Universe 25: The Mice Had Food. They Ran Out of Roles

The most famous mouse colony in psychology did not run out of food or space. It ran out of things to be. That is a better lesson for a fast-growing company than the one usually told about it.

The experiment is John B. Calhoun's Universe 25. The popular version says crowding drove the mice mad, and by extension that cities will do the same to us. Calhoun's own paper says something narrower and, for anyone running an organization, more useful.

What happened in Universe 25?

A closed enclosure, eight mice, unlimited food and water, and a peak of 2,200 animals at day 560, after which the colony never grew again.

The details are in Calhoun's 1973 paper "Death squared," published in the Proceedings of the Royal Society of Medicine. On July 9, 1968, at the US National Institute of Mental Health, he put four pairs of mice into a square pen divided into 16 cells with 256 nest boxes. Fifteen mice could rest comfortably in one box. By his arithmetic, shelter would not become a limit until the population passed 3,840, and food not until 9,500. There were no predators.

The first litters came after 104 days. Then the population doubled about every 55 days. At around 620 weaned mice, growth suddenly slowed to a doubling every 145 days or so. On day 560 it stopped. The last litter that survived was born on day 600.

After that it was all decline. On June 22, 1972, there were 122 mice left, 22 males and 100 females. By the paper's final edit that November, 27. Calhoun projected that the last male would die around May 1973.

Growth stopped at 2,200 mice, with shelter and food still to spare

Universe 25 by phase, days since colonization and population. Source: Calhoun (1973), "Death squared," Proc. R. Soc. Med. 66

1. Day 0Four pairs of mice go inJuly 9, 1968. 256 nest boxes, enough shelter for 3,840 mice; food for 9,500
2. Day 104Phase A ends, first littersThe first 104 days were an adjustment period
3. Days 104 to 315Phase B: doubling every 55 daysAt about 620 weaned mice, growth abruptly slows
4. Days 315 to 560Phase C: the roles fill upDoubling now takes about 145 days. Surplus young mice compete for roles in a full social system; those who lose withdraw
5. Day 560 onPhase D: peak of 2,200, then only declineLast surviving litter on day 600; 122 mice in June 1972, 27 in November

The colony never ran short of resources. At its peak it used 57% of its shelter.

What ran out was roles

Calhoun's explanation is in the paper. The available social roles filled up first, and the next generation had nowhere to go.

In phase C, very few young mice died. Far more of them reached adulthood than the colony had places for. In the wild, young animals with no place can leave. Here they couldn't. In Calhoun's words, they had to remain, and they contested for roles in the filled social system.

The ones who lost behaved in specific ways. Some males withdrew, physically and psychologically, and gathered in large pools near the center of the floor, where they occasionally attacked each other. Some females moved to the upper nest boxes and stopped reproducing.

And then there were the males Calhoun called "the beautiful ones." They never approached females and never fought. Their behavior shrank to eating, drinking, sleeping and grooming. They competed for nothing, so they took part in nothing.

His conclusion was that when the opportunities to fill roles fall far short of the number of individuals capable of filling them, violence and the breakdown of social organization follow.

Does Universe 25 prove anything about people?

Not much. It is a good metaphor and weak evidence, and it should be used that way.

The method has problems. The Scientist's account notes that the work was largely observational and subjective, that it lacked quantitative measures such as stress hormones, and that it has not been replicated. Calhoun himself said it was not normal science.

The interpretation has problems too. In a 2009 paper in the Journal of Social History, historians of science Edmund Ramsden and Jon Adams showed how the popular message, crowding causes pathology, overshadowed the more nuanced and more positive one Calhoun wanted to spread. Later in his career he stopped treating population growth as bad in itself. He argued that humans can enlarge their "conceptual space" through ideas and technology, and his focus moved from the number of contacts toward roles, networks and creativity. His own reading of the experiment changed over time.

That said, the human analogy was his before it was anyone else's. The paper's first sentence is "I shall largely speak of mice, but my thoughts are on man." Critics at the time accused him of anthropomorphism. Human crowding studies since then have produced inconsistent results.

So I use Universe 25 for the question it raises and not for an answer. Plenty of psychology studies managers like to quote fail when someone runs them again. This one has never been run again at all.

The phase C company

Here is the question worth borrowing. What happens to an organization when headcount grows faster than the number of real roles?

A common pattern in foreign companies' Japan entities goes like this. Year one, a handful of people, each owning one area. Everyone knows who is responsible for what. Year two, headquarters approves a budget and the entity hires to twenty or thirty. Salaries, desks and laptops all arrive. Resources are fine. What doesn't grow is the number of seats where one person owns one piece of work end to end.

The big boxes on the org chart stay the same. More people get put inside each box.

Then you start to see things that look a little like the pen.

Titles inflate. When there aren't enough seats, titles fill the gap. More managers and heads of things appear, with no reports, no budget and nothing they can decide on their own.

Young people can't see the next seat. Ask someone eighteen months in what role they are aiming for next, and they can't say, because there isn't one. In Japan, far more people want to change jobs than actually do (I looked at that gap here). People who can neither leave nor move up are living in something close to a closed pen.

Friction collects at the borders. Two teams chase the same customer. Three people have conflicting opinions on one document. It looks like a personality problem. Usually nobody drew the line.

And then the beautiful ones. On time, no trouble, fast replies to email, no proposals, never in front. Executives call it quiet quitting. I read it as someone who looked for a place, didn't find one, and stopped looking.

What to watch, once a quarter

Counting people tells you little. Count roles.

Count the seats where one person owns a result that can be attributed to them, and compare that number to headcount. If headcount doubled in a year and this number didn't move, you are in phase C.

Count titles. Of the people with a manager-type title, how many have reports, a budget, and decisions they can make without asking? A title with none of the three was handed out in place of a seat.

Ask people 12 to 18 months in one question: what is your next seat here? The share who can't answer will tell you about a problem about six months before attrition does.

Listen to meetings. When the arguments shift from how to do something to whose job it is, the borders were never drawn.

The fix is unglamorous. Cut the roles before you hire the people. Split big jobs into units one person can own end to end, and write down what each owns and what each can decide. It is the same order as scoring the role before the person.

The obvious hole in the analogy is that mice can't resign and people can. I keep coming back to that. But in Japan the gap between wanting to leave and leaving is wide, and the ones who can leave first tend to be the ones you most wanted to keep. The rest stay, and keep competing around seats that are already taken.

So if you can't carve out the seat, wait on the hire.

---