It’s all about numbers.
Specific genes, and the traits that come with those genes spread through a population through an increased number of surviving offspring. Those genes which give an advantage to a specific individual, or group of related individuals, will eventually become distributed through the population becoming the rule rather than the exception.
Let’s create a little example. Because the selection processes have changed somewhat for Homo sapiens, becoming less obvious through our use of technology, we’ll use a pre-technological community.
Start with a population of 10 people, 5 males, 5 females who pair off into 5 sets of parents.
Let’s say the average survivability of offspring, that is the likelihood a child will grow into an adult capable of reproducing, is 50%, mostly due to circumstance and chance.
However, we have one pair who because both members have a specific gene, the survivability of their offspring is 80%. In the community, 40% of the members have the gene. It doesn’t matter for our little exercise what the gene does, it could make them immune to a common disease, it could give them better eyesight, or increased stamina, or something else. It doesn’t matter. What matters is that it’s a recessive gene that increases the chances offspring will survive long enough to have offspring of their own.
In this community, the average number of offspring born is 5. Obviously some families will have more and some will have fewer but it simplifies the example if all families have the same number of children born.
We start with 40% of the initial 10 members of the population with the gene.
Family A (1 has gene) has 5 children, 3 live to reproductive age, 2 inherit the gene.
Family B has 5 children, 2 live to reproductive age, 0 inherit the gene.
Family C has 5 children, 3 live to reproductive age, 0 inherit the gene.
Family D (1 has gene) has 5 children, 2 live to reproductive age, 1 inherits the gene.
Family F has 5 (2 have gene) children, 4 live to reproductive age, 3 inherit the gene.
We now have 14 reproductive age adults. Six of them have inherited the gene so we’re up to 43% of members with the gene.
From that we get seven families, two with both parents having the gene.
Family A (1 has gene) has 5 children, 3 live to reproductive age, 1 has the gene.
Family B has 5 children, 2 live to reproductive age, 0 have the gene.
Family C (1 has gene) has 5 children, 3 live to reproductive age, 2 have the gene.
Family D has 5 children, 2 live to reproductive age, 0 has the gene.
Family E has 5 children, 2 live to reproductive age, 0 has the gene.
Family F (2 have gene) has 5 children, 4 live to reproductive age, 3 have the gene.
Family G (2 have gene) has 5 children, 4 live to reproductive age, 3 have the gene.
We have 20 adults, 9 have the gene. We’re up to 45%.
It becomes pretty obvious that increasing the number of offspring with the gene reaching reproductive age will also give the gene itself a higher survivability.
If we change the benefit from the offspring themselves to the offspring’s support system – parents, grandparents, great grandparents – that increased survivability doesn’t change for either the offspring or the gene itself.
If the gene increases the productive years of parents so that they become grandparents, even though they themselves are past the age of reproduction, it will enable them to spend more time caring for the offspring, increasing the likelihood those offspring will reach reproductive age.
The very act of increasing the number of active care-givers will result in the gene remaining and increasing in the population.
Having a small percentage of non-reproducing adults who can contribute to offspring survivability may stabilize the percentage of a gene within a population assuming the non-reproductivity is related to the gene (or the developmental environment) and the benefit of the extra care outweighs the costs of having fewer members participating in reproduction.
This can also be applied to non-reproductive homosexual members. Homosexual members can still provide protection, food and support to their nephews and nieces carrying the gene, giving them a better chance of reaching adulthood.
The idea that homosexuality will result in a population dying out, or will result in the gene for homosexual preferences dying out is not based on how genetics works or on a larger scale how selection works.
That said, in the pre-technological population we’re using for the example, the percent of homosexuals in the community is self regulating because eventually the costs of having non-reproductive members will exceed the benefit of having extra care givers.