Bryan Caplan  

Evolutionary Psychology on Crusonia

PRINT
Krugman and Blinder: Welcome t... Everest and the Quest for Stat...
Suppose two 20-year-olds wash up on a the desert island of Crusonia.  One is male, the other female.  They are both from the same country, but are otherwise randomly selected.  Both are convinced they have no hope of escaping the island.

Two questions:

1. What fraction of castaways pair bond over the next ten years?

2. If the castaways are unexpectedly rescued and return to their home country, what fraction of pair bonds remain together over the following ten years?

Please show your work.



COMMENTS (8 to date)
Sam writes:

Given some time virtually all of them will pair bond at least once.

I suspect a high fraction carry their affections home. There are a decent number of normal 20 somethings in long term relationships. That's the base rate. Then add living together on a desert island for a decade. That's a memorable first date.

shecky writes:

Now, suppose the two 20-year-olds are the same sex...

Andrew Pearson writes:

Part one:

First, let's distinguish exactly what we mean. Unless one or both has a religious or other moral objection then it seems highly unlikely that they would not start sleeping together. Actual romantic attachment is a different issue, but sleeping together may make it more likely.

I see two approaches. The first is to use a methodology similar to that of Peter Backus to work out what proportion of couples will be compatible. Backus considers education levels (probably somewhat important - if you can't have a mutually satisfying conversation then this kind of bond is unlikely to develop), physical attractiveness (I suspect this would not be very important with only one option, particularly given that on a desert island looks seem unlikely to stick around) and a variety of criteria which are covered by the assumption of two people of opposite sexes on the same island. To this I'd add the probability that both are either straight or bisexual, and I'm not certain how to adjust for personality so, this not being an academic paper, we'll assume that personality exactly cancels out the fact that they are having sex whether or not there is a bond.

We can model education in a variety of ways. I'd suggest either we model it as:
(a) There are certain "levels" of education and people will tend to bond with people of the same level.
(b) Education follows a normal distribution and people need to be within a certain distance.

I won't go into the maths here, because it's far too complicated for a simple comment on a blog.

If 97% of the population is straight or bisexual, then there is a near-enough 94% probability that both of our hypothetical couple are.

Multiply this all out and you get a figure, potentially very variable between different countries. Suppose we can divide into classes with proportions 15% (school dropouts), 35% (school, no university), 40% (university, but not excelling within that context) and 10% (university plus) and people bond iff they are of the same class then 31.5% of random pairs will bond. Call that 30% due to the chance that one or both are gay. Of course, these figures don't necessarily accurate reflect any actual country.

The second way we could attack the problem is to look at surveys of people involved in speed-dating. IIRC, men tend to express interest in meeting about 30% of women again; women tend to express interest in meeting about 10% of men again. The key thing which makes this useful is that this figure holds up regardless of the quality of the people of the opposite sex at the event.

If speed-dating is a reasonable proxy for general compatability (and of course that's a massive if) then that gives our poor couple only a 3% chance of compatibility. It's probably higher than that - looks probably cease mattering, there's presumably some correlation between being interested in each other - but I'd guess that an actual romantic relationship is unlikely to develop. Let's say between 5% and 20% as a rough guess.

I can't say much regarding the second except to suggest looking at the general attrition rate for couples, and suggesting adjusting this up/down based on the shared-experience-bringing-together effect versus external risks to the relationship (e.g. families not getting on with the other partner).

It's also worth considering that the twenties tend to be formative years. The two are likely to be influenced by the other, which may well lead to them becoming more compatible than any two random strangers.

Conclusion: highly uncertain, but I'll guess that a large minority of pairs in this situation will form a pair bond.

sourcreamus writes:

1. 100%, all 20 year olds want to be in a romantic relationship, there are no alternatives to consider so they use the only available options.
2. 25%, 36% of people married at 20 end up divorcing anyway. When you take people back to their lives they are going to want to spend time with family and lose touch with the person they were marooned with. So double the normal rate.

Hazel Meade writes:

same answer as sourcereamus, except that for 2, I give it maybe 1%.

When marooned on a desert island, all social context such as class, ethnicity, peer groups, etc, are eliminated. 100% of people marooned will pair up since there are no social pressures pushing them not to, and there is nobody else to pair up with.

However, once returned to their home country, underlying differences in socioecomic status will become apparent. It will turn out they have very different backgrounds. One has a college educaiton, the other is a high-school dropout. One likes dogs, the other like cats. One is a country hick, the other is a city dweller, etc.
The family and friends of the richer, higher status, partner will disapprove of the match. The higher status partner will start to feel ashamed of dating the lower status person. That will cause an even higher breakup rate. Probability of saying together will depend on the likelihood that both individuals are members of the same socioeconomic group, and have similar backgrounds and interests in the "real" world. Which is to say, very low.


Hazel Meade writes:

Also should add, physical attractiveness becomes a much larger factor when you have other people to compare to.

It's well known that people who rank at the same level of physical attrractiveness date eachother. 10s date 10s. 7s date 7s. etc.
So you get a pair up of a 10 and a 3 on the island, but as soon as they get back to land the 10 is going to have a lot of other options.

The chances of both partners being equally attractive, from the same social class, with similar interests and backgrounds, and having similar education levels is almost zero.

MingoV writes:
Please show your work.
I am not aware of any studies of such situations, or even if such situations ever existed. The nature of the island is important: can it support more than two people? It would be illogical to pair bond and have children if someone must die.

Assuming the island has abundant food, my guesses are:

1. Nearly 100% will pair bond and almost all will have children.
2. At least 75% will stay together.

My rationale for #2 is that the combination of long isolation with pair bonding and children will produce a stronger bond than in most marriages. The divorce rate will be lower than for the general population.

NZ writes:

Is this an economics question?

I'm impressed by the cynicism of other commenters. Many of you apparently don't think that 10 years of struggling to survive with someone and even--by your own suggestion--sleeping with them, would amount to any lasting bond. Add to this fact that the couple would be in a "natural" environment, removed from all the forces in 21st century telling them that splitting up is okay and that life begins at 30 anyway. Incredible.

Anyway, I think I remember reading about some of Josef Mengele's experiments where he'd put a man and a woman together in some imminent-death situation (like, lock them together in an icy chamber for days at a time) and see if they would have sex. You can probably find more on that somewhere.

Comments for this entry have been closed
Return to top