This article is the long, in-depth version of the piece we wrote for Queer Majority, diving deeper into the data and exploring additional insights from our study.
Recently, a Gallup poll came out that compares levels of LGBTQ identification in younger versus older generations. The poll showed that the numbers have roughly doubled with each generation - 2.3% of Baby Boomers, born 1956-1964, identify as LGBTQ; 4.5% of Generation X (1965-1980); 9.8% of Millenials (1981-1996).
In Generation Z, born 1997-2005, these numbers continue to skyrocket. Gallup’s poll showed that 22.3% of Generation Z identifies as LGBTQ - a gain of more than 12 percentage points over the previous generation.
What could be causing this rise? Could it be mainly due to identification with non-traditional gender identities? Or be primarily more people identifying as gay or lesbian? Nope.
When Gallup broke down the data, they found that the numbers for gay, lesbian, and transgender identification increased by small amounts (in terms of total number of people). For instance, in Gen X 1.3% of people identified as gay; in Millenials, 1.6%; in Gen Z, 2.6%.
But the number of people identifying as bisexual (i.e., attracted to more than one gender), has had a meteoric rise. 1.9% of Gen X’ers consider themselves bi. That triples to 5.9% of Millennials, and almost triples again for Gen Z, to 15.3%.
This presents a mystery: why are so many people now identifying as bisexual?
Politically, those on the right have tended to argue that increases in these kinds of identities are a social fad - leftists jumping on the bandwagon for social clout, or a chance to feel special and wave a flag. Those on the left, on the other hand, have a tendency to argue that these identities are fundamental. The rising numbers of people identifying with non-heteronormative identities are just a positive sign of greater acceptance: that in the younger generations, in large part because of their tolerance of and protections for identity, there is more space for people to be who they are.
There is also a third possibility: could it be that people are genuinely becoming more bisexual over time?
We ran a study to find out. We polled 608 respondents from different generations (using our participant recruitment platform Positly, as well as social media) to understand why it is that bisexuality has risen so much. We used our data to test three different theories, or as we call them, “bi-potheses”:
Social Bandwagoning - People might be identifying as bisexual because it's trendy or they think it’s a cool thing to be. If this is true, younger generations would show a much larger rise in their level of identification with bisexuality than their levels of bisexual attraction.
Increasing Acceptance - People might be feeling more free to identify themselves as something they already are. If this is true, we should find high levels of bisexual attraction consistently across all the generations, but a spike in self-identification in the younger generation.
Increasing Bisexual Feelings - People might be becoming more bisexual. If this is true, we would find a rise in attraction, identity, and behavior at the same time (without identity rising disproportionately). If this is the case, it would raise more questions about why bisexual feelings are on the rise.
Based on your experience, which of the above do you think is true about the rise of bisexuality? Or, do you have a hypothesis of your own? We recommend taking a moment to consider what you believe, before we show you what our study found.
Our eight measures of bisexuality
BBisexuality is an inherently fuzzy concept. It means different things to different people and has meant different things at different times. As such, we tried to capture as much of that complexity as we could by measuring bisexuality for our study using 8 different definitions simultaneously. In all cases we eliminated any respondents who identified as gay or lesbian, to be sure we were testing only for bisexuality. We also eliminated respondents who had never had a sexual partner, since not all definitions apply to them. Here are the eight definitions of bisexuality we applied:
Self-identified Bisexual - this is simply whether they assign the "bisexual" label to themself.
Sometimes Same Sex Attracted (1-99%) - this means they report that at least 1% of people they feel sexually attraction toward are men, and at least 1% of people they feel sexual attraction to are women. This is a very low bar since even a small amount of attraction to their own sex would land them in this bucket.
Fairly Often Same Sex Attracted (10-90%) - this raises the bar a bit on what it means to be bisexual, requiring that at least 10% of the people they feel sexual attraction to are men, and at least 10% are women
Very Often Same Sex Attracted (25-75%) - this is an even more stringent definition of bisexual feelings, requiring that at least 25% of the people they are sexually attracted to are men, and at least 25% are women.
Attracted to Same-Sex Body Parts - this counts a person as bisexual if they are a man who is sexually attracted to penises or if they are a woman who is sexually attracted to vaginas or breasts.
Wants Future Same-Gender Sex Encounter - this definition considers a person bisexual if they want to have a same gender sexual encounter in the future
Most-Recent Partner Is of Same Gender - this definition requires that their current (or, if currently single, most recent) sexual partner was of their same gender
Has Had Same-Gender Sex - finally, this definition simply requires that they have ever had sex with someone of the same gender
Each of these definitions is binary, in the sense that it assigns each study participant either a 1 (they are bisexual by that definition) or a 0 (they aren't).
Every one of these definitions has its own limitations. There is no one "true" definition of bisexuality, hence the need for multiple measures to tease apart what is really changing across the generations.
In addition to the bisexuality questions, we asked each person about their gender and political affiliation to test whether bisexuality is related to either.
The study was anonymous, conducted digitally, and respondents did not know the research questions we were testing. Thus, while lying was possible (e.g., out of embarrassment) there was no incentive to lie - so we expect that the vast majority of respondents were answering honestly. Of course, if people are lying in greatly different proportions in different generations that would distort our findings, but given the anonymity and lack of reason to lie we think that's an unlikely explanation for our results.
A note before we present our findings. Some of our questions ask about attraction to the same sex, and others about attraction to the same gender. This might lead to situations where, for instance, members of a same-sex partnership who identify as genders other than those assigned to them at birth might not mark their partnership in the way the study intended. However, this discrepancy is unlikely to change the conclusions of the research because cases like these reflect a very small fraction of all responses.
Findings
Let’s look at what the study found. First of all, if we construct a composite measure, measuring the average of our eight definitions, we get a result that looks like Gallup’s findings. Each generation does indeed seem to be more bisexual than the ones before. In fact, using our composite measure, we find even higher rates of bisexuality than in Gallup’s poll! This makes sense, because many of our criteria are more inclusive than self-identification, which is what Gallup used - as well as for another reason, which we'll mention in a moment.
Note that the aggregate findings of increases across the generations still hold up if we apply statistical controls (controlling for gender, political views, and the source of study participants to adjust for potential biases in our data) - except for the difference between millennials and Gen Z which becomes small (and potentially non-existent).
Aggregate data doesn’t tell us much about what contributes to this rise. Let’s break it down by each definition.
Bisexual Identity
For bisexual identity, our trend closely matches the results of Gallup, but our respondents showed even higher rates than Gallup’s did: 3% for Baby Boomers (vs. 0.6%), 8% for Gen X (vs. 2%), 17% for Millennials (vs. 6%), and 23% for Gen Z (vs. 15%). As we'll see, the higher number we received may have to do with a progressive skew in our data - whereas Gallup attempts to achieve a nationally representative U.S. sample, ours is somewhat more likely to be progressive than random Americans.
Bisexual Attraction
We asked respondents to rate what percent of the people they find sexually attractive are the same sex as them vs. are the 'opposite' sex as them. When we measured the results, we set cutoffs at 1%, 10%, and 25% levels of attraction to the same sex, plus measuring people’s attraction to same-sex body parts.
Amazingly, almost 60% of Gen Z said at least 1% of people they are sexually attracted to are of the same sex. About 40% of respondents in older generations said they were attracted to the same sex with at least this frequency.
This number lowers if we consider people attracted to their own sex more often. About a third of Gen Z said that at least 10% of people they are attracted to are of their sex, and nearly a quarter said at least 25% of people they are attracted to are of their sex. The mean of previous generations is 17% and 6%, respectively.
Another (very different) way to investigate sexual attraction is to ask what body parts people find sexually attractive. Somewhat similar to these numbers, 26% of Gen Z and 14% of older-generation respondents across the spectrum say they are attracted to the body parts of their same sex.
If you take the mean of non-Gen Z generations, each of these attraction-based ratings has increased by at least 53% between previous generations and Gen Z. The number of people identifying as attracted to the same sex 25% or more of the time has increased a whopping 280%.
Testing the Bi-potheses
So, based on these results, what can we conclude about the three hypotheses for why bisexuality is increasing so much? That is, which bi-pothesis is true?
If social bandwagoning explained the massive increases in bisexual identity, we would see bisexual identity going up, but the other numbers would show significantly less change. People might say they were bisexual more often, but they wouldn’t rate higher on attraction to their own sex or the body parts of their own sex compared to earlier generations. Since we are seeing much higher levels of sexual attraction (and not just identity), it suggests the increasing bisexuality is unlikely to just be a bandwagon effect.
If increasing acceptance is the reason for this huge increase in bisexual identity, the rise in self-identification in younger generations would be much steeper than the rise in attraction. As we can see in the chart above, 23% of Gen Z respondents self-identified as bisexual, vs a 9% mean of other generations (a 146% rise). Attraction to the same sex at least 10% of the time went from 17% to 32% (an 88% rise), and attraction at least 25% of the time went from 6% to 23% (a 280% rise). Lastly, attraction to same-sex bodyparts went from 14% to 29% between generations (a rise of 98%). Thus, self-identity has indeed risen between generations, but not especially more than attraction - both have risen a great deal.
If increasing bisexual feelings is the reason for the rise in bisexual identity, we’d see a steep rise in both attraction and identity. Remarkably, we do! The numbers suggest that more people are identifying as bisexual because more people actually feel attracted to the same sex.
So this data suggests a resolution to the mystery! Younger people today appear to be having a lot more bisexual feelings than older generations, and hence more bisexual identification is exactly what we'd expect.
But, if this were really true, wouldn't we also see more bisexual behavior among young people?
Behavior
The answer is…mostly.
Of all the metrics we measured, we received the lowest overall numbers for the question of whether respondents’ most recent partner was or is of their same gender. This makes sense if you consider that a bisexual person is likely to have a partner of either gender at the time of assessment.
However, compared to previous generations, Gen Z was still more than twice as likely to have their most recent partner be of the same gender.
Interestingly, this is one of the only two metrics that doesn’t rise across all generations. Baby boomers are slightly more likely to be living with a partner of the same gender than are Gen X or Millennials, although Gen Z again eclipses them by orders of magnitude.
The other measure that doesn’t consistently rise is the proportion of each generation who has had sex with someone of the same gender. We see an almost linear curve of same-gender sex through Baby Boomer, Gen X, and Millenial generations. In fact, more people of those generations have had sex with the same gender than the number of people who identify as strongly attracted to people of the same sex!
Why does the curve dip down when we get to Gen Z? Does this disprove the hypothesis that bisexual attraction is rising about as fast as self-identification?
Probably not. Consider the fact that the oldest Gen Z’ers are 27 years old, and the youngest are only 12. They haven’t had as much time to explore their sexuality as the older generations, so it makes sense that they are less likely to have had a same gender partner, because they've simply had fewer partners.
Finally, it's time to look at the last set of correlations. How do these results change across self-identified gender and political orientation? Get ready for some impressive variation!
Bisexuality and Gender
While designing this study, our founder Spencer Greenberg posted this poll and this poll on X (formerly Twitter), asking people to estimate what proportion of men vs women are bisexual. Most commonly, respondents thought that 5-10% of women and 5% or less of men are bisexual.
Respondents were right that more women than men identify as bisexual, 29% and 12% respectively if we take the average of all eight of our metrics. (The number of non-binary people responding to the survey was too small to make meaningful measurements for that group.) But, the average of our metrics is approximately three times higher than the most common responses given in the polls!
Bisexuality and Politics
Our study looked at one last metric. How does bisexuality vary across the political spectrum?
The results here are interesting. While about 27% of people left of center identified as bisexual by an average of all eight of our metrics, only about 10% of those right of center identified the same way. This held true across most individual metrics: attraction, identity, and behavior all decreased as conservativeness rose. The only metric that showed a different pattern was the number of people currently with a same-sex partner, which stayed constant across the political spectrum.
Are conservatives less likely to explore their sexual preferences? Less likely to admit to them? Do they marry younger to partners of the 'opposite' sex? More research is needed here.
In Summary
Our study suggests that bisexuality actually seems to have increased across generations - not merely identification, but actual bisexual attraction as well. By measures of attraction, identity, and behavior, people are more bisexual today than they were back in the past. Why? It's hard to say for certain - perhaps greater social acceptance of bisexuality actually leads to more bisexual attraction (or the reverse: lack of social acceptance, or shame around such attraction, might lead to people nipping in the bud such sexual attraction and not exploring it, so it doesn't end up developing further or growing). But, that's just one hypothesis. If you have your hypotheses as to why we found these results, we’d love to hear them!
Appendix Charts
All Results
Gender
Political Orientation
Testing for confounding variables
This analysis was done to make sure that our results were not due to confounding variables (generation, conservatism, gender)...
Note for table above: Data is for 608 participants. Each row shows one linear regression predicting one of our 8 bisexuality measures, using as independent variables gender, conservativeness, and isPositly variables (indicating if the participant was recruited via our Positly.com recruitment platform - if not, they were volunteers from the internet).
Generations were one-hot encoded with Baby Boomers as the reference group (i.e., anyone not in the other three generations is a Baby Boomer). The analysis excludes those who’ve never had a romantic partner before as well as 13 participants who are missing "politicalSelfPositionConservatism". All variables — dependent and independent — were normalized (mean = 0; stdev 1), resulting in an intercept of 0 for all regressions.
Residual Plots
Note: The residuals shown in these plots represent how much more or less bisexual each participant is compared to what would be predicted based solely on their political views, gender, and data source (Positly vs. online volunteer). We calculated these residuals by running linear regressions for each bisexuality definition, using political conservatism, gender (man/woman), and data source as predictors, then subtracting the predicted values from the actual values. Points above 0 indicate participants who are more bisexual than predicted by these factors alone, while points below 0 indicate those who are less bisexual than predicted.