Which states have the highest and lowest estimated transgender population in 2020?
Executive summary
California is routinely reported as the state with the largest estimated number of transgender residents by sheer count, while analyses that look at percentages for specific age groups show different leaders and laggards — for example, New York had the largest estimated share of transgender teenagers and Wyoming the smallest in the data reported around 2020 [1] [2] [3]. These headline rankings come with important methodological caveats: national and state estimates rely on combining multiple surveys and statistical modeling, and therefore ranks by count and by percent (and by age cohort) do not always match [4] [3].
1. California leads in absolute numbers; larger-population states dominate counts
Multiple outlets summarizing the Williams Institute modeling and related population estimates identify California as the state with the highest number of people estimated to be transgender — driven largely by its overall population size — with media accounts putting that number in the six-figure range [1] [2]. Reporting from data compilers and news outlets makes the straightforward point that states with larger total populations tend to have larger transgender populations in absolute terms, a pattern visible in the Williams Institute’s state-level tables and in secondary summaries [1] [4].
2. Different measures produce different “winners” — percentages, youth vs. adults
When researchers report percentages rather than raw counts, and when they split youth from adults, the state rankings shift; the New York Times highlighted that New York had the largest estimated share of transgender teenagers (3 percent) while Wyoming had the smallest (0.6 percent) in the available school-survey–based data for 2017–2020 [3]. The Williams Institute explicitly uses multiple sources — adult BRFSS surveys and youth YRBS surveys — and statistical modeling to produce separate state estimates, so a state that ranks high for youth share may not rank highest by total adult count [4] [3].
3. Methodology matters — phone surveys, school surveys and modeling underpin the estimates
The most-cited state estimates use Behavioral Risk Factor Surveillance System (BRFSS) phone-survey data for adults and Youth Risk Behavior Survey (YRBS) data for teenagers, combined with statistical modeling to fill gaps where states did not ask the question every year; the Williams Institute’s approach and the coverage in outlets such as Axios underline that these are modeled estimates rather than direct complete counts [4] [5] [3]. As a result, small-state rankings are particularly sensitive to sampling variability and modeling assumptions, and comparisons should be read in that light [4].
4. Lowest estimates vary by metric — teen percentages vs. adult counts
The clearest low-end statement supported in the reporting is that Wyoming had the lowest estimated share of transgender teenagers (0.6 percent) in the 2017–2020 school-survey data cited by The New York Times [3]. For lowest absolute counts across all ages, the publicly summarized materials in this set point to the general pattern that less-populous states will have the smallest numbers, but the specific state-by-state lowest adult-count assertions are not explicitly detailed in the provided snippets, and the Williams Institute tables referenced would be the direct source for those precise rankings [4] [6].
5. The political and reporting context shapes how rankings are used
Media and advocacy outlets often deploy state rankings to make policy and political arguments — for instance, emphasizing where protections are needed or where legislative attacks are concentrated — and those agendas can influence which metric (percentage vs. count; youth vs. adults) is foregrounded in a story [2] [7]. Readers should weigh the source (academic modeling at the Williams Institute versus secondary news or advocacy summaries) and remember that headline state “most/least” claims can mask complex age, racial and methodological heterogeneity in the underlying data [4] [3].