Researchers are working with sharply different estimates of how many LGBTQ+ adults live in the United States, while federal agencies are removing questions that help measure identity and inequality. The debate matters well beyond the headline numbers: without consistent, trusted data, it becomes harder to understand who is affected by policy, crime and gaps in services.

Essential Takeaways

  • The estimates vary: HRC puts the LGBTQ+ share of US adults at 12 per cent, compared with Gallup's 9 per cent and the Williams Institute's 5.5 per cent.
  • Methods shape results: Surveys differ in timing, questions and how they count responses, so their findings aren't directly interchangeable.
  • Federal questions are disappearing: A Williams Institute analysis found around 360 federal data collections lost at least one sexual-orientation or gender-identity measure during the administration's first year.
  • Missing data has real consequences: Researchers say losing consistent questions can make it harder to track disparities, including transgender people's experiences of crime victimisation.
  • Trust matters: Confidentiality and confidence in how information is used can influence whether people take part in surveys.

Why LGBTQ+ population estimates don't match

The figures are strikingly different, but they aren't necessarily competing answers to a single, simple question. The Human Rights Campaign's analysis estimates that 12 per cent of US adults are LGBTQ+, using Household Pulse Survey responses collected from 2021 to 2024. Gallup's 2025 estimate is 9 per cent, while the Williams Institute put the figure at 5.5 per cent using earlier data.

The gap is even wider in estimates of transgender and gender-expansive adults. HRC places that group at about 2.3 per cent of adults; Gallup's estimate for adults identifying as transgender is 1.1 per cent, and the Williams Institute's 2025 estimate is 0.8 per cent. The definitions, questions and periods used differ, so comparing the numbers needs a little care.

HRC's Charleigh Flohr describes Census Bureau data as a particularly strong source because households are selected through random sampling. Online responses may also make it easier for some people to answer personal questions. Still, participation isn't perfectly even, and decisions about skipped or uncertain answers can affect who appears in the final results.

Federal survey changes could leave lasting gaps

The dispute over estimates arrives as US federal agencies remove measures of sexual orientation and gender identity from data collection. A Williams Institute analysis found that about 360 federal data collections lost at least one such measure during the administration's first year. The removals affect more than population counts: governments, researchers and community organisations use federal statistics to understand the needs of different groups.

The consequences can be difficult to reverse quickly. Researchers rely on repeated, comparable surveys to spot changes over time and investigate disparities. Andrew Flores of UCLA School of Law's Williams Institute has used federal data to examine higher rates of crime victimisation among transgender people; without questions identifying transgender respondents, that comparison becomes harder to make.

The Williams Institute also reported that most of the removals it examined were made through administrative change requests that didn't require formal public notice or a public comment period. That speed contrasts with the careful testing normally involved in adding questions to a survey, and raises concerns about what information may no longer be collected or published.

Why the questions matter beyond a headline number

Population estimates help researchers put other findings into context. HRC uses demographic breakdowns to adjust survey results, while federal statistics are also used by public agencies, universities, journalists and non-profits. If the underlying data become patchy, even well-designed research can lose some of its ability to show who is being reached and who is being missed.

There are also questions about information already gathered. Erin Meyer of the Inter-university Consortium for Political and Social Research has raised concerns about education data on students whose gender was recorded outside male and female categories, saying the figures weren't included in released data. It's a reminder that data policy isn't only about what future surveys ask; it can also shape what the public gets to see.

For anyone reading a population estimate, the most useful checks are straightforward: look at when the data were collected, how the survey defined LGBTQ+ identity, and whether respondents could skip questions. Those details won't settle every disagreement, but they make comparisons fairer and the findings easier to interpret.

Trust is part of the counting process

People are more likely to answer sensitive questions when they believe their information will be kept confidential and used responsibly. That confidence can be particularly fragile when policies and official practices change quickly. A survey may have robust privacy protections on paper, but people still need to feel safe enough to respond honestly.

Researchers have warned that rebuilding a consistent record would take time even if the questions returned. In the meantime, lost years of comparable information could leave gaps that future studies can't simply fill by starting again. That's why the quality of a count depends not only on the question itself, but also on the trust behind it.

LGBTQ+ people are part of everyday life, from classrooms and workplaces to clinics and public services. Better data won't explain every experience, but it can help make those experiences visible when decisions are made.

A more complete picture starts with asking carefully, and keeping the answers safe.

Source Reference Map

Story idea inspired by: [1]

Sources by paragraph: