According to the selfish gene theory, it seems like because identical twins sometimes get produced, a mutation to a gene that says, "if you have an identical twin, be fully altruistic towards them" would get selected for when ever it happens to arise in an identical twin. After humans evolve to be fully altruistic towards identical twins, wouldn't a mutation to a gene that makes a woman more likely to produce identical twins also be selected for because if it happens to appear in a a child who has a non-identical sibling, that gene might be discontinued from reproduction by the partial selfishness of the sibling but if it appears in an identical twin, the other identical twin will preserve that gene with their altruism.
What you describe could have happened under the right conditions. However, there are a few things you haven't considered.
- Because humans are especially altricial, always having twins would double the cost of children on parents.
- The benefit of sexual reproduction is immune diversity. So a population like this could be far more vulnerable to disease.
So as far as this thought experiment goes, genes can't be too selfish or they dramatically reduce their own fitness. It's something worth simulating computationally to see what happens.
From jzx's answer, I thought of a possible answer to my own question. Maybe not producing identical twins all the time is a paradoxical evolutionary stable strategy for the following reason:
Until recently in evolutionary history, the population remained constant when parents only had enough food to feed 2 children. Siblings didn't evolve to be fully altruistic towards each other so if a parent had 3 children in case of the small chance of having enough food to feed 3, none of them would willing let themself starve to make sure the others survive, risking making only one survive, so the parents would have had an evolutionary advantage in only producing 2 children in the first place. Although definitely producing identical twins is an evolutionary advantage, a mutation that makes the chance of producing identical twins slightly higher than it already is is an evolutionary disadvantage because a woman with an increased chance of producing identical twins might have produced 1 child on her first pregnancy and identical twins on her second pregnancy, giving a total of 3 children and increasing the risk of only rearing 1.
If we produced identical twins all the time then we'd completely lose genetic diversity which would, I assume, overwhelmingly compromise our adaptive capabilities. Don't forget, those identical twins then need to reproduce with other identical twins of others families. Moreover, and this is the more obvious explanation, as a consequence deleterious recessive disorders would rapidly increase in rate.
You might also like to study frequency dependent selection for why it is sometimes advantageous to have identical twins, albeit at a low rate.