What did a ruling change—and what did the public actually say?
Policy comparisons
10 worked examples pair specific rulings with earlier survey responses. These selected cases illustrate the comparison; they do not measure overall Court–public agreement.
How to read this comparison
Direct comparison: a question addresses a named policy component. Context only: related preferences, not approval of the specific holding. Missing data: no suitable earlier observation in this selection. Procedural rulings can be not comparable.
We use the latest eligible survey year per source within three years, keeping tied question variants. Same-year surveys are included only when verified fieldwork ends before the ruling. Percentages retain each survey’s wording and denominator; they are not pooled. Court interpretations are explicit source-based assessments, not automated classifications. Votes are not treated as policy magnitude.
These are historical outcomes. No causal, predictive, or statistical-significance claim is made. Missing matches do not mean no relevant survey exists elsewhere.
Marriage equality · A closer look
Should same-sex couples be allowed to marry?
Public support and legal protection changed over time. But permission to marry, recognition of a marriage, and equal treatment within marriage are different questions.
What the surveys tell us
Support rises between the selected 2012 and 2016 observations in CES, GSS and PRRI. The levels differ across sources, whose wording and response categories are not identical.
What the rulings tell us
General marriage preferences match Obergefell’s licensing ruling. Targeted Gallup and ABC/Post questions also address federal and interstate recognition. Birth-certificate rules still have only broad marriage-attitude context.
Public opinion over time · 1988–2024
Share supporting or agreeing with same-sex marriage in 28 audited source/question-year observations. Each bar runs from 0 to 100%. Sources are kept separate; missing years are not interpolated. This history includes surveys before and after the rulings; only eligible earlier surveys appear in the case comparisons below.
CES
Explore 5 observations · 2012–2016
2012 · gaymarriage_legalize53.7%
Wording and denominator
Do you favor or oppose allowing gays and lesbians to marry legally?
Homosexual couples should have the right to marry one another
Two wording variants kept distinct; Stata variable label is not full questionnaire context.
Positive-weight respondents with substantive, cannot-choose or nonresponse item codes; inapplicable/year-unavailable/system-missing excluded.
2004 · marsame29.9%
Wording and denominator
Homosexual couples should have the right to marry one another
Two wording variants kept distinct; Stata variable label is not full questionnaire context.
Positive-weight respondents with substantive, cannot-choose or nonresponse item codes; inapplicable/year-unavailable/system-missing excluded.
2006 · marsame135.4%
Wording and denominator
Homosexual couples have right to marry
Two wording variants kept distinct; Stata variable label is not full questionnaire context.
Positive-weight respondents with substantive, cannot-choose or nonresponse item codes; inapplicable/year-unavailable/system-missing excluded.
2008 · marsame138.9%
Wording and denominator
Homosexual couples have right to marry
Two wording variants kept distinct; Stata variable label is not full questionnaire context.
Positive-weight respondents with substantive, cannot-choose or nonresponse item codes; inapplicable/year-unavailable/system-missing excluded.
2010 · marsame146.3%
Wording and denominator
Homosexual couples have right to marry
Two wording variants kept distinct; Stata variable label is not full questionnaire context.
Positive-weight respondents with substantive, cannot-choose or nonresponse item codes; inapplicable/year-unavailable/system-missing excluded.
2012 · marsame147.1%
Wording and denominator
Homosexual couples have right to marry
Two wording variants kept distinct; Stata variable label is not full questionnaire context.
Positive-weight respondents with substantive, cannot-choose or nonresponse item codes; inapplicable/year-unavailable/system-missing excluded.
2014 · marsame156.0%
Wording and denominator
Homosexual couples have right to marry
Two wording variants kept distinct; Stata variable label is not full questionnaire context.
Positive-weight respondents with substantive, cannot-choose or nonresponse item codes; inapplicable/year-unavailable/system-missing excluded.
2016 · marsame160.4%
Wording and denominator
Homosexual couples have right to marry
Two wording variants kept distinct; Stata variable label is not full questionnaire context.
Positive-weight respondents with substantive, cannot-choose or nonresponse item codes; inapplicable/year-unavailable/system-missing excluded.
2018 · marsame166.6%
Wording and denominator
Homosexual couples have right to marry
Two wording variants kept distinct; Stata variable label is not full questionnaire context.
Positive-weight respondents with substantive, cannot-choose or nonresponse item codes; inapplicable/year-unavailable/system-missing excluded.
2021 · marsame64.0%
Wording and denominator
Homosexual couples should have the right to marry one another
Two wording variants kept distinct; Stata variable label is not full questionnaire context.
Positive-weight respondents with substantive, cannot-choose or nonresponse item codes; inapplicable/year-unavailable/system-missing excluded.
2022 · marsame66.2%
Wording and denominator
Homosexual couples should have the right to marry one another
Two wording variants kept distinct; Stata variable label is not full questionnaire context.
Positive-weight respondents with substantive, cannot-choose or nonresponse item codes; inapplicable/year-unavailable/system-missing excluded.
2024 · marsame63.1%
Wording and denominator
Homosexual couples should have the right to marry one another
Two wording variants kept distinct; Stata variable label is not full questionnaire context.
Positive-weight respondents with substantive, cannot-choose or nonresponse item codes; inapplicable/year-unavailable/system-missing excluded.
PRRI
Explore 9 observations · 2011–2024
2011 · Q8B48.2%
Wording and denominator
Allowing gay and lesbian couples to marry legally.
All positive-weight wave respondents, including unsure/refused/skipped and missing.
2012 · legssm48.4%
Wording and denominator
Allowing gay and lesbian couples to marry legally.
All positive-weight wave respondents, including unsure/refused/skipped and missing.
2013 · Q14A51.9%
Wording and denominator
Allowing gay and lesbian couples to marry legally.
All positive-weight wave respondents, including unsure/refused/skipped and missing.
2014 · Q12B48.6%
Wording and denominator
Allowing gay and lesbian couples to marry legally.
All positive-weight wave respondents, including unsure/refused/skipped and missing.
2015 · Q9A59.5%
Wording and denominator
Allowing gay and lesbian couples to marry legally.
All positive-weight wave respondents, including unsure/refused/skipped and missing.
2016 · Q22E64.2%
Wording and denominator
Allowing gay and lesbian couples to marry legally.
All positive-weight wave respondents, including unsure/refused/skipped and missing.
2017 · Q17A61.9%
Wording and denominator
Allowing gay and lesbian couples to marry legally.
All positive-weight wave respondents, including unsure/refused/skipped and missing.
2023 · Q40_E65.9%
Wording and denominator
How much do you favor or oppose each of the following? Allowing same-sex couples to marry legally.
All positive-weight wave respondents, including unsure/refused/skipped and missing.
2024 · Q10_C66.1%
Wording and denominator
How much do you favor or oppose each of the following? Allowing same-sex couples to marry legally.
All positive-weight wave respondents, including unsure/refused/skipped and missing.
VOTER
Explore 2 observations · 2016–2020
2016 · view_gaymar_201649.9%
Wording and denominator
Do you favor or oppose allowing gays and lesbians to marry legally?
Positive-weight administered wave respondents, including unsure/skipped; not-asked and system missing excluded.
2020 · view_gaymar_2020Nov56.4%
Wording and denominator
Do you favor or oppose allowing gays and lesbians to marry legally?
Positive-weight administered wave respondents, including unsure/skipped; not-asked and system missing excluded.
CES percentages exclude missing answers and use support/oppose respondents only. Other sources retain additional categories; differences between sources are not necessarily changes in opinion. See the response breakdowns below.
Why is VOTER lower in 2016?
VOTER records 49.9% favor, 35.1% oppose, 14.6% unsure and 0.3% skipped, using its supplied cross-sectional weights. Among favor/oppose respondents only, support is 58.7%; CES reports 61.6% on that narrower denominator. This explains most of their apparent difference, but does not make the surveys equivalent.
We verified VOTER’s raw response counts and weights against its survey guide. It interviewed returning panelists in November–December 2016. Differences from GSS and PRRI remain; timing, question format and sample design may matter, but we have not isolated their effects. The bars retain the original denominators.
What changed in court?
Selected decisions, not a complete legal history. Court outcomes are described in words, not positioned on a numerical ideology scale.
Appeal dismissed for want of a substantial federal question.
Summary disposition left the challenged Minnesota judgment in place; not a new nationwide legislative prohibition.
Broad licensing-policy direction is comparable in principle, but this survey selection has no eligible pre-1972 observation.
No matching earlier survey response in this selection.
Holding source and scope
409 U.S. 810, October 10, 1972, docket 71-1027; PDF page 10. Obergefell Part III discusses and overrules Baker.
Summary dismissal has limited merits effect on the issues necessarily decided, unlike a standing dismissal. It does not adopt every rationale below. See Mandel v. Bradley, 432 U.S. 173, 176 (1977): https://www.law.cornell.edu/supremecourt/text/432/173 . Baker was later overruled in Obergefell Part III.
Marriage benefits for federal employees legally married to same-sex partners.
Gallup 2013 weighted response percentages
Response
Percent
Favor
54.0%
Against
39.0%
No opinion
8.0%
Published weighted percentages; includes no opinion. Rounded totals may equal 101%; values are not renormalized.
Question details
Question summary (paraphrased). See the original report for exact wording and questionnaire context. Report gives full-survey n and split forms; item-level sample size is not established.
Zero or omitted categories do not establish that the original questionnaire lacked other answer options. Rounding may prevent totals of exactly 100%.
Holding source and scope
Per curiam opening and closing paragraphs; dissent heading for author and joiners.
An unsigned opinion does not supply a named majority-joinder roster. Broad marriage approval is not measured approval of this birth-certificate holding.