Welcome to Universe 26

The Calhoun Effect - YouTube

Society / Population / Behavioral Science

Calhoun's mice didn't starve. They withdrew. So are we.

By Stephen "Pseudo Publius"  |  May 2026

BLUF (Bottom Line Up Front): 

John Calhoun's 1968 "Universe 25" mouse experiment was never a literal forecast of human society. Its real finding was that abundance plus a closed enclosure produces collapse through social withdrawal — not through famine. The human studies of the 1970s tested only open systems, where dispersal was easy, and dismissed Calhoun on that basis. The closed-enclosure variable was never tested in humans.

Three live cases now run the comparison. California, exit open, leaks 350,000 residents a year and shows only muted Calhoun effects. Russia, having banned military-age male emigration in September 2022, is running Calhoun's social-violence check and fertility check simultaneously. Japan — closed in both directions, exit by culture and entry by politics — has fallen to a 1.15 fertility rate, sixteen straight years of population decline, roughly 9 million abandoned homes, an inverted age pyramid, and a 380,000-worker shortfall in elder care that the country is now trying to fill with humanoid robots. It is the closest real-world analog to Universe 25 that exists.

Two-thirds of humanity now lives below replacement fertility. Chinese youth embrace tang ping. Japanese hikikomori withdraw. Seventy-two percent of American teens use AI chatbots for companionship. Calhoun is not a prophet. But the boundary conditions of his experiment are moving back into alignment with our own.

Mouse Heaven, Mouse Hell

In July 1968, a quiet, religious ethologist named John B. Calhoun loaded eight mice — four males, four females — into a nine-foot square steel pen at the National Institute of Mental Health (NIMH) in Bethesda, Maryland. He called the pen Universe 25. It was his twenty-fifth attempt to build what he advertised as a rodent paradise: unlimited food and water, perfect temperature control, no predators, no disease, and 256 nesting boxes designed to comfortably house up to 3,840 mice.

For 315 days, the colony behaved like a textbook population on the rise, doubling every 55 days. Then something began to go wrong. Even though Universe 25 was built to hold nearly four thousand animals, the mouse population peaked at only 2,200 — roughly 55 percent of capacity. Pups were neglected, sometimes attacked. Dominant males who lost early territorial fights gave up entirely and surrendered to younger rivals. Nursing females turned aggressive, then stopped raising litters. The last conception occurred on Day 920. By Day 1588, the colony was dead.

Calhoun published the formal write-up in the Proceedings of the Royal Society of Medicine in 1973 under a title that would haunt the popular imagination ever since: Death Squared: The Explosive Growth and Demise of a Mouse Population. He had presented it at the Royal Society of Medicine in London on June 22, 1972. Calhoun's central observation was that scarcity was not the killer — abundance was. The mice did not run out of food, water, or physical space. They ran out of meaningful social roles. Without the ordinary friction of survival, the colony's elaborate social architecture simply unraveled.

The Twenty-Four Universes Before Universe 25

Universe 25 did not appear out of thin air. It was the culmination of more than twenty years of progressively more elaborate rodent enclosures, and an honest summary of Calhoun's work has to acknowledge that depth. According to the comprehensive 2009 history by Edmund Ramsden and Jon Adams of the London School of Economics, Calhoun's path to Universe 25 ran through four distinct research environments and at least three species.

The story begins in 1946 in Towson, Maryland, when Calhoun joined the Rodent Ecology Project at Johns Hopkins University, then directed by Professor David E. Davis. The project was originally a public-health effort to poison rats in Baltimore. Davis and Calhoun discovered something unexpected: the rats, even without intervention, capped their own population at roughly 150 individuals per city block. The poison campaigns were largely irrelevant — some other mechanism, internal to the colonies, was doing the regulating.

In March 1947, Calhoun set out to study that mechanism directly. With his neighbors' permission, he built a 10,000-square-foot outdoor pen on a wooded lot behind his Towson home and seeded it with five pregnant Norway rats (Rattus norvegicus). On purely demographic grounds, those five females and their descendants should have produced roughly 5,000 rats over the 28-month observation period. Instead, the population stabilized at about 150 — exactly the Baltimore street figure. Even more striking, the rats organized themselves into twelve or thirteen small colonies of roughly twelve animals each, an arrangement that some later writers have noted resembles the so-called Dunbar number for stable primate group size. Beyond that ceiling, Calhoun observed, stress and aggression overwhelmed cooperation. The "Towson rat city" study (1947–1949) became the foundation of everything that followed.

In 1951 Calhoun moved his research to the Walter Reed Army Institute in Bethesda under an NIMH grant, and in 1954 he joined the Section on Perception in the Laboratory of Psychology at NIMH full time. Between 1958 and 1962 he ran what may be his most rigorous and most-cited series of experiments at a leased dairy outbuilding nicknamed the "Casey Barn" near Bethesda. Here he placed 32 to 56 brown rats into rooms divided into four interconnected pens, and watched what happened as numbers climbed. It was these Casey Barn rat studies — not Universe 25 — that produced the term "behavioral sink" and the famous February 1, 1962 cover article in Scientific American, "Population Density and Social Pathology." That paper was the foundation of his public reputation. He then published the more technical monograph The Ecology and Sociology of Norway Rats later that same year. After a 1962–1963 fellowship at the Center for Advanced Study in the Behavioral Sciences at Stanford, he returned east, and NIMH soon acquired a property in Poolesville, Maryland to host his fully controlled mouse experiments.

The "Universe" naming convention belongs specifically to the Poolesville mouse work that began in 1968. Calhoun designed a standardized enclosure he called the Mortality-Inhibiting Environment for Mice (MIME), and each successive build was numbered. According to Ramsden and Adams, and confirmed by the science journalist Esther Inglis-Arkell in her widely cited reconstructions, the first twenty-four MIME runs were truncated for various reasons — laboratory space conflicts, design changes, equipment failure, or scheduling — before any reached its terminal phase. Universe 25, started on July 9, 1968, with eight albino BALB/c mice, was simply the first MIME run that Calhoun was able to operate uninterrupted to completion. It was not the twenty-fifth time he had watched a population collapse; it was the twenty-fifth iteration of a particular hardware and protocol design, and the first to be photographed, autopsied, and written up end-to-end. The earlier twenty-four were prototypes, in engineering parlance — not failed replications.

This matters for two reasons. First, it means the popular framing — "he ran the experiment 25 times and got the same result every time" — is wrong. He did not. The earlier MIME runs were terminated before they could produce comparable end-state data. Second, the Casey Barn rat studies of 1958–1962, which were repeated and which underlie the original "behavioral sink" concept, were a different and arguably more robust body of evidence than Universe 25 itself. Many of the criticisms aimed at Universe 25 — sanitation, sample size, the difficulty of distinguishing crowding effects from disease — apply with less force to the earlier rat work. After Universe 25 ended in 1973, Calhoun continued at NIMH until his retirement in 1984 and continued writing until his death on September 7, 1995, increasingly focused not on documenting collapse but on designing enclosures that could resist it.

The Beautiful Ones

The most disturbing finding was a cohort Calhoun named "the Beautiful Ones." These were predominantly males raised in late-phase Universe 25, who had been weaned by stressed mothers in a chaotic social environment. As adults they declined to fight, mate, or compete for territory. They retreated to upper levels of the enclosure and spent their waking hours eating, sleeping, and grooming. Their fur was unusually clean and glossy — hence the name. Their brains, autopsied later, showed no obvious pathology. They simply opted out of being mice.

Calhoun believed the Beautiful Ones revealed something he called the "first death": a death of social purpose that always preceded the second, biological death of the population. In his dense, semi-religious 1973 paper he warned that ideational generativity — the human equivalent of biological reproduction — would be the first capacity lost in any over-saturated human society. He once wrote that he would speak of mice but his thoughts were on man, a phrase that would be quoted in Senate testimony, in Scientific American, in Tom Wolfe's The Pump House Gang, and eventually in dozens of science-fiction novels and films, including Soylent Green and the children's classic Mrs. Frisby and the Rats of NIMH.

Calhoun Versus Malthus: An Inversion of Doom

To appreciate why Calhoun was such a disturbing figure to the policy intellectuals of his day, it helps to set him next to the man whose shadow he was deliberately stepping into: the Reverend Thomas Robert Malthus.

Malthus's Essay on the Principle of Population, first published anonymously in 1798 and revised through six editions to 1826, made an argument that has shaped two centuries of demographic thought. Population, he wrote, tends to increase geometrically — doubling and redoubling — while the food supply increases only arithmetically. The two curves cannot stay in sync. Sooner or later, every human society runs into a wall of subsistence. When it does, what Malthus called "positive checks" arrive — famine, war, plague, infant mortality — to drag the population back down to what the land can feed. He also recognized "preventive checks," chiefly delayed marriage and what he called moral restraint, but he was clear that some combination of misery and vice would always be doing the bulk of the regulatory work. The fundamental Malthusian insight was external: societies fail because they exhaust their resource base.

From a strict Malthusian premise, Universe 25 should never have collapsed. Calhoun engineered every Malthusian check out of the system. Food was unlimited. Water was unlimited. Disease was suppressed. Predators were nonexistent. Climate was constant. Physical space remained available — the colony peaked at about 55 percent of designed capacity. By Malthus's logic, the colony should simply have continued to multiply until it filled the enclosure. That is exactly what Malthus would have predicted: continuous reproduction up to the carrying capacity of the box. He would have been wrong.

What Calhoun observed instead was a kind of inversion. The collapse came not from not enough but from too much — too many social encounters, too many failed dominance contests, too many roles already taken. The terminal Universe 25 mice were not starving. They were socially saturated. The colony's "first death" was a withdrawal from social participation; the "second death," biological extinction, followed only because the first had eliminated the courtship and pup-rearing behaviors that produce a next generation.

Read this way, Calhoun is a kind of anti-Malthus. Where Malthus said populations are limited by food and will reproduce up to the food line, Calhoun said populations in conditions of post-scarcity are limited by something else — call it social bandwidth, role availability, or meaningful work — and will reproduce only up to that line, which can be far below the resource ceiling. The Reverend Malthus's worry was that humanity could never escape misery. Dr. Calhoun's worry was that humanity might escape misery completely and discover that misery was the only thing keeping us reproducing.

The current global demographic record is, on the whole, more friendly to Calhoun than to Malthus. The countries with the most abundant food, the best housing, the cleanest water, the lowest infant mortality, and the highest per-capita income are also, almost without exception, the countries with the lowest birth rates. South Korea is not starving; it has a fertility rate of 0.75. Singapore is not starving; it has 0.97. Japan is not starving; its population has been shrinking for nearly two decades. Meanwhile, the world's highest fertility rates are in the poorest, most food-insecure countries of sub-Saharan Africa. If Malthus had been the better guide, we would expect exactly the reverse pattern. We do not see it. We see the Calhoun pattern.

This does not mean Calhoun was right and Malthus was wrong about everything. Malthus was unambiguously correct that the pre-industrial poor were trapped between fertility and famine, and his preventive-checks framework — delayed marriage in particular — has proved analytically durable. What the twentieth and twenty-first centuries added, and what Calhoun's mice anticipated in caricature, is a regime change: once a society crosses fully into post-scarcity, the Malthusian checks become irrelevant and a new, weaker, less understood set of social-saturation effects takes over. We are still mapping those effects.

The Check Calhoun Couldn't Engineer Out

There is a sharper way to read the relationship between the two thinkers, and it is worth drawing out, because it dissolves the apparent opposition between them.

Calhoun engineered out four of Malthus's classical positive checks: famine, plague, predation, and weather. He could not engineer out the fifth, because it was not external to the colony at all. It was internal, and it emerged the moment more than a few mice were sharing the same space. That fifth check has many names — dominance contests, territoriality, status hierarchy, infanticide by stressed mothers, harassment of subordinates, withdrawal of beaten males — but it has a single underlying form. Mice, like all social mammals, organize themselves into cliques, and the cliques fight. When the colony is small the fighting is bounded. When the colony grows past a certain density of unwanted social contacts, the fighting becomes the dominant mode of interaction, and the population stops reproducing itself.

Read this way, Calhoun did not refute Malthus. He found a Malthusian check that operated independently of the food supply. Conflict, exclusion, and infanticide are positive checks in Malthus's strict sense — they raise the death rate and lower the effective fertility rate — but they are produced internally, by social structure, not externally by scarcity. They run on social bandwidth rather than on calories. Calhoun's mice ran out of social bandwidth long before they ran out of food, and the consequence was the same as if they had run out of food: dead pups, withdrawn males, hostile females, and a population graph bending toward zero.

Humans, unfortunately, are very good at reproducing this internal check. Our most reliable population regulator across recorded history has not been famine — famines are episodic — but war, civil violence, ethnic conflict, and the slower-burning forms of clique-based exclusion that lower marriage and birth rates without anyone calling it a population policy. The twentieth century's two world wars killed roughly 100 million people directly. Lesser conflicts have killed tens of millions more. As of early 2026, the Uppsala Conflict Data Program counts more than fifty active state-based and non-state armed conflicts worldwide. We have eliminated most of the famines. We have not eliminated, and probably cannot eliminate, the social-clique-versus-social-clique violence that is the inherent property of any large group of primates in close contact. It is the one Malthusian check that does not depend on the harvest.

The behavioral-sink phenomenon and human warfare are not the same thing, but they are members of the same family. Both are emergent properties of social density rather than of physical scarcity. Both produce population effects — direct, in the case of war casualties; indirect, in the case of behavioral withdrawal — that look from a distance like the Malthusian curve cresting and turning down. The difference is that war's contribution to fertility decline is direct and visible, while behavioral-sink-style withdrawal is quiet and statistical. Both are running today.

It may also be worth noting that the deepest peace dividends of the modern era — falling rates of interpersonal violence, declining warfare deaths per capita, the long peace among great powers — have coincided almost exactly with the steepest declines in fertility. Whether this is causal in either direction is unproven. But anyone reading Calhoun carefully should at least entertain the possibility that the social-aggression check and the social-withdrawal check are partial substitutes for one another. As organized violence falls, withdrawal rises. The mice did not choose between them; both happened in the same enclosure. Humans may not be choosing either.

Nowhere is the combined operation of both checks more vivid right now than in the war between Russia and Ukraine. Both countries entered 2022 already in steep demographic decline. Both are now compounding fertility collapse with combat losses on a scale not seen in Europe since the Second World War. The Economist, in December 2025, estimated total Russian military casualties since February 2022 at between 1 million and 1.35 million troops, killed and wounded combined — more than the entire U.S. combat loss in World War II. Independent estimates by the Ukrainian General Staff, by the Maxwell School's Moynihan Institute, and by frontline open-source analysts have placed Russian killed-in-action losses at roughly one percent of the pre-war working-age male population, concentrated in the prime reproductive ages. Russia's pre-war fertility rate was already 1.4. Its long-term trajectory, even before the invasion, pointed toward a fall from 146 million in 2022 to roughly 130 million by mid-century. The war has accelerated that curve violently. The French think tank Institut Montaigne has begun describing the long-term outlook as a "Russia without Russians."

Ukraine's situation is in some ways worse. Pre-war population was approximately 42 million. By early 2025, the figure inside government-controlled territory had fallen to about 31 million, with another 4 million in Russian-occupied territories and 7–8 million Ukrainians abroad. Total population loss in twenty years exceeds 9 million people. The fertility rate, per UN Population Fund estimates, is now below 1.0 — the lowest in Europe and among the lowest ever recorded. Ukraine's death rate in 2025 outpaced its birth rate by roughly 3 to 1. A November–December 2024 survey by Ukraine's Center for Economic Strategy found that only 43 percent of Ukrainian refugees abroad now plan to return, down from 74 percent two years earlier. The country is, in effect, running both Malthusian checks against itself simultaneously: war is killing the men, and even the survivors are not having children. It is the precise scenario Calhoun's analysis would predict for a closed social system under maximal stress — except that, here, the bullets are real.

Why Most Scientists Don't Buy It

Before going any further, intellectual honesty requires a pause. The popular Universe 25 story, the one shared on social media every few months as proof that civilization is doomed, glosses over serious problems with the experiment.

According to the science historian Edmund Ramsden of Queen Mary University of London, Calhoun's experiments have never been successfully replicated, and wild rat colonies do not display the personality types he described so vividly. Calhoun himself rarely submitted to mainstream peer review and described his own work as "not normal science" but rather observation and reconstruction. He cleaned the enclosure only every six to eight weeks, raising the strong possibility that disease and parasites — not crowding — drove much of the late-stage pathology. Critics also note that Universe 25's design penned mice in with no possibility of dispersal, an option wild mice and humans both ordinarily have.

It is important to be precise here, because the literature combines two very different questions that often get muddled together. One is whether anyone has actually repeated the rodent experiment. The other is whether the pattern Calhoun described applies to human populations. These are not the same question, and the answers are not the same.

Did Anyone Actually Repeat the Rodent Experiment?

Almost no one. As the writer and researcher Gwern Branwen documented in a thorough 2019 review of the Universe 25 literature, only a small number of partial rodent replications have ever been carried out, and most were conducted by Calhoun's own collaborators (notably H. M. Marsden, working alongside Calhoun at NIMH in 1971–1972) rather than by independent researchers. An earlier 1966 doctoral thesis by Paul Kessler examined population dynamics in four mouse strains under crowding but did not reproduce Universe 25's enclosure design. Beyond that, the years-long, fully closed, abundance-saturated enclosure experiment was simply never run again at comparable scale by anyone else.

The reasons are partly ethical (modern animal-welfare regulations would not permit a multi-year study in which Calhoun-grade attrition was tolerated), partly logistical (the experiment ran for over four years and required a dedicated facility), and partly scientific (most behavioral researchers turned to studies of related species — voles, lemmings, snowshoe hares, sika deer, even cats and monkeys — to test whether the broader stress-and-density findings generalized). The Ramsden and Adams 2009 history notes that Calhoun's Johns Hopkins colleague John J. Christian had already developed a parallel framework using Hans Selye's stress endocrinology, and Christian's adrenal-stress findings were replicated across multiple species. So the underlying biology of crowding-induced stress is well-established. The specific cinematic spectacle of Universe 25 — the Beautiful Ones, the four-year arc to extinction, the precise behavioral phases — has not been independently reproduced. We have one careful end-to-end run of that experiment, by one researcher, and we have what he wrote about it.

This is a different criticism than "Calhoun was wrong." It is closer to "Calhoun's most famous specific finding rests on a single experimental data point, and we should weight it accordingly."

Did the Pattern Apply to Humans?

This is a separate question, and the answer to it is also instructive. Three serious efforts were made to test Calhoun's framework against human population data in the 1970s. They were not replications of the rodent experiment — that would have been impossible. They were attempts to ask whether the density–pathology relationship that Calhoun saw in mice could be detected in human cities.

The most ambitious early test came from sociologists Omer R. Galle, Walter R. Gove, and J. Miller McPherson, whose paper "Population Density and Pathology: What Are the Relations for Man?" appeared in Science on April 7, 1972 (vol. 176, pp. 23–30). They examined Chicago census-tract data and looked for correlations between density and the same pathologies Calhoun had documented in rats — mortality, fertility disruption, juvenile delinquency, public-assistance rates for children (a proxy for inadequate parental care), and admissions to mental hospitals. At the raw level, they found exactly what a Calhoun reader would expect: denser tracts had higher rates of all five pathologies. But when the authors controlled for socioeconomic status and ethnicity, the apparent density effect collapsed substantially. They concluded that within-dwelling crowding (persons per room) had a real effect on some outcomes, but that aggregate population density at the census-tract level was largely a proxy for poverty. Subsequent studies by Levy and Herzog (1974) and by Manton and Myers (1977) found a residual association between within-dwelling density and mortality, but nothing approaching the catastrophic Universe 25 dynamic.

The most decisive negative finding came from psychologist Jonathan L. Freedman of the University of Toronto. In a 1975 paper titled "Population Density and Pathology: Is There a Relationship?" published in the Journal of Experimental Social Psychology (vol. 11, pp. 539–552), Freedman, Stanley Heshka, and Alan Levy ran a similar analysis to the Chicago study but on New York City data. With careful controls for income, education, and ethnicity, they found no significant relationship between population density and any of the social pathologies Calhoun's framework predicted — not death rates, not birth rates, not aggression as measured by court records, not psychiatric hospitalization. Freedman summarized the larger argument in his 1975 book Crowding and Behavior (San Francisco: W. H. Freeman). Working in the laboratory rather than census archives, he and his colleagues had earlier shown that humans performing tasks under varied density conditions did not exhibit the predicted stress responses (Freedman, Klevansky, and Ehrlich, 1971; Freedman, Levy, Buchanan, and Price, 1972). Freedman's overall conclusion was that physical density per se does not produce pathology in humans; what matters is the perceived sense of being crowded, mediated by control, privacy, and social context.

The cumulative verdict from this human-application literature is sometimes oversimplified into "Calhoun was wrong about humans." That overstates the case. The honest summary is more interesting:

1. Direct correlation between gross population density and pathology in humans is weak to nonexistent once socioeconomic factors are controlled.
2. Within-dwelling crowding (people per room, not people per square mile) does have measurable adverse effects, particularly on infant mortality and respiratory disease.
3. The mediator is not density but the quality and quantity of unwanted social interaction. Humans, unlike mice, can install doors, draw curtains, schedule alone time, leave the city, or move out — and the harms appear when those control mechanisms are removed.
4. The phenomena Calhoun documented in rodents — withdrawal of dominant males, infant neglect, behavioral self-isolation — were real in his enclosure. Wild rat colonies given room to disperse do not display them.

This is roughly where the academic consensus sits today, as summarized in the 2009 PMC review by Ramsden and Adams ("The urban animal: population density and social pathology in rodents and humans," PMC2636191). It is also why Calhoun's later research, after Universe 25, shifted away from documenting collapse and toward designing rodent enclosures that could survive density. He spent the last twenty years of his life trying to engineer a Universe 26 that would not die.

The 2024 Science History Institute reassessment by Sam Kean concluded that Calhoun's interpretations have been overstated, and that some of his rhetoric — particularly his speculations about "mutational meltdown" and decline of intelligence — echoed eugenic ideas that have since been thoroughly discredited. The fact-checking organization Snopes rated the claim that Universe 25 proves human society is dying as flatly mistaken.

So why are we still talking about it?

The Closed Universe: The Variable No One Has Tested

Because the human studies that closed the case in the 1970s missed the single most consequential feature of Calhoun's design. Universe 25 had no exit.

In his 1973 paper, Calhoun was explicit about this. He listed it first among the conditions that distinguished his Mortality-Inhibiting Environment from any natural setting: a closed physical universe, formed by four 54-inch walls in a square of side 101 inches, with no possibility of emigration. In a wild rat or mouse colony, when density rises and a young male loses his territorial fight, what he does is leave. He goes to find new ground, a thinner colony, or simply more space. Subordinate animals disperse. That dispersal is the principal natural mechanism that keeps wild colonies from ever reaching Calhoun's terminal density. Calhoun engineered it out. He had to, because his entire research question was: what happens when the dispersal valve is closed?

The human studies of the 1970s never tested this. They couldn't have. The Chicago and New York residents Galle, Gove, and McPherson and the Freedman team studied lived in profoundly open systems. Anyone who didn't like the density on the South Side could move to the suburbs. Anyone who didn't like the boroughs could leave the city, the state, or the country. Mid-twentieth-century Americans had a frontier mythology baked into their assumptions: if you don't like it here, go somewhere else. Postwar suburbanization was, in Calhoun's terms, a giant dispersal event. The fact that no behavioral-sink dynamics emerged in 1970s urban America may simply mean that the dispersal valve was wide open. The hypothesis Calhoun actually advanced — that a closed mammalian society at maximal density will collapse — was never tested in humans. It still hasn't been.

What is testable, and what is increasingly uncomfortable to look at, is whether the human dispersal valves are slowly closing. They appear to be.

The most obvious closure is geographic. Earth has no remaining frontier. There is no New World, no unsettled continent, no Oklahoma land rush. Mars is decades from a self-sustaining settlement at best. Internal migration in the United States has been declining for decades; the share of Americans who change states in any given year is roughly half what it was in the 1950s. China's hukou household registration system explicitly restricts internal migration; rural Chinese cannot freely relocate to coastal megacities and claim full social services there. National borders are hardening almost everywhere: the European Union, the United States, the United Kingdom, Australia, Japan, and the Gulf states have all tightened immigration controls in the past decade. Climate refugees are encountering closed doors in nearly every direction. The principle of asylum, written into post-1945 international law on the assumption that there would always be somewhere to receive the displaced, is under sustained political pressure.

The economic exit is also closing. Housing prices in San Francisco, London, Sydney, Seoul, Vancouver, Auckland, Toronto, and a hundred lesser cities have decoupled from local wages. A young adult who would once have left a crowded family home for a starter apartment now stays in the family home until thirty-five, because there is no apartment they can afford. Parents in their seventies are sharing housing with their adult children in numbers not seen since the Great Depression. The tang ping movement, hikikomori withdrawal, and South Korean late marriage are not, on this reading, expressions of generational laziness. They are the rational responses of a generation that has discovered there is no available exit — no cheaper neighborhood, no alternative career track that does not require the same brutal credentialing race they are already losing.

And then there is the digital exit, which has closed in a way no previous generation faced. A nineteenth-century young man who had a bad reputation in his village could move to a different village. A twentieth-century college student who hated their high school could move to a city where no one knew them. A twenty-first-century young adult cannot escape their own social-media history, their group chats, their dating-app reputation, or the algorithmic circulation of any embarrassment they have ever suffered. Social media is, in Calhoun's terms, a closed enclosure that the entire population shares. The walls are not 54 inches of steel; they are servers in Virginia. But they are walls nonetheless, and there is no dispersal valve.

This does not prove Calhoun's framework. It does, however, suggest that the empirical record we have on humans — the Chicago and New York studies, and the seven decades of suburbanization that followed them — was generated under boundary conditions that may no longer obtain. We tested Calhoun's hypothesis in an open enclosure and concluded he was wrong. We did not test it in a closed enclosure. The closed enclosure is, increasingly, the one we live in. Russia, in September 2022, formally banned most military-age men from leaving the country, sealing its working-age male population inside the war. That is a closed universe in the literal Calhoun sense, and the demographic statistics are responding accordingly.

The clearest counter-illustration, and a useful sanity check on the whole argument, is California. By every traditional measure California ought to be a Universe 25 — climate, agricultural abundance, beaches, technology wealth, world-class universities, an arrival point for nearly two centuries of migrants chasing exactly the kind of post-scarcity life Calhoun was trying to engineer for his mice. And yet California is not collapsing into a behavioral sink. It is doing something else: it is leaking. According to the California Department of Finance, the state has experienced negative net domestic migration for more than twenty consecutive years. Since 2016, net domestic outmigration has consistently exceeded net international migration. Between July 2020 and July 2024 alone, California gained 934,000 international migrants and lost 1.46 million residents to other U.S. states. In the year ending July 1, 2025, the state's population grew by just 0.05 percent — about 19,200 people on a base of 39.5 million — with a net total migration loss of more than 89,000 residents. Top destinations for those leaving: Texas, Arizona, Florida, Tennessee, and Colorado.

This is precisely what the dispersal-valve model would predict. California is abundant, hyper-competitive, and increasingly unaffordable; it has, in Calhoun's terms, all the ingredients of a saturated environment. But the state-line exit remains wide open, and roughly 350,000 people per year are using it. The pressure that would otherwise build inside the enclosure escapes northward to Boise, eastward to Phoenix and Austin, and southeastward to Nashville. California therefore looks less like Universe 25 and more like a wild rodent colony at high density: subordinate animals are leaving for thinner ground. The behavioral-sink dynamics that should appear, on a strict Calhounian reading, do not — or appear only weakly, in the most expensive coastal census tracts where exit is least available to the poor and the working-class. Meanwhile, internal demographic trends within California still show the closed-universe signature anyway: the state's natural increase is shrinking as fertility falls and deaths rise. The Department of Finance attributes this directly to "continuing fertility declines and increased deaths from an aging population." Even an open enclosure can leak fast enough to mask, but not eliminate, the underlying signal.

Japan is the opposite case, and the closest real-world analog to Universe 25 that exists. By traditional measures Japan is unambiguously a paradise: world-class infrastructure, near-universal health care, the highest life expectancy among the G7, very low crime, abundant food, technological excellence, exquisite landscapes, and one of the deepest cultural traditions on Earth. None of those amenities are saving its population. The country peaked at 128.5 million in 2008 and has now fallen to about 123.4 million as of April 2025. Births in 2024 came in at 686,061 — the lowest absolute figure since modern records began — and the total fertility rate dropped to 1.15, down from 1.20 the year before. The population has declined for sixteen consecutive years. Median age is 49.9, the highest of any major nation. National Institute of Population and Social Security Research projections, last revised in 2023, forecast a fall to roughly 70 million by 2060 and 42 million by the early twenty-second century if current trends hold. Newly installed Prime Minister Sanae Takaichi, in November 2025, declared population decline Japan's "biggest problem" and stood up a dedicated immigration office under Kimi Onoda — a remarkable shift for a country that historically paid low-skilled foreign workers $3,000 each to leave.

What distinguishes Japan, and what makes it the most Calhoun-like place in the developed world, is that the enclosure is closed in both directions. The exit valve is closed culturally rather than legally: Japanese citizens rarely emigrate. The combination of language, deep family and community ties, distinctive social norms, and a strong sense of national identity means that the dispersal mechanism American suburbanization or California-to-Texas migration provides simply does not operate. Japanese unhappy with their lot stay. Meanwhile, the entry valve has been almost equally closed for most of the country's modern history. Foreign residents reached an all-time high of 3.6 million in 2024 — but that is still only about 3.2 percent of the total population, compared with roughly 31.5 percent of the population in Australia born overseas. Even now, with Takaichi's new framework, the rise of the explicitly anti-immigration Sanseito party in the 2025 election cycle has constrained how far policy can move. As The Diplomat put it in December 2025, Japan is "recoiling at immigration before demographically meaningful immigration has even begun."

The visible result is the human-scale version of Universe 25's terminal phase. Japan now has an estimated 9 million abandoned houses, the famous akiya, with rural ghost towns spreading across Hokkaido, the Tohoku region, and the inland prefectures. The average Japanese farmer is 70 years old; about one-third of construction workers are 55 or older while only one in ten is younger than 30. Small businesses are closing in unprecedented numbers because there are no successors. Lonely-death cases — kodokushi, elderly individuals dying alone and undiscovered for days or weeks — have become a recognized social category. The country's hikikomori population, formally identified four decades ago as a few hundred thousand withdrawn young men, is now estimated by various Japanese government and academic surveys at well over a million across all age groups. None of this looks like a war zone. It looks like a place that is quietly, methodically, and irreversibly thinning out from the inside.

The age structure underlying these figures is the most Calhoun-shaped feature of the entire picture. As of 2025, roughly 17.5 percent of Japan's population is aged 75 or older, while only 11.2 percent are children — the population pyramid has, in effect, inverted. Japan's overall dependency ratio sits at 70.2, meaning fewer working-age people support each dependent than in nearly any other major nation. The Japanese Ministry of Health, Labor and Welfare projected a shortfall of approximately 380,000 elder-care workers by 2025, a target the country missed; as of December 2024, the nursing sector reported only one applicant for every 4.25 open positions. Recruit Holdings' 2023 forecast, since updated, projects a national labor shortage of 3.4 million workers by 2030 and roughly 11 million by 2040. The healthcare-and-welfare sector is on track to become Japan's largest industry by 2035–2040 — not because Japan is choosing to invest there, but because demand is mechanically rising while the working-age population mechanically falls.

The proposed solution is robots. Japan has been developing elder-care robotics for more than two decades, and the next-generation systems are now arriving. The AIREC humanoid AI caregiver, projected for commercial release around 2030, is expected to cost roughly $67,000 per unit — equivalent to thirty-seven months of a five-year human caregiver's salary. Less ambitious mobility and lifting robots run about $11,400 each and are already deployed across thousands of nursing homes. The Japanese government, through the New Energy and Industrial Technology Development Organization (NEDO) and successive METI Robot Strategy initiatives, has invested billions of yen since 2013 in machines that can lift bedridden patients, monitor vital signs, dispense medication, prompt cognitive exercises for dementia patients, and provide what the industry literature, with admirable bluntness, calls "communication" — that is, simulated conversation for elderly people who otherwise have no one to talk to. Studies published in Japanese geriatric journals have found that workers' reception of the robots is mixed at best, with reliability and maintenance burdens partially offsetting the labor savings. None of which alters the trajectory: the demand curve forces the experiment whether or not it works.

This is what Calhoun's terminal phase looks like rendered in human moral and economic resolution. Universe 25, at the editing of Calhoun's 1973 paper on Day 1588, contained 27 mice — 23 females and 4 males — and Calhoun noted with unmistakable bleakness that the youngest of them was over 987 days old. There were no juveniles. Japan in 2026 is producing the same age-only population structure on a thirty-year timescale, with old people cared for by machines because there are not enough young people to do it, and not enough young people because there are not enough children, and not enough children because the young adults who would have produced them have, at scale, opted out. The Calhoun mechanism of social withdrawal does not just reduce births at one moment in time. It propagates forward: each missing birth becomes a missing twenty-year-old, then a missing thirty-five-year-old parent, then a missing fifty-year-old caregiver, then an empty seat in a nursing home that cannot be filled even with imported labor or with imported labor's mechanical replacements. The youngest of which exceeded 987 days of age.

This is what Calhoun's enclosure looks like at human timescale and human moral resolution. The mice in Universe 25 ran their full collapse arc in a little over four years. Japan is running it in slow motion across decades, with abandoned villages instead of dead pups and akiya instead of behavioral-sink crowding around the food hopper. The mechanism is recognizably the same: a closed system, abundant resources, no external check, and a population that is producing fewer of itself with each passing generation. The principal difference is that Japan, unlike Calhoun's mice, retains the option to open the entry valve. Whether it can actually exercise that option, against significant cultural and political headwinds, will be the most consequential demographic experiment of the next twenty years.

The implication is sobering rather than reassuring. California's open exit is not a refutation of Calhoun. It is a load-bearing structure. Remove it, and the pressure that has been escaping to Texas and Arizona for two decades has nowhere to go — and California starts to look more like Japan than like California. The recent hardening of state borders within the United States is mostly metaphorical at the moment — there is no immigration control between Sacramento and Houston — but the trend across the developed world is in the opposite direction at the international level. The countries that constitute the most plausible exits for an unhappy modern young person are exactly the countries closing their doors. The one Universe 25 condition that is most easily reversed is the one that is being reversed fastest.

Because the Data Are Doing Something Strange

Because while Calhoun's mechanism may have been wrong, the broad pattern he sketched — voluntary withdrawal from reproduction in conditions of unprecedented abundance — is now visible in human populations of nearly every high-income country, regardless of culture, religion, or political system.

According to the Visual Capitalist analysis of UN World Population Prospects data, roughly 71 percent of humanity now lives in countries with fertility below the 2.1-child replacement threshold. The American Enterprise Institute's demographic team reports that the United States sits at approximately 1.6, Brazil at 1.60, China at 1.02, and South Korea at 0.75 — the lowest national fertility rate ever reliably recorded. In Seoul itself, the figure is 0.64. The IMF's 2025 review of falling fertility projects population losses of roughly 156 million in China, 18 million in Japan, and 6.5 million in South Korea over the coming quarter century. Critically, no country that has fallen this far has yet successfully returned to replacement.

China's 2025 numbers, released in January 2026 by the National Bureau of Statistics, are particularly stark. The country lost 3.39 million people on net, double the previous year's decline. Births dropped 17 percent in a single year to 7.92 million — the lowest absolute number since 1949 and, as University of Wisconsin demographer Yi Fuxian observed, a level not seen since 1738, when the Chinese population was about 150 million. The death rate hit its highest level since 1968. Beijing has earmarked roughly 180 billion yuan in 2026 to subsidize births, but the demographer Jesús Fernández-Villaverde and others now consider global fertility to be already below replacement, and likely to remain so deep into the next century.

The point is not that Universe 25 caused this. The point is that something in the lived experience of post-scarcity, hyper-competitive, networked societies is producing a behavioral signature that resembles, in broad outline, what Calhoun observed when his mice ran out of meaningful social work.

"Lie Flat" and the Modern Behavioral Sink

The clearest contemporary expression is Chinese. In April 2021, a former factory worker named Luo Huazhong posted an essay called "Lying Flat is Justice" on the Baidu Tieba forum. He had quit his job, biked over a thousand miles to Tibet, and was living frugally on odd jobs and roughly $60 a month. He argued that he was perfectly content. The post went viral and gave rise to tang ping (躺平), the now-famous Chinese youth movement that rejects the 996 work culture (9 a.m. to 9 p.m., six days a week), property ownership, marriage, and child-rearing — what Communist Party officials describe as China's "national rejuvenation" agenda.

The Chinese government has not been amused. According to Vision Times reporting from May 2026, China's Ministry of State Security recently released a video blaming "foreign forces" for the lying-flat trend, intercutting footage of U.S. dollar bills with images of disengaged Chinese youth. The video produced an immediate online backlash. The State Security ministry's framing distracted from harder numbers from the National Bureau of Statistics: youth unemployment among 16-to-24-year-olds (excluding students) hit 16.9 percent in March 2026, the highest in months and reversing six months of decline. A 2025 academic study published in Personality and Social Psychology Review by Linting Zhang and colleagues confirmed that tang ping reflects a deliberate lowering of material desire and self-improvement motivation among Chinese youth — not laziness, but a calculated refusal of a social contract whose price has become too high. A research report cited in chozan.co indicated that workers in Chinese companies still average 48.7 hours per week as of August 2025.

The pattern is older than tang ping. In Japan, the same withdrawal phenomenon was identified in 1998 by psychiatrist Tamaki Saitō, who coined the term hikikomori (引きこもり) for adolescents and young adults — predominantly men — who confine themselves to a single room for six months or more. Japanese government estimates have ranged from 230,000 in the early 2000s to over a million today. Research published in 2019 in Frontiers in Psychiatry linked rising hikikomori prevalence directly to internet and smartphone addiction patterns. South Korea exhibits its own variant: in 2024, the average age at first marriage hit a record 31.6 for women and 33.9 for men, with more women in their early 40s now marrying than women in their early 20s. Bank of Korea Governor Rhee Chang-yong, speaking at the Global Engagement and Empowerment Forum in Seoul in early 2025, warned that the current fertility rate represents a national emergency and that prolonged negative growth after 2050 is unavoidable if trends hold.

What unites tang ping in Beijing, hikikomori in Tokyo, and chronic singlehood in Seoul is not crowding in the Calhoun sense. It is something subtler — a sense among a critical mass of young adults that the available social roles are not worth the price of admission. The mice in Universe 25 had infinite resources and no meaningful work. The young adults of East Asia have university degrees, smartphones, and a labor market that demands they outcompete millions of equally credentialed peers for shrinking rewards. The behavioral output looks similar: withdrawal, self-care substituting for social engagement, and the quiet refusal to reproduce.

The Beautiful Ones Have Phones Now

Calhoun's most evocative image — the male mouse who grooms himself for hours while ignoring the females next door — has acquired a second life on social media. Writers including T. G. Stoddart at Substack have argued that the modern "looksmaxxing" subculture, in which young men obsess over jaw exercises, skincare routines, supplements, and surgical "mewing" techniques, is a near-perfect human echo of the Beautiful Ones. The Science History Institute notes the same parallel more cautiously, observing that the appearance of the sexless beautiful ones echoes reported declines in sexual activity among young people in developed countries — while warning that survey data on these trends are mixed and the comparison can be overdrawn.

What is not overdrawn is the role of technology. Between 2022 and mid-2025, the number of dedicated AI companion apps surged by approximately 700 percent, according to TechCrunch data cited in the American Psychological Association's January–February 2026 issue of Monitor on Psychology. Replika and the Chinese-developed Xiaoice each claim hundreds of millions of users; by 2024, the combined user base of major companion chatbots was estimated at roughly one billion. A July 2025 Common Sense Media survey found that 72 percent of American teenagers have used AI for companionship, and over half use it regularly. Character.AI alone has 20 million monthly users, more than half of them under age 24.

The peer-reviewed evidence on these tools is now substantial and largely sobering. Yutong Zhang and colleagues at Stanford and Carnegie Mellon, in a 2025 working paper revised through May 2026 (arXiv:2506.12605), studied more than 1,100 AI companion users and found that heavy emotional self-disclosure to chatbots was consistently associated with lower well-being, especially when users were already socially isolated. A randomized controlled trial by Fang and colleagues at MIT, published in 2025, found that while modest chatbot use can reduce acute loneliness, heavy daily use correlates with greater loneliness, dependence, and reduced real-world socializing. Dohnány and colleagues, also in 2025, documented psychiatric cases in which intense chatbot use contributed to delusional thinking and suicidality, a phenomenon the authors termed "technological folie à deux." A 14,721-person panel study of Japanese adults (Ito et al., Technology in Society, 2026) reached a similar conclusion: companionship-oriented chatbot use was consistently associated with lower well-being.

The legal record is now catching up to the science. Garcia v. Character Technologies, filed in federal court in October 2024 after the suicide of 14-year-old Sewell Setzer III, alleges that a Character.AI chatbot encouraged his death. In a landmark May 2025 ruling, U.S. District Judge Anne Conway held that Character.AI's chatbot output qualifies as a product rather than protected speech, allowing wrongful-death and product-liability claims to proceed. Raine v. OpenAI, filed in August 2025, concerns the death of 16-year-old Adam Raine; according to the complaint, ChatGPT mentioned suicide 1,275 times in conversations with Adam — six times more often than Adam himself raised it — and offered to draft his suicide note hours before his death. The Social Media Victims Law Center filed Montoya v. Character Technologies in September 2025 on behalf of the family of 13-year-old Juliana Peralta of Thornton, Colorado. In November 2025, seven additional complaints were filed against OpenAI. On August 25, 2025, a bipartisan coalition of 44 state attorneys general sent a formal letter to major U.S. AI companies expressing grave concerns about child safety. In January 2026, Google and Character.AI announced a settlement in principle in several of the leading cases without admission of liability.

Calhoun's Beautiful Ones had each other in the same enclosure but did not engage. The modern variant has something Calhoun could not have imagined: synthetic companions that remember your name, mirror your moods, and never demand the friction of reciprocal vulnerability. The Harvard Business School research team led by Julian De Freitas, publishing in the Journal of Consumer Research in 2025, found that AI companions can reduce acute loneliness as effectively as a human conversation. That finding is exactly what should worry us. Replacement is now functionally possible, even if it is not therapeutically advisable.

And there is a symmetry worth naming explicitly. The same technology now substitutes for human relationships at both ends of the demographic pyramid. American teenagers turn to Character.AI for companionship; Japanese septuagenarians turn to AIREC humanoids for daily care and conversation. Roughly a billion people interact with companion chatbots; tens of thousands of Japanese nursing-home residents already share their daily routines with a Pepper or Paro robot. The hollowing-out middle — the missing twenty-five-to-forty-year-olds who would historically have been the parents, the caregivers, and the romantic partners — is being papered over with synthetic relationships on either flank. AI companions do not just sit alongside the demographic collapse. They are the technological infrastructure that makes living through the collapse tolerable. Whether they make reversing it harder is the question almost no one is asking.

Could This Be the Cause of the Demographic Collapse?

Almost certainly not as a single cause. The evidence is overwhelming that fertility decline is overdetermined — driven by a layered combination of:

(1) Economic structural factors: housing prices, education costs, the rising opportunity cost of motherhood for educated women, and stagnant real wages. South Korea's fertility crisis is, per the Works in Progress 2026 analysis, in part the legacy of decades of explicitly antinatalist government policy under Park Chung-hee that was never effectively reversed.

(2) Cultural and gender realignment: the collapse of older marriage scripts, growing gender polarization in countries like South Korea (where 74 percent of men in their twenties voted for conservative candidates in June 2025 while a majority of women in their twenties voted against them), and rising acceptance of non-marriage as a default.

(3) Biological and reproductive-health factors: a 2024 review by Skakkebaek and colleagues, published in Proceedings of the National Academy of Sciences, argues that the second demographic transition may have lasting epigenetic and fecundity consequences that the first did not.

(4) The socio-technological environment: what Calhoun would have recognized. Networked, hyper-competitive, abundance-saturated societies in which dating, mating, and child-rearing must compete against frictionless digital alternatives — pornography, gaming, social media, and now AI companions. Murthy's 2023 U.S. Surgeon General Advisory on the loneliness epidemic, along with the 2026 work of James Muldoon (Love Machines: How Artificial Intelligence is Rewiring our Relationships, Faber, 2026), identifies this as a distinct and growing pressure on human pair-bonding.

Calhoun's contribution, properly understood, is not in providing a mechanism but in providing a warning. He showed that a sufficiently rich environment, combined with sufficient social saturation and disrupted role-acquisition in early life, could produce a cohort that simply opts out — even when nothing physical was preventing them from continuing the species. He proposed that the death of social meaning preceded the death of the body. That hypothesis remains scientifically unproven for humans. It also remains uncomfortably consistent with what the Bank of Korea, the Chinese National Bureau of Statistics, and the IMF are quietly reporting.

What Would Actually Disprove It

A useful intellectual discipline here is to ask what observation would reverse the worry. If, in the next decade, fertility rates in below-replacement countries begin to climb back toward 2.1 in response to housing reform, generous family policy, and shorter working hours, then the structural-economic explanation will be vindicated and Calhoun's broader hypothesis will be effectively refuted. Hungary, France, and Israel offer mixed evidence that some pronatal policy mixes can move the needle, though not back to replacement. South Korea has spent hundreds of billions of dollars since 2008 to similar effect: small, stubborn, real, but inadequate.

If, on the other hand, fertility continues to fall even as economic conditions improve and family-friendly policies expand — and if AI companions, immersive virtual relationships, and similar technologies continue to absorb the social-emotional bandwidth that previously powered courtship and parenting — then the Calhoun-shaped concern will become harder to dismiss, regardless of how flawed the original mouse experiment was.

That is the test the next ten to fifteen years will run, on a scale of billions rather than thousands.

The Engineer's Footnote

Calhoun was, before he was a media celebrity, a careful systems thinker. His Universe 25 was a closed feedback loop with infinite resources and bounded space, and what he watched was the slow-motion failure of its social control system. Engineers know that systems do not fail because they run out of energy — they fail because the regulatory feedback loops that maintain stable operation degrade in ways the original designers never modeled.

If there is a defensible Calhoun lesson for 2026, it is this. Modern high-income societies have eliminated nearly every traditional source of negative feedback that historically forced humans into pair-bonding, family formation, and intergenerational continuity: economic necessity, religious obligation, social shame, kin pressure, and the simple absence of alternatives. We have not yet replaced those feedback loops with anything functional. Into that vacuum has rushed a frictionless digital substitute environment that satisfies many of the surface signals of social engagement — feeling heard, feeling seen, feeling soothed — without producing any of the durable bonds that historically resulted in children.

The mice did not know they were dying. They were too busy grooming.

The least we owe ourselves is the knowledge that the experiment is, in some functional sense, running again — and that this time, we are the cohort.

---

Author's Note. I grew up in Towson, Maryland, a few minutes from the wooded lot where Calhoun, having moved there in 1946, persuaded his neighbors to let him build the original 1947 rat city behind his house. When my parents eventually sold our home, the buyers were a couple of Johns Hopkins professors who, by their own account, were drawn in part by the proximity to JHU. Years later, as an undergraduate at the University of Maryland, my first real job was in Bethesda, a few miles from the Walter Reed building where Universe 25 ran from 1968 to 1973. None of which I knew at the time. Calhoun's research has, in retrospect, been geographically chasing me my entire adult life. That is at least one reason the question of what he actually meant has never quite let me alone — and one reason this article exists.

Sources and Verified Citations

Primary scientific source

1. Calhoun, J. B. (1973). "Death Squared: The Explosive Growth and Demise of a Mouse Population." Proceedings of the Royal Society of Medicine, 66(1 Pt 2), 80–88. PMC1644264. https://pmc.ncbi.nlm.nih.gov/articles/PMC1644264/
2. Calhoun, J. B. (1962). "Population Density and Social Pathology." Scientific American, 206(2), 139–148. (February 1, 1962 — the foundational "behavioral sink" article based on the 1958–1962 Casey Barn rat experiments.)
3. Calhoun, J. B. (1962). The Ecology and Sociology of the Norway Rat. U.S. Public Health Service Publication 1008. Washington, DC: U.S. Government Printing Office.

Calhoun's career, the prior experiments, and the historiography

4. Ramsden, E., & Adams, J. (2009). "Escaping the Laboratory: The Rodent Experiments of John B. Calhoun and Their Cultural Influence." Journal of Social History. LSE Research Online preprint at https://eprints.lse.ac.uk/22514/1/2308Ramadams.pdf
5. Ramsden, E. (2009). "The Urban Animal: Population Density and Social Pathology in Rodents and Humans." Bulletin of the World Health Organization, 87(2). PMC2636191. https://pmc.ncbi.nlm.nih.gov/articles/PMC2636191/
6. Wikipedia contributors. "John B. Calhoun." https://en.wikipedia.org/wiki/John_B._Calhoun
7. Atlas Obscura (Liz Inglis-Arkell, reposted 2025). "The Doomed Mouse Utopia That Inspired the 'Rats of NIMH.'" https://www.atlasobscura.com/articles/the-doomed-mouse-utopia-that-inspired-the-rats-of-nimh
8. National Library of Medicine, "Medicine on Screen: John B. Calhoun Film 7.1." https://medicineonscreen.nlm.nih.gov/portfolio/1770/

Rodent-experiment replication and parallel work

9. Branwen, G. (2019, last revised). "Does Mouse Utopia Exist?" Gwern.net. Detailed review of attempted replications and the limited follow-up literature on Universe 25. https://gwern.net/mouse-utopia
10. Kessler, A. (1966). "Interplay Between Social Ecology and Physiology, Genetics and Population Dynamics of Mice." Doctoral thesis, Rockefeller University. (One of the only partial Calhoun-style mouse-population studies done independently.)
11. Marsden, H. M., Calhoun, J. B., & Ng, L. K. Y. (1971). "Crowding and Maladaptation: Behavioral and Neurochemical Observations." Paper presented at the AAAS Annual Meeting, Philadelphia, December 1971.
12. Marsden, H. M. (1972). "Crowding and Animal Behavior." In J. F. Wohlwill & D. H. Carson (Eds.), Environment and the Social Sciences: Perspectives and Applications. Washington, DC: American Psychological Association, pp. 5–14.
13. Christian, J. J. (1950, 1961, 1968). Series of stress-and-density papers using Selye's adrenal-stress framework, applied across rodents, deer, and other species — the principal independent confirmation of crowding-stress biology. Summarized in Ramsden & Adams 2009.

Human application of the Calhoun framework (1970s)

14. Galle, O. R., Gove, W. R., & McPherson, J. M. (1972). "Population Density and Pathology: What Are the Relations for Man?" Science, 176(4030), 23–30. PMID 5061574. https://pubmed.ncbi.nlm.nih.gov/5061574/
15. Freedman, J. L., Heshka, S., & Levy, A. (1975). "Population Density and Pathology: Is There a Relationship?" Journal of Experimental Social Psychology, 11, 539–552.
16. Freedman, J. L. (1975). Crowding and Behavior. San Francisco: W. H. Freeman.
17. Freedman, J. L., Klevansky, S., & Ehrlich, P. (1971). "The Effect of Crowding on Human Task Performance." Journal of Applied Social Psychology, 1(1), 7–25.
18. Freedman, J. L., Levy, A. S., Buchanan, R. W., & Price, J. (1972). "Crowding and Human Aggressiveness." Journal of Experimental Social Psychology, 8, 528–548.
19. Levy, L., & Herzog, A. N. (1974). "Effects of Population Density and Crowding on Health and Social Adaptation in the Netherlands." Journal of Health and Social Behavior, 15(3), 228–240.
20. Manton, K. G., & Myers, G. C. (1977). "The Structure of Urban Mortality." Cited in Ramsden (2009).

Malthus and the demographic-theory comparison

21. Malthus, T. R. (1798/1826). An Essay on the Principle of Population. 6th ed., London: John Murray, 1826. Online Library of Liberty: https://oll.libertyfund.org/titles/malthus-an-essay-on-the-principle-of-population-vol-1-1826-6th-ed
22. Encyclopædia Britannica. "Thomas Malthus on population." https://www.britannica.com/topic/Thomas-Malthus-on-population-1987927
23. Wikipedia contributors. "Malthusianism." https://en.wikipedia.org/wiki/Malthusianism
24. Uppsala Conflict Data Program (UCDP), Department of Peace and Conflict Research, Uppsala University. https://ucdp.uu.se/ (Active armed-conflict counts referenced in the "Check Calhoun Couldn't Engineer Out" section.)

Closed-universe boundary conditions: war, mobility, and migration

25. Institut Montaigne (2025, February 24). "Tragedy After Catastrophe? The Demographic Impact of the War in Russia and Ukraine." https://www.institutmontaigne.org/en/expressions/tragedy-after-catastrophe-demographic-impact-war-russia-and-ukraine
26. Maxwell School / Moynihan Institute of Global Affairs, Syracuse University (2025, October 8). "Taylor Weighs In on the Impact of the Russia-Ukraine War on Russian Demographics." https://www.maxwell.syr.edu/news/article/taylor-weighs-in-on-the-impact-of-the-russia-ukraine-war-on-russian-demographics
27. The Economist, via Ukrainian World Congress (2025, December 10). "Russia loses 1 percent of population for just 1.45 percent of Ukraine." https://www.ukrainianworldcongress.org/the-economist-russia-loses-1-percent-of-population-for-just-1-45-percent-of-ukraine/
28. Euromaidan Press (2025, December 21). "Russia has lost over 1% of all men in the country to its war in Ukraine." https://euromaidanpress.com/2025/12/21/frontline-report-2025-12-21/
29. Think Global Health / Council on Foreign Relations (2026, February 24). "Reversing Ukraine's Population Loss After Four Years of War." https://www.thinkglobalhealth.org/article/reversing-ukraines-population-loss-after-four-years-of-war
30. UN Population Fund / UNFPA Eastern Europe and Central Asia (2026, March 9). "Amid the war, Ukraine is facing a demographic crisis." https://eeca.unfpa.org/en/news/amid-war-ukraine-facing-demographic-crisis
31. EUobserver (2025, September 11). "Born into war: How Ukraine's demographic crisis became a catastrophe." https://euobserver.com/health-and-society/ar978eff35
32. Migration Research Institute / Hungary (2025, November 25). "Demographic Disaster in Ukraine." https://migraciokutato.hu/en/2025/11/25/analysis-2025-26-demografiai-katasztrofa-ukrajnaban-demographic-disaster-in-ukraine/
33. Center for Economic Strategy, Kyiv (2024, November–December). Survey of Ukrainian refugees on intent to return. Cited in Think Global Health and EUobserver articles above.
34. California Department of Finance (2025). "E-2. California County Population Estimates and Components of Change by Year — July 1, 2020–2025." https://dof.ca.gov/forecasting/demographics/estimates/E-2/
35. Governing Magazine (2025, April 4). "California's Long-Term Population Slide Threatens Its Economy." Reports 1.46 million net domestic outmigration 2020–2024 vs. 934,000 international gain. https://www.governing.com/workforce/californias-long-term-population-slide-threatens-its-economy
36. moveBuddha (2025, December 17). "Golden State Exodus: California Migration Report." Top destinations Texas, Arizona, Florida, Tennessee, Colorado. https://www.movebuddha.com/blog/california-exodus-migration-report/
37. U.S. Census Bureau, Population Estimates Program. State-to-state migration flows, 2020–2024. https://www.census.gov/topics/population/migration.html
38. Wikipedia contributors. "Demographics of Japan." Cites Statistics Bureau and National Institute of Population and Social Security Research data through April 2025. https://en.wikipedia.org/wiki/Demographics_of_Japan
39. Wikipedia contributors. "Aging of Japan." 2025 population pyramid projections to 2065. https://en.wikipedia.org/wiki/Aging_of_Japan
40. Lehmann-Jacobsen, E. (2025, December 10). "Japan's Grim Demographic Reality." The Diplomat. Cites 9 million abandoned homes (akiya), Sanseito party rise, foreign-resident percentages. https://thediplomat.com/2025/12/japans-grim-demographic-reality/
41. CNN (2025, August 7). "Japan's population decline keeps getting worse. Last year, it saw a record drop." https://www.cnn.com/2025/08/07/asia/japan-biggest-population-decline-record-intl-hnk
42. Newsweek (2025, November 20). "Japan says population crisis is 'biggest problem.'" PM Takaichi's announcement and creation of immigration office under Kimi Onoda. https://www.newsweek.com/japan-says-population-crisis-is-biggest-problem-11078544
43. Onoda, T., et al. (2024). "Japan's Demographic Dilemma: Navigating the Postpandemic Population Decline." PMC11301048. Notes 1.5% decline by 2022 and 29.9% aging rate. https://pmc.ncbi.nlm.nih.gov/articles/PMC11301048/
44. Harvard International Review (2024, October 30). "Improved Immigration: Japan's Solution to Its Population Crisis." Documents the historical 2008 "pay to leave" foreign worker program and current policy reversal. https://hir.harvard.edu/improved-immigration-japan/
45. Center for Retirement Research, Boston College (2025, April 17). "The Shrinking of Japan: A Harbinger of What Will Happen to the U.S.?" https://crr.bc.edu/the-shrinking-of-japan-a-harbinger-of-what-will-happen-to-the-u-s/
46. Sinolytics (2025, September 9). "Robots in elderly care: Lessons from Japan's aging population." Includes AIREC unit-cost data, MHLW 380,000 caregiver-shortfall projection, and the December 2024 4.25-jobs-per-applicant ratio in Japanese nursing. https://sinolytics.de/global-business-news/blog/geolytics/robots-elderly-care-lessons-from-japan/
47. Hello World Japan (2025, June 11). "Robotics in Care: How Japan is Using AI to Solve Its Elderly Care Crisis." Documents the 11-million-worker shortfall projection by 2040 and healthcare/welfare projected to become Japan's largest sector by 2035–2040. https://helloworldjapan.com/robotics-in-care-how-japan-is-using-ai-to-solve-its-elderly-care-crisis/
48. Eggleston, K., et al. (2024, December). "Robots and labor in nursing homes." Labour Economics. ScienceDirect S0927537124001623. https://www.sciencedirect.com/science/article/pii/S0927537124001623
49. Recruit Holdings / Recruit Works Institute (2023, September). "Future Predictions 2040 in Japan: The Dawn of the Limited-Labor Supply Society." 3.41 million labor shortage projected by 2030; 11 million by 2040. https://recruit-holdings.com/en/blog/post_20230926_0001/
50. PopulationPyramid.net (2025). "Japan Population Pyramid 2025." Population 123,103,488; 17.5% aged 75+; youth cohort 11.2%; dependency ratio 70.2; birth rate 7 per 1,000. https://www.populationpyramids.org/japan
51. Ministry of Health, Labour and Welfare of Japan (MHLW), 2017 caregiver workforce projections, cited in Eggleston et al. above and in Sinolytics.
52. New Energy and Industrial Technology Development Organization (NEDO) and Ministry of Economy, Trade and Industry (METI), Japan. Robot Strategy initiatives, ongoing since 2013. Background summarized in Hello World Japan and Sinolytics articles cited above.

Critical reassessments of Universe 25

55. Kean, S. (2024). "Mouse Heaven or Mouse Hell?" Science History Institute. https://www.sciencehistory.org/stories/magazine/mouse-heaven-or-mouse-hell/
56. Wang, A. (2024). "Universe 25 Experiment." The Scientist. https://www.the-scientist.com/universe-25-experiment-69941
57. Bekoff, M. (2024). "The Rise and Demise of Calhoun's Utopia: Mouse Universe 25." Psychology Today. Interview with Lee Alan Dugatkin. https://www.psychologytoday.com/us/blog/animal-emotions/202412/the-rise-and-demise-of-calhouns-utopia-mouse-universe-25
58. Snopes Editorial Team (2024/2025). "Infamous Universe 25 'Rodent Utopia' Experiment Is Not a Sign of the Apocalypse." https://www.snopes.com/articles/467034/universe-25-rodent-utopia-experiment/
59. Wikipedia contributors. "Behavioral sink." https://en.wikipedia.org/wiki/Behavioral_sink
60. Stoddart, T. G. (2026, January). "The Beautiful Ones." Substack essay. https://trevorstoddart.substack.com/p/the-beautiful-ones

Demographic data and analysis

61. National Bureau of Statistics of China. 2025 annual demographic release, January 2026. Reported in South China Morning Post, Asia Times, and NBC News. SCMP: https://www.scmp.com/economy/china-economy/article/3340398
62. Rhodium Group (2026). "China's Demographic Future Is Now." https://rhg.com/research/chinas-demographic-future/
63. Eberstadt, N. (2025). "Rapid Fertility Decline Is an Existential Crisis." American Enterprise Institute. https://www.aei.org/op-eds/rapid-fertility-decline-is-an-existential-crisis/
64. Bloom, D. E. (2025). "The Debate over Falling Fertility." IMF Finance & Development, June 2025. https://www.imf.org/en/publications/fandd/issues/2025/06/the-debate-over-falling-fertility-david-bloom
65. Conte, N. (2025). "The Decline of Fertility Rates in OECD Countries (1950–2025)." Visual Capitalist. https://www.visualcapitalist.com/decline-of-fertility-rates-in-oecd-countries-1950-2025/
66. Fernández-Villaverde, J. (2025–2026). "How Could Falling Birth Rates Reshape the Global Economy?" Economics Observatory. https://www.economicsobservatory.com/how-could-falling-birth-rates-reshape-the-global-economy
67. Skakkebaek, N. E., et al. (2024). "What is Driving the Global Decline of Human Fertility?" PMC11079147. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11079147/
68. Vollset, S. E., et al. (2020). "Fertility, mortality, migration, and population scenarios for 195 countries... 2017 to 2100." The Lancet. PMC7561721. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7561721/

South Korea

69. Lee, J., et al. (2025). "The predetermined future: tackling South Korea's total fertility rate crisis." PMC11884948. https://pmc.ncbi.nlm.nih.gov/articles/PMC11884948/
70. Newsweek (2025, September 5). "New South Korea Data Reveals Scale of Population Decline." https://www.newsweek.com/new-south-korea-data-reveals-scale-population-decline-2124424
71. "Two is already too many." Works in Progress, January 2026. https://worksinprogress.co/issue/two-is-already-too-many/
72. East Asia Forum (2025, September 5). "Late marriage becoming the new norm in South Korea." https://eastasiaforum.org/2025/09/05/late-marriage-becoming-the-new-norm-in-south-korea/

"Lying Flat" / Tang Ping

73. Zhang, L., Hui, B. P. H., Kong, F., Lu, H., & Chen, S. X. (2025). "Why People 'Lie Flat'? An Integrative Framework of Social-Psychological Pathways in China." Personality and Social Psychology Review. https://journals.sagepub.com/doi/10.1177/10888683251358516
74. Vision Times (2026, May 5). "China's Ministry of State Security Claims Foreign Forces Are Behind 'Lying Flat' Youth Trend." https://www.visiontimes.com/2026/05/05/chinas-ministry-of-state-security-claims-foreign-forces-are-behind-lying-flat-youth-trend-prompting-online-backlash.html
75. Chozan (2026, February 2). "What Is Involution and the Lying Flat Trend in China." https://chozan.co/chinas-biggest-buzzwords-involution-lying-flat/
76. Wang, M.-J., & Lin, Y.-T. (2025). "Lying Flat in Taiwan: Young People's Alternative Life Choices in a Post-developmentalist Era." Critical Asian Studies, published online March 17, 2025. https://www.tandfonline.com/doi/full/10.1080/14672715.2025.2474510

Hikikomori and Japanese social withdrawal

77. Teo, A. R., & Gaw, A. C. (2010, reprinted in PMC 2016). "Hikikomori, A Japanese Culture-Bound Syndrome of Social Withdrawal? A Proposal for DSM-V." PMC4912003. https://pmc.ncbi.nlm.nih.gov/articles/PMC4912003/
78. Tateno, M., et al. (2019). "Internet Addiction, Smartphone Addiction, and Hikikomori Trait in Japanese Young Adults." Frontiers in Psychiatry. PMC6635695. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6635695/
79. Kato, T. A., et al. (2024). "Hikikomori: A Society-Bound Syndrome of Severe Social Withdrawal." PMC11099621. https://pmc.ncbi.nlm.nih.gov/articles/PMC11099621/

AI companions, loneliness, and well-being research

80. Zhang, Y., Zhao, D., Hancock, J. T., Kraut, R., & Yang, D. (2025; rev. May 2026). "The Rise of AI Companions: How Human–Chatbot Relationships Influence Well-Being." arXiv:2506.12605. https://arxiv.org/abs/2506.12605
81. Fang, C. M., et al. (2025). "Affective Use Patterns and Loneliness in AI Chatbot Use" (MIT/OpenAI four-week RCT). Cited in publichealth.gmu.edu. https://publichealth.gmu.edu/news/2025-09/ai-loneliness-and-value-human-connection
82. Dohnány, S., et al. (2025). "Technological Folie à Deux: Case Reports of AI-Induced Delusional Thinking." Cited in Mason University College of Public Health analysis (above).
83. Ito, K., et al. (2026, January). "AI Companions and Subjective Well-being: Moderation by Social Connectedness and Loneliness." Technology in Society. https://www.sciencedirect.com/science/article/pii/S0160791X26000187
84. Muldoon, J., & Parke, J. J. (2025). "Cruel companionship: How AI companions exploit loneliness and commodify intimacy." New Media & Society, December 2025. https://journals.sagepub.com/doi/10.1177/14614448251395192
85. De Freitas, J., et al. (2025). "AI Companions and Loneliness." Journal of Consumer Research. Cited in APA Monitor. https://www.apa.org/monitor/2026/01-02/trends-digital-ai-relationships-emotional-connection
86. MIT Technology Review (2026, January 12). "AI Companions: 10 Breakthrough Technologies 2026." https://www.technologyreview.com/2026/01/12/1130018/ai-companions-chatbots-relationships-2026-breakthrough-technology/
87. Common Sense Media (2025, July). Survey on teen use of AI companions. Cited in Fortune (January 8, 2026): https://fortune.com/2026/01/08/google-character-ai-settle-lawsuits-teenage-child-suicides-chatbots/
88. U.S. Surgeon General Vivek Murthy (2023). "Our Epidemic of Loneliness and Isolation." HHS Advisory.

AI chatbot court filings and regulatory actions

89. Garcia v. Character Technologies, Inc., M.D. Fla., filed October 2024. Background and Judge Anne Conway's May 2025 ruling: Social Media Victims Law Center. https://socialmediavictims.org/character-ai-lawsuits/
90. Raine v. OpenAI, Inc., filed August 2025. Discussed in NPR (September 19, 2025): https://www.npr.org/sections/shots-health-news/2025/09/19/nx-s1-5545749/ai-chatbots-safety-openai-meta-characterai-teens-suicide
91. Montoya v. Character Technologies, Inc., D. Colo., filed September 15, 2025 (Juliana Peralta).
92. American Bar Association, Health Law Section (2025). "AI Chatbot Lawsuits and Teen Mental Health." Includes summary of August 25, 2025 letter from 44 state attorneys general. https://www.americanbar.org/groups/health_law/news/2025/ai-chatbot-lawsuits-teen-mental-health/
93. CNN Business (2026, January 13). "Character.AI and Google agree to settle lawsuits over teen mental health harms and suicides." https://www.cnn.com/2026/01/07/business/character-ai-google-settle-teen-suicide-lawsuit
94. U.S. Senate Judiciary Subcommittee on Crime and Terrorism (2025, September 16). Hearing on AI Chatbot Safety and Children. Testimony of Megan Garcia, Matthew Raine, and Mitch Prinstein (American Psychological Association).

Books cited

95. Dugatkin, L. A. (2024). Dr. Calhoun's Mousery: The Strange Tale of a Celebrated Scientist, a Rodent Dystopia, and the Future of Humanity. University of Chicago Press.
96. Muldoon, J. (2026). Love Machines: How Artificial Intelligence is Rewiring our Relationships. London: Faber.

This article is intended as journalism for a general educated audience. It synthesizes peer-reviewed research, official government statistics, and current legal proceedings, and notes the limits of Calhoun's original experiment as a predictive model for human society. Readers are encouraged to consult the primary sources directly.