Americans Have Moved On From Transgender Culture War Issues, But Politicians Have Not

Austin Botelho
Cybersecurity for Democracy
8 min readAug 17, 2023

Transgender Rights, as well as social acceptance of Transgender Americans, has been an ongoing topic of conversation in the Republican Primary this year. To understand how ordinary Americans were engaging with news and political content on this topic, we used CrowdTangle to measure user engagement with news content on Facebook and Instagram, as well as data from Meta’s Ad Library to measure impressions of ads about Transgender Rights. We found that organic interest in the topic spiked around the Bud Light controversy, but decreased since late April, while political ad spending has grown and remains high.

Transgender Americans are at the forefront of a contested political discourse that often veers into the dehumanizing. Pew found that, as of June 2022, the United States public was divided on whether society has gone too far in accepting transgender people: 38% agreed, 36% disagreed, and 23% thought society was adequately accepting. As a consequence, in 2023, 79 anti-trans bills were passed. To better understand how this topic is discussed, we studied content about this issue in the US news ecosystem on Facebook and in paid ads on Facebook and Instagram between April 1, 2023 and July 1, 2023. (For full methodological details, see the Methodology Note at the end of this post.)

We found that:

  • There was a spike in organic interest in transgender issues during April around the Bud Light Boycott controversy.
  • Right-leaning news sources had 4x the engagement with content about transgender issues than left-leaning sources, and right-wing political advertisers spent twice as much on the topic.
  • Significant paid engagement on this topic (i.e., ads) did not begin until May, peaking in late May but continuing at elevated levels throughout June and July.
  • Our prior research has found that sources with a reputation for spreading misinformation get more engagement, a feature we have called a ‘misinformation premium’. On this topic, this feature was even more pronounced; the misinformation engagement premium was 33% higher than normal and more pronounced on the political fringes.
  • Despite significant spending on this topic since May, organic interest (i.e., organic engagement with content on this topic) has not returned, remaining similar to the pre-April 2023 baseline.

“Political sugar”: April Spike in interest short lived

The conversation around transgender issues loomed large in television broadcasts as well as online. Over three months from April 1, 2023 to July 1, 2023, they were mentioned 3.5k times, or 38 per day, across eight national broadcasting news channels. On Meta, 2,702 unique media pages posted 32,503 times gaining 3,450,397 engagements (any of likes, shares, comments, or other reactions) while 359 advertisers spent $303,207.5 to earn 22,738,517 impressions on 1,985 ads, or 59.5 per dollar.

full data

Both Television and organic interest was heavily concentrated in April and declined over the remaining observation period. Television coverage had a smaller uptick again in June for Pride month.

A primary driver of the elevated April engagement on Facebook was the Bud Light Boycott controversy. On April 1, 2023, transgender influencer Dylan Mulvaney shared a sponsored post advertising Bud Light. This prompted a conservative backlash calling to boycott the brand. There are 4,590 (14%) posts mentioning this event in the dataset comprising nearly a third (31%) of all engagement, mostly concentrated early April.

The decline in overall engagement was primarily driven by a decline in per-post engagement. The average per-post engagement in April was 94% higher than average overall news engagement, but by June it was 13% lower. (Left-leaning misinformation bucks the trend, but does not represent a large portion of the data).

The decline was most pronounced for sources without a reputation for misinformation. To further interrogate this trend, we pulled an additional two weeks of data before and after the original set to examine the full breadth of data we had available at the time of analysis. For each partisanship-misinformation combination, we calculated the Spearman correlation to see whether average engagement significantly decreased over time. After a Bonferroni correction, left-no (r = -0.82 , p < 0.001), right-no (r = -0.60 , p = 0.005), and center-no (r = -0.58 , p = 0.007) source engagement all had significantly negative correlations with time.

Conservative fixation

Across all of the data sources examined — broadcast news, paid digital, organic digital — conservative voices out-perform others. On Facebook and Instagram, conservative advertisers spent more than twice ($176k vs $79k) as much as liberal ones on the matter. They tended to sponsor fewer, big money ads compared to liberal advertisers. This earned them 70% more impressions (12.4m vs 7.1m). However, liberal advertisers were charged a lower dollar per impression rate, a roughly 21% discount. Conservative-Liberal differences were even wider in organic content. Conservatives posted 10% more resulting in 4x as many engagements. Specifically, far right sources were engaged with more than all other known sources combined.

All of the top ten pages by total organic engagement were conservative. America First News, a news show hosted by former Newsmax pundit Grant Stinchfield, posted 3x more (1,266 times) than the next most prolific page.

Misinformation thrives

Misinformation impacts engagement even more than partisanship. Posts from sources with a history of misinformation garnered 5.7x more engagement on average (438 vs 77).

The misinformation premium is larger for left-leaning posts. Users engaged with posts from misinformation-spreading right-leaning sources 2.4x more than posts from their higher quality right-leaning counterparts. On the left, this gap is 6.2x. That is because left-leaning factually reliable sources are vastly less engaged with than other partisan sources.

Cybersecurity for Democracy has previously reported this misinformation premium before on Facebook. To see if our results are an artifact of a bias toward misinformation in general or transgender issues being a particularly fertile ground, we compared it to a baseline dataset for the same time period. The misinformation premium was 33% higher for transgender issues (5.7x vs 4.3x) than the overall baseline. This premium was larger on the political fringes: 16% higher on the far right (2.3x vs 1.98x) and 180% higher on the far left (8.1x vs 4.5x).

Political ads failed to generate organic attention

In response to the April spike in interest in transgender issues, political advertisers pumped money into ads in May.

While this induced higher ad impressions, organic engagements continued to decline.

Conclusion

Conservatives have latched onto transgender issues as a cultural wedge that animates their base. Across broadcast news, paid, and organic Facebook content, they more often referenced, funded, and engaged with them. This attention asymmetry is worrisome for trans-acceptance especially as it interacts strongly with factuality to promote right-wing misinformation. Despite organic engagement with transgender issues declining from its April high, political advertisers unsuccessfully pumped money into trans-focused ads.

Methodology

We collected three months of English-language Meta data from April 1, 2023 to July 1, 2023 mentioning trans, transgender, transition gender, biological sex, basic biology, biological male, biological female and no other words starting with trans (e.g transition, transportation, transit, transform, transparency). The paid dataset contained ads while the organic dataset contained posts. The baseline dataset was a random sample of 1% of the total English-language Meta news volume for the same time period. Television news coverage mentioning the word “transgender” was collected over the same time period from the The GDELT Project.

We labeled the partisanship of all advertisers airing more than five ads in the ads dataset and the top 100 pages by post and engagement count in the post dataset. Where available, we used the Media Bias Fact Check or Newsguard ratings. Following this methodology, we ascertained which pages in the organic dataset have previously promoted misinformation. In total, more than 74% of the data and 87% of the total engagement and impressions received ratings.

The data and code to replicate these findings are here.

Appendix

Broadcast media mentions of transgender topics

Conservative broadcasting channels (Fox News and Fox Business) contributed 40% of the mentions followed by centrist ones at 32% (CSPAN, CSPAN2, and CSPAN3) and liberal ones at 28% (CNN, MSBNC, and CNBC). Fox News, in particular, concentrated on the topic mentioning it more than MSNBC and CNN combined. CSPAN ranked highly because of its coverage of the Hearing on Transgender Rights and the House Rules Debate on Transgender Athletes.

Partisan transgender topic ad examples

Conservative broadcaster PragerU spent an additional $55,000 on May 21st alone on an ad campaign for their children’s content.

Many conservative ads, like the one below, depict the existence of transgender people as a liberal “ideology”. The implication is that it is a fabricated phenomenon being forced upon people, especially children. This narrative is pushed despite the medical community’s overwhelming support of evidenced-based gender-affirming care for youth.

Example of a conservative ad on transgender issues

Liberal ads, by contrast, often tried to rally support against anti-trans legislation like the one from Amnesty International below.

Example of a liberal ad on transgender issues

Misinformation premium for trans topic

We ran Negative Binomial regressions controlling for subscriber count with and without a partisanship-misinformation interaction term to isolate the impact of both misinformation and partisanship. The most parsimonious model according to the deviance measures (model 1 below) finds that all included regressors, except for left partisanship, are statistically significant. Misinformation provides a 47% boost in engagement holding all other factors constant while the boost from being right-leaning is 201%.

About NYU Cybersecurity for Democracy

Cybersecurity for Democracy is a research-based, nonpartisan, and independent effort to expose online threats to our social fabric — and recommend how to counter them. It is a part of the Center for Cybersecurity at the NYU Tandon School of Engineering.

Would you like more information on our work? Visit Cybersecurity for Democracy online and see how tools, data, investigations, and analysis are fueling efforts toward platform accountability.

--

--