Thread by @mattsclancy

Are bigger populations more innovative? Let's look at a few papers emphasizing the distant past to see what they say. #ThursdayThreads

There are a few reasons to think they would be:
- more brains ➡️ more ideas
- more people ➡️ more specialization (?)
- more consumers ➡️more profit for innovating

@JoHenrich even has a famous model arguing that you need big populations to prevent technological regress.

Suppose tech progress comes from random tinkering, plus retention of good modifications. To get progress, you need to transfer knowledge across generations.

But transmission is noisy, so you usually end up with inferior copies. Errors steadily accumulate unless you have occasional "masters" who show up at random and improve tech. Small populations risk a run of bad luck, where none of these type show up and tech is lost.

Derex et al (2013) test the basic idea in a lab. They get groups of 2-16 to construct "arrowheads" (a simple task) and "fishing nets" (complex) on a computer. Performance is scored, and it's possible to tweak designs and get higher scores. nature.com/articles/natur…

After being shown how to do it, the groups repeat. Each round they can observe a group member, pick which object to "make," and then they get a score.

After 15 rounds, big groups are more likely to discover improved arrowheads, and to maintain (but not improve) fishing nets.

Stepping outside the lab, lots of people have tried to see if there's a correlation between cultural complexity and population size in the distant past, with mixed results. pnas.org/content/113/16…

A challenge is that populations are rarely totally isolated. If a small population is in contact with a big one, it can access the big one's technology (and vice versa). In effect, we don't really have two populations then, just one really big one.

A cool exception is Kline and Boyd (2010), who look at the link between indigenous technology and population for a set of 10 (relatively) isolated islands in Oceania. 10 observations isn't much, but it's better than one! royalsocietypublishing.org/doi/full/10.10…

KB find the bigger populations tend to have more tools, and that they tools they have are more complex (as measured by the number of components comprising them).

Can we go beyond labs and islands though? Kremer (1998) gives is a shot. Rather than try to measure technological complexity though, he assumes better tech allows a given resource base to support more people. ssc.wisc.edu/~walker/wp/wp-…

Since bigger populations also have more ideas, this leads to a self-catalyzing cycle, where more people begets more ideas, which begets more people.

That predicts that the rate of global population growth should be increasing in the absolute size of the global population.

Not bad, at least until the demographic transition!

Next, in an argument that anthropologists/archeologists probably find exasperating, he asks us to suppose everyone had access to roughly the same set of technologies at the end of the ice age, since the major land-masses were all connected.

Once the ice melts though, sea levels rise and land bridges between Africa-Eurasia, the Americas, Australia, Tasmania, and Flinder's Island are severed. Given access to the same technology, they can all support roughly the same number of people per square kilometer.

But some of them have more land, and therefore support more people. Kremer’s model predicts these regions will have faster technological progress and hence higher population density, in the long run.

Hey, they do!
Granted this is 5 observations so we're grading on a curve here. But not bad for a theory that seems obvious, but is kind of hard to actually test.

Tentative conclusion: yep, more people ➡️ more innovation (all else equal).

PS - if you liked this thread, I write about research on innovation on Thursdays (search #ThursdayThreads). There's also a newsletter (mattsclancy.substack.com).
And an RSS feed! (mattsclancy.substack.com/feed/)