The bottom line

On August 26, 2026, HHS released a commissioned study based on a preregistered survey of 1,208 U.S. adults. The authors introduced a five-item Clinical Gender Affirmation Scale (CGAS) and reported that higher scores were associated with higher left-wing authoritarianism scores and greater stated justification of six political-violence scenarios. HHS itself makes two critical limitations explicit: the study cannot establish causal direction, and it did not measure violent behavior or intent to commit violence.

What population did the study actually examine?

The analytic sample consisted of 1,208 adults recruited through CloudResearch Connect. Only 1.1% identified as transgender or gender nonconforming. That means the study is not evidence about whether transgender people are more violent. Its central variable was agreement with statements about supporting and affirming transgender people among a general adult sample.

This distinction matters because public debate can easily collapse “support for transgender people” into “transgender people themselves.” Methodologically, those are different variables.

Reliability is not the same as construct validity

The CGAS showed very high internal consistency (Cronbach’s alpha = .95). That is useful evidence that the five items moved together. It does not, by itself, establish that the scale validly measures a broad construct such as “gender ideology,” “radical gender ideology,” or a general extremist orientation.

Best-practice guidance for scale development typically requires evidence on content validity, dimensionality, reliability, convergent validity, discriminant validity and criterion validity. The HHS report presents internal-consistency evidence, but not a full psychometric validation program for the new scale.

There is also a substantive point: the five visible items focus on support, gender expression, chosen names, pronouns and gender exploration. They do not directly ask about surgery, hormones, puberty blockers or other specific medical interventions. Any broader label applied to the total score therefore requires independent justification.

Correlation, regression and causation are not interchangeable

The report found positive associations between CGAS scores and the justification of six political-violence scenarios. It also used ordinal regression models that adjusted for political identity. These are legitimate statistical analyses. They still do not turn a cross-sectional design into a causal one.

The report itself acknowledges several causal possibilities: endorsement of gender-affirming positions could influence authoritarian attitudes; people with stronger authoritarian attitudes could be more attracted to those positions; or an unmeasured third factor could affect both. The authors state that a follow-up experimental study is intended to address causal direction.

Justifying violence is not the same as committing violence

HHS explicitly states that the survey did not measure whether respondents had committed political violence or intended to do so. It measured how justified respondents considered six actions. Attitudes may matter, but attitudes, intentions and behavior are not the same outcome.

This distinction becomes especially important when findings are translated into policy narratives. Evidence about stated justification cannot be treated as direct evidence about dangerousness without an additional empirical bridge.

Outcome selection shapes what can be inferred

The six scenarios included killing Charlie Kirk, killing Donald Trump, killing Brian Thompson, burning ICE facilities, damaging wealthy people’s property and violence against police. Critics quoted by The Guardian noted the absence of politically symmetrical scenarios aimed at left-wing figures or transgender people. That does not invalidate the observed associations, but it limits how confidently the battery can be interpreted as a general measure of political-violence propensity.

Government communication is part of the scientific problem

HHS used language such as “radical gender ideology” and “sex-rejecting clinical beliefs.” When a government agency adopts a politically loaded label for a newly created scale, the wording can influence how the statistical result is understood by the public. Construct naming is not neutral when the label itself embeds an interpretation.

For public policy, the principle should be proportionality: the greater the potential impact of a government claim on rights, services and social legitimacy, the stronger the need for validation, replication, independent review and careful causal language.

Peer review and public-sector evidence

On September 7, The Guardian reported that the paper had not undergone external peer review before HHS released it. A coauthor said the team planned to submit it to a journal. Peer review does not guarantee truth, but it is an important independent quality-control step, especially when a contested finding is being promoted as official scientific evidence.

HHS also maintains an information-quality framework tied to federal peer-review standards for influential scientific information. Whether this specific report legally falls within those requirements is a separate administrative question, but the governance issue is legitimate: evidence used by government to frame a national controversy should face scrutiny proportionate to its likely public impact.

Eight questions for reading any official study about a social group

1. Who was actually studied?
2. What exactly did each variable measure?
3. Was the new scale validated, or only internally consistent?
4. Were outcome scenarios politically symmetrical?
5. Does the design establish temporal order?
6. Which confounders were considered?
7. Does the construct label describe the measure, or already interpret it?
8. Was independent review adequate for the claim’s public impact?

Conclusion

The strongest position is not to dismiss the HHS findings, nor to inflate them. The study found associations that merit further research. It did not demonstrate that transgender support causes authoritarianism or political violence, and it did not measure violent behavior. The scientifically responsible next steps are psychometric validation, balanced outcome measurement, independent replication and causal designs before the findings are used to justify broad public-policy claims.

Institutional authority can amplify a finding. It cannot convert correlation into causation.

Selected sources

1. U.S. Department of Health and Human Services. August 26, 2026. HHS release and commissioned report.
2. Fedida, O., Finkelstein, D., Wright, C., Jussim, L., & Finkelstein, J. (2026). Where the Clinic Meets the Movement.
3. Boateng et al. (2018). Best practices for developing and validating scales for health, social and behavioral research.
4. Methodological guidance on cross-sectional research designs.
5. The Guardian, September 7, 2026.
6. HHS Information Quality Peer Review framework.