Could gay marriage reduce the number of homosexuals? The General Social Survey has been asking Americans since 1988 about the sex of people they sleep with. (I use this as my measure of homosexual orientation.) I calculated the completed number of children people have--assuming folks are finished by age 40--for each sexual orientation for two time periods, and listed them below (N = 5,694):


Mean number of children

Gay men
pre-1995 1.46
2000s .94

Lesbians
pre-1995 1.17
2000s 1.58

Straight men
pre-1995 2.46
2000s 2.06

Straight women
pre-1995 2.57
2000s 2.13

To get a sufficiently large sample of gay men and lesbians in the earlier period, I had to look at all people 40 and over (which means that some people were beginning their childbearing years as early as the second decade in the last century).

First, there is no surprise to see large drops among both heterosexual men and women. Gay men have followed the same trend of significantly fewer kids, and they were low in the first place. Lesbians, in contrast, have more now. I assume that this is due to adoption and artificial insemination being more widely available now.

These findings are interesting in connection with the issue of homosexual prevalence and also gay marriage. Greater acceptance is causing fewer gay men to stay in the closet and have children with their girlfriends or wives. The experts tells us that homosexuality is inherited, so if this is true, we might see fewer and fewer gay men in the future. Such a trend would likely be accelerated if the gay lifestyle were completely legitimated through same-sex marriage.

The implications are more complicated for lesbians. Having their relationships normalized evidently has led more of them to openly cohabit with a partner and to have more children than when lesbianism was not acceptable. But many of these kids are adopted, and babies born to one woman won't be biologically related to the other, so even if the typical couple has two children born to one partner (a high estimate), the average completed number of children is only one since zero and two average to one. This is a smaller number of children than the 1.17 average from the older period, so it appears that lesbians could become less prevalent over time as well. (I realize that all this is an oversimplication with lots of assumptions, so input is welcomed).

UPDATE: It didn't take me long to find evidence that contradicts my line of thinking. A study found that male homosexuality appears to be inherited through the maternal line. The research did not examine lesbians.

If this finding is true, then it doesn't matter so much how many kids a gay man has; what matters is the fertility of his sisters and nieces.