolt on ramps afterward, one accommodation per squeaky wheel. Or you can assume the full range of travelers in the basic design. Accommodation makes every difference an exception to process, so inclusion depends on who asks, and when, and the structure fails the first case nobody foresaw. Universal design multiplies pathways by default, so that nobody has to file a request to belong, and the structure holds diversity it never predicted. The curb cut was designed for wheelchairs and turned out to serve strollers, delivery carts, travelers with luggage, and every kid on a bike. Accessibility is resilience. They are the same property viewed from different angles.

The obvious objection is that somebody still has to decide what the bridge must carry, and whoever holds that pen can flatten the landscape they claim to be designing for. One answer is a mechanism rather than a magic number: many crossings, no single owner. Overlapping authorities, decisions pushed to the lowest competent level, and local variation within shared commitments. Elinor Ostrom spent a career, most famously in Governing the Commons, documenting how polycentric systems again and again outperformed central designers at exactly this kind of problem. No one entity decides for every river.

Repair Before Reproach

Demanding trust is the one move that cannot produce it. Trust is downstream of trustworthiness the way a bridge’s reputation is downstream of its load-bearing record, and no civics curriculum substitutes for a deck that holds.

The asymmetry in the current ask makes it worse. The loudest demands that ordinary citizens trust public institutions tend to come from the people best equipped to skip them: private alternatives in schooling, medicine, security, and law that let them avoid the failing crossing entirely. I don’t read that mainly as hypocrisy. I read it as physics. When the people with the most political power exit, the constituency for maintenance goes with them, the repair budgets turn raidable, and the feedback that should drive the redesign stops arriving. A bridge the decision-makers never ride is a bridge that quietly stops being maintained.

So the radical moderate’s position is the unglamorous one: not rubble, not reassurance, but the repair crew, arriving with the inspection report and the rebar. Yes, distrust can overshoot, and some of the loudest critics would rather have rubble than repair. But institutions earned most of this distrust honestly, and tearing down every span out of spite would strand the people who have no helicopter first. The repair crew is the only one of the three that builds anything.

And the bridge in my town? The state finally tore it down this summer. After years of netting and reassurance, the demolition crew showed up and admitted in steel and dust what every driver stopped at that light had known for a decade. I won’t pretend it didn’t feel like vindication. It was also the most trustworthy thing the bridge’s owners ever did, because for once the official action matched the visible evidence. Nobody needed a fact sheet to believe that wrecking ball.

Now there’s a gap where the span used to be, and the rare thing failing institutions almost never get: a clean sheet. The design question waiting in that gap is the one this whole post has been circling. Who is the new bridge for? Build it for the actual range of travelers, Mariana and the families waiting on her included, with the sidewalk in the blueprints instead of bolted on in thirty years, and then let it be boring. Boring is what trust feels like from the inside. A thousand crossings where nothing happens.

Your Turn

Which institution is your bridge: the one you used to cross without thinking and now brace on? And what would it actually take, concretely and structurally, for you to trust it again? Not reassurance. What repair? Tell me in the comments; the best examples are never the ones I’d think of. And if you liked what you read above, subscribe and share! This idea is the foundation of my next book and I’d love to hear all the thoughts - good, bad and ugly - as I work through these ideas in real time.

Leave a comment

Subscribe now

Share

In my last post I argued that America, the country and the people and the communities and the values, is in far better shape than our despair suggests. The argument only works, though, if you hold one thing firmly apart from it: the national executive is not all right, and pretending otherwise would make the first claim dishonest.

I’ll start with the obvious. Do I think the President is enriching himself at the expense of ordinary Americans? Yes. Do I think he’s a man of bad character? Yes. If that were the whole problem, it would be a manageable one. Bad men have held high office before and the republic absorbed them. The self-enrichment is the symptom. The disease is more serious and more durable: the deliberate dismantling of the constitutional limits that are supposed to make the office safe no matter who holds it.

This is the part that should worry a conservative or a libertarian or a classical liberal more than it worries anyone else. The whole point of our constitutional architecture, the thing Madison and Montesquieu and the authors of the emoluments clauses were building, was a government that doesn’t depend on the virtue of the person in charge. As Madison argued in Fed 51, this is basic common sense: “If men were angels, no government would be necessary.” Our constitutional design assumes the office will sometimes be held by someone grasping and self-interested, and it builds limits so that the grasping stays contained. When the limits hold, a bad president is a bad few years. When they’re hollowed out, a bad president becomes a precedent, a new and lower floor that every future occupant gets to stand on. The real story is constitutional decline, not the money. The money is the most legible evidence that the limits aren’t holding.

Three episodes, read as constitutional questions

Abstraction is how this stuff gets waved away, so here are three well-documented episodes from the current term. And remember, the point isn’t the dollar figures. It’s what each one does to a structural limit. It’s also the fact that while other kinds of political malfeasance can at least plausibly be explained away as policy differences (sometimes), in the case of self-enrichment it’s much harder to see any legitimate claim for the breach. I’ll move through each case and then come back to the bigger issue.

The Qatari jet (May 2025). Qatar’s royal family offered, and the administration accepted, a Boeing 747-8, a “flying palace” valued at roughly $400 million, to be retrofitted as Air Force One, with ownership slated to pass to Trump’s presidential library foundation after his term. The Foreign Emoluments Clause exists for exactly this. It forbids officeholders from accepting presents from foreign states without the consent of Congress, precisely because the founders feared foreign influence over American leaders. Legal scholars across the spectrum called acceptance without congressional consent a likely constitutional violation. One noted it “looks precisely like the kind of gift that the clause was written to prohibit.”

The administration’s defense runs like this. The jet was given to the Department of Defense, not to Trump personally. It serves as government transport during the term and then passes to a presidential library foundation rather than into his private hands. And gifts to the United States government, as opposed to gifts to an individual officeholder, have a long and accepted history. On that reading the emoluments clause simply isn’t triggered, because the recipient is the institution and not the man. That’s a coherent position, and a careful reader should sit with it rather than wave it away.

But it doesn’t hold, and the workaround is the more troubling part. The clause’s purpose is to prevent foreign influence over American decision-making, and that danger doesn’t evaporate because the title sits with an agency instead of a person. A president who flies in a palace a foreign monarchy provided, and whose library foundation will inherit it, has received something of value whatever the paperwork says. And if “route it through an agency” is a valid way around the Foreign Emoluments Clause, then the clause is functionally dead, because almost any future foreign gift can be structured the same way. A limit you can satisfy by relabeling the recipient isn’t a limit. It’s a formality. Formalities are exactly what the founders were trying not to rely on.

The crypto venture (ongoing through 2025). World Liberty Financial, the crypto venture launched by Trump and his sons, became a direct channel between the President’s policy power and his family’s fortune. His own financial disclosure reported more than $57 million from it in a single year. The structural problem isn’t that he made money, it’s the mechanism. When the President holds assets whose value rises and falls on his own administration’s regulatory and enforcement choices, the line between governing and self-dealing disappears. That’s the domestic emoluments concern the founders also wrote into the Constitution, and it’s why every modern predecessor divested. The norm of divestiture wasn’t a courtesy. It was the practical machinery that kept the office from becoming a profit center.

The Changpeng Zhao pardon (October 2025). This is the one that should trouble anyone who cares about the rule of law, and it troubled The Dispatch, a conservative outlet, as much as anyone. Zhao, the founder of Binance, who had pleaded guilty to failing to maintain anti-money-laundering controls, received a full and unconditional pardon on October 23. The context: Binance had reportedly helped build the code behind World Liberty Financial’s stablecoin (Binance disputes this), and a $2 billion foreign investment into Binance had been routed through that very stablecoin, vaulting the Trump family venture, in one reporter’s phrase, from small time to the big leagues. The pardon power is one of the least checkable powers in the Constitution. The founders made it nearly absolute, gambling that presidents would use it with a sense of public trust. Used as an instrument of personal business interest, it converts a tool of mercy into a tool of self-dealing, and there’s almost no formal check left to stop it. That isn’t a scandal about one man. It’s a demonstration that a structural assumption baked into Article II no longer holds.

Make institutional arguments, not partisan ones

Notice what the three have in common. In each case the wrong isn’t that the President got richer. It’s that a limit, a foreign-gift rule or a divestiture norm or a pardon-power trust, was tested and found not to bind. Once a limit is shown not to bind, it stops being a limit for everyone who follows. We have seen this shape before: a norm survives only until someone proves it optional, and then the next occupant inherits the lower floor.

This is why I find the right-of-center critics more clarifying than the watchdog tallies. When The Dispatch treats the pardons as cronyism, or when Reason argues that the worst of this term flows from a rejection of the constitutional and libertarian principles that constrain power, they aren’t making a partisan complaint. They’re making the conservative one: that the structures matter more than the man, and that the gravest damage a bad officeholder does is what he normalizes, far more than what he takes. A libertarian writer at R Street put the disappointment plainly. The people he expected to be the bulwarks, the conservatives and libertarians and people of faith, mostly waved it through.

Whether any of these episodes is a prosecutable violation is genuinely contested. The courts have historically dodged emoluments suits on standing grounds, and the pardon power really is close to absolute. But “probably can’t be litigated” is not the reassurance it sounds like. It’s the actual problem. A limit that can’t be enforced is a limit that depends entirely on the character of the person it’s meant to constrain, which is precisely the thing constitutional design is supposed to make unnecessary.

The trust problem underneath

This connects back to the people I wrote about in the first post, the voters who told ethnographers a president would be justified in ignoring court orders for the good of the country. It would be easy to read that as a tolerance for lawlessness. I think it’s closer to the opposite. It’s what happens when people conclude the lawful institutions stopped working for them a long time ago. If you believe the courts and the agencies and the universities and the press and the healthcare system are already tilted against you, already not neutral, already not trustworthy from where you stand, then “deviate from the norms” doesn’t feel like an attack on the system. It feels like the only correction available to a game you’d already decided was rigged.

The norm deviation is downstream of the distrust. And the distrust, in many cases, is earned. The institutions aren’t as captured as the most cynical account claims. But a great many Americans have had real experiences of being unheard, condescended to, treated as a problem to be managed rather than a constituency to be served, and that’s enough.

Trump’s constitutional-limit-jumping works because of that distrust, not in spite of it. When people no longer believe the institutions are fair, the limits those institutions embody start to look like obstacles protecting a rigged game rather than safeguards protecting everyone. Erode trust first, and the limits fall almost on their own. Nobody rushes to defend the referee once they’ve decided the referee was bought. That’s the mechanism. The self-enrichment gets the headlines. The distrust is the solvent dissolving the constraints.

So the response can’t only be “defend the limits.” You cannot lawyer your way back to legitimacy. The deeper task is to rebuild institutions that are trustworthy regardless of your politics, institutions a conservative in rural New York and a progressive in Brooklyn can both look at and believe will treat them fairly. That is genuinely hard, harder now than it has been in my lifetime, because polarization means we increasingly don’t even share a baseline account of what “fair” looks like. But it’s the only durable answer, because limits that aren’t backed by trust are just paperwork, exactly as the emoluments workaround demonstrates.

This is what I mean by deep pluralism. It isn’t the thin hope that we’ll all eventually agree. It’s the harder commitment to building resilient institutions that can hold the loyalty of people who don’t agree and never will. Pluralism isn’t a temporary condition to be resolved on the way to consensus. It’s the permanent weather. Institutions are either built to function in it or they aren’t, and the ones that aren’t will keep generating exactly the distrust that makes norm deviation feel reasonable.

How is “America all right”?

The reconciliation isn’t a rhetorical trick; it’s the whole point. I can say America is all right and this administration is doing real damage because they operate at different levels. The country, its civic instincts and its communities and its broadly shared belief in fair play and limited government, is the deep, slow-moving thing. The conduct of one administration is the fast, shallow thing on the surface. And the evidence that the deep thing still works is everywhere in this very story. ProPublica can still publish the records. Watchdogs across the spectrum can still name the conduct. Conservative writers can still call it what it is. Courts still exist as partial checks. You and I can still read about all of it and say so out loud. We are not (yet) in a place where corruption is the unquestioned operating system. We’re in a place where it’s being attempted in the open and contested in the open. That contest is the institutional work.

It’s the same argument I keep making about pluralism. Resilient institutions don’t depend on good people being in charge. They’re built to survive bad ones. The task in front of us isn’t to despair that a bad actor reached the top. That was always going to happen eventually; the design assumed it. The task is to make sure the limits hold, so that when this term ends, the floor hasn’t been permanently lowered for whoever comes next, of either party.

America is all right. The job is to keep our institutions worthy of it.

If you haven’t read the first half of this argument, the case that the country itself is sound, it’s here: “America Is All Right”. And I’d genuinely like to hear from the conservatives and libertarians reading this. Does the institutional version of this worry land differently than the partisan one? The comments are open.

Your Turn

I love lively debate, disagreement, and anything else you can throw at me. What am I missing? Am I wrong about America? Am I wrong about Trump? Let me know in the comments. And if you like what you read, please share! (And subscribe if you haven’t already.) Reader shares are still the major way new readers find me.

Leave a comment

Share

Subscribe now

Our town held its Little League opening parade a couple of weekends ago. The volunteer fire station band played and hundreds of people lined the street, parents and little leaguers and a few grandparents in lawn chairs. The varsity baseball team turned out to welcome the smallest kids, and at the end the Boy Scouts ran the flag up the pole. I stood on the curb watching all of it and thought, a little to my own surprise: we’re okay.

That is probably not the thought I’m supposed to be having right now. The thought I’m supposed to be having and the one the headlines and my social media feeds hand me every morning is that the country is coming apart. A lot of our national despair comes from a particular slippage in logic. We watch the national government strain and fracture, and we quietly conclude that the rest of the country must be fracturing with it. But the American public is much larger than its national government. When you look at the local level, at the people and the communities and the things most Americans actually care about, there is a great deal of evidence that we are in better shape than the news suggests.

This is easy to misread as Pollyanna optimism but I don’t think it is. Nationally there is plenty to be frightened about, and I spend an entire companion post being frightened about it (hang in there until tomorrow when that post drops!). But I’ve started to think we might be turning a corner, and I don’t think I’m fooling myself.

A few weeks ago I was in DC and got to talking with the Ethiopian driver taking me across town. I asked him what his favorite thing about America was, expecting something small and polite. He didn’t hesitate: “freedom and the equality.” The first concrete thing he named was that here he can work, earn a living, and keep what he earns. In Ethiopia, he said, they take it. The second was that here he can criticize the government, which he then proceeded to do, cheerfully, from the front seat. I told him a lot of us are pretty worried about American politics right now, and that I nonetheless suspected we might be all right. He laughed and nodded. “Yes,” he said. “We’re still good.”

He might have believed it more than I did right then, but I still think I’m pretty optimistic.

Now I know a parade and a cab driver are vibes, not evidence. The evidence turns out to be more interesting than the vibes.

Conservative voters are complex, like everyone else

The trip I was in DC for was a launch symposium for an ethnographic study from ReD Associates and Johns Hopkins’s SNF Agora Institute, and it reframed how I think about this moment.

The study began from a deliberately uncomfortable premise. The researchers went looking for people who agreed with a statement along the lines of: President Trump is justified in ignoring court orders for the good of the country. The researchers call this “democratic norm deviation,” or a willingness to break from democratic and limited-government norms when you feel the country requires it. The point wasn’t to condemn these people. It was to understand them.

The assumption I hear constantly, especially from people on the left is that these folks are just bigots. They’re fine with terrible things happening to their neighbors. If that narrative were true, if half the voting public were comfortable with fascism, then yes, you should be terrified. We would be in a genuinely catastrophic place.

Fortunately for all of us, the research doesn’t show that and it’s not even close. What the ethnographers found is that these voters operate from a baseline picture of the American republic that is, frankly, optimistic. They believe in limited government. They believe in faith and family and the importance of place: of having a community you grew up in and love and want to protect. Their view is that the country overcorrected on some things, that wokeism overreached and the border genuinely got out of control. The uncomfortable part, from where I sit, is that each of those concerns has some real basis.

Understanding doesn’t require agreement

To be fair, my disagreements with those voters are real. I don’t accept the norm-deviation argument. You don’t get to abandon democratic norms because you feel the country needs you to. And I think these voters misjudge two things. First, I think they’re misjudging the severity of specific problems, both how dangerous the border situation actually is and how dangerous wokeism actually is. Second, and more important, the solution. Handing sweeping power to Trump is a profound mistake, because he is not a good-faith actor and he is not looking out for the average person. This is, by my read, the most corrupt presidency in American history when it comes to self-enrichment on the backs of ordinary Americans.

So we don’t agree, but the shape of the disagreement matters. Their sense that something is rotten in the state of Denmark is not wrong. Most of us, across the spectrum, can see that a lot of our institutions are failing, higher ed and the media and healthcare among them. And the pronounced left-lean of many of those institutions does leave conservative-leaning Americans feeling like they have nowhere to be heard. When people feel unheard, they go looking for other ways to be heard. That isn’t pathology. That’s human.

This is actually good news! Sort of.

I find real hope here, for three reasons.

The first is simply that these people have a good set of values. We are not locked in a struggle against half of America that wants fascism. We are trying to understand a large group of fellow citizens who have felt unheard, especially in elite spaces. That’s a completely different problem, and a far better one to have.

The second is that it hands us something to actually do. If the real issue is failing institutions and collapsing trust, that is fixable. Progressives and conservatives and libertarians can all agree that much of our healthcare system and much of our media are broken in ways that make them hard to trust. So the question turns constructive: how do we build institutions that are trustworthy to people regardless of their politics, institutions a Republican and a Democrat can both rely on? That points toward real work on viewpoint diversity, institutional neutrality, and better ways of bringing conservative Americans into these conversations. We are doing that work badly right now. (A related point I keep making: both liberals and conservatives often behave as though the other perspective will eventually just disappear. They won’t. We need institutions resilient enough to hold political pluralism now and permanently.)

The third is the method itself. The researchers went into people’s homes. They asked questions. They got curious. They didn’t walk in assuming these were racists, and they didn’t walk out agreeing with them. What you get from that kind of encounter isn’t consensus but understanding. You can still come away thinking someone is wrong about the world while knowing they’re operating in good faith, that they love their country and their neighbors and are trying, like the rest of us, to figure out how to move forward. This pattern shows up well beyond one study. More in Common’s Hidden Tribes work points the same direction: most people who voted for Trump are not MAGA hardliners, and even most of the hardliners are, in their own way, acting out of real care for the country.

We keep studying the bullet holes

There’s a famous story from the Second World War that relates here. Analysts looked at the bombers coming back from missions, mapped where they were riddled with bullet holes, and proposed adding armor to those spots. A statistician named Abraham Wald pointed out the mistake. The planes in front of them were the ones that made it home. The bullet holes they were studying marked the places a plane could be hit and still survive. The spots that actually needed armor were the ones with no holes at all, because the planes hit there never came back to be studied.

We read the country the way those analysts first read the planes. We study the damage, the failures and the scandals and the worst day in each news cycle, because that’s what comes back to us, that’s what makes the headlines. And we miss the parts holding the whole thing up. The Little League parade does not trend. My aunt and uncle, who left this week for a year and a half of mission work in a small, hot town in Guatemala, spending a good chunk of their own savings to make strangers’ lives a little easier, will not generate a single news story. (Maybe I’ll write one.) People like them are everywhere, doing quiet and unglamorous good, and almost none of it is legible to a feed optimized for alarm.

The week of the parade I also watched my daughter take her first communion. I sat in the pew and watched a whole congregation come forward (a woman in bike shorts, a grunge t-shirt and a formal lace mantilla among them), everyone taking the sacrament in their own slightly different way, all of us inhabiting and creating one community. I felt exactly what I felt on the curb at the parade. We’re okay. Not finished. Not without serious things to fix. But okay.

We need reform, not despair

None of this is a reason to relax. Authoritarianism is easy to animate when people feel unheard and under attack and convinced no one is listening. We have to take the research seriously and think hard about how we reconnect with voters, conservatives especially, because we are failing at it.

But the underlying material is strong, and that’s the thing I keep returning to. America is more than its national government. We are stronger than the moment suggests, in the people we’ve raised and the communities they hold up and the things the average American actually cares about.

America, I think, at least for right now is all right.

Tomorrow I’ll post the companion to this post is about the part that isn’t ok: the Trump administration, and why the danger it poses is real even when the country underneath it is sound. I’ll link that here when the post is live.

Your Turn

If you’ve got a Little League parade or a cab driver or an aunt in Guatemala of your own, I’d love to hear about it in the comments. What, if anything, has you convinced we’re ok? What am I missing? And as always, subscribe and share if you find my work valuable. Reader shares are one of the main ways my work finds new readers.

Leave a comment

Subscribe now

Share

Last week I called Gene Lokken’s death a case of structural hobbling. While I focused on AI algorithms given the Pope’s encyclical on AI, I was a bit disingenuous in how much blame I placed on the AI, which I noted in a footnote. The truth is that while faceless algorithms do create the potential for real harm, the Lokken case actually points to something even more disturbing: the algorithm that cut off his skilled nursing coverage was almost certainly correct on its own terms.

The research is reasonably clear that frail elderly patients with the kind of clinical picture Lokken presented rarely benefit from rehabilitation in the technical sense. They often decline functionally during the rehab stay itself. In one cohort of patients seventy-five and older admitted to a geriatric rehabilitation unit, roughly one in six measurably lost function between admission and discharge, leaving in worse shape than they arrived. UnitedHealth’s nH Predict almost certainly identified a real pattern when it flagged Lokken’s case for early termination.

While this doesn’t excuse the denial, it changes what the denial is a denial of. And notice what the denial could and could not do. If rehab really was the wrong setting, the humane response would have been to move Gene Lokken somewhere better: home, with paid support, or hospice alongside comfort care. The algorithm had no power to do any of that. Its only lever was the checkbook. So even granting that it read the clinical picture correctly, the one action available to it was financial. It could stop paying. It could not move him to a better setting, and it did not. He stayed exactly where he was, now on his family’s dime. Being right and doing good came apart completely. That distinction matters, because it points to a deeper problem than any of the available villains can account for, and to a slower fix than the lawsuit Lokken’s family is rightly pursuing will produce.

The algorithm did not create the harm

A skilled nursing facility was almost certainly not the right place for Gene Lokken to spend the last year of his life. The trouble is that, given the structure of Medicare and the structure of the post-acute care market in the United States, an SNF was very plausibly his family’s only option that anyone was willing to pay for.

Custodial home care, the level of care most elderly people actually need at the end of life, is not covered by Medicare.

Hospice, which would have covered comfort-focused care at home, requires that the patient surrender further curative treatment for the same diagnosis, and Medicare regulations prohibit the use of hospice and SNF benefits simultaneously for the same condition. Hospice also typically covers only 4-8 hours of care per day, which is much less than working families need to care for a terminally ill loved one.

Hospitalization, the alternative covered by Medicare without restriction, is one of the most dangerous places an elderly person can be: roughly a third of hospitalized patients over seventy acquire a new disability during the stay that they did not have when admitted, often even when the illness that put them in the hospital is successfully treated. To even qualify for the SNF coverage Lokken received, his family first had to clear Medicare’s three-day rule, which requires a three-day inpatient hospitalization before SNF benefits trigger at all.

What we end up with is a complex web that captured Gene Lokken more effectively than any intentionally built net could have: An elderly patient with a fracture cannot receive paid-for care at home. He cannot combine comfort-oriented hospice with rehabilitation. He must be hospitalized for three days to qualify for the SNF that is his only post-acute facility option, and he will likely leave that hospital with new disabilities he did not arrive with. The structure offers, for most families, exactly one realistic post-acute pathway, and that pathway is the one nH Predict’s data correctly indicated was not, in the strict clinical sense, where the patient needed to be.

The nH Predict algorithm answered one question: does continued rehabilitation clinically benefit this patient? It answered correctly. But the structure had quietly made that question stand in for a different one: should anyone keep paying to care for this man at all? Those are not the same question. A frail ninety-one-year-old can be past the reach of rehabilitation and still need round-the-clock care that someone has to fund. By collapsing the second question into the first, the structure let a valid clinical finding settle a question it does not actually answer. The algorithm was right about rehab. It was never asked the question that mattered.

The obvious objection is that no insurer should fund care that does not work. That objection is right. The harder thought is that no humane society should let an elderly man with a broken leg discover that the only paid-for care available is the kind that does not work. Both are true at once. Medicare is structured so that the second truth has no leverage on the first.

This is not the algorithm’s failure. The algorithm read the structure correctly. The structure is the failure.

Structural hobbling and rigidity cascades

In the academic world, I write on a pattern I call structural hobbling: the constellation of well-meaning, procedurally legitimate rules that, in the aggregate, produce outcomes no one designed and no one is positioned to fix. Jennifer Pahlka, whose work I leaned on in last week’s post, calls a related dynamic the cascade of rigidity: policy intent, however flexible at the top, becomes more rigid at every step of the bureaucratic descent until the person who actually has to act, whether the discharge planner, the SNF administrator, the family member, or the algorithm, has no authority to recognize a misfit.

Both patterns are visible in the bind that killed Gene Lokken’s options. The three-day rule was written in the 1960s, when the structure of post-acute care looked entirely different than it does today; it has survived because rescinding it would require Congress to do something rather than the absence of action. The hospice-and-SNF exclusion was written to prevent double-billing, which is a reasonable thing to prevent; the people who designed it did not intend to force families to choose between rehabilitation and comfort care for a dying parent, but in operational reality that is what it does. The absence of custodial home care coverage in Medicare reflects the original statutory line between “medical” and “non-medical” services, a line that made sense in the political environment of its drafting and makes no sense in the clinical environment of a twenty-first-century aging population.

Every one of these rules has a defensible origin story. Every one of them is followed conscientiously by people doing their jobs. The aggregate effect of all of them, on the family of a ninety-one-year-old man with a broken leg, is that the only humane option is one the algorithm correctly identifies as clinically inappropriate. The cascade does not produce this outcome through any single bad decision. It produces it through the patient summation of defensible ones.

Four dimensions on a single case

In my public work I talk about four-dimensional thinking (4D for short). It’s a discipline for refusing the false binaries that political and policy conversation keeps trying to force onto situations that have more sides than that. Lokken’s case is a good one for it.

The first dimension is where you are standing. Gene Lokken was ninety-one, recovering from a fracture, frail, in nursing care his physicians had asked for and his family was paying for after his coverage ended. From where he was standing, the question was not “rehab or not” in the abstract. It was “rehab or hospital readmission or inadequate home support or a hospice transition his family had not been counseled toward and probably did not know was on the table.” Each option had a different sticker price, a different probability of survival at one year, a different probability of returning to the hospital, and a different probability of dying with people who loved him in the room.

The second dimension is the breadth of the social world the decision sits inside. The same decision looked different to everyone touching it: a family juggling jobs and a crash course in healthcare bureaucracy; a discharge planner inside a workflow built to move patients out on Medicare’s clock; an SNF administrator inside a reimbursement structure that pays more for rehabilitation than for hospice referral (and research suggests SNFs are reluctant to refer dying residents to hospice partly because it cuts their revenue); physicians trained to default to cure because cure is what the system pays for; and an algorithm contracted to flag patients unlikely to benefit from more rehab. The next section returns to these actors one at a time.

The third dimension is the pits and peaks, the actual tradeoffs, which do not line up the way intuition expects. Comfort and continued life are not the same thing; surviving longer in a setting that produces hospital-acquired disability is not always better than surviving less long in your own bed. The clinical literature now talks routinely about rehab and death, the pattern in which elderly patients are rehabilitated up to and through their last days because the structure does not support transitioning them to comfort care. Every actor in Lokken’s case was navigating those tradeoffs without the vocabulary or the incentives to surface them honestly.

The fourth dimension is time. The question for Gene Lokken in the spring of 2022 was not the question for Gene Lokken in the spring of 2023. The question for the Lokken family is not the question for the next family in their situation. The question for Medicare in 1965, when the post-acute care system was being built, is not the question for Medicare today, with a population whose median life expectancy is more than a decade longer and whose end-of-life trajectory is differently shaped. Each of these time horizons points to a different decision. The system is calibrated to almost none of them.

In four dimensions, the situation does not have a right answer that anyone in it failed to find. It has a structural shape that makes the right answer almost unreachable, regardless of the integrity, training, or judgment of any individual inside the structure.

Three goods: no way to honor all three

The four-dimensional view helps us see what was true about the case from multiple angles. It does not yet name why each actor in the case kept making the decision they made. To get to that, you have to name the three competing goods that were live in every step of Gene Lokken’s discharge.

The first good is getting it right in the clinical sense. Ninety-one-year-olds with Lokken’s profile rarely benefit from rehabilitation. The algorithm’s job, on its own terms, was to identify exactly that. From this angle the algorithm was honoring a real good. Refusing to put frail elderly people through treatment that will not help and may harm them is a moral commitment, and it is the commitment geriatricians have been arguing for in the medical literature for decades.

The second good is honoring the expectations of patients and the people who treat them. When a patient’s own physician recommends a course of care, the patient and the family build their planning around the assumption that the care is necessary and that the system will pay for it. Neither of those things may be true or justifiable in the cosmic or even macro-moral sense, but expectations matter. Overriding that recommendation in software, without conversation, with no human reachable to push back, breaks an implicit contract that runs deeper than what any insurance policy spells out. From this angle the algorithm was failing a real good, the one the physician was honoring by making the recommendation in the first place.

The third good is making the most humane choice available given the system you actually have. The discharge planner, the SNF administrator, and Lokken’s family were not choosing between rehab and a better option. They were choosing between rehab and repeat hospitalization, between rehab and inadequate home support, between rehab and a hospice transition the family may not have been able to manage or have been counseled toward. From this angle the SNF stay was the most humane realistic path even if it wasn’t a terribly good one, and cutting it off was an attack on the only working option the family had.

These three goods are all legitimate. None of them is a mask for self-interest (at the very least it’s complicated). Clinical correctness is what good geriatric medicine looks like. Honoring expectations is what good doctor-patient relationships look like. Choosing the most humane option under constraint is what conscientious caregivers and discharge planners do every day.

The problem is that the system Gene Lokken was inside made it impossible to honor all three at once. Every actor was honoring the good closest to their role, and the aggregate outcome honored none of them. The algorithm got the clinical question right and the relational question grotesquely wrong. The physician got the relational question right and was clinically constrained by what the system would pay for. The family got the humane-choice question as right as they could and paid for it out of money they had hoped to leave to grandchildren. The system as a whole produced an outcome that no one in it could defend.

This is what the combination of structural hobbling and the cascade of rigidity produces inside an actual human life. It is not bad people choosing wrongly among clear options. It is good people choosing rightly among options each of which sacrifices real good(s). The structure is the only actor in the room with the standing to make a different choice, and the structure does not have a face you can address.

Lokken’s case indicts much more than the algorithm

This is the part of the argument that the lawsuit cannot reach, and that the encyclical I wrote about last week is in fact pointing toward when it talks about subsidiarity and the design tests for institutional accountability. In fact, it points to the reality that the principles of the encyclical go far beyond AI. The problem in Gene Lokken’s case is not that an algorithm overrode a doctor. The problem is that the institutional structure inside which both the algorithm and the doctor were operating produces the same outcome under any allocation of authority that respects the rules. Give the doctor full authority and the doctor is still constrained by what Medicare will pay for. Give the discharge planner full authority and the discharge planner is still constrained by SNF capacity and reimbursement structures. Give the family full authority and the family still runs out of money in three months. Replace the algorithm with a human reviewer and you get a human reviewer making the same call slightly more slowly, with no more authority to recommend anything better.

This is the pattern that physicians caught in prior authorization, teachers caught in IEP compliance, social workers caught in CPS protocols, and front-line workers in nearly every American institution will recognize on first read. Good people. Defensible rules. Structurally produced harm. Pahlka’s career is built on the observation that the answer to this is not to find a better person or write a better algorithm. The answer is to reach back into the architecture that produces the bind and rebuild it, slowly, with the people who actually have to live inside it as partners in the redesign. But (sadly) that’s not how American institutional reform usually works. It is how it would have to work for Gene Lokken’s grand-niece, in 2055, to face a fundamentally different choice than his family did in 2022.

I’ve spent most of my academic career on the cluster of questions this case sits inside. My book The Medicalization of Birth and Death (Johns Hopkins University Press, 2019) argues that we have made medical settings the default home for the first and last weeks of American life, and that the structure of insurance, professional licensure, and liability has made the alternatives nearly impossible to choose. A bioethics article I just shipped makes the structural-injustice version of the same argument: the four principles of medical ethics, autonomy, beneficence, nonmaleficence, and justice, are taught and applied as if they were independent inputs that a thoughtful clinician can balance.

They are not.

The structure forces tradeoffs that consume all of them simultaneously, and good ethical reasoning inside a broken structure produces broken outcomes regardless of how careful the reasoning is.

That is the harder claim, and it is the one this case has been making to me since I first read about it. In my bioethics work I keep returning to a figure I call Joseph, a ninety-one-year-old who wants to die at home and cannot find a way to do so because the web of regulations and rules makes the most simple and most justified request impossible. Gene Lokken is one more Joseph, and there will be others. The bind was not built by the algorithm, or the clinician, or the family. It was built by the structure of the system and the structure is often the thing we avoid talking about.

Structural vs. process reform: No villains, just rebuilding

Last week I argued that the encyclical’s extension of subsidiarity to private digital power was the move that mattered, because it gave us a vocabulary for naming what is wrong with bureaucracies whose decisions cannot be reached. I stand by that, but now I want to add to it.

The vocabulary of subsidiarity, of human-scale accountability, of legibility and reachability, is necessary. It is not sufficient.

There is no version of the Lokken case in which putting a human on the appeal line, or making the algorithm legible to the family, would have produced a humane outcome.

The humane outcome required, among other things, that custodial care at home be paid for, that hospice be available alongside continued rehabilitation or that hospice covered more than four hours of care a day, that the three-day hospitalization rule not exist in its current form, and that SNFs have palliative training and incentive structures that don’t penalize them for transitioning dying residents to comfort care. Each of these is a structural reform that has to be won at the level of Medicare statute, agency rulemaking, professional training, and reimbursement design.

This is slower work than naming a villain, and not a substitute for it. The villain account is satisfying because it points at someone we can demand a different decision from. It’s also deeply satisfying to blame Big Tech because they’re a villain everyone agrees to villainize. The structural account is harder because the people we would have to demand different decisions from are, in the relevant sense, all of us, slowly, over the course of years, in the rule-making processes and professional bodies and statute revisions that almost no one is paying attention to. High-profile litigation is often the precondition for the harder structural fix, and my quarrel is not with naming a villain but with stopping there and with treating the verdict as the end of the work rather than the moment that makes the rest of it possible.

I have been making this argument in academic venues for a decade. It is the argument most likely to get a piece rejected from the journals that publish on AI ethics, on the grounds that it’s “not about AI.” It is also the argument most likely to be correct. The encyclical does not say it in this form, but the encyclical’s deeper claim, that the moral character of an institution is in its design choices and not only in its outputs, points the same direction.

Gene, and the families we have not yet failed

Gene Lokken’s family is in litigation against UnitedHealth, and they should win. They will win on a narrower (and in some ways inaccurate) claim than the one I have been making here. They will win on the claim that an algorithm should not override a treating physician’s judgment, which is mostly true, and which is the claim the lawsuit is best positioned to vindicate. The claim I am making, that the structure inside which the algorithm and the physician were both operating produces the wrong answer either way, is the claim that the lawsuit cannot vindicate and that will not be vindicated by any single court decision.

This isn’t a counsel of despair. It’s a counsel of where the actual work is. The structural reform of Medicare’s post-acute care provisions, of the hospice and SNF benefit boundary, of the three-day rule, of custodial care coverage, of palliative training in nursing facilities, and of the reimbursement structures that make all of the above durable, is exactly the kind of work that civic institutions, scholarly research, professional associations, and the slow grind of bureaucratic reform are positioned to do. None of it will produce a headline, but at least some of it would be transformational for the next ten thousand families in Gene Lokken’s situation.

The algorithm did not kill Gene Lokken. The structure inside which the algorithm and his family and his physicians were all making defensible decisions did. That is the worse story. It is also the true one.

Next week I’ll walk through what trustworthy institutions actually take, and why the strategy of telling Americans to trust more is the institutional version of victim-blaming. But for now, I’ll leave some links to additional reading in the Notes and Citations section below. And before you get too engrossed in that work, leave me a comment below!

Notes and citations

• On the previous post: “An Algorithm Denied Gene Lokken’s Care. The Pope Just Named the Pattern.” — link to be added at publish.

• On the pattern of rehabilitation up to and through end of life, see Singh et al., “Rehab and Death: Improving End-of-Life Care for Medicare Skilled Nursing Facility Beneficiaries” (Journal of the American Geriatrics Society, 2026), and the related work on rehabilitation therapy for nursing home residents at end-of-life.

• On functional decline during a geriatric rehabilitation stay, see “Functional decline in geriatric rehabilitation ward”. On hospitalization-associated disability in older adults more broadly, see Covinsky, Pierluissi, and Johnston, “Hospitalization-Associated Disability” (JAMA, 2011).

• On Medicare’s three-day rule for SNF coverage, see the CMS guidance. On Medicare’s general non-coverage of custodial / long-term care, see Medicare.gov on long-term care.

• On inappropriate SNF payments and quality concerns, see the HHS Office of Inspector General reports and the Center for Medicare Advocacy on opportunistic nursing home ownership.

• On the broader argument about medicalization of birth and death and its costs to patients and families, see my book The Medicalization of Birth and Death (Johns Hopkins University Press, 2019).

• On structural hobbling, see my article “Structural Hobbling: Regressive Harm, Diffuse Responsibility, and Structural Injustice” (Social Philosophy and Policy, 2025).

• On the cascade of rigidity, see Jennifer Pahlka’s Recoding America (2023) and her AI Meets the Cascade of Rigidity (Niskanen Center, 2024).

Your Turn

If you have been a family member navigating a parent’s post-acute care, you have lived inside the structure I am describing. What did you have to fight for that, in a sane system, should have been the default? Drop it in the comments. I read them.

Leave a comment

Subscribe now

Share

In the spring of 2022, Gene Lokken, a 91-year-old man in Minnesota, fell and fractured his leg and ankle. He was admitted to a skilled nursing facility, and his doctors recommended a course of rehabilitation. Two and a half weeks into the rehabilitation, his Medicare Advantage plan, administered by UnitedHealthcare, cut off his coverage. The decision did not come from a doctor at the insurance company. It came from an algorithm called nH Predict, developed by a UnitedHealth subsidiary, that had estimated, based on patterns in the histories of other patients, that Lokken should not need any more rehabilitation. His treating physicians disagreed. The algorithm prevailed. Lokken’s family paid out of pocket, up to $14,000 a month, for the next year of his care. He died still in that care, in nursing his physicians had asked for and his insurance, advised by software, had refused to cover.1

Lokken is one of the named plaintiffs in a federal class action against UnitedHealth Group that is, as I write this, working its way through discovery in the District of Minnesota. In March of this year, the judge in that case ordered UnitedHealth to disclose a broad range of internal documents about how the algorithm was designed, how it was used in coverage decisions, and whether it was meant to override the clinical judgment of treating doctors. The plaintiffs cite a striking number from the algorithm’s track record: of denials that were appealed, roughly nine in ten were ultimately reversed. The structure was producing the wrong answer almost every time, on a population that had standing to push back. The Senate Permanent Subcommittee on Investigations reached a similar conclusion in a 2024 report covering several of the largest Medicare Advantage insurers, and in January of this year the federal Medicare program launched a new model that brings AI-driven prior authorization further into the way decisions are made about elderly patients’ care.

As in many cases of institutional or algorithmic failure, the institutions and algorithms themselves are not necessarily malicious. They are not even necessarily badly run. But they are structured so that no person you can reach is responsible for what is happening to you, and so that the cost of pushing back through the available channels is higher than the cost of giving up. And they’re also structured in a way where the interactions are so complex that finding the actual point where the decision went wrong is almost impossible. I call this pattern structural hobbling, and I’ve argued in a recent article that it is one of the most regressive features of modern American life, because the people who can route around it (with lawyers, with connections, with time) are not the people who can least afford to be hobbled in the first place.

On Monday, Pope Leo XIV published an encyclical called Magnifica Humanitas, his first major social encyclical. It’s being covered as a document about artificial intelligence, which it is. But it also a restatement of a political truth that crosses religious and secular lines, extending a centuries-old Catholic doctrine called subsidiarity into the digital age. In the process of both it names a pattern that both political theorists and government reformers need to take seriously.

When Catholics, the founders, and institutional economists agree

Subsidiarity, as the Catholic tradition has used the term since at least Pius XI’s 1931 encyclical Quadragesimo Anno, is the principle that decisions should be made at the level closest to the people affected. Individuals, families, neighborhoods, parishes, professional associations, and local governments do what they can. Higher levels of authority step in only to do what the lower levels genuinely cannot, and even then only to support, not to displace. The principle has a Hayekian sibling (the knowledge problem: distant decision-makers cannot see what local actors can), an Ostromian one (polycentric governance: complex problems usually do better with many overlapping centers of authority than with one) and a deeply American one (federalism as the safety valve for democracy). It is one of the few intellectual frameworks I know of that travels easily across the Catholic tradition, the classical liberal tradition, and the empirical institutional-economics literature.

Pope Leo connects AI and subsidiarity throughout the encyclical, but paragraph 71 lays out where he sees the real danger: “Here, the highest level is not the State, but rather major economic and technological actors that exercise de facto power over the conditions of everyday life. This level, which monopolizes expertise, data and decision-making authority, involves companies and platforms that define conditions for access, rules of visibility, forms of interaction, and even economic opportunities.”

Pope Leo XIV is clarifying what folks from all political viewpoints should appreciate, namely that the problem is not either government power or corporate power, but power writ large. In this case, the entity from which power must be devolved, in our era, is not the State alone. It is the small set of actors (private or public, but in the case of AI largely private) who, between them, set the terms on which hundreds of millions of people work, learn, communicate, borrow money, find jobs, and reach each other. AI and tech companies more broadly effectively operate as bureaucracies, with the same opaque procedures, the same diffuse responsibility, and the same absence of any human at the other end of the line.

The principle of subsidiarity offers a roadmap for navigating AI governance precisely because it protects against the two great centralizing dangers: state power deployed to eliminate competing viewpoints and corporate monopoly that forecloses meaningful choice. The deep pluralism that characterizes human life requires that people at every level retain the ability to use technology on their own terms and subsidiarity is the structural principle that makes that possible. In the AI space, this means allowing multiple models to flourish, preserving individual agency over data, devolving decision-making authority to local communities and making the costs and responsibilities of different approaches legible and accountable. What we’re really talking about is the same old fight between centralization and pluralism, playing out on a new technological terrain.

Algorithms and structural hobbling: not identical, but connected

Gene Lokken did not know what subsidiarity was. He also did not need to. What his family experienced, in the language of my own work, is the textbook case of structural hobbling: a constellation of well-meaning rules and procedurally legitimate decisions that produced a result no one in particular intended, with no one in particular accountable, and with the cost falling on people who could not absorb it. The Medicare Advantage rules were defensible in the aggregate. The algorithm was sold to the insurer as a tool to help “inform providers, families and other caregivers about what sort of assistance and care the patient may need.” The internal target the company set for its employees, that patient stays should stay within one percent of the algorithm’s prediction, was the kind of operational goal that would look reasonable on a slide deck. The decision to discharge Gene Lokken’s coverage after seventeen days was made through a process that, viewed step by step, was not obviously malicious at any single step. None of that helped his family pay the bill.

In my academic work I argue that this pattern is one of the most under-theorized sources of injustice in contemporary life, and that the standard left-of-center frame (which locates injustice in either individual bad actors or in the inactivity of the state) systematically misses it. Pope Leo XIV, in paragraph 95 of the encyclical, echoes the 1891 encyclical Rerum Novarum of Pope Leo XIII, who clearly inspires his own work and his own papacy: “In many cases within the digital context, control over platforms, infrastructure, data and computing power does not rest with States, but with major economic and technological actors. These entities effectively set the conditions for access, determine the rules of visibility and shape the very possibilities for participation.” While one could see this as a condemnation of corporate power itself, that’s not really the point. As he makes clear earlier in the encyclical, “In practice, however, technology is never neutral, because it takes on the characteristics of those who devise, finance, regulate and use it.” The term “regulate” here emphasizes the often complex relationship between government and corporate power, with the two frequently compounding each other. (If I were going to quibble with the Pope on anything in this encyclical it’s that he underemphasizes the way government power fuels corporate power, but more on that in a later post.)

My structural hobbling framework describes what he is naming as four interlocking features of tech broadly but AI more specifically. The first is scale: the platform mediates the lives of enough people that exit is not realistic. The second is opacity: the rules are not legible to the people they govern. The third is diffuse responsibility: no one decision-maker can be identified, summoned, or replaced. The fourth is regressive impact: the people best equipped to push back through the available channels are not the ones who most need to. Pope Leo’s encyclical names all four, in its own vocabulary, in the span of about two pages.

This is a similar argument to that made by government reformers like Jennifer Pahlka, whose work I’ve discussed before. Her description of a cascade of rigidity produces the same brittleness of institutions and citizens that I’ve diagnosed here. That a technocratic reformer, the Catholic Church, and a political theorist arrived at the same diagnosis is not coincidence. To me at least it provides strong support for a very real phenomenon and indicates that we need a much more expansive conversation about the complex interaction of private and public governance rules rather than seeing the two as distinct sets of rules with different goals.

What subsidiarity requires of Big Tech

It would be one thing if the encyclical simply named the problem. Naming a problem and prescribing a response are different acts, and most of what is written about AI right now does the first and not the second. The encyclical is unusual in that it does both, and it does the second by translating a centuries-old principle into concrete design tests for a new kind of institution.

The relevant passages run from paragraphs 71 through 72 and from 108 through 110. Stitched together and translated out of the encyclical’s vocabulary, they propose something like the following: Power should be held at the level closest to the people affected, which means that algorithmic decisions affecting livelihoods, credit, healthcare, or political participation should not be made in jurisdictions and on terms that the affected community has no representation in. Decision-making processes should be legible, which means that a person whose access is denied should be able to understand the basis for the denial in language a non-engineer can follow. Responsibility should be identifiable, which means that an actual human being, with a name and a job description, should be reachable when the system errs. The community affected should have a real voice in design, not be informed of the design after the fact (”rather than confining their role to mere oversight after the standards have been set elsewhere,” in the encyclical’s words at paragraph 109). And the people who do the unseen labor that makes these systems work (the data labelers, the content moderators, the workers in the resource-extraction chains) should be visible in the accounting. Incidentally, these are all principles of good governance that are violated almost more often than they’re followed in our current system.

I would add a sixth criterion from my own work that the encyclical does not state in this form, although it gestures at it. The system should be auditable. If an algorithm is doing the work of an institution, the algorithm itself, including its training data and its failure modes, should be open to inspection by parties the institution cannot capture. We have legal and institutional vocabularies for this in the public sector. Inspectors general. Independent audits. Statutory disclosures. There is no good reason these vocabularies should stop at the edge of the platform.

Read together, these criteria amount to a scorecard. You can apply it to a hospital network, a federal agency, a state DMV, a content platform, a credit-scoring service, a hiring algorithm, or a school district’s automated grading system. The scorecard cuts across partisan identity, because it describes what any person of any background needs in order to extend trust to an institution that mediates their life. It also doesn’t treat public and private power as though they are different animals with different intentions. It assumes that centralized power always has the capacity for rigidity if not abuse and therefore requires vigilance. This is what subsidiarity actually requires, in concrete terms, when you take the doctrine seriously and apply it to the institutions that now do the work that institutions used to do.

Subsidiarity and AI

A few months ago I wrote that talking about “AI policy” was a lot like talking about “computer policy”: the category is too broad to support useful regulation, and the right approach is to identify specific harms in specific use cases and target the response with precision. I stand by that argument. The encyclical does not, in my reading, require its abandonment.

What the encyclical does require is that the trust-but-verify framework be applied not only to the use cases on the demand side, but to the design choices on the supply side. In paragraph 104 Pope Leo says, in plain language, “we cannot consider AI to be morally neutral... every technical tool embodies choices and priorities through what it measures, ignores and optimizes.” This is the move that I think will sit with me longest. The targeted-regulation approach I defended works for deployment. It does not, by itself, work for design. By the time a tool reaches deployment, the anthropology of the human person that the designers held has already been encoded in what the model is trained to maximize, what it is trained to ignore, and what counts as an error in the loss function. We can either ask designers to make that anthropology visible, or we can let it shape our lives invisibly. The encyclical is right that those are the two options.

What this changes, in practice, is that the regulatory conversation needs an upstream half. Not in place of the targeted-harm approach but as a complement to it. We need a vocabulary for what it means for a tool, before it is deployed at all, to have been built in conformity with the kind of subsidiarity Pope Leo is describing. We do not yet have that vocabulary in American public life. The encyclical is one place to start building it. But it’s also not the case that formal government regulation is the only way to do this upstream work. It might even be the worst way to do this work. Subsidiarity gives us a range of options for holding companies accountable that don’t require bloated government bureaucracies to make design decisions about the design of private tech bureaucracies.

It’s no accident that Anthropic’s co-founder Christopher Olah was invited to the presentation of the encyclical and to speak afterward. He ended his remarks with a telling plea, calling for a partnership between the Church and tech companies, arguing: “We need informed critics who will tell the labs when we are failing. We need moral voices that the incentives cannot bend.” His remarks demonstrate two things: that it’s not impossible for tech companies to self-regulate (but they need support to do so) and that it’s not just the government who can exert pressure in this space. These are both important lessons worth remembering.

Takeaways for non-Catholics

A document like this can be hard to engage with for readers who do not share its theological starting points, and I have some sympathy for the impulse to set it aside on that basis. I want to push back gently on the impulse for two reasons.

The first is that Catholic social teaching is one of the most underused intellectual resources Americans have for thinking about institutional design across partisan lines. Subsidiarity has been worked out as a principle for over a century by people who were trying to defend the integrity of families, parishes, neighborhoods, and small associations against both the encroachments of the state and the encroachments corporations. It is one of the few frameworks in our intellectual landscape that takes the middle layer of society seriously as a thing in itself, not as a residual variable between the individual and the state. You do not have to be Catholic to find it useful. Hayek, Ostrom and our founding fathers all got to a related place from entirely different starting points, and the convergence is itself a reason to take the principle seriously.

The second is that this encyclical, more than most, is structured to be read by people outside the tradition. Pope Leo is explicit, in his opening chapter, that he understands the document as an offering for dialogue, and he treats the Catholic vocabulary as a contribution to a shared conversation rather than a closed system. The institutional theory holds whether or not you accept the theological frame underneath it. Take what you can use.

For Gene and the families who come next

Gene Lokken did not get his coverage back. The structure that took it away is still standing, and the lawsuit his family is part of is still in discovery. Whatever the lawsuit ultimately produces, the algorithm is still running, the Medicare Advantage incentives that made the algorithm attractive are still in place, and a new CMS model expanding AI-driven prior authorization is now live in six states. The structure has a name now, in a vocabulary that Catholics and Hayekians and a number of other people share. The principle for taking the structure apart has a name too. None of this fixes the next denial, and none of it puts a human being on the other end of the appeal that Gene Lokken’s physicians could not get answered in time.

A human on the other end is a smaller thing than the encyclical is offering. It is also more than most of the documents I have read on this subject manage. For now, at least, I’ll take it, but I’ll have a lot more to say on the failure of humans on the other end next week too.

Your Turn

Until next week, I’d love to hear reader thoughts on the encyclical: what it got right, what it missed (if anything), and how we should be thinking about both AI design and deployment in the coming days. Drop ideas, comments, and criticism in the comments! And please share and subscribe if you like what you read.

Leave a comment

Subscribe now

Share

A reader, Lori, left a comment on a recent post asking me to apply my “radical moderate” lens to the urban/rural divide. They don’t think it actually exists. Most of the demographers and urbanists they know agree. Here’s my attempt.

Thinking about urban vs. rural as one of those pesky false binaries I love parsing is a great question. And like many other binaries before it, destroying it requires holding two things in tension at once, which is exactly what radical moderation is for.

The short version: the urban/rural divide is one of the most powerful predictors of political behavior in American life right now. And the categories “urban” and “rural” are so unstable and so politically weaponized that treating them as a clean binary almost guarantees you’ll misunderstand what’s actually happening.

Both things are true and they have to be held together.

Steelmanning the “It’s Not Real” Side

The commenter who prompted this is in good company. Anyone who works seriously with the data (demographers, geographers, and urban planners alike) will tell you that “rural” and “urban” aren’t natural categories. They’re administrative ones, and the administration can’t agree on what they mean.

By the USDA Economic Research Service’s count, more than half of America’s “rural” population lives inside a metropolitan county. That means that the majority of rural Americans, by one official definition, live in metro areas by another official definition.

The underlying concepts are doing different work for different agencies, and the agencies don’t agree on what they mean. The Census Bureau defines urban as densely developed territory with at least 2,000 housing units or 5,000 people, with rural as the residual (”everything that isn’t urban”). The Office of Management and Budget operates at the county level, talking about “metropolitan” and “nonmetropolitan.” The USDA’s Economic Research Service uses a nine-category Rural-Urban Continuum Code. The National Center for Education Statistics has its own scheme on top of all that.

And “suburban”? The category most Americans actually live in? The Census Bureau doesn’t officially recognize it at all. Suburbs get folded into either “urban” or “metropolitan” depending on the system, even though when Pew asks Americans to self-identify, 43 percent say they live in a suburb. The largest community type in America has no official definition.

Stack on top of this the fact that, when researchers compared self-identification to official classifications, they found significant mismatches: only 45 percent of people officially classified as “urban” identify as urban, and only 48 percent of officially “suburban” residents identify as suburban. People in identical census tracts will describe their communities completely differently. The category is doing identity work as much as it’s doing geographic work.

The Carsey School at the University of New Hampshire, whose demographers Kenneth Johnson and Dante Scala have done some of the best work in this area, titles their election analyses with a thesis: “Rural-Urban is a Continuum, Not a Dichotomy.” They break the country into eight categories along that continuum. The political behavior of, say, a small metro suburb is dramatically different from a non-adjacent rural town, both of which would get binned together in a crude urban/rural split.

The categories are contested and constantly drifting. Treating them as natural kinds (as if “urban people” and “rural people” were two species) is bad social science.

Steelmanning the “It’s Very Real” Side

And yet.

You should be suspicious here. I just argued the categories are contested and constantly drifting. Now I’m about to cite a 40-point gap measured using those same categories. How can both be true? Stick with me.

In the 2024 presidential election, people who say they live in a rural area voted 69 to 29 for Trump. People who say they live in a city voted 65 to 33 for Harris. That gap has grown sharply over the last two decades. In 2000, Republicans held a narrow 6-point edge among rural voters; by 2024 the rural edge was over 40 points, with a similar margin in reverse among urban identifiers.

The number is built on self-identification, not on government classification. It’s not a fact about geography. It’s a fact about who claims a rural identity, and what that identity has come to mean politically.

This is not nothing. Whatever the definitional problems, something is sorting people along this dimension powerfully enough to produce one of the most reliable predictors of vote choice we have.

You can construct a similar pattern with education. College-educated voters favored Harris by 16 points; non-college voters favored Trump by 14 points. And education sorts geographically in ways that reinforce the urban/rural pattern. College graduates concentrate in metro areas; non-college voters are more dispersed.

You can do it with religious attendance, with occupational structure, with attitudes toward institutions, with media diets. The pattern is consistent enough that you have to take it seriously even while you’re picking apart the categories.

So which is it? Is the divide real or fake?

The Reframing: It’s Not Where You Live, It’s How You See Yourself

Here’s where 4D thinking helps, and where the most interesting recent political science meets the question.

The single most clarifying body of work I’ve encountered on this is by Kristin (Krissy) Lunz Trujillo, now at Boston College. She has spent years dismantling the assumption that “rural” is primarily about geography at all. Her argument, developed across her dissertation and a series of papers in Political Behavior, Political Research Quarterly, and elsewhere, is that rural identity is a social identity, not a location. And the political attitudes we attribute to rural America track that identity far more reliably than they track actual residence.

That’s a big distinction, and it does real work.

In her earlier survey work, Lunz Trujillo found that rural identifiers are not disproportionately lower- or working-class, and that rural social identity does not significantly differ between racial and ethnic groups. Translation: the standard journalistic image of “rural America” as essentially a synonym for the white working class is empirically wrong. Plenty of professionals, college graduates, and people of color identify strongly as rural. Plenty of poor white people in declining industrial towns do not. The Venn diagram has been collapsed in popular discourse in a way that distorts both circles.

Her 2022 paper in Political Behavior showed that rural social identification specifically (psychological attachment to being from a rural area or small town) predicts anti-intellectualism, because rural identifiers tend to view experts and intellectuals as an out-group. Notice what this is doing: the political effect runs through felt belonging, not through geography per se. A rancher who doesn’t identify strongly as rural and a college professor who grew up rural and still feels it deeply may not behave the way their addresses would predict.

And in a 2025 piece with Jennifer Lin in Political Research Quarterly, Lunz Trujillo finds that rural identity predicts political values more strongly than rural residence does. Rural identifiers place particular weight on morality and freedom as core values. People who live in rural areas but don’t identify that way? They look much more like everyone else.

This reframes a lot. The political “urban/rural divide” is, more precisely, a divide between people who hold rural identity as a meaningful part of who they are and people who don’t. That identity tracks geography imperfectly. It can persist after migration. It can be absent in people who’ve lived their whole lives in farming counties. It can be present in people who moved to the suburbs decades ago but still feel rooted in the small town they came from.

There’s a follow-up question this opens up, one I’m still working through. Why has rural identity become so politically resonant in the first place? My best guess is that it names something real: a steady hollowing of the everyday institutions of community life that urban America, for all its problems, has mostly been spared. Hospitals are closing. Local newspapers have folded. The factories left. The schools consolidate. While the story is more complex than we often hear, the opioid crisis was yet another symptom of rural vulnerability. Kathy Cramer’s The Politics of Resentment and Robert Wuthnow’s The Left Behind both make versions of this argument from on-the-ground ethnographic work, and recent quantitative work finds that rural Americans report lower trust in government regardless of which party holds the White House. That’s what you’d expect if the alienation is structural, not partisan. It’s also clear that their distrust, like I’ve argued here, isn’t unjustified.

Add a second crucial finding: a 2024 study by Suzanne Mettler and colleagues at Cornell, published in Politics, Groups, and Identities, found that the growing rural-urban political divide is driven almost entirely by white voters. Black voters in rural areas vote about the same as Black voters in cities, overwhelmingly Democratic. Latino voters show modest variation but nothing like the chasm among whites. Asian American voters likewise.

Putting these two findings together: the geographic divide we keep talking about is, mechanically, a particular pattern in how some white Americans construct rural identity and what that identity has come to mean politically. The white-working-class-rural archetype that dominates media coverage is a flattening of at least three things (race, class, geography) that don’t actually overlap as much as the archetype assumes.

This matters enormously. If you think the cause is geography, your prescriptions are about geography: broadband subsidies, rural revitalization, or urban renewal. If the cause is identity and meaning, the prescriptions look completely different. You can’t fix an identity-based cleavage with infrastructure spending. And you can’t understand an identity-based cleavage until you understand why that identity matters to people.

What else might be sorting people, beyond identity? Density (a smooth gradient, not a binary). Education (the most powerful single sorter we have). Economic structure (manufacturing vs. service economies behave differently regardless of population). David Goodhart’s “Somewheres vs. Anywheres” distinction between place-rooted and mobile cosmopolitan identities. None of these map cleanly onto “urban” and “rural” as administrative categories.

The Political Weaponization

Now we get to why the divide gets flattened into a binary, despite the data screaming “continuum, multi-dimensional, mostly about something else.” Both parties have strong incentives to maintain the fiction.

For Republicans, the rural-vs-elite framing converts a relatively small share of the electorate (about 14 percent of voters lived in nonmetropolitan counties in 2020) into a moral majority. It positions a coalition that is, in fact, ideologically diverse (agricultural producers, exurban professionals, declining-industry workers, evangelical Christians, and libertarians) as a unified bloc of “real America” against an alien urban elite. The flattening serves coalition management.

For Democrats, the urban-vs-backward framing converts what is in reality a fragile and shrinking metropolitan coalition into a self-image of educated, cosmopolitan modernity opposed to ignorant rural reaction. It papers over enormous tensions inside that coalition, between working-class urban voters of color and college-educated professionals, between immigrant communities and progressive activists, between renters and homeowners who share an address but not an interest. The flattening also serves coalition management.

No American benefits from the truer story: that we are looking at a complex sorting along multiple dimensions, that white Americans in particular have polarized along these dimensions in ways that don’t fit cleanly into geographic categories, and that the suburbs (where most Americans actually live and where elections are actually decided) are politically heterogeneous in ways that don’t fit either narrative.

The media reinforces this because the binary makes better content. “Two Americas” is a clean story. “A continuous gradient of density, education, rootedness, and economic structure interacting with a particular cleavage among white voters” is true but doesn’t fit in a chyron.

The Radical Moderate Take

So where does this leave us?

The divide is real, and we shouldn’t pretend otherwise. The gap between people who describe themselves as rural and people who describe themselves as urban is large and growing, and it’s one of the most reliable predictors of vote choice we have. People living in different places really do vote, worship, work, and live differently. To wave this away as “the data shows you’re wrong” is to dismiss the lived experience of millions of Americans who feel, accurately, that their communities are being shaped by political forces that don’t understand them.

But the categories are largely fiction, and the binary is largely manipulation. There is no clean urban/rural line. There are gradients, definitional inconsistencies, suburbs that don’t fit anywhere, and a divide that is mostly about white American identity in ways the geographic frame obscures.

The radical moderate move is to take both seriously: to honor the felt reality of difference while refusing the political utility of pretending it’s a simple binary. To ask: what is actually sorting people? What policies would address those actual sorters rather than the misleading proxy? What conversations could happen between, say, a small-town pharmacist and a city schoolteacher if they weren’t being told they belong to opposing tribes? What institutional repair do we need to regain the trust that’s been lost? How can we address the loss of community, hospitals, educational institutions, and public health infrastructure? This question is particularly relevant for my field of higher ed, given that consolidation will hit smaller suburban and rural institutions far harder than any others.

I suspect the conversations would surprise everyone. Most Americans, when you take them off the team jerseys, want versions of the same things: stable communities, meaningful work, kids who do better than they did, and institutions that don’t lie to them. The fiction of the divide is what makes us forget that.

The divide is real. The categories are mostly fake. And the most radical thing we can do is hold both truths at once and refuse the simpler stories on offer.

Your Turn

Let me know what you think in the comments! Do you identify as rural, suburban, or urban? Why? What does that identity meant to you? What kinds of reforms would help rebuild trust among rural-identifying white populations that have low trust in government? Are those (hopefully) the kinds of reforms that would build trust for other communities too? How can we prioritize making institutions more trustworthy instead of addressing trust as a marketing problem? What did I miss? And as always, if you like what you read, please share widely and subscribe!

Leave a comment

Subscribe now

Share

Someone asked me a question at a talk last week that I’ve gotten before (a lot, actually):

The question usually relates to something like conflict in Gaza or Ukraine or maybe rising fascism.

It’s a good question and dismissing it too quickly is its own kind of failure.

And in truth, in some moments, when the violence has already started and the institutions have already cracked, I’m not sure 4D thinking saves you. A framework for navigating complexity doesn’t do much good after you’ve already driven off the cliff.

But that’s exactly the point.

We Get to Crisis Because We Stop Thinking in 4D

Existential crises don’t materialize out of nowhere. They’re built, slowly, out of years of brittle institutions, zero-sum thinking, and the steady hollowing out of the spaces where people once worked through disagreement. Foreign policy is a useful example here. Power politics has always been the dominant logic of international relations, for all kinds of structural reasons. But even there, we’ve actively made things worse. When we hollow out the State Department, to take one recent example, we don’t just lose bureaucratic capacity. We lose the four-dimensional networks of diplomats and regional experts whose relationships on the ground are actually what prevents conflicts from becoming conflagrations. Those relationships take decades to build. They get dismantled in a budget cycle.

The result is that we arrive at genuine binaries, genuine moments where the choices feel like fire or gasoline, having spent years creating exactly the conditions that make those moments inevitable.

Stanislav Petrov and Resistance

The story has become almost a cliche, but it’s still worth retelling (most cliches have an element of this to them). On September 26, 1983, a Soviet lieutenant colonel named Stanislav Petrov was on duty at a command center outside Moscow when the early warning system flashed a single word: LAUNCH. Five American missiles, allegedly inbound. Protocol said to report it immediately. Escalation would have been automatic.

Instead of escalating though, Petrov paused. He thought about the 4D picture: a genuine American first strike would involve hundreds of missiles, not five. The system was known to be glitchy. Ground radar wasn’t confirming anything. He concluded it was a false alarm and sat on the information.

He was right. The satellites had mistaken sunlight reflecting off clouds for a missile launch. His pause, that refusal to let the binary logic of nuclear doctrine override everything else he knew, likely prevented nuclear war.

Petrov wasn’t practicing radical moderation as a political philosophy. He was practicing it as a cognitive habit. It’s a commitment to asking: is this actually as simple as it looks? Even under maximum pressure, even with every institutional incentive pointing toward escalation, he held the four-dimensional picture in his head long enough to see that the situation wasn’t what it appeared to be.

We need more of that, especially when the pressure to collapse into a binary is most intense.

The Danger of Behavioral Consistency

One obvious objection asks: isn’t this just status quo bias dressed up in philosophical language? A disposition to prefer existing arrangements over disruptive ones, to counsel patience when what the moment demands is action?

No, and the Petrov case is the reason why. He didn’t act slowly. He acted with extreme urgency. The difference is that the urgency he brought to that moment was the urgency of refusing to collapse, not the urgency of escalating. He tested whether the binary was real before he let it govern his decision. That test took seconds. But it’s what mattered most in that moment.

That’s what 4D thinking actually demands: not gradualism as a default and not revolution as a default, but a genuine structural test of whether the situation warrants the response being demanded of you. A radical moderate will sometimes look like a gradualist and sometimes look like a revolutionary. What she won’t do is collapse into the binary without first testing whether the binary is real. That can look like ideological inconsistency from the outside. Edmund Burke got that criticism so frequently there’s an entire literature on whether he had any principles at all (spoiler: he did). What looked like behavioral inconsistency was theoretically coherent: it’s not the principles that shift, it’s the context those principles are being applied to.

Petrov Is Each One of Us

Petrov’s case was unusual, though. He faced a collective action problem compressed into a single decision made by a single person in a single room. Most of the crises we actually face don’t work that way. Institutional decay, civil conflict, and democratic backsliding are diffuse processes driven by thousands of individual decisions made by people who often don’t know each other and aren’t coordinating. No single Petrov moment saves a hollowed-out legislature or a captured media ecosystem. The structural problems are real and they require structural responses (more on that in a future post on institutional decay).

But it’s also still true that each of us is making Petrov-like decisions every single day. When you decide whether to share the outrage-maximizing headline or pause long enough to read past it. When you decide whether to write off your MAGA uncle or stay in the relationship. When you decide whether to treat the algorithm’s version of your political opponents as accurate, or go find out what they actually think. None of those individual decisions averts nuclear war. But aggregated across millions of people, they either feed the brittleness or resist it. Petrov didn’t save the world by having a better system. He saved it by refusing, in one critical moment, to let the system think for him.

That move is available to all of us, more often than we think.

What Happens After

The other thing that doesn’t get enough attention is that even when crises do tip over into genuine conflict, the 4D thinking we bring to the aftermath determines whether that conflict ends or just pauses.

We know some things about this with reasonable confidence. Zero-sum solutions, where one side is heavily penalized and forced to absorb the full cost of the conflict, tend to produce the conditions for the next conflict. The standard reading of how World War I became World War II is basically that story, at least as one of the common arguments: the punitive terms of Versailles didn’t resolve the underlying tensions, they just compressed and deferred them. The Marshall Plan worked because it operated on a different logic: Reconstruction instead of punishment and a recognition that you still have to live with these people.

Restorative justice practices operate on the same insight, and there’s actually decent empirical backing for this now. Multiple meta-analyses have found that restorative justice programs produce higher victim and offender satisfaction, greater restitution compliance, and meaningful (if modest) reductions in recidivism compared to traditional punitive approaches. The research on post-conflict applications is more mixed, and context-dependence matters enormously. But the broad pattern holds: processes built around acknowledgment, accountability, and eventual reintegration outperform pure punishment on almost every outcome we actually care about, including whether the conflict stays over. South Africa’s Truth and Reconciliation Commission and Rwanda’s gacaca courts aren’t perfect models, and scholars debate their limits at length. But they’re attempts to do the harder thing: hold complexity, keep the 4D picture in view, and ask what it would actually take for people to live together afterward. They require leaning into all four dimensions, but particularly the fourth.

None of this is soft or naive. Genuine reconstruction requires you to hold complexity, to think about what comes next, and to keep asking the harder questions even when everything in you wants to just call it done. Punishment is emotionally satisfying and structurally simple. Restorative approaches are neither. They require exactly the kind of 4D thinking that crisis conditions make most difficult.

The Danger of Existential Inflation

It’s also the case that we’re systematically overusing the word “existential.”

Not every political conflict is a civilizational emergency. The habit of framing normal political disagreement in apocalyptic terms has real consequences. It licenses crossing institutional guardrails. It makes compromise feel like collaboration with evil. It gives political actors permission to do things they would otherwise know better than to do, because this time, surely, the stakes are too high for normal rules to apply.

I was talking to a friend recently who pointed out that the “punch a Nazi” rhetoric does precisely this. Sure, punch Nazis if you’re living in 1945 Germany or Poland. But in non-apocalyptic situations, we should not be punching people simply because they have terrible views. Down that road is a cliff and a particularly ugly one.

It’s not just political violence we have to fear though. Well before we get there, we have to deal with people quiet quitting politics because it sucks so badly when every debate is framed as existential.

Existential threats that never manifest exhaust people. When everything is existential, nothing is. People tune out, retreat into their own lives, and stop engaging with the political world altogether. Which, conveniently, leaves the field to exactly the people most committed to keeping the crisis narrative alive.

Adam Smith has a passage in The Theory of Moral Sentiments I think about a lot. He describes what he calls “the man of system,” the political leader so enamored of his own ideal plan that he cannot tolerate any deviation from it. Smith writes that such a person imagines he can arrange the members of society “with as much ease as the hand arranges the different pieces upon a chess-board,” failing to reckon with the fact that in the great chessboard of human society, every piece has a principle of motion of its own. Force the pieces where they don’t want to go, and the game “will go on miserably, and the society must be at all times in the highest degree of disorder.”

Much human suffering has been caused by people who think that pushing other people around like chess pieces and (often) using existential threat narratives to do so is a legitimate way to win or keep power.

That’s a reason for serious caution about radical solutions, even when they’re emotionally compelling. It’s an argument for building and maintaining the institutions that metabolize conflict before it becomes catastrophic, rather than waiting to find out whether the catastrophe was worth it.

So What Do We Do?

We watch for brittleness. In universities. In media. In political institutions. Brittleness is detectable before it becomes crisis, if you’re paying attention and honest about what you see. It often takes the form of false binaries, zero sum solutions, and tribalistic impulses. When we see that developing, we need to take a Petrovian pause.

We also build the four-dimensional networks, the diplomatic relationships, the community trust, the restorative mechanisms, that give people off-ramps before they need them and paths back after they’ve gone too far. This is a way of building out the landscape so that people have truly pluralistic ways of navigating it and where there are fewer direct conflicts over resources, power or whatever else. This is not just about avoiding zero sum binaries, but building the actual non-zero sum solutions.

And finally we resist the habit of reaching for the existential frame as a reason to stop thinking carefully. Because the moments when someone tells you there’s no time for complexity are usually exactly the moments when complexity is most necessary. Don’t let anyone drive you off the cliff. Chances are very good there’s an off ramp if you look for it.

There are very few situations where the options are genuinely reduced to a single binary. What’s much more common is that we’ve stopped being able to see the other dimensions. The work of radical moderation, a lot of the time, is just the work of insisting on looking.

Your Turn

What do you think? Are there cases where you’ve seen the existential framing shut down exactly the kind of thinking we needed most? Or cases where it were actually warranted? How can we tell the difference? As always, I’d love to hear from you in the comments!

If you liked this post, you might like this one from earlier this year: Is Trump a Fascist?

Leave a comment

Subscribe now

Share

This is another entry in the False Binary Alert series, but it also updates my radical moderation framework in important ways. If you like this post, I link to some additional pieces you might like at the end. Subscribe to stay tuned for additional posts on 4D thinking, deep pluralism, and radical moderation.

As with most of our most pressing policy challenges, the biggest disagreement about immigration isn't between left and right. It's between people who think of policy making as a two-sided fight to be won and people who recognize it as a complex landscape we have to map and navigate together. That gap is what 4D thinking, deep pluralism, and radical moderation help close. None of them works without the other two.

Here’s the TL;DR version for how they fit together:

4D thinking is how we see.

Deep pluralism is what’s actually there once we look.

Radical moderation is how we act given what we see.

The first is epistemic. The second is descriptive. The third is political and moral.

Take any one away and the other two collapse. Without 4D thinking, deep pluralism looks like a problem to be solved: the other side just needs to disappear. Without deep pluralism as a descriptive fact, radical moderation collapses into split-the-difference centrism. Without radical moderation, 4D thinking becomes virtue signaling: a way to sound sophisticated without doing the hard work of implementation.

I’m going to focus on immigration as a case study to show you how the three fit together. Immigration works well in this space: it’s very easy to polarize around because it plays into in-group/out-group dynamics at a basic level, it’s a permanent issue (humans have been moving around and arguing about it for as long as we’ve been human), and it’s a space where the underlying tensions don’t resolve. They just get managed. I’ll use immigration to walk us through the framework, but you can insert any other policy debate into it, which is what makes it powerful.

Let’s start with what I mean by each dimension.

The First Dimension: You

Length is where you’re standing. It includes your values, goals, talents, immediate situation. Every decision has to start here, because you can’t engage the landscape without knowing where you are on it.

On immigration, this looks different depending on who you are. A farmer in California whose crops rot without migrant labor is in one place. A construction worker in Ohio whose wages have stagnated for twenty years is in another. A teacher in a district absorbing a wave of new ELL students is in a third. A child of immigrants whose grandmother came through Ellis Island is in a fourth.

These are first-dimension facts. They’re not opinions. They’re where you’re starting.

And this is where 1D thinking fails immediately. The political binary tells the farmer and the construction worker they’re on opposite sides: that their interests are fundamentally opposed, that one has to lose for the other to win. The 4D view tells you both are standing on real ground, both have legitimate concerns, and any workable policy has to take both of them seriously.

Your first-dimension facts are yours. They tell you where you’re starting. They’re not arbitrary, but they don’t make you right about the whole landscape.

The Second Dimension: Everyone Else

Breadth is social networks, connections, who’s in the room and who isn’t. This is where we start mapping the actual terrain.

On immigration, the second dimension is massive. It includes the people who employ immigrants and the people who compete with them for jobs. The communities that receive them and the communities they left. Churches running resettlement programs. Hospitals treating them. Schools enrolling their kids. ICE officers arresting them. Families split across borders by our current policies, and families who feel their neighborhoods changing in ways they didn’t choose.

It also includes people who aren’t in the room. DREAMers who can’t vote. Asylum seekers stuck in Mexico. The ten-year-old citizen whose parent just got deported. The farmworker who died in the heat last summer because nobody would report the unsafe conditions.

Second-dimension thinking forces us to notice who the system is structured to exclude from the conversation about how the system should work. It’s also about how network effects play out in real life, changing the shape of the landscape we’re all navigating.

Most political fights on immigration happen without ever surveying the second dimension. We argue about whether we’re “for” or “against” immigration as though the question were one-dimensional. But nobody is for or against immigration in the abstract. People have views about specific populations, specific policies, specific communities, and specific tradeoffs. The binary flattens all of it.

The Third Dimension: Pits and Peaks

Height and depth are about tradeoffs, risks and benefits: what you gain, what you lose, what it costs, and who pays.

This is where the work gets hard, because the third dimension is where honest people disagree most. Even people who agree on a big picture slogan like “open borders” might disagree vehemently about who shoulders specific kinds of costs or how costs and benefits should be distributed. I’ve written before that our current immigration policy is uniquely terrible: a cobbled-together mess that serves almost no one well and hurts the most vulnerable. Even granting that, the question of what to replace it with lives at dimension three.

Start with the peaks of the landscape, because every honest accounting includes them. Immigration drives economic growth: labor force expansion, entrepreneurship rates, sectoral capacity in agriculture, construction, healthcare, and tech. Family reunification is a real moral good in its own right; people get to live with the people they love, building live together and supporting children with stable connections. Demographic vitality matters more every year as fertility declines and the working-age share of the population shrinks; immigration is one of the few levers we have. These aren’t ideological claims. They’re the gains side of the ledger.

The pits are most often where the disagreement actually lives.

Some pits on the restrictive side: people dying in the desert. Families separated. An entire class of undocumented workers living in fear and unable to plan their own lives. Labor markets distorted by the existence of a shadow workforce. Rule of law eroded by selective enforcement. Huge swaths of the economy quietly dependent on people the law pretends don’t exist.

Some pits on the permissive side: wage pressure on workers who are already economically vulnerable. Real strain on local services when the pace of arrivals outstrips what communities can absorb. Legitimate concerns about cultural cohesion when newcomers arrive faster than integration can happen. Political backlash when people feel policy is being imposed on them rather than negotiated with them.

Both sets of pits are real. Pretending either set doesn’t exist is 1D thinking dressed up as moral seriousness.

Critically, noticing that both sides have pits does not mean the pits are equivalent. Some are deeper. Some trap more people. Some are easier to climb out of than others. People dying in the desert is not the same as wage pressure on native-born workers. A teacher struggling with new ELL students is not facing the same stakes as a teenager whose parent just got deported. The asymmetry matters. I’ve argued elsewhere that the restrictive pits are almost always deeper (Trump Voters Weren’t Wrong, I Was Wrong About Trump).

But pits being unequal doesn’t mean the smaller ones don’t exist. And it definitely doesn’t mean a policy that ignores them will succeed. This is where a lot of well-intentioned people get tripped up: they correctly identify the worst injustice, then design policy as though the other concerns don’t matter. That isn’t moral seriousness. That’s 1D thinking with better vocabulary. It also leads to what I call brittle policy (more on this later).

Real injustices deserve priority. Sustainable policy has to account for the full landscape anyway, including the parts we’d rather ignore.

The Fourth Dimension: Time

Time is my favorite dimension (as a Burkean) but also the most complex: it’s evolutionary, developmental, generational, historical, and future-oriented. It’s the dimension we ignore hardest, because it asks us to think about things we can’t see.

Evolutionary time on immigration, for example, is about in-group/out-group bias, a real feature of human cognition. We’re wired to share resources more readily with people we perceive as part of our tribe, and immigration policy has been weaponizing that wiring for as long as we’ve had nations. You can’t reason someone out of a hardwired cognitive bias by calling them xenophobic. You have to design policy that takes the bias seriously while refusing to be ruled by it.

Developmental time is about how individual immigrants and their children integrate over years and decades. The second generation usually outperforms the first. The third usually outperforms the second. Policy that evaluates immigration on a one-year timeline misses what actually happens.

Generational time is about what we owe people we’ll never meet. The grandchildren of the people we admit, or don’t, will be American citizens or not in part because of what we decide now.

Historical time is about how we got here. American immigration policy has been a kludgy mess assembled mostly out of xenophobic panic since the 1880s. Every “immigration crisis” in American history has eventually resolved into a generation of Americans whose grandparents were the problem. This pattern is not subtle.

Future-oriented time is about what we’re building toward. Demographic trends, labor shortages, fertility declines, and climate migration should be part of any equation. But the world arriving in twenty years isn’t the world our current immigration system was designed for. Democratic institutions aren’t well-suited to prepare policy for it either.

Skip the fourth dimension and you end up with policy designed for this week’s news cycle (which is, not coincidentally, exactly the policy we have).

Deep Pluralism as Feature, Not Bug

Once you’ve actually worked through the four dimensions on an issue like immigration, something uncomfortable becomes obvious. The people who disagree with you aren’t stupid, evil, or uninformed. Their first dimension is different from yours. Their second dimension includes different people. They’re weighing different pits on different timescales.

They’re standing at different points on the landscape.

That isn’t a flaw in them. It’s a feature of human life. People with different neurology, different experiences, different social networks, and different histories will see the same landscape differently. They always have. They always will.

That’s deep pluralism. It’s the reality 4D thinking reveals.

This is why the fantasy that the other side will eventually disappear is so destructive. It assumes deep pluralism is temporary: that if we educate enough people or wait out enough elections the landscape will collapse back into a single dimension where everyone agrees with us. But 4D thinking shows us the pluralism isn’t going anywhere, because it’s generated by the dimensions themselves. People will always occupy different points on the landscape. That’s just what it means to be a landscape.

Which Is Why We Need Radical Moderation

If deep pluralism is permanent, we have two options.

Option one: try to win. Build institutions designed to defeat the other side. Breach guardrails, animate tribal loyalty, treat every election as existential, and wait for the other side to disappear. This is what both sides are doing. It isn’t working.

It also produces brittle institutions and brittle citizens. Brittle institutions cave with every political shift. Brittle policies get overturned by the next president’s pen. They rely on zero-sum outcomes and fail to address root causes. They fail epistemically and morally: by ignoring the 4D world we actually live in, they make durable reform impossible.

Meanwhile, brittle citizens are fragmented and polarized, their civic and personal lives splintered across ideological lines. Or they’re just so exhausted that they give up on politics altogether.

There’s a better way though, which leads us to…

Option two: build institutions strong enough to hold people who are genuinely, permanently different from each other. That’s radical moderation. It isn’t the middle of the 1D line. It’s the refusal of the 1D line. It’s the commitment to navigating the 4D landscape rather than pretending we’re on a tightrope.

Approaching policy in this way helps create both resilient institutions and citizens, instead of brittle ones. Resilient institutions bend without breaking. They’re resilient precisely because 4D thinking and accounting for deep pluralism helps generate outcomes are more likely to be nonzero-sum and more likely to win broad-based support. They not only tolerate deep pluralism, but their very structure is a kind of alchemy, taking deep pluralism and creating new opportunities for engagement and flourishing. Policies grounded in deep pluralism are less likely to fail. And resilient citizens connect across partisan lines, build strong communities that help people navigate disruption, and have a strong sense of both place and self.

Going back to immigration, radical moderation doesn’t give you a single policy answer. It gives you a set of institutional commitments that are each resilient on their own, but that compound when working together:

A legal pathway with enough capacity and clarity to actually channel people through it. Right now the waitlist is off the charts: Indian nationals on the employment-based green card track face a backlog of over a century. That’s not a system. That’s a paper fiction nobody can meet.

Enforcement that’s predictable rather than arbitrary. That means legislatures legislating. We haven’t passed serious immigration reform since 1986. Four decades of presidents ping-ponging executive orders is exactly what brittle institutions produce.

Absorption infrastructure that takes receiving communities seriously rather than dismissing their concerns as bigotry: school capacity, housing, workforce boards, local civil society.

A status-regularization process for the undocumented population already here, because they’re already members of our communities in every way that matters except the one Congress refuses to acknowledge.

An acceptance that tradeoffs exist on immigration and on every policy issue we can think of.

This is harder than “open the borders” or “close the borders.” It’s also the only approach with any chance of producing policy that lasts longer than the next election cycle. We need institutions, not slogans. I’ve made the same case on a different front in Culture Wars Colonized the Family, Higher Ed Doesn’t (Just) Have a Trust Problem and in a number of other posts over the years.

What Radical Moderation Isn’t

It’s hopefully obvious by now, but worth repeating: radical moderation is not centrism.

Centrism tends to assume there’s a meaningful middle point between two coherent positions and tries to find it. Most simplistically, you just split the difference, take a little from column A and a little from column B, call it compromise. But most of our political fights aren’t between two coherent positions. They’re between two oversimplified slogans that flatten a 4D landscape into a 1D line. There’s no meaningful middle between “open borders” and “close the borders” because neither describes a real policy. Radical moderation isn’t about finding the midpoint on the line. It’s about getting off the line entirely.

The Loop

Radical moderation, once you commit to it, forces you back into 4D thinking. And you can’t navigate deep pluralism without the map. The moment you stop mapping the landscape, you slide back into some version of “my side is right and the other side is an obstacle,” which is where we started.

So the three concepts feed each other. 4D thinking uncovers deep pluralism. Deep pluralism demands radical moderation. Radical moderation requires 4D thinking to identify which institutions really produce robust and resilience change that maps onto the real world.

When the three work together you get accurate maps, effective policy, and flourishing citizens. Remove one and the whole thing falls over. Remove all three and you’re not on a stool at all, you’re back on the 1D line, where our politics has been stuck for decades: flat, polarized, brittle, and unable to hold the weight of any real problem.

But we can do better than that.

The landscape is already there.

We just have to be willing to look at it together.

Your Turn

As always, I’d love to hear what you think about this updated framework in the comments. I read every one and they often shape where my writing goes next. Let me know if there’s an issue wehre you’d like to see this framework applied or if I’ve missed something!

The next piece in this series will most likely take on AI in higher ed and how we’re using (or just panicking about) AI in classrooms, syllabi, and scholarship. Predictably, these conversations are a masterclass in 1D thinking — and the institutional questions it raises are exactly the ones radical moderation is built for.

Subscribe if you want that one in your inbox rather than having to come find it.

Subscribe now

If this piece helped you see a familiar fight differently, I would be grateful if you’d restack it. The people who most need the three-legged stool are almost never on your side of the feed. That’s the whole point.

Leave a comment

Share

Related posts you might like:

• Clowns to the Left of Me, Jokers on the Right: On the fantasy that the other side will eventually disappear.

• Trump Voters Weren’t Wrong: On why ignoring pits leads to brittle policy and brittle citizens.

• Are Sanctuary Cities Unconstitutional: On institutions and immigration.

• A Convert Mulls Christian Values: On anti-immigrant Christianity.

• Governor Abbott is a Jerk: On forgetting that humans matter.

Not counting spam or my personal email account, I received around 600 work emails last week. I sent 173.

Those numbers are about average.

That’s over 100 emails a day to read, sort, and assess and 30 emails a day to draft, revise, and send. Most of them involve clarifying policy, explaining curriculum flow, or answering questions that have already been answered eighty-five times in various memos or on our website. The rest are travel details, research queries, reading group scheduling info and other research admin.

The number itself is mundane but what it represents is not. Every one of those emails occupies a small piece of working memory — is this urgent, does it require a decision, does someone need a follow-up — and working memory is not renewable on demand. It depletes. By early afternoon on a typical day I have spent a significant portion of my finite cognitive resources on work that, if we’re honest, does not require me - or really any particular human - to do it.

We can call this the life admin tax. You pay it on emails and meeting prep and scheduling and the seventeen-step process required to book travel reimbursement. You pay it on the mental overhead of tracking what’s in flight across four or five different domains of your life simultaneously. The tax is not random. It falls hardest on people who are doing the most: associate deans (yes, we do actually do work), working parents, researchers juggling teaching and writing and service, and anyone whose life has more plates spinning than there are hands to spin them.

Unfortunately, the loud takes on AI right now from both sides miss what this tax actually is.

A Radically Moderate Position on AI

The critics say AI is eroding our humanity: outsourcing our thinking, hollowing out our writing, replacing what makes us most ourselves.

The boosters say AI is replacing us: that the work humans do is essentially the same work AI does, just slower and more expensively.

Both positions share a hidden assumption, namely that what AI can do and what humans should do are the same activities.

They are not.

Those 600 emails are not human work. They are the scaffolding around human work: the substitute teaching coordination, the room reservation cascade, the policy clarification chain. They are what displaces the human work. They are the tax. And the radical-moderate position on AI, the one I’ve been working out in practice for the better part of a year, is that thoughtful AI use is precisely how you stop paying the tax so the human work has somewhere to go.

My 4D framework connects here as elsewhere. We see breadth easily — the surface area of our obligations, the daily list, the inbox. We don’t see what breadth costs in depth (the quality of the thinking we should be doing), height (what we could be contributing if we weren’t triaging), or time (the cumulative effect over the years of a career). When you treat life admin as an annoying feature of your job, you accept the tax. When you treat it as a structural problem worth engineering a solution to, the math changes.

Here is what engineering a solution has actually looked like, in case the abstraction is unhelpful.

My Comprehensive Claude Assistant
for Exhausted (but Happy) People

One caveat here that I want to get out of the way: I love my job. It’s intellectually fulfilling and I make real differences in people’s lives. I recently won a mentoring award because I genuinely love connecting to other human beings, supporting their goals, removing harmful friction, and creating intellectual community. But I also have firm boundaries after years of mentoring new faculty in how to survive in the academy: I rarely work evenings or weekends and when I do it’s usually pre-planned because I took time off for other stuff during the week. In my spare time I read garbage novels, workout, hang with my family, travel, organize kid clothing (a Sisyphisean and inhuman task) and garden. I’m not a workaholic by any stretch of the imagination.

What follows is not a system that urges people into some kind of post-industrial productivity hellscape where every waking moment must have profitable outputs. Instead, this is a system that helps me remain more human: more connected as a mother, author, colleague, nature-lover, and a mentor.

Your mileage may vary. The goal here though is to outsource the life admin tasks. What you do with that free time is up to you.

The Shape of the System

I have two current systems within Claude, both organized differently. For productivity and task management, I use Claude Cowork. For writing, I use Claude Chat and organize it around specific Projects. I’ll talk about both separately, but the two overlap quite a lot and I’ve been using Cowork more for writing lately too.

For productivity and life admin, I keep a single project folder on my computer that Claude can read and write to. Inside it are a monthly plan, a weekly plan, a running tasks file, and a handful of project-specific documents — talk outlines, article drafts, a parking lot of half-formed ideas I don’t want to lose. Claude can open any of these, pull from any of these, and update any of these as we work.

This is the equivalent of a personal assistant with a fantastic memory, all my work context at its fingertips and who knows exactly how to pressure me to finish projects (because that’s included in its context memory).

You can see a screenshot of my scheduled tasks below. The main screen are all my scheduled tasks (there are a couple I’ll probably sunset, but for now I’m using all of them). On the lower left hand pane you can see some recent queries and chats, including one where Claude is waiting for my response.

The “input dump nudges” are periodic reminders to give Claude the context it needs to do the briefings accurately. These are curated lists of screenshots, calendar inputs, kids’ practice schedules, travel and flight info, and random other things I add weekly or quarterly so Claude knows what’s up.

The life admin assistant in particular sounds simple and it is simple. The trick for that one is that everything lives in plain text, in one place, where Claude can see it. The folder is the memory. The assistant is the interface. The calendar and email integrations are the eyes and ears. Keep those three roles straight in your head and the whole thing becomes much easier to build.

Voice memos into drafts

For writing and admin analysis, I have a number of chat Projects broken out by domain/topic: radical moderation, structural hobbling, associate dean work, and so on. I upload entire manuscripts, policy docs, and other critical context to these projects so Claude’s responses are both accurate and targeted. This limits hallucinations (as does clear prompting).

My primary writing pipeline is voice memo to transcript to draft. The commute, the dog walk, or the ten minutes between meetings: I dictate. I had an inspiration the other day while doing laundry and shot off a quick voice memo that otherwise would have been lost to the sands of my mind.

Later, I paste the transcript into Claude with context — this is a Substack post, this is a memo for the college, these are talking points for a radio segment — and Claude organizes it into a working draft in my voice. Standing style rules and writing samples (including both my book manuscripts) live in the project folder, so they don’t have to be re-stated each time. Lately, Cowork has done a better job drafting than Chat; I need to figure out why, but for now I use both interchangeably and sometimes find they critique each other’s work in interesting ways, perhaps in part because they have different context documents.

What this does, cognitively, is convert drafting into revising. Two very different tasks. The blank page is the part of writing I find most expensive. With voice memos as raw material, the thinking is already there. I’m shaping it, not generating it from scratch.

Admin Tasks

I’ve also been outsourcing administrative busy work to Claude in a variety of ways. Just this week, Claude has set up Excel sheets for me (including formulas, which I never learned to do myself), updated PowerPoints for three upcoming talks simultaneously, provided an analysis of curricular overlap and potential for new programs, helped me update my CV, and formatted a book chapter for an edited volume while cross-checking references.

Some newer things I’ve been playing with the last week:

• Claude Design: works in Canva and other platforms and designs in the app itself; just launched (available to all paid subscribers). This is a little glitchy so far, but pretty cool and will likely get better very quickly. It’ll eat up your usage though, so save it for the weekends if you don’t need it for actual work.

• Claude in PowerPoint: Can create and organize content and text from an article or draft, and develop templates from images and verbal inspiration. It allows me to focus on the content and my message instead of screwing around trying to choose colors or get shapes to align. As in Excel, this plug-in works directly within the presentation, so you’re not uploading documents all over the place.

• Claude in Excel: Can create and update spreadsheets, fix formulas, and brainstorm ways to visualize data across a range of tasks. As in PowerPoint, this plug-in works directly within the presentation, so you’re not uploading documents all over the place.

• Claude in Chrome: can read websites, pull data, and can make changes to websites with sign-in permission (I plan on trying this in SquareSpace to update my website, but have not done it yet). It’s been great for organizing my Substack drafts and recommending which can be deleted. I also plan on using it to scrape a publishing website for info on agents accepting new queries.

• Scheduled tasks in Claude Co-work: I dump calendar, deadline and other data to create travel itineraries, weekly planning sessions, and team-based reports. Claude syncs with Monday.com (and other task management systems) and can pull data from there for weekly/daily briefings. This is what’s laid out in more detail below.

Daily Briefings for Cognitive Offloading

My absolute favorite thing though is to have Claude work as a personal assistant and accountability buddy. Every morning Claude generates a short daily briefing for me that pulls from my calendar, my task files, and Claude’s memory of what’s in flight. It tells me what’s anchoring the day, what’s urgent on the admin side, what’s on the radar but not today, and offers one gentle nudge about whatever small thing is most likely to stall if I don’t handle it before noon.

Here’s a clip of my daily briefing from yesterday:

It goes into much more detail (at my request), including my top priorities and suggestions on how to balance things and top priorities for the day. You get to choose the outputs here. You can choose a super basic daily briefing or a much more detailed one like I do. The personal assistant is totally customizable, including tone and how judgy you want it to be (I want mine mildly judgy). As you can see below, I swapped in an AI post for an immigration post (that’ll drop next week) after Claude and I chatted a bit about Substack strategy, so the daily briefing is editable as needed.

The briefings are not elaborate and they take about two minutes for Claude to generate from the files and calendar. The value is not in the format. It’s in the anticipation and the context. A good assistant is not one that answers well. It is one that anticipates well. Claude does that well if I give it enough context (and it reminds me to do that), which I’m happy to do (within obvious privacy parameters, which I might discuss in a later post).

Weekly Retros and Pre-Travel Briefings

On Fridays or Sundays Claude creates a short structured check-in: What shipped, what slipped, what next week looks like against the calendar. Before a trip Claude develops a pre-travel packet with what I’m speaking on, who I’m meeting, what’s queued at home, what needs to happen before I leave (and another list for what I need to do when I get back).

I have it update my workout schedule to ensure I have time set aside and it assigns me one household/personal maintenance task each week from an annual master list I’ve uploaded. These include things like change the oil or find a lawyer for estate planning: things that never get done because there’s no firm deadline and no external accountability to do them.

I also have it schedule quarterly check-ins and data dumps so I can update calendars, projects, activities, inboxes, and all the other places where deadlines and goals go to die without something keeping them alive. This is an extension of the quarterly strategic planning I’ve done for many years now, but now instead of having to remember to find and pull out my strategic plan for the quarter to keep it fresh, it lives in Claude and my Claude assistant prompts me to dump info, reminds me to break down projects into tasks, and find the time to fit things in (and helps me figure out what to jettison when it doesn’t all fit, which it inevitably doesn’t).

I resisted setting these up for months, but they take Claude two minutes to generate. These check-ins and briefings are now the backbone of how the week holds together.

It’s also super easy to update: when my kids get late updates to softball practice schedules, I can dump those into Claude, ask it to re-run my weekly plan and/or daily briefing, and it’ll re-run and let me know what, if anything, needs to move or be adjusted as a result. It’s kind of like magic.

I Am (and Don’t Want to Be) a ChatBot

There’s a structural insight buried inside all of this that I want to name, because it’s the part that changed how I think about the whole project. I carry an enormous amount of institutional knowledge in my head — about programs, policies, people, processes — and spend hours a week retrieving it on demand for colleagues who need it.

Similarly, I carry an enormous burden of personal and family knowledge and what David Allen calls “open loops” in my head: doctor’s appointments that need to be scheduled, softball practices that need to get on the calendar, and car maintenance that needs to be booked.

But the retrieval is not the value I add. The thinking is the value I add. As Allen also noted, the brain is a thinking machine, not a storage machine.

Before Claude, I was the chatbot of my work and personal life and it was exhausting.

But Claude can now do the retrieval; I should be doing the thinking.

This generalizes well beyond academic administration. If you are the person who knows where everything is, who remembers the context, who tracks what’s in flight: a meaningful share of your cognitive load is retrieval, not judgment. Returning that share to a tool is not a loss of humanity. It is a recovery of the parts of your work that actually require you: a distinct person with distinct talents who needs more time to share them with the world.

What the Time is For

This is where my radically moderate position gets specific.

The hours the system returns do not return to “more productivity,” at least not in the way the productivity industry usually means the word. They return to the longer thinking required to actually write a book, rather than perpetually drafting a chapter and losing the thread. They return to direct presence with my kids on the days I am home, instead of half-presence with my phone in my hand because I haven’t processed the morning’s email. They return to real teaching in the classroom rather than the administrative scaffolding that surrounds it. They return to conversations with colleagues that go past logistics to the kinds of decisions that actually deserve my judgment and creative thinking.

This is what thoughtful AI use returns when it’s working. Not less humanity, but more room for the activities humans should be doing in the first place.

The critics are right that there is a version of AI use that hollows people out. There is. It’s the version where you outsource judgment, presence, and craft, and keep the inbox. The position I’m working out is the inverse: outsource the inbox, keep the judgment, presence, and craft. The math only works in that direction.

Your Turn

I’m going to keep writing about this as the system evolves. Later this week I’ll post the tactical side of this post: specific prompts and set-up for various use-cases, mostly aimed at writers, researchers, and academics, but maybe useful for others in admin (home and work) roles too.

Next week I’ll be getting back to the political stuff, with the 4D framework applied to immigration ahead of a talk I’m giving in DC.

For now, let me know if you have questions about my setup, ideas of your own for how you use AI to make yourself more human, or anything else! Drop comments in the chat and I’ll respond to all of them. And if you found this post helpful, please consider restacking. Readers sharing my ideas are how new readers find me.

Leave a comment

Subscribe now

Share

You can (and should!) read my other posts on higher ed here:
Part I: The Higher Ed Death Spiral
Part II: What About the Wage Premium?
Part III: Halting the Higher Ed Death Spiral
Disrupting Higher Ed Before the Future Does it For Us
Universities as Radically Moderate Cartographers

Yale's Committee on Trust in Higher Education spent a year consulting faculty, students, alumni, politicians, and critics to figure out why the public doesn't trust higher education anymore. The committee was composed entirely of tenured Yale faculty. (I'll give you a moment with that.)

Before I push back too hard, Yale’s Committee on Trust in Higher Education, which released its findings earlier this month, produced a serious and thoughtful document. The committee spent a year consulting faculty, students, alumni, administrators, journalists, politicians, and critics. They hosted listening sessions and a speaker series. They read the scholarship. And they were unsparing about some things that institutional committees almost never say out loud. It’s worth reading in full, even if it’s incomplete.

But I’ve been writing about higher education’s structural failures for the better part of three years now, and the Yale report lands squarely in the awkward space between “finally” and “is that all?”

What the Report Gets Right

The committee identifies three main drivers of declining public trust: cost, admissions opacity, and campus speech culture. These are real problems with real data behind them. But they’re kind of like individual mushrooms atop a vast but hidden fungal body underground. The issues go much deeper and these three aren’t actually the causes of the trust decline in any real sense. Proximate causes, maybe. But not ultimate.

On cost: over the past three decades, the average sticker price of undergraduate tuition, adjusted for inflation, has more than doubled. Nearly a quarter of all Americans with federal student loans are currently in default, the highest rate since the government began tracking. The committee calls out the “high tuition-high aid” model as a trust disaster even when it functionally works, because what families see first is the sticker price and the process of learning what you’ll actually pay is, as the report puts it, “laborious, frustrating, and unpredictable.” That’s accurate.

On admissions: the committee is surprisingly direct about the fact that preferences for legacies, recruited athletes, and donors’ children distort a process supposed to be built on academic merit. A widely cited paper by Chetty, Deming, and Friedman finds that, after controlling for SAT scores, applicants from families in the top 1 percent are substantially more likely to be admitted to selective private colleges than comparable middle-income applicants. The report recommends Yale “reduce preferences for special classes of applicants” and establish a visible minimum academic threshold. That’s a meaningful recommendation.

On speech and self-censorship: nearly a third of Yale undergraduates in a 2025 survey disagreed that they felt free to express their political beliefs on campus, up from 17 percent in 2015. More than half of college students nationally report feeling intimidated sharing opinions in class. The committee also notes that registered Democrats outnumber Republicans in Yale’s Faculty of Arts and Sciences, Law School, and School of Management by 36 to 1. They call this a genuine problem for the quality of scholarship. While that should have been obvious decades ago, I guess better late than never?

The sections on grade inflation and governance are also pretty solid. In 1963, 10 percent of Yale grades were A or A-. In 2022-23, 79 percent were. The median Yale student now receives an A. Grades have ceased to function as information or signals. The committee rightly calls this a collective action problem: no individual wants to be the strict grader whose students suffer relative to peers. The institution can’t solve it by willing harder. It requires structural coordination. The report recommends establishing a college-wide grading mean and computing course percentiles on transcripts. These are sensible ideas that will face enormous faculty resistance and may actually happen anyway at Yale, given who wrote the report. (That last parenthetical matters, as I’ll get to.)

What the Report Misses

Despite the “brutality” of the report (per the NYT), the report still misses a lot. The report frames the trust collapse primarily as a product of the last decade: political polarization, admissions controversy, free speech failures. Fix those, and trust can be rebuilt.

I’ve argued here that this is too optimistic, because the structural problems in higher education predate the current political moment by decades. Some of them are the kinds of perverse incentive structures that survive any change in political climate. I made a similar point in response to two university chancellors who wrote an op-ed last year calling for depoliticization as though the pre-polarization status quo was unquestionably solid. It wasn’t then, and that argument hasn’t gotten stronger.

The student loan system is the clearest example. The report acknowledges that higher education costs too much. It recommends Yale raise its income threshold for free tuition and improve financial aid transparency. These are good things for Yale students. But the report doesn’t name what I think is the central driver of tuition inflation across the entire sector: the availability of cheap, easily accessible federal credit to 18-year-olds, which has allowed universities to spend in ways only loosely connected to educational outcomes. When the tap is nearly unlimited and the borrowers are young people who don’t yet understand what $80,000 in debt will cost them in their 30s, universities face every incentive to build the climbing wall and hire the consultants and launch the new initiative. I wrote about this when student loan debt crossed $2 trillion. The report mentions the default crisis in passing. It doesn’t trace the causal chain.

The wage premium story is similarly more complicated than the report acknowledges. I dug into this at length last fall: when you adjust for debt, time to completion, and differential outcomes across income and racial groups, the financial case for a four-year degree is much weaker than the aggregate numbers suggest. A quarter of college graduates see no wage premium at all. For Black students, the wealth premium disappears entirely compared to non-attending peers once you factor in debt. This matters because the “college for everyone” narrative shapes enrollment, borrowing, and the pipeline of students who arrive at institutions that aren’t equipped for them and can’t serve them well.

I’ve also written about the replication crisis, the incentive structures that reward publication volume over quality, and the institutional resistance to change that defines academic culture across the country. The report acknowledges all of this. But it treats these as context for the trust problem rather than as causes of it. The deeper problem is that these aren’t just trust failures. They’re mission failures. Universities don’t know what they’re for, and no amount of admissions transparency or grade deflation fixes that.

The 4D Problem With the Report’s Solutions

My analytical framework for thinking about complex problems asks us to work in four dimensions. Length captures the individual level: idiosyncratic individual preferences, goals, and talents. Breadth captures social networks, connections, and ripple effects across communities and institutions. Height and depth together capture the full landscape of tradeoffs, risks, benefits, pits, and peaks that any policy or decision creates. Time captures the full sweep of causes and consequences: genetic and developmental, generational, historical, and likely futures. (An updated explainer on this framework is coming soon. Subscribe if you want the full version.)

Subscribe now

The Yale report does real (though still incomplete) work in the height and depth dimension. It names the pits: students drowning in debt, admissions systems rigged toward the already advantaged, classrooms where grades no longer mean anything. It describes real tradeoffs in governance, in the cost structure, in the tension between academic freedom and institutional accountability. That’s not nothing.

But the other three dimensions are largely absent.

On length: the report’s recommendations are almost entirely universal. They apply to “Yale students” and, by implication, to students everywhere. What they don’t do is grapple seriously with the variation in individual starting points, goals, and aptitudes that makes a single model of higher education so poorly suited to the actual population it serves. The student who should be at Yale and the student who would be better served by a two-year technical credential occupy completely different positions in this dimension. The “college for everyone” narrative collapses that difference, and the report, despite its many virtues, mostly leaves it collapsed.

On breadth: the decentralized nature of American higher education means Yale’s reforms don’t travel. The social networks that make a Yale degree valuable are themselves unequally distributed, and the report’s solutions largely take that inequality as a given rather than as something to address structurally. A first-generation student at a regional public university is embedded in a completely different network of connections and institutional supports than a legacy admit at an elite private school. The breadth dimension of this problem isn’t solved by Yale improving its admissions transparency.

On time: this is where the report’s gaps are most consequential. The time dimension runs in multiple directions at once. Developmentally, we are asking 18-year-olds to make $80,000 decisions about their futures based on labor market data that may be obsolete before they graduate. Generationally, the student loan crisis is already a multi-generational problem, and the reforms the committee proposes do nothing to address the debt already held by the 43 million Americans who borrowed for degrees they may or may not have finished. Historically, the structural failures I’ve described here predate the political moment the report was written to address by at least two decades. And the future goes almost entirely unaddressed: an 18-year-old enrolling this fall is looking at a job market that AI is actively reshaping, in fields where the wage premium has already plateaued. The professional fields commanding the highest college wage premiums - business, communications, parts of engineering - are exactly the fields where AI displacement will likely be most significant over the next decade. The committee’s recommendations are mostly structural reforms on a 5-10 year timeline. The students enrolling this fall will graduate into a workforce that doesn’t yet exist.

Top-Down and Bottom-Up

The report is necessarily a top-down document. Yale’s president commissioned it; Yale’s tenured faculty wrote it; Yale’s governance structures will implement whatever parts actually get implemented. And some of what it recommends - mission clarity, grading reform, financial aid transparency - can be done that way. Yale has the endowment ($44 billion), the institutional prestige, and the concentrated faculty governance to actually move on some of these things. That’s not nothing.

But American higher education is not a top-down landscape. It’s roughly 4,000 institutions, decentralized by design, with different missions, different student populations, different financial models, and different relationships to state and federal funding. What works at Yale is mostly not transferable to a regional public university facing declining enrollment and tuition-dependence or a large private STEM school like mine. The committee acknowledges this honestly. It is also a significant limitation.

In my own city of Rochester, there are six-ish colleges and universities within roughly a ten-mile radius. Any serious regional approach to higher ed reform would ask: what if those institutions coordinated? What if they stopped competing for a shrinking 18-year-old population and instead specialized, shared faculty lines, built joint programs that served regional workforce needs? We could address a lot of what the Yale report recommends on a regional scale, not because a national body mandated it, but because the math of survival would make it rational.

It’s not happening. And it’s not happening because the incentive structures of federal funding, accreditation, and institutional prestige all push against it. The report gestures at accreditation reform as “a long-term goal.” I’d argue it’s one of the most important structural levers we have, and it gets about two sentences.

Real bottom-up change looks like experimentation: institutions that lean hard into a distinctive mission, stop trying to be everything to everyone, and build models that others can learn from. Some will fail. The ones that don’t will demonstrate what’s possible. That process is slower and messier than a 20-recommendation report from Yale’s faculty. It’s also more likely to actually change anything. I laid out what that might look like concretely here.

How Trust Gets Lost and How It Comes Back

The report opens with a number: trust in higher education has declined faster over the past decade than trust in almost any other American institution. Gallup puts confidence in higher education at 36 percent in 2024, down from 57 percent a decade earlier. That’s not a great signal as we face an enrollment cliff, cratering international enrollments thanks to political dysfunction, and the AI disruption ahead.

Trust is lost through two related mechanisms. The first is promise-breaking. Universities promised accessibility but charged tuition requiring generational debt. They promised meritocracy but ran admissions systems that advantaged legacies, athletes, and donors. They promised truth-seeking but produced research ecosystems that rewarded volume over quality and disciplinary cultures that resisted viewpoint diversity. They promised preparation for meaningful lives but often graduated people who couldn’t write a clear paragraph and didn’t know what to do next.

The second mechanism is what I’ve called cartographic failure. I’ve argued at length that universities are supposed to be neutral cartographers of the human landscape, mapping reality as accurately as possible so that people can navigate it. When cartographers manipulate the map to suit their own ideological commitments, or simply stop caring whether the map matches the terrain, people get lost. They fall into pits. And they stop buying maps from that cartographer.

You wouldn’t buy a map from a cartographer who doesn’t believe in mountains.

Rebuilding that trust requires more than better communications or restored grading norms, though those things matter. It requires doing the actual work of being accurate and honest: in research, in teaching, and in self-presentation. And it requires the willingness to say, publicly, that the current model cannot continue in its current form.

The Yale report takes real steps toward that honesty. It is more willing to say hard things than most institutional documents. But it is still, in the end, written by people who benefit from the current system and who are proposing changes that will preserve as much of it as possible. That’s not a criticism so much as a structural observation. It is the condition under which all institutional self-reform occurs. It’s also why I’m skeptical that 20 recommendations from a Yale committee, however good, will be sufficient.

Where We Go From Here

At the federal level, Congress needs to seriously examine what the current student loan system incentivizes. Not eliminate it - pulling the tap immediately would be catastrophic for millions of students and hundreds of institutions - but restructure it so the incentives align with actual educational outcomes rather than enrollment numbers. This probably also requires rethinking credentialing requirements across the economy. Across industries, we mandate four-year degrees for jobs that don’t need them, which inflates the student population and the debt load simultaneously, and consigns students who don’t belong in a four-year program to a high-cost, high-risk gamble. This is slowly changing, but too slowly.

At the accreditor level: accreditors currently measure inputs - faculty credentials, facilities, student-to-faculty ratios - far more than outcomes. An accreditation system that tracked whether students were actually learning, completing degrees, finding meaningful work, and managing debt would create genuine bottom-up pressure for reform. The Yale report calls this a long-term goal. I’d move it up the list considerably. But it also requires accurate assessment, which at many institutions ends up being the tail wagging the dog. Data is highly decentralized, very difficult to access, and we don’t do randomized controlled trials of student outcomes for obvious reasons. Tracking success is much harder than it looks and accreditors often don’t even try to ensure the data is reasonably rigorous.

At the institutional level: mission clarity is real. Universities that can honestly answer “what are we here for, and how do we know if we’re doing it?” will navigate what’s coming better than those still trying to be everything to everyone. That probably means some consolidation, genuine specialization, and willingness to let programs and facilities go. The institutions that figure this out will survive. The ones that keep building new dorms to attract a shrinking applicant pool will not.

At the faculty level: the collective action problem on grades is a useful model for thinking about every other coordination failure in higher ed. No individual wants to be the strict grader. No department wants to cut the popular but educationally indefensible program. No Faculty Senate chair wants to look hard at tenure and how much full professors really contribute to the academy in research or teaching. These are genuine coordination problems. Good intentions don’t solve them. Incentive structures do, and occasionally so do governance mandates from people who have actually earned the institutional standing to impose them.

The trust problem in higher education is, at its root, a clarity problem. Universities are not trusted because people are uncertain what universities are for and whether they’re delivering on whatever that is. The answer is not a better communications strategy. It’s institutions actually knowing what they’re for, doing it well, and being honest when they fall short.

The Yale report is a useful document from an institution that has the resources and the standing to act on some of what it recommends. What it can’t do is solve the deeper structural problem, which is that most of American higher education doesn’t have Yale’s endowment, Yale’s name, Yale’s student body and alum base, or Yale’s capacity for internal reform.

The people falling into the pits our higher ed system has created - pits of loan debt, poor outcomes, and opportunity costs - aren’t mostly at Yale. They never were.

And until we confront the much broader structural and incentive problems that higher ed faces, we’ll continue failing the students who need us the most.

Your Turn

For all my academics out there: what did I miss? What needs to be on this list? For the parents and students out there: what do you look for in a college? Why? For employers: What should universities really be focusing on? What skills are we failing to teach? And how does K-12 ed fit into this set of problems? As always, I’d love to hear your thoughts in the comments. And if you like what you read, please subscribe and share!

And don’t forget these great reads:

Part I: The Higher Ed Death Spiral
Part II: What About the Wage Premium?
Part III: Halting the Higher Ed Death Spiral
Disrupting Higher Ed Before the Future Does it For Us
Universities as Radically Moderate Cartographers

Leave a comment

Subscribe now

Share