Disinformation Report: July 1-15
Updated: Oct 4
How Anti-Trans Attacks Are Landing Across TFA’s Priority States — and How Campaigns Should Answer
The loudest attack in any priority state this period was a manufactured one: Libs of TikTok‘s claim that a Wisconsin Democratic campaign staffer had called for violence, folded into the governor’s race within a day. The period’s biggest opening arrived on its final day, when Defense Secretary Pete Hegseth announced testosterone screening and replacement therapy for troops — from an administration that barred trans service members — and handed Democrats their only reach-winning day of the month. This briefing maps how the period’s attacks landed across the thirteen states TFA is tracking: who drove each story, who it reached, and which counter-messages test best now.
July 1 – July 15, 2026
Data: Trans Rhetoric Social Media Dashboard
77K+ channels | 20 platforms | 984K+ posts
13 priority states | 665 state-tagged posts | 3.5M engagements | c4 / Organize Action
Topline Finding
The map inverted this period. Last cycle a single national event — the Supreme Court’s sports ruling — localized into every priority race at once; this cycle the trans conversation ran overwhelmingly national and cultural (the response to Elliot Page’s role in The Odyssey, and the July 15 Pentagon hormone story), and mostly bypassed the states. Priority-state reach thinned across the board, and what remained was concentrated: Texas again carried the most volume, with the Talarico Senate race the only live candidate fight on trans terrain; Wisconsin took the second-highest reach in the field on the strength of a single manufactured story (a Libs of TikTok claim about a Democratic campaign staffer, 236K); and New Hampshire produced the period’s one concrete state-policy move, Gov. Kelly Ayotte signing SB 430, the school gender-identity disclosure law. The modeled audience shifted with the content mix: anti-trans state content this period reached a roughly gender-balanced audience (51% female), not June’s young-male skew — the entertainment-driven month pulled a broader crowd into the conversation. The opening is the final day: the Pentagon testosterone announcement and the CHIPS Act remark gave Democrats their only reach-winning day of the period, on a hypocrisy contrast no campaign yet owns.
Source: Trans Rhetoric Dashboard | issue-location layer, off-topic virality removed | reach labeled for states above 100k
The nation’s heaviest reach sat outside the priority list — New York (the Odyssey premiere), Louisiana (the Chili’s pronoun-firing story), and South Carolina (the Lindsey Graham allegation) — while priority-state reach thinned: the fight ran national, not local.
Source: Trans Rhetoric Dashboard | issue-location layer, off-topic virality removed
Reach is concentrated in a handful of states — and this period even the leaders are thin. A quiet map is the moment to pre-position, not to relax.
Key Findings
The fight ran national-cultural, not state-electoral. The period’s dominant trans stories — the Odyssey casting response and the July 15 Pentagon hormone announcement — played out as national culture war, and priority-state reach thinned across the board. A quiet state map is the window to pre-position responses before the August primaries.
Wisconsin’s spike was one manufactured story. Libs of TikTok’s claim that a Democratic transgender campaign staffer had called for violence against Republicans (236K) carried nearly all of the state’s reach, and Tom Tiffany folded a girls’-sports pledge into the governor’s race the same news cycle. The staffer-association attack — a campaign’s own hire as the villain — is the format evolution to watch.
New Hampshire made the period’s one concrete policy move. Gov. Kelly Ayotte signed SB 430, requiring school employees to answer parents’ written questions about a child’s gender identity within 10 business days — overriding a 2024 state supreme court privacy ruling. The schools/parents axis is now statute in a state with Senate and Governor races.
Texas is the only state where the pro-trans side is contesting at scale. The Talarico Senate race kept Texas the highest-volume state, and stance there split nearly even (0.44M anti / 0.40M pro) — attacks on Talarico’s church alongside teacher-testimony counter-content. Everywhere else the pro-trans side is thin or absent.
July 15 handed Democrats their only reach-winning day. The Pentagon’s testosterone-screening announcement and the CHIPS Act remark — both July 15 — drove the only day this period that left-leaning accounts nationally out-engaged right-leaning ones (12.1M to 9.1M). The hypocrisy contrast (banning trans care while funding hormone optimization for troops) is repeatable, and no campaign owns it. A companion memo covers the announcement.
The audience shifted — this period’s attacks reached a balanced crowd, not young men. The modeled audience of anti-trans priority-state content was ~51% female this period, against June’s 78% male — the entertainment-heavy mix pulled a broader audience into the conversation. Audience composition tracks the content mix; re-check it each cycle rather than building for last cycle’s crowd.
The message-testing picture flipped with the narrative. On this period’s dominant healthcare narrative (federal subpoenas of trans teens’ medical records), the empathetic register now tests best (58.3%), “politicians in Washington” is the strongest antagonist (66.8%), and naming the criminalization of doctors adds the largest lift of any ingredient (+29). The June playbook (practical register, “the government”) is stale.
North Carolina previews the Cooper attack. The NCGOP and national party accounts ran Roy Cooper’s vetoes of sports and minors’-care bills as a Senate-race frame — low volume, but disciplined and on-message. The sports attack on Cooper is built and waiting for scale.
Rep. Juan Ciscomani is now posting the attacks himself. Per the standing tracking request: Ciscomani (AZ-06) personally posted a girls’-sports attack on Democratic challenger JoAnna Mendoza (July 7, ~10K organic engagement). No sign of paid-scale amplification in this window’s organic data; the watchlist flags his account on every new post.
The Battleground at a Glance
Two layers drive this briefing. The issue-location layer tags which state each post is about; the account-location layer tags where the messenger is based. A post can be about a state without originating there (a national account attacking Roy Cooper counts toward North Carolina), so the reach figures below measure attention on a state, not local activity alone. Off-topic virality that keyword-matched a state (foreign-language tabloid items, unrelated news) has been removed by hand.
State | Primary | Posts | Reach | vs last | Anti / Pro | Top messenger |
Texas | TBD | 162 | 1.4M | ▼ 28% | 445K / 398K | BuzzFeed |
Maine | 6/9/26 | 106 | 709K | ▲ 14% | 18K / 134K | Libby Emmons |
Wisconsin | 8/11/26 | 99 | 432K | ▼ 17% | 72K / 5K | Libs of TikTok |
Georgia | 5/19/26 | 16 | 351K | ▼ 12% | 2K / 339K | 👉 M-Û-R-Č-H 👈 |
New Hampshire | 9/8/26 | 75 | 182K | ▼ 66% | 52K / 69K | Michael Knowles |
Michigan | 8/4/26 | 39 | 118K | ▼ 75% | 20K / 8K | Amy Mek |
Minnesota | 8/11/26 | 61 | 94K | ▼ 92% | 30K / 42K | Megyn Kelly |
Kansas | TBD | 18 | 77K | ▼ 71% | 0 / 1K | Reuters |
Pennsylvania | 5/19/26 | 27 | 72K | ▼ 92% | 7K / 2K | Reuters |
Arizona | 8/4/26 | 51 | 67K | ▼ 58% | 9K / 26K | Owen Gregorian |
Nevada | 6/9/26 | 23 | 23K | ▼ 40% | 1K / 170 | Jesse Singal |
North Carolina | 3/3/26 | 8 | 12K | ▼ 66% | 1K / 275 | NCGOP |
Iowa | TBD | 5 | 2K | ▼ 90% | 0 / 1K | Matthew Yglesias |
“vs last” compares reach with the previous reporting period (≈ means within ±10%). A rising state deserves attention before its number is large.
Source: Trans Rhetoric Dashboard | stance classifier; "neutral" includes hostile-leaning news coverage
Tone is a directional read from the stance classifier; “neutral” absorbs a large amount of hostile-leaning news coverage, so anti-trans shares are conservative — Wisconsin’s largest attack post, for example, is stance-coded neutral, so its hostile share understates. Georgia’s favorable read rests on a single viral post and should not be read as a trend. Texas is the one state with a genuinely contested stance split.
Who’s Driving the Conversation
Right-Leaning Drivers
Libs of TikTok — 5.8M downstream reach, 21 seeded stories
Again the clearest originator in the field — the Wisconsin staffer story and the Louisiana pronoun-firing story both surfaced here first, then traveled to officials and national accounts. The point in the chain where a state story can still be anticipated.
Tom Tiffany — Wisconsin governor’s race
Converted the staffer story’s news cycle into a girls’-sports campaign pledge within a day — the cleanest example this period of an attack moving from aggregator to candidate.
NCGOP and national party accounts — Cooper veto record
Ran disciplined, repeated framing of Roy Cooper’s vetoes (sports, minors’ care, parental rights) as Senate-race material — low reach now, but the attack is built and waiting for ad spend.
Rep. Juan Ciscomani — AZ-06, direct attacks
Now posting girls’-sports attacks on challenger JoAnna Mendoza from his own account (per the standing tracking request). Organic reach is modest; the thing to watch is paid rotation.
Left-Leaning Drivers
Rep. Pramila Jayapal — 784K engagements, 8.73 performance score
The highest-scoring elected official on the July 15 Pentagon story — framing the hormone announcement as hypocrisy from an administration that barred trans troops. An elected official carrying a clean contrast remains the left’s highest-leverage messenger type.
Aaron Rupar — 2.0M engagements, 4.03 score
His clip of the CHIPS Act remark (1.6M) anchored the fact-check wave — the simple, shareable correction that let everyone else pile on.
PinkNews — 2.3M engagements, 0.21 score
Carried the New Hampshire disclosure-law story to a national LGBTQ+ audience — but its heavy trans posting badly underperformed its own baseline (0.21), a volume floor rather than a resonance driver.
How the Attacks Travel — Originators vs. Amplifiers
Two different jobs create anti-trans reach. A small set of accounts originate stories — surfacing a local incident that then spreads; a larger set amplify, putting the most eyeballs on stories someone else started. The two call for different campaign responses: pre-empt originators, rapid-respond to amplifiers.
Top originators (seed stories that travel): Libs of TikTok (5.8M, 21 seed stories); E! News (3.5M, 4 seed stories); Aaron Rupar (2.2M, 4 seed stories); Fox News (2.1M, 10 seed stories); Benny Johnson (1.7M, 5 seed stories).
Top amplifiers (most reach on others’ stories): E! News (2.1M, 4 reposts); Tommy Robinson 🇬🇧 (1.2M, 1 repost); Congresswoman Pramila Jayapal (739K, 2 reposts); Entertainment Tonight (715K, 2 reposts).
Who These Attacks Reach
The dashboard has no engager-level data, so audience is modeled — inferred from the demographics of the accounts that produce and amplify content, weighted by engagement, under a homophily assumption. It blends US Census surname data, Social Security first-name records, each platform’s known user base, and where accounts are based. For the roughly 57% of anti-trans reach carried by accounts with a real name or seed match, this is a genuine read; the rest falls back to a national baseline. Treat it as directional, not measured — but the split between the two sides is stark enough to act on.
Audience of… | Gender | Top age cohorts | Race (modeled) | Evidence | Conf. |
Anti-trans attacks | 51% female / 49% male | Millennial 32% · Gen Z 26% · Gen X 23% | White 71% · Hispanic 15% · Black 9% | 34% | 0.49 |
Pro-trans content | 59% female / 41% male | Millennial 36% · Gen Z 28% · Gen X 22% | White 76% · Black 9% · Hispanic 9% | 49% | 0.50 |
All state content | 52% female / 48% male | Millennial 36% · Gen X 24% · Gen Z 22% | White 79% · Hispanic 9% · Black 8% | 57% | 0.55 |
The headline this period is the change: anti-trans state content reached a roughly gender-balanced audience (51% female / 49% male) — a sharp shift from June’s 78%-male skew. The period’s content mix explains it: an entertainment-led month (the Odyssey story, carried by celebrity outlets) pulled a broader, more female, slightly older audience into the same conversation the attacks run in. Two implications: audience composition tracks the content mix and should be re-checked each cycle rather than assumed; and this period’s attacks were reaching persuadable general audiences, not a young-male niche — which raises, not lowers, the stakes of answering them. “Evidence” is the share of reach from accounts with a real name/seed match rather than a national-average fallback; “Conf.” is the mean modeled confidence. Confidence is lower this period (0.49–0.55) than last — read the split directionally. The racial estimate is the weakest layer and should be read most loosely.
Case Studies
Case Study 1: Wisconsin — The Staffer as the Villain
What happened. Nearly all of Wisconsin’s reach — the second-highest in the field — came from one story: Libs of TikTok claimed a Democratic transgender campaign staffer had called for violence against Republicans (236K). The claim is the account’s characterization and is not verified here. Within the same news cycle, local conservative media alleged a “blackout” of the story, and Tom Tiffany — running for governor — posted a girls’-sports-and-spaces pledge that rode the wave without naming it.
Why it matters. This is the individual-villain playbook with a sharper electoral edge: the villain is a campaign’s own hire, which converts a smear into a guilt-by-association attack on the candidate and makes every Democratic campaign’s staff roster a target surface. It is cheap, it requires no policy hook, and the aggregator-to-candidate handoff took under a day.
Takeaway. Campaigns need two things before this lands on them: a rapid factual posture for staff-based smears (verify, respond to the person and the facts, never the population), and the discipline not to let an unverified aggregator claim set the terms of a news cycle.
Case Study 2: New Hampshire — The Disclosure Law as Terrain
What happened. Gov. Kelly Ayotte signed SB 430, requiring public-school employees to answer parents’ written questions about a child’s gender identity “completely and honestly” within 10 business days — overriding a 2024 state supreme court ruling that had upheld a district’s student-privacy approach. Critical coverage (PinkNews, 42K) and supportive recirculation of a 2024 trans-athlete lawsuit carried the story; it was the period’s only enacted state policy in the field.
Why it matters. New Hampshire has a Senate race (Pappas) and an open Governor’s race, and the schools/parents axis is now statute there rather than rhetoric. The law’s mechanics — compelled answers, a shifting “imminent risk” exception — will generate implementation stories all fall, each one a fresh attack surface or a fresh opening depending on who frames it first.
Takeaway. Answer on family ground, not process ground. The tested counter for exactly this terrain (see Messaging) leads empathetic — families and their doctors/teachers, not Washington or Concord politicians, decide what’s right for a child — and it wins with the persuadable middle. Litigating the statute’s mechanics concedes the frame.
Case Study 3: July 15 — The Hypocrisy Contrast, Field-Tested
What happened. On the period’s final day, Defense Secretary Pete Hegseth announced annual testosterone screening for service members 30 and over, with testosterone replacement therapy offered to those flagged — from an administration that barred transgender people from serving. The same day, at an event in Carlisle, Pennsylvania, President Trump claimed CHIPS Act grants had required recipients to be transgender (“if you weren’t transgender, you didn’t qualify”) — a claim with no basis in the statute, clipped by Aaron Rupar (1.6M). Together they produced the only day of the period when left-leaning accounts nationally out-engaged right-leaning ones (12.1M to 9.1M), led by elected officials — Rep. Pramila Jayapal posted the period’s highest-scoring official content (8.73).
Why it matters. The contrast — banning gender-affirming care for trans people while launching hormone optimization for troops — is not a one-day story; it is a repeatable structure. And the reach pattern shows it works: the left’s only winning day of the period came from offense on this contrast, not from defending against an attack. A separate one-page memo on the announcement accompanies this briefing, since the period boundary truncates its coverage here.
Takeaway. Pre-position the contrast so campaigns can deploy it the hour the next contradiction surfaces instead of catching one by luck. The tested language leans empathetic, names “politicians in Washington,” and carries the criminalization-of-care fact — the exact combination that leads the current MaxDiff panel.
Attack Formats to Watch
The staffer-association smear. Wisconsin’s evolution of the individual-villain story: the villain is a campaign’s own hire, converting an unverified claim into guilt-by-association against the candidate. Assume every Democratic campaign’s staff roster is a target surface; pre-build the verify-and-redirect posture.
Forced-outing / parental-rights statutes. New Hampshire’s SB 430 is enacted; Kansas’s district-level fights and federal “student privacy” enforcement threats run the same axis. Expect copycat bills and implementation stories all fall — each one is contested ground the empathetic family frame can win.
The veto-record attack. North Carolina’s disciplined framing of Cooper’s vetoes (sports, minors’ care, parental rights) is a template for running an executive’s record as a Senate attack. Built now, scaled later — the response should exist before the ad buy does.
The national culture-war import. The Odyssey mockery wave (and its local echoes) invites candidates to take a bait fight with no campaign payoff. The discipline is to decline it — hostile-visibility spectacle has no ad-migration path, and engaging elevates it.
Which Counter-Messages Land — Message Testing (MaxDiff)
This period’s message testing was fielded July 13 on the period’s dominant healthcare narrative — federal subpoenas of hospitals for transgender teenagers’ medical records and the CMS funding-rule retreat. A synthetic MaxDiff panel showed candidate messages in head-to-head pairs to voter archetypes weighted to the electorate; “selection rate” is the share of match-ups a message or ingredient wins. The panel’s winners shifted with the narrative — the June sports-ruling playbook does not carry over, and the differences are the finding.
1. By frame — re-weighted to the audience you reach
The frame-level view, re-weighted to the gender and age of the audience the attacks reach (the young-male skew from the previous section). ▲/▼ mark frames that over- or under-perform with that audience versus a generic likely-voter panel.
Counter-frame | Win % | Read |
Criminalization — prosecuting the doctors | 71.9% | Top frame with likely voters |
Federalism — this is for states, not Washington | 71.2% |
|
Corrects misinformation (“may, not must”) | 62.4% |
|
Science / evidence | 59.0% | Over-indexes with reached audience ▲ |
Distraction — they’d rather talk about this than prices | 55.9% |
|
Family / freedom — parents decide, keep government out | 52.3% |
|
Discrimination | 45.3% | Under-indexes with reached audience ▼ |
Opposition research | 43.6% |
|
Government overreach | 41.3% | Weakest-testing |
2. The best-testing messages, verbatim
The strongest messages pair the subpoena/criminalization fact with either a federalism correction or a family-and-doctors frame (Independent selection rate shown alongside the overall).
Message (excerpt) | Overall | Ind. | Frame · register |
“The Department of Justice is subpoenaing hospitals for the complete personal medical files of transgender teenagers, and the employment records of the doctors who treated them. Whatever you think about the underlying care, federal agents collecting children’s medical charts by name should stop everyone cold.” | 87.0% | 100.0% | overreach · outrage |
“States regulate the practice of medicine. That has been the law since Medicare was created — the statute says plainly that Medicare and Medicaid cannot be used to control how medicine is practiced inside a state. This rule ignored that. It was never going to survive a courtroom, and the administration knew it.” | 71.9% | 64.1% | federalism · practical |
“A family in a state where this care is legal, following the advice of the pediatrician they have trusted for years, should not have to read a federal register notice to find out whether their kid’s doctor is still allowed to see them. Let families and their doctors decide. Washington has enough to do.” | 71.9% | 76.6% | family freedom · empathetic |
“The federal government tried to tell every hospital in America that it would lose all of its Medicare and Medicaid money — the funding that keeps the emergency room and the cancer ward open — unless it followed Washington’s orders about one specific group of patients. That is not how medicine works in this country. Doctors and parents make these decisions, not politicians.” | 71.2% | 76.6% | overreach · practical |
The top message — leading with federal agents collecting children’s medical charts by name — wins 87% overall and 100% with Independents in the panel. Note what it does: it concedes nothing about the underlying care debate and still wins, because the overreach fact does the work.
3. What makes a message work
Holding the topic constant, three variables move a message more than its frame label does: register, the antagonist it names, and the argument ingredients it carries.
Register — lead empathetic, not outrage (a flip from June):
Register | Avg win % |
|
Empathetic | 58.3% | Best-testing register |
Practical | 54.6% |
|
Aspirational | 46.9% |
|
Outrage | 45.9% | Weakest — avoid leading here |
Antagonist — name “politicians in Washington,” not “Republican politicians”:
Villain named | Avg win % |
|
“Politicians in Washington” | 66.8% | Best-testing antagonist |
“The government” | 52.1% |
|
Congressional Republicans | 42.0% |
|
Naming Trump | 39.9% |
|
“Republican politicians” | 35.2% | Weakest |
Argument ingredients — the lift each adds:
Ingredient in the message | Lift |
|
Criminalization of doctors (“prosecute physicians”) | +29.0 | strongest lift |
States’ rights / federalism (“it’s a state choice”) | +21.3 |
|
Government-overreach label | +19.5 |
|
Parental authority (“families decide”) | +19.3 |
|
Slippery-slope / “erasure” framing | +18.9 |
|
Medical-authority appeal | +7.1 |
|
Corrects misinformation (“may, not must”) | +5.7 |
|
Equity / class dimension | −10.4 | — hurts |
Criminalization of doctors is the strongest single ingredient (+29.0) — naming that physicians face prosecution moves voters more than any frame label does. Federalism (+21.3), the government-overreach label (+19.5), and parental authority (+19.3) all still add real lift. Two changes from the June panel deserve attention: slippery-slope framing now helps (+18.9) — on a subpoena narrative, “this will reach further” reads as documented fact rather than speculation — and the partisan antagonist (“Republican politicians”, 35.2%) now tests weakest, so keep the villain institutional. The equity/class dimension still hurts (−10.4).
4. The bottom line for messaging
Build the counter as an empathetic, family-grounded story that names “politicians in Washington” and carries the criminalization fact: a family, their pediatrician, and federal agents collecting a child’s medical chart by name. Lead with the family, land the overreach fact, and keep the partisan label out of the villain slot — the institutional framing is what travels with Independents. Outrage-first constructions now test weakest even though the material invites them. Modeled signal — pair with live polling before ad spend.
Looking Ahead — The Primary Calendar
[DEVELOPING] The Pentagon hormone story is one day old. The announcement landed on this period’s final day, so its full arc — implementation details, official responses, whether any campaign picks up the contrast — belongs to the next briefing. The companion one-page memo covers what is known now. [DEVELOPING] The Odyssey is in US theaters as of July 17. Expect a renewed mockery wave through the opening weekend and awards chatter. The guidance holds: no campaign payoff in the bait fight; decline it.
August primaries (Michigan 8/4, Arizona 8/4, Minnesota 8/11, Wisconsin 8/11). Wisconsin now enters its primary with the staffer story in circulation and Tiffany running on girls’ sports; the Arizona AG race and AZ-06 (Ciscomani–Mendoza) are the other contests most likely to be run on trans issues. Expect creative to sharpen in the next two weeks.
New Hampshire 9/8 — SB 430 goes operational when school starts. The disclosure law’s first implementation stories (a compelled answer, a discipline case, a family’s complaint) will land during the fall semester, squarely in the Senate and Governor race window. The family-ground answer should exist before the first story does.
Ciscomani paid-rotation watch (standing request). He is posting girls’-sports attack content organically; the open question from the July 8 ask is whether it enters paid rotation. The watchlist flags his account per post; the next c4 will report whatever the ad-library check finds.
Strategic Recommendations
COUNTER | PIVOT AWAY | LEAN INTO |
• The staffer-association smear (Wisconsin): a verify-and-redirect posture for staff-based attacks — answer on the person and the facts, never the population, and don’t let an unverified aggregator claim set the cycle. • Forced-outing implementation stories (NH SB 430, fall semester): the family-ground answer, pre-drafted before the first compelled-disclosure story lands. | • The Odyssey bait fight — hostile-visibility spectacle with no ad-migration path; engaging elevates it. • Outrage-led constructions — the material invites them and they now test weakest (45.9%); the partisan villain label (“Republican politicians”) tests weakest of all antagonists. | • The hypocrisy contrast, pre-positioned — July 15 proved it wins the day when it exists; the goal is to stop needing an accident to trigger it. (See the companion memo.) • The empathetic family frame carrying the criminalization fact — the tested combination that wins Independents on the healthcare terrain now in play. |
The bottom line. A quiet state map is not a safe one. This period the attacks ran national and cultural, and the state-level machinery — a manufactured staffer story in Wisconsin, an enacted disclosure law in New Hampshire, a veto-record attack built in North Carolina, a congressman posting his own girls’-sports creative in Arizona — kept assembling quietly underneath it. The offensive finding is the clearest one this cycle has produced: the left’s only reach-winning day came from a pre-packaged hypocrisy contrast it did not have to manufacture, and the message testing says exactly how to carry it — empathetic, family-grounded, Washington-villained, criminalization fact attached. Build that before the August primaries; the attacks will not stay quiet through them.
Limitations
Geography is measured from two dashboard layers: issue-location (which state a post is about; ~10% of posts carry a state tag) and account-location (where a channel is based; ~2% of channels). A post can be about a state without originating there, so state reach measures attention, not local activity. Off-topic virality that keyword-matched a state (foreign-language tabloid items, unrelated news) was removed by hand. Stance (anti/pro/neutral) is a directional classifier read; “neutral” absorbs much hostile-leaning news, so anti-trans shares are conservative.
The audience read is modeled, not observed: the dashboard has no engager-level data, so audience is inferred from the demographics of the accounts producing and amplifying content, engagement-weighted, under a homophily assumption (producer ≠ audience is the foundational caveat). It blends Census surname data, SSA first-name records, platform user base, and account location; the “evidence” column reports the share of reach with a real name/seed match versus a national fallback. The racial estimate is the weakest layer (only ~2% of channels have a state) and is hedged hardest. The originator/amplifier read is modeled from near-duplicate text clustering plus native amplifier edges captured for ~36% of posts. The MaxDif frame fit is modeled × modeled — a synthetic panel re-weighted to a modeled audience — a directional signal, not a poll. Engagement is not persuasion.
Methodology
Data source. Trans Rhetoric Social Media Dashboard, built for TFA on the Junkipedia social listening API — 77,000+ channels across 20 platforms, 984,000+ posts spanning January 2024 to present, collected on 90+ keywords (topic terms, framing terms, and named trans public figures). Period: 2026-07-01 – June 30, 2026. Priority states are drawn from Adrienne’s “TFA priority states-races 5.18.26” sheet, plus Iowa (Jun 9 request) and Texas/Talarico. Per-state signal uses two geolocation layers — issue-location (which state a post concerns) and account-location (where a channel is based) — filtered to trans-relevant posts carrying engagement, with a manual audit of the high-engagement tail to remove off-topic keyword matches.
Message Testing (MaxDiff)
The messaging results come from a MaxDiff (maximum-difference scaling) exercise fielded July 13 on the period’s governing healthcare narrative — the federal subpoenas of hospitals for transgender minors’ medical records and the CMS funding-rule retreat. The instrument is a set of 32 candidate messages written to vary systematically along four dimensions — frame (criminalization, federalism, family freedom, overreach, discrimination, distraction, science, opposition research), the antagonist named (“politicians in Washington,” “the government,” Congressional Republicans, Trump, Republican politicians, or none), rhetorical register (empathetic, practical, aspirational, outrage), and eleven binary argument-elements (e.g. criminalization-of-doctors, states’-rights, parental authority). Opposition messages are included and written with equal craft, not as strawmen, to anchor the low end of the scale.
The instrument is fielded against a panel of 21 voter archetypes weighted to the national electorate (race ≈ White 77% / Black 10% / Hispanic 8% / Asian 4% / Native 1%; age-weighted toward ~35% 65-plus; with party, education, and ideology crossed in). Each archetype evaluates message pairs in character and selects the more convincing of the two, across roughly 2,700 head-to-head comparisons in a single pass with no optimization toward target numbers. A calibrated 10–15% “unexpected-choice” rate is built in to reproduce real-respondent inconsistency and to hold the top-to-bottom spread to the 30–40 points observed in fielded MaxDiff studies rather than an artificially wide split.
A message’s selection rate is the share of its head-to-head appearances it wins; rates are then aggregated by frame, antagonist, register, and argument-element, and the lift of each ingredient is the difference in average selection rate between messages that carry it and those that do not. Results are reported two ways: likely-voter-weighted (the message, register, antagonist, and ingredient cuts) and re-weighted to the engagement-reached audience (the frame cut), the latter by applying the audience model’s gender/age distribution for the content these attacks actually reach. This is a synthetic panel — archetype responses are model-simulated, not live human interviews — calibrated against prior human benchmarks and held to realistic spreads. Read it as message triage that prioritizes what to put into live testing, not as a substitute for a fielded poll.
Diffusion: Originators vs. Amplifiers
The originator/amplifier analysis separates the accounts that start a story from those that give it the most reach. It runs in two passes. First, near-duplicate clustering: every post’s text is normalized (URLs, @-handles, casing, and punctuation stripped) and grouped with near-identical posts, so the same claim or clip is tracked as it reappears across accounts — this window’s ~24,500 posts collapse into roughly 3,000 recurring-story clusters covering about 40% of volume. Second, native amplifier edges: the platform-level quote, reply, and repost relationships (present on about 36% of posts) link each amplifying post to the item it echoes.
Within a cluster, the originator is the earliest poster (or, for a retweet/quote chain, the account being quoted); downstream reach is the total engagement the story accrues beyond the originator’s own post; amplifiers are the other accounts in the cluster, ranked by the reach they add to stories they did not start. Originators are ranked both by the number of distinct stories they seed and by the downstream reach those seeds generate. Caveats: only content with a captured native edge (~36%) contributes to the edge-based view; text clustering misses heavy paraphrase and cross-language restatement; and “originator” is the earliest post in the dataset, which can trail an off-platform or unmonitored origin.
Audience Model
Because the dashboard captures no engager-level data, audience is modeled from the accounts that produce and amplify each slice of content, engagement-weighted, under a homophily assumption (producer ≠ audience is the foundational caveat). It blends US Census surname distributions (race), Social Security first-name records (gender and age), each platform’s known user base, and account location. The evidence share reports the fraction of reach carried by accounts with a real name or hand-labeled seed match rather than a national-average fallback; confidence is the engagement-weighted mean of per-account model confidence. Race is the weakest layer and is hedged hardest. Legal characterizations of the June 30 ruling follow the decision and TFA’s published message guidance: it permits state sports bans, does not require them, does not reach Title IX outside sports, and does not decide whether schools may choose inclusion.
