Groningen, a.d. x Kal. Junias MMDCCLXXI A.U.C.
One of the most baffling things I have encountered in social science is what might be dubbed the rationality paradox: put a few rational individuals together and watch the world burn, yet put irrational people together and they can collectively be quite rational. Examples abound, put to keep this legible I’ll try to keep things short.
An example of collectively irrational behaviour is the infamous tragedy of the commons. Imagine you and your neighbour share a pond with fish. You both like to eat fish so you both go fishing quite often. However, because neither of you can prevent the other from fishing, both of you fish as much as possible, you really like fish after all. After a while the pond is empty, and neither of you can eat fish in the long term. While this almost seems too stupid to be a real life scenario, we can see this type of behaviour all around, fish literally gets caught in such quantities that there will be no fish in the future. The Dutch, Norwegians and British wreaked havoc on the whale population in the North Atlantic already in the 17th century. We burn more fossil fuels than the planet can deal with. Nevertheless, for each individual it is rational to burn fuel and eat that fish: our individual consumption does not really affect the planet directly, while we in turn cannot prevent others from consuming too much, and fish tastes really amazing.
Meanwhile, two irrational yet trusting individuals would have an easy solution for the previous problem: they would simply trust each other. Seeing neither would try to catch fish in spite of their promises, they would both have fish indefinitely. Sounds like a pretty optimal outcome to me. In social science these type of explanations for cooperative behaviour do recur: People cooperate because they simply are cooperative. So, collectively these two people are much more rational than they are individually. See here the rationality paradox: irrational individuals reach a better outcome than rational individuals could.
However, I do not believe that individuals are that irrational. I think neighbours learn to trust each other. Tit-for-tat behaviour (I’m willing to trust you as long as you act trustworthy) is not the same as irrational behaviour. Many of the problems of collectively irrational behaviour have to do with a lack of trust among people, or a lack of communication which could foster this trust. Tit-for-tat behaviour allows people to gain trust and thus prosper together. You are more likely to be scammed by someone you will never see again (I’ll never forget that taxi driver in Tehran) than by someone you have to cooperate with on a daily basis. Indeed, humans evolved in small enough groups to know each other, in which it helped to see whether others were to be trusted, in which it helped to cooperate and create trust, but in which competition also played a role. Hence we are much better at solving these type of collective action problems in small groups than in large groups. Indeed, the modern world is much more complex than a small tribe roaming nature. We often need rules to deal with our collective irrationality.
So, plain irrationality (I always trust people, for instance) does not work well as a theory. This basically denies agency to individuals, it pretends that everyone is just getting along with life constantly showing the exact same behaviour. That is not how humans work though, we ain’t no ants. Like I said before, in some situations we cooperate, in other situations we compete. It is this which theories proposing irrational humans often cannot grapple with: the changing behaviour according to circumstances. At the very least theories of irrational behaviour need to explain what causes changes in behaviour, something which humans do often, but which these theories often cannot explain
So, while the rationality paradox is very real in theory, in real life we mostly need to deal with explaining what induces cooperation amongst individuals, rather than that we see irrational individuals showing ant-like cooperative behaviour.
Bottom Line: Rational individuals can behave irrationally collectively, but irrational individuals can behave rationally collectively. However, it is questionable whether irrational behaviour of that type really exists, because we ain’t no ants.