online opt-in
CECooperative Election Study (CES)
Cooperative Election Study (CES) 2020 Common Content · Sep 29, 2020 – Dec 14, 2020 · US national (large enough for state-level estimates)
Methodology
- Pollster
- Cooperative Election Study (CES) academic
- Sponsor
- Harvard University (CES/CCES consortium of university research teams)
- Fielded
- Sep 29, 2020 – Dec 14, 2020
- Population
- adults
- Sample size
- 61,000
- Sample frame
- online opt-in
- Panel
- YouGov (with Disqo, Prodege and Generation Lab panels in some years)
- Mode
- online - YouGov matched random sample: a target sample is drawn from the American Community Survey / Current Population Survey frame and matched to YouGov opt-in panelists, then weighted by propensity score and raked
- Margin of error
- not disclosed
- Weighting
- commonweight (2017: weights_common) - propensity-score matched sample weighted and raked to age, gender, education, race, Hispanic origin, voter registration status and presidential vote
- Geography
- US national (large enough for state-level estimates)
- Base notes
- Pre-election wave 29 Sep - 2 Nov 2020; post-election wave 8 Nov - 14 Dec 2020. CC20_355d was on the pre-election wave.
Sources
Verification note. Wording copied verbatim from CES20_Common_pre_qx.pdf (grid CC20_355, row CC20_355d). Weighted percentages computed by this agent from the public-use microdata CES20_Common_OUTPUT_vv.csv (n=61,000) using the released weight commonweight; the raw unweighted counts (18,134 support / 42,487 oppose / N 60,621) match the counts published in 'CCES Guide 2020.pdf' exactly, which cross-validates the extraction.
Figures below are transcribed from these documents. Where they disagree with the source, the source governs.
Questions asked (2)
CES variable CC20_355d - ban transgender people in the military (pre-election wave)
For each of the following tell us whether you support or oppose these decisions.
Ban Transgender People in the Military
Wording notes: bare policy label | no middle or don't-know option | embedded in an executive-orders battery
neutral — plain policy label; battery framing is the main caveat
Subgroup figures (6)
CES demographic variable measuring transgender status of the respondent
Have you ever undergone any part of a process (including any thought or action) to change your gender / perceived gender from the one you were assigned at birth? This may include steps such as changing the type of clothes you wear, name you are known by or undergoing surgery.
No
94.8%
Prefer not to say
2.8%
Yes
2.4%
Wording notes: extremely broad wording ('any thought or action')
unclear — The pre-2021 CCES wording asks whether the respondent has 'ever undergone any part of a process (including any thought or action)' to change their gender, which is far broader than transgender self-identification and produces prevalence estimates of 2-4%, several times higher than every other national survey. It should not be read as a transgender population estimate.