For the first time, the Census Bureau has collected representative population-wide data on gender identity and sexual orientation in the latest two weeks of the Household Pulse Survey. Here are the results by age, one including people who didn't know/refused to respond, one not.
Given the stable age gradient on unclear or refuse responses, I prefer to use the method dropping those respondents to estimate a population parameter, i.e. this graph.
Here's how the non cis-het groups look as line graphs for more clarity on those trends. You can see all such identities have gotten a lot more common in younger cohorts. However, the pace of increasing prevalence is not identical.
For example, younger cohorts have only become slightly more likely to be female-born woman-identifying people attracted almost exclusively to women (lesbian).

There has been a large increase among female-born woman-identifying people attracted to men AND women (bisexual female).
Likewise among male-born man-identifying people: there has been a modest increase in gay-identifying, but a rather large increase in bisexual identification.
And we can also see a large increase in the population share which does not identify with the gender identity most typically conforming to birth sex (transgender). I state it nebulously like this because the actual categories used are somewhat complex.
Here are the rates among birth males and birth females.
HPS asks three questions: sex at birth (male or female), gender now (male, female, transgender, none of these, and refuse), sexual orientation (gay/lesbian, straight/heterosexual, bisexual, something else, don't know, and refuse).
I then collapse the categories down, making all male-male-straight and female-female-straight cisgendered heterosexuals, making all mismatch of birth and current ID transgender, and then coding non-trans remainder by orientation.
This procedure is debatable. You could I suppose assume that sexual orientation is the prior category and then code out from there. But I think coding it as first birth, then identity, then orientation is reasonable. Also that's the actual order HPS asks the questions in.
Anyways, back to the splits by natal sex.... folks those splits are *wild*.
Here's the trends in non-cisgendered ID by age and natal sex. Pretty similar though for youngest ages seems higher among natal females.
Pretty stable gender splits on gay/lesbian identity, and also males are WAY more likely to be gay than natal females are to be lesbian.
And finally, bisexuality. This is clearly the dominant category of non cis-het sexual identity in absolute size and rate of growth.
So what's interesting here is that after decades of massive political activism related to homosexuality, there has been very little actual change across cohorts in homosexuality (tho there may have been increases within cohorts!).
However, despite no similar parallel salience-raising political campaign related to transgender identity or bisexuality until comparatively quite recently, there have been rather large increases in those groups.
This is all sort of strange. In a simple stigma story you'd actually expect rather large cohort effects for the groups with the biggest and clearest change in stigmas: gay and lesbian people.
So you have to tell a story of how the sexual changes of the last 30 years created massive cohort effects in bisexuality, strong-but-smaller effects in non-cisgender identity, and small-to-no effect in lesbian or gay identity. I'm not sure what theory makes those predictions?
To the extent we can "date" changes, it seems like the rise in bisexual identity began with the cohort now turning 50 or so, and amplified for those turning 35 or so: so the birth-cohort turning points are 1971 and 1986. Age 18 around 1989 and 2004.
For gay/lesbian identity, there's maybe a jump around the early 40s, so birth cohorts in the early 80s, adulthood around 2000.

For non-cisgender identity, the big inflection point is for those around 30-32 today, so born around 1989-1992, adulthood around 2010.
look, im here for a theory of lifting-bros-destroyed-masculinity.
This is a viable theory for factors impacting lesbian/gay ID.... but it would tend to mean rates for gay people older than 35 or so should be HIGHER, which suggests even LESS increase in lesbian/gay identity in reality!
anyways don't mind me i just have to go re-weight all my surveys now
I saw the note. I understand the concern. However, the whole point of collecting data is to use it! If we don't intend it to be used, don't collect it! Obviously the data could be wrong, but it's probably the least-wrong data we have.

• • •

Missing some Tweet in this thread? You can try to force a refresh
 

Keep Current with Lyman Stone 石來民 🦬🦬🦬

Lyman Stone 石來民 🦬🦬🦬 Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @lymanstoneky

16 Sep
he's right

(but also if manchin isn't gonna budge having a simple binary work requirement impacts a fairly small share of kids and is still a huge improvement over status quo ante)
getting worried that on the CTC what is going to happen is GOP won't budge at all because there's a million other bad things in the bill and also Dems will fail to agree on something because of perfect-universality-extremists.
whereas im over here saying, giving more money to kids is good, and getting a permanent child allowance for every family with any earnings/employment at all in the prior year would be a massive improvement over what we get when the expansion expires
Read 4 tweets
16 Sep
officially over 1,000 responses to my survey of Lutherans, and over 200 for my broader survey of religion. For all Facebook ads, mass mailing, everything, by the time it's all said and done, I'll have spent ~$2,000, meaning my cost per response is going to be $1.70 or less.
which is hilarious since I was quoted $95/response by a major firm and lots of people told me that was actually not crazy.
now of course my sample isn't random!

but when you're sampling a group which is <1% of the population to begin with via an online sample, *that's not a random sample anyways*
Read 38 tweets
15 Sep
war with China would absolutely not be easy to win and could plausibly result in our defeat, which is why it is vital that we prepare more intensively for such a war, and why it is extremely concerning that Milley may have been back-channeling relevant US plans to China.
Milley may have thought China thought we were about to attack. Whether China actually thought that is unclear. And obviously we were *not* about to attack.
Here's the thing though:

Had Milley *not* sent this message, China would have had to debate the matter internally. And they would not have launched a pre-emptive strike because, seriously, that would have been insane.
Read 7 tweets
15 Sep
I think it's a bit much to call it an ideology of masculinity when a lot of it is wildly differential rates of diagnosis for ADD, ADHD, autism, and a gajillion other learning disorders. Differential rates of violent death by sex can be identified in pre-Neolithic remains!
A theory of sex differences in education which doesn't account for the factors we know are OVERWHELMINGLY the most predictive of educational performance (diagnosed learning or attention issues and documented disciplinary issues) seems kind of weird.
And attributing it to a specific masculine ideology is also odd. There may be ideology involved, but the reason for differential rates of male diagnosis for ADHD may be related to ideology, but not an ideology we can call machismo.
Read 14 tweets
15 Sep
Here are estimates of births in Georgia (country, not state) using various methods.

In my view, the student enrollment data by age and the official vital stats data are the most reliable, followed by the reconstructed 2014 census data.
The key thing to understand here is that this implies practically a 10% undercount of recent births in the 2014 Georgian census, which is a massive undercount.
The 2002 census was widely believed to be a huge overcount of adults. It's not clear if this undercount issue might expand to adults as well.
Read 9 tweets
15 Sep
the reason we conduct experiments and collect evidence is so that we can reveal those who disagree with us to be cranks and fools, to threaten them with serious costs, penalties, and shame if they persist in the offense of disagreement.
this will be a controversial reading of how science works but i think it's basically correct.

the objective of experiments is to show as clearly as possible that objections *are untenable,* that a rational person *must* agree
that is, they carry with them the not-too-subtle threat that to the extent the new evidence is credible, disagreement marks a person as a cook or a crank of some kind
Read 28 tweets

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal Become our Patreon

Thank you for your support!

Follow Us on Twitter!

:(