Interesting patterns in who identifies themselves as homosexual, bisexual, or transsexual
TV shows showing homosexual stars has played a role in this. It would be hard to do empirical work on this in the US because you would only have purely time series data to deal with, but I wonder if it could be done across countries. The relatively low income shown for those in this category is due in significant part to the large percentage of young people who identify themselves as having these preferences. What is perhaps more surprising is that blacks and Hispanics are more likely than whites to identify themselves this way.
There is some other interesting data at a dating website (though this obviously suffers from self selection problems). This information indicates that about half of the people who identify themselves as bisexual are really heterosexuals. The other interesting feature is how the percents change over time. Men are much more likely to be homosexuals when they are young and women when they are older. Is this result really surprising given that men find it more difficult to obtain sex when they are young and because of the relative shortage of men the same is true for women when they are older?
It would also be interesting to see whether those who might be less attractive to the opposite sex are more likely to be homosexual. For example, low income males or those with significant prison histories might fall in this category (it also raises the question of whether criminal records explains the higher rates among blacks and Hispanics). On the last point there is the whole issue of whether prison turns some people to be BLGT or whether their unattractiveness to others because of their criminal activity is what causes them to be this way. I wonder if any of these various Gallup surveys that asked this question on homosexuality had information on criminal records.