probability panel
PRRI
PRRI 2023 Politics of Gender Survey ('The Politics of Gender, Pronouns, and Public Education'); also AVA March 2023 wave · Mar 9–23, 2023 · US national
Methodology
- Pollster
- PRRI independent
- Sponsor
- PRRI
- Fielded
- Mar 9–23, 2023
- Population
- adults
- Sample size
- 5,438
- Sample frame
- probability panel
- Panel
- Ipsos KnowledgePanel
- Mode
- online, self-administered - Ipsos KnowledgePanel (probability, address-based sampling from the USPS Delivery Sequence File), supplemented with Ipsos opt-in panel respondents calibrated to the KnowledgePanel weights to reach minimum sample sizes in smaller states
- Margin of error
- +/- 1.5 percentage points
- Weighting
- pre-stratification weights to the March CPS supplement; PPS sample selection; post-stratification raking (IPF) on gender by age, by race/ethnicity, Census division, metro area by race/ethnicity, education by race/ethnicity, income by race/ethnicity, language proficiency (ACS), party affiliation (Pew NPORS); separate state-level weights; opt-in cases calibrated to the KnowledgePanel weights
- Geography
- US national
- Base notes
- 5,046 Ipsos KnowledgePanel respondents plus 392 recruited from opt-in panels. Q3c/Q3d and Q4a-Q4c were split-sample (Form 1 N=2,726; Form 2 N=2,712; Q4a N=1,361; Q4b N=1,365). PRRI labels this fielding 'Mar. 2023' in the AVA LGBTQ trend tables, so this record replaces a separate March 2023 AVA wave rather than duplicating it. This is PRRI's single richest transgender instrument: it carries the pronoun-comfort items, the social-contagion item, the split-sample school-harm experiment and the age-graded curriculum battery that appear nowhere else in PRRI's series.
Sources
Verification note. toplines, party breaks and intensity splits read from the PRRI report page; PRRI describes but does not print the literal item stems on that page, so wordings are reconstructed from PRRI's own descriptions || Every distribution copied from the Politics of Gender topline PDF; n, dates, MOE and design effect from that topline's methodology section; party breaks on the gender-binary and curriculum items from the PRRI report page. || WORDING RECOVERY SWEEP AND MATERIAL CORRECTION. The database carried this as a two-option forced choice ('there are only two gender identities... or there is a range of many'). PRRI's own topline shows a four-point strength-of-belief scale with a rotated option order. Percentages and stem transcribed from PRRI's 2023 Politics of Gender topline.
Figures below are transcribed from these documents. Where they disagree with the source, the source governs.
Questions asked (17)
whether there are only two gender identities or a range of many
Which comes closer to your view: there are only two gender identities, man and woman, or there is a range of many possible gender identities?
Wording reconstructed from: reconstructed from PRRI's description of the item; literal stem not printed on the report page
Only two gender identities
65%
Only two gender identities - feel strongly
50%
A range of many gender identities
34%
A range of many gender identities - do not feel strongly
20%
Only two gender identities - do not feel strongly
15%
A range of many gender identities - feel strongly
14%
Wording notes: forced binary between 'only two' and 'a range of many' with no middle option | intensity is captured with a follow-up, allowing a 4-point read
neutral — both positions are stated in the respondent's own terms and the intensity follow-up avoids collapsing weak and strong holders
Subgroup figures (3)
comfort using gender-neutral pronouns
How comfortable would you be using the gender-neutral pronoun 'they' to refer to someone who asked you to?
Wording reconstructed from: reconstructed from PRRI's description; literal stem not printed
Uncomfortable
40%
Comfortable
35%
Doesn't matter
23%
Wording notes: offers an explicit 'doesn't matter' option, which about a fifth of respondents take
neutral — comfort framing rather than policy framing; symmetric comfortable/uncomfortable options
comfort using pronouns that do not match a person's perceived gender
How comfortable would you be using pronouns that do not match the gender you perceive a person to be?
Wording reconstructed from: reconstructed from PRRI's description; literal stem not printed
Uncomfortable
43%
Comfortable
36%
Doesn't matter
21%
Wording notes: frames the pronoun as mismatched with the respondent's own perception rather than with the person's identity - a framing that measurably depresses comfort relative to the 'they' item
tilts_restrictive — 'do not match the gender you perceive a person to be' centers the respondent's perception as the reference point rather than the person's stated identity, which prompts a mismatch reading
belief that there are only two genders vs. a range of gender identities
When it comes to gender, do you ...
Feel strongly that there are only two genders, man or woman
50%
Think there is a range of many gender identities, but do not feel strongly about it
20%
Think there are only two genders, but do not feel strongly about it
15%
Feel strongly that there is a range of many possible gender identities
14%
Skipped/Refused
1%
Only two genders (net) 65%Range of gender identities (net) 34%
Wording notes: strength of view built into the answer set | no middle or 'depends' option | response order flipped for half the sample
neutral — the two positions are stated in parallel with matched intensity gradations and the order is rotated
Subgroup figures (3)
comfort with a friend disclosing they are transgender (revised answer set with a 'doesn't matter' option)
Q2. How comfortable or uncomfortable do you think you would be with a friend telling you the following? [RANDOMIZE]
That they are transgender
Very comfortable
25%
Doesn't matter either way
25%
Somewhat uncomfortable
17%
Somewhat comfortable
16%
Very uncomfortable
16%
Skipped/Refused
1%
Wording notes: adds a 'Doesn't matter either way' option absent from PRRI's 2019 and 2021 comfort battery, so it is not directly comparable to that trend
neutral — balanced comfortable/uncomfortable scale; the added neutral option arguably improves the measure by not forcing an affective response
comfort with a friend who uses gender-neutral pronouns
Q2. How comfortable or uncomfortable do you think you would be with a friend telling you the following? [RANDOMIZE]
That they use gender neutral pronouns, for example, using "they" instead of "he" or "she"
Doesn't matter either way
23%
Very uncomfortable
21%
Very comfortable
20%
Somewhat uncomfortable
19%
Somewhat comfortable
15%
Skipped/Refused
1%
Wording notes: gives a concrete example of the pronoun rather than leaving it abstract
neutral — balanced scale with a neutral option; the worked example reduces respondent confusion about what is being asked
comfort using a friend's pronouns when they do not match the respondent's perception of the friend's appearance
Q2. How comfortable or uncomfortable do you think you would be with a friend telling you the following? [RANDOMIZE]
That they would like you to use pronouns like "he" or "she" that do not match your perception of their appearance
Very uncomfortable
22%
Very comfortable
21%
Somewhat uncomfortable
21%
Doesn't matter either way
21%
Somewhat comfortable
15%
Skipped/Refused
1%
Wording notes: anchors the item to the respondent's own visual judgment ('your perception of their appearance')
loaded_terms — 'do not match your perception of their appearance' invites the respondent to treat their own read of someone's appearance as the reference point for that person's gender, which is a substantive premise rather than a neutral description
whether Americans spend too much time talking about gender and pronouns
Q3. How well, if at all, do you think each of the following statements describes America today? [RANDOMIZE]
People spend too much time talking about their gender and pronouns
Completely
37%
Somewhat
25%
A little bit
19%
Not at all
18%
Skipped/Refused
1%
Completely or somewhat (net) 62%
Wording notes: one-sided statement with no counter-statement offered | 'describes America today' framing invites agreement with a familiar complaint
tilts_restrictive — the statement is written from the annoyed-observer side; there is no matched item asking whether Americans spend too little time on it, so the 62% who say it describes America at least somewhat is not a symmetric measure
belief that young people are being peer pressured into being transgender
Q3. How well, if at all, do you think each of the following statements describes America today? [RANDOMIZE]
Young people are being peer pressured into being transgender
Not at all
39%
Completely
22%
Somewhat
21%
A little bit
16%
Skipped/Refused
1%
Completely or somewhat (net) 43%
Wording notes: states a contested empirical claim (social contagion) as something to be rated for descriptive accuracy | no counter-statement offered
loaded_terms — 'peer pressured into being transgender' embeds the social-contagion thesis in the stem; a respondent has no way to reject the premise other than by choosing 'not at all', so the item measures endorsement of a frame rather than a belief the respondent volunteered
whether public schools give students harmful information about gender and sexual orientation (Form 1)
Q3. How well, if at all, do you think each of the following statements describes America today? [RANDOMIZE]
Public schools are giving students harmful information about gender and sexual orientation
Not at all
34%
Completely
27%
Somewhat
21%
A little bit
15%
Skipped/Refused
2%
Completely or somewhat (net) 48%
Wording notes: 'harmful' is asserted in the stem | Form 1 of a split sample testing an age qualifier
tilts_restrictive — the stem asserts harm and asks only how well it describes reality; the paired split-sample design is nonetheless valuable because it isolates how much of the agreement is driven by the age qualifier rather than by the content itself
whether public schools give students harmful information about gender and sexual orientation at too young an age (Form 2)
Q3. How well, if at all, do you think each of the following statements describes America today? [RANDOMIZE]
Public schools are giving students harmful information about gender and sexual orientation at to young an age
Not at all
35%
Completely
33%
Somewhat
16%
A little bit
15%
Skipped/Refused
1%
Completely or somewhat (net) 49%
Wording notes: 'harmful' asserted in the stem plus an age qualifier | Form 2 of the split sample
tilts_restrictive — same asserted-harm problem as Form 1; adding 'at too young an age' raised 'completely describes' from 27% to 33% while leaving the net roughly flat (48% vs 49%), which is exactly the kind of wording sensitivity this database exists to record
grade level at which schools should discuss that some people do not consider themselves a man or a woman
Q4. Currently, there is a debate about what is appropriate to teach children in public schools about gender, sexuality and sexual education. Please tell us at what point you think it's appropriate for schools to discuss each of these issues, or if it's never appropriate: [RANDOMIZE]
That some people do not consider themselves to be a man or a woman
This topic is never appropriate to teach or discuss in public K-12 schools
36%
In middle school (6th to 8th grade)
26%
In high school (9th to 12th grade)
18%
In later elementary school (3rd to 5th grade)
11%
In early elementary school (Kindergarten to 2nd grade)
7%
Skipped/Refused
1%
Wording notes: age-graded response set rather than a simple allow/ban binary | preamble names the topic as a 'debate', which cues controversy
neutral — the graded answer set is a genuine improvement over allow/ban items because it separates 'never' from 'not yet'; the parallel heterosexual-relationship item gives a comparison baseline (24% never vs 36% never here)
Subgroup figures (3)
grade level at which schools should discuss that some people are transgender
Q4. Currently, there is a debate about what is appropriate to teach children in public schools about gender, sexuality and sexual education. Please tell us at what point you think it's appropriate for schools to discuss each of these issues, or if it's never appropriate: [RANDOMIZE]
That some people are transgender, in other words not the gender that corresponds to their sex at birth
This topic is never appropriate to teach or discuss in public K-12 schools
34%
In middle school (6th to 8th grade)
27%
In high school (9th to 12th grade)
18%
In later elementary school (3rd to 5th grade)
12%
In early elementary school (Kindergarten to 2nd grade)
8%
Skipped/Refused
1%
Wording notes: includes a definitional gloss ('not the gender that corresponds to their sex at birth') | preamble names the topic as a 'debate'
neutral — graded answer set with a plain definitional gloss; the gloss is descriptive rather than evaluative
Subgroup figures (3)
LGBT nondiscrimination protections in jobs, public accommodations and housing (item explicitly names transgender people)
Q6. Do you favor or oppose each of the following? [RANDOMIZE STATEMENTS]
Laws that would protect gay, lesbian, bisexual, and transgender people against discrimination in jobs, public accommodations, and housing.
Favor 38%
Wording notes: bundles transgender people with gay, lesbian and bisexual people in one item | three policy domains combined in one question (jobs, public accommodations, housing)
neutral — symmetric favor/oppose scale, no loaded terms; the design limitation is that it is double-barreled across three policy domains and four identity groups, so transgender-specific support cannot be isolated
laws preventing parents from allowing a minor to receive gender-transition medical care
Q6. How much do you favor or oppose each of the following? [RANDOMIZE]:
Laws that prevent parents from allowing their child to receive medical care for a gender transition.
Favor 20%
Wording notes: frames the ban as constraining parents rather than minors or providers | does not name specific interventions (puberty blockers, hormones, surgery) | 'medical care for a gender transition' rather than 'gender-affirming care'
tilts_affirming — the parental-authority frame ('prevent parents from allowing') suppresses support for restriction relative to wordings that foreground the minor or the specific medical procedure; PRRI's readings run notably less restrictive than pollsters who ask about 'gender transition medical treatment for minors' directly
knows someone who is transgender
Q7. Do you know someone who:
is transgender
No
63%
Yes, another acquaintance
25%
Yes, close friend
5%
Yes, family member
5%
Yes, myself
1%
Wording notes: 'Do you know someone who' is a broader stem than PRRI's earlier 'close friend or family member' wording, so levels are not comparable to the pre-2021 trend
neutral — factual contact item
knows someone who uses gender-neutral pronouns
Q7. Do you know someone who:
uses gender neutral pronouns, for example, using "they" rather than "he" or "she"
No
62%
Yes, another acquaintance
23%
Yes, close friend
6%
Yes, family member
6%
Yes, myself
1%
Wording notes: multiple response; categories do not sum to 100
neutral — factual contact item