Skip to content
Contents
The Public Mind

    Type to search. Planned chapters match by title only.

    Chapter 6.1

    Confidence in thirteen institutions

    The average gap between Democrats and Republicans in the share with a great deal of confidence in thirteen institutions rose from 4.8 points in 1973-1980 to 6.7 in 2011-2018 and was 9.4 in person in 2022-2024, and the gaps run in both directions.

    Affiliation
    NYU Stern
    Published
    DraftFigures and numbers may change as the analysis is rebuilt.what is still being checked

    The bootstrap intervals for the mean absolute gap resample respondents and ignore the survey's sampling design; a design-based version of those intervals is still to be run.

    Among respondents interviewed in person in 2022-2024, Democrats are more confident than Republicans in science, medicine, education, the press, Congress and television, and Republicans are more confident in the military, the Supreme Court and organized religion, in each case with an interval that excludes zero. The executive branch gap takes the sign of the party that holds the White House, and the intervals for banks, major companies and organized labor include zero. The average size of the thirteen gaps, ignoring direction (the mean absolute gap), was 4.8 (95% CI 4.3 to 5.5)4.8 percentage points (mean over institutions of the absolute Democratic-minus-Republican gap)Estimandmean over thirteen institutions of the absolute Democratic-minus-Republican gap in the share with a great deal of confidence; in-person interviews only; seeded respondent bootstrap interval (2,000 replicates)Interval95% CI 4.3 to 5.5n10,281Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period1973-1980SourceGeneral Social Survey points in 1973-1980 and 9.4 (95% CI 8.0 to 11.8)9.4 percentage points (mean over institutions of the absolute Democratic-minus-Republican gap)Estimandmean over thirteen institutions of the absolute Democratic-minus-Republican gap; in-person interviews only; seeded respondent bootstrap interval (2,000 replicates)Interval95% CI 8.0 to 11.8n1,980Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022-2024SourceGeneral Social Survey points in 2022-2024 among in-person respondents.

    The General Social Survey (GSS) asks about "the people running" each institution and offers three answers: a great deal of confidence, only some, or hardly any [1]. This chapter tracks the share answering "a great deal" among respondents who identify with a party or lean toward one (leanersPartisan leanerA respondent who first says they are independent but, when asked, says they lean toward one party. Most figures group leaners with the party they lean toward, and each chapter says which grouping it uses. are counted with the party they lean toward). A gap is the Democratic share minus the Republican share, in pointsPercentage pointThe unit for the difference between two percentages. A change from one share to another is stated in points, not in percent, so that absolute and relative change are not confused., so a positive gap means Democrats are more confident and a negative gap means Republicans are. Every number and figure uses this sign. Where an estimate has a confidence intervalConfidence intervalA range around an estimate that would contain the true value in 95% of repeated samples if the sampling model held. The book reports 95% intervals unless it says otherwise., the text prints it in parentheses after the estimate. The GSS asked about twelve institutions in 1973 and 1974 and about thirteen from 1975, when banks were added.Shares are weightedSurvey weightsNumbers that scale each respondent so the sample matches the population on known traits such as age, race, education and region. A weighted average counts a respondent with a large weight for more than one with a small weight.. Intervals use the Kish effective sample sizeEffective sample sizeThe number of equally weighted respondents that would give the same precision as the weighted sample. It is smaller than the raw count when the weights vary a lot, and it sets how wide the confidence interval is. [2] and treat the weighted sample as a simple random sample, and an interval for a share near zero is cut off at zero. Cells below the suppression floor (fewer than 150150 respondentsEstimandsuppression floor (constant)Weightsnone (count)Populationminimum size of a party-by-year cellPeriod1973-2024, survey years with the confidence itemsSourceGeneral Social Survey respondents or an effective sample under 100100 respondents (Kish effective sample size)Estimandsuppression floor (constant)Weightsnone (count)Populationminimum effective size of a party-by-year cellPeriod1973-2024, survey years with the confidence itemsSourceGeneral Social Survey) are not shown, and the smallest cell shown has 198198 respondentsEstimandcount of respondentsWeightsnone (count)Populationsmallest party-by-year cell shown in the figuresPeriod1973-2024, survey years with the confidence itemsSourceGeneral Social Survey respondents. The estimates rest on 41,79741,797 respondentsEstimandcount of respondentsWeightsnone (count)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6), with at least one confidence answerPeriod1973-2024, survey years with the confidence itemsSourceGeneral Social Survey respondents in 3333 survey yearsEstimandcount of survey yearsWeightsnone (count)Populationyears in which the confidence items were fieldedPeriod1973-2024, survey years with the confidence itemsSourceGeneral Social Survey survey years.

    The classic account of falling confidence in business, labor and government in the 1960s and 1970s [3] described the level of confidence. This chapter describes the difference between the parties. Chapter 6.3, Trust in science (planned) takes up the science gap in depth, and here science is one institution of thirteen.

    The way the GSS interviews respondents changed during the period. It was in person, with some telephone interviews, through 2018. In 2021 it invited an address-based sample to answer online [4], in 2022 it interviewed in person and on the web side by side [5], and 2024 has both modes again. The web accounted for 87.3%87.3% percent of respondents interviewed on the webEstimandshare of respondents whose interview mode is web (unweighted count)n4,032Weightsnone (count)Populationall GSS respondents in the year, not only party identifiersPeriod2021SourceGeneral Social Survey of 2021 respondents, 46.1%46.1% percent of respondents interviewed on the webEstimandshare of respondents whose interview mode is web (unweighted count)n3,544Weightsnone (count)Populationall GSS respondents in the year, not only party identifiersPeriod2022SourceGeneral Social Survey of 2022 respondents and 53.2%53.2% percent of respondents interviewed on the webEstimandshare of respondents whose interview mode is web (unweighted count)n3,309Weightsnone (count)Populationall GSS respondents in the year, not only party identifiersPeriod2024SourceGeneral Social Survey of 2024 respondents. A change in a gap after 2020 can therefore come from a change in survey modeSurvey modeHow respondents were interviewed: in person, by phone, on paper or online. Mode can change answers, so a change in mode can look like a change in opinion.. The text compares 2022-2024 with earlier periods using in-person interviews, because every earlier survey was in person or mostly so.

    The average gap rose after the 1990s

    The mean absolute gap averages the sizes of the thirteen gaps and ignores their direction. Among in-person respondents it was 4.84.8 percentage points (mean over institutions of the absolute Democratic-minus-Republican gap)Estimandmean over thirteen institutions of the absolute Democratic-minus-Republican gap in the share with a great deal of confidence; in-person interviews only; seeded respondent bootstrap interval (2,000 replicates)Interval95% CI 4.3 to 5.5n10,281Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period1973-1980SourceGeneral Social Survey points in 1973-1980 and 4.14.1 percentage points (mean over institutions of the absolute Democratic-minus-Republican gap)Estimandmean over thirteen institutions of the absolute Democratic-minus-Republican gap in the share with a great deal of confidence; in-person interviews only; seeded respondent bootstrap interval (2,000 replicates)Interval95% CI 3.6 to 4.8n12,086Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period1991-2000SourceGeneral Social Survey in 1991-2000, then 6.26.2 percentage points (mean over institutions of the absolute Democratic-minus-Republican gap)Estimandmean over thirteen institutions of the absolute Democratic-minus-Republican gap in the share with a great deal of confidence; in-person interviews only; seeded respondent bootstrap interval (2,000 replicates)Interval95% CI 5.5 to 7.2n10,035Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2001-2010SourceGeneral Social Survey in 2001-2010, 6.76.7 percentage points (mean over institutions of the absolute Democratic-minus-Republican gap)Estimandmean over thirteen institutions of the absolute Democratic-minus-Republican gap in the share with a great deal of confidence; in-person interviews only; seeded respondent bootstrap interval (2,000 replicates)Interval95% CI 6.0 to 7.5n6,791Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2011-2018SourceGeneral Social Survey in 2011-2018 and 9.4 (95% CI 8.0 to 11.8)9.4 percentage points (mean over institutions of the absolute Democratic-minus-Republican gap)Estimandmean over thirteen institutions of the absolute Democratic-minus-Republican gap; in-person interviews only; seeded respondent bootstrap interval (2,000 replicates)Interval95% CI 8.0 to 11.8n1,980Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022-2024SourceGeneral Social Survey in 2022-2024 (Figure 1). The intervals for 2001-2010 and 2011-2018 lie above the interval for 1991-2000, and the interval for 2011-2018 lies above the interval for 1973-1980.

    Across all modes in 2021-2024, when 63.9%63.9% percent of respondents interviewed on the webEstimandshare of the respondents behind the 2021-2024 estimates whose interview mode is web (unweighted count)n5,321Weightsnone (count)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6), with at least one confidence answerPeriod2021, 2022 and 2024 pooledSourceGeneral Social Survey of the respondents were interviewed on the web, the mean absolute gap was 10.6 (95% CI 9.8 to 11.5)10.6 percentage points (mean over institutions of the absolute Democratic-minus-Republican gap)Estimandmean over thirteen institutions of the absolute Democratic-minus-Republican gap; all interview modes (the pooled 2021, 2022 and 2024 surveys, mostly web); seeded respondent bootstrap interval (2,000 replicates)Interval95% CI 9.8 to 11.5n7,993Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2021-2024SourceGeneral Social Survey, and among web respondents in 2022-2024 it was 12.0 (95% CI 10.6 to 13.6)12.0 percentage points (mean over institutions of the absolute Democratic-minus-Republican gap)Estimandmean over thirteen institutions of the absolute Democratic-minus-Republican gap; web interviews only; seeded respondent bootstrap interval (2,000 replicates)Interval95% CI 10.6 to 13.6n2,412Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022-2024SourceGeneral Social Survey.

    The pooled in-person value for 2022-2024 averages the surveys of 2022 and 2024, which differ. In person the average was 11.3 (95% CI 9.7 to 14.2)11.3 percentage points (mean over institutions of the absolute Democratic-minus-Republican gap)Estimandmean over the institutions that pass the suppression floor of the absolute Democratic-minus-Republican gap; in-person interviews only; seeded respondent bootstrap interval (2,000 replicates)Interval95% CI 9.7 to 14.2n1,158Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022SourceGeneral Social Survey in 2022, and its interval lies above the interval for 2011-2018. In 2024 it was 7.3 (95% CI 5.6 to 10.7)7.3 percentage points (mean over institutions of the absolute Democratic-minus-Republican gap)Estimandmean over the institutions that pass the suppression floor of the absolute Democratic-minus-Republican gap; in-person interviews only; organized labor is left out because its party cells fall below the suppression floor, so twelve institutions; seeded respondent bootstrap interval (2,000 replicates)Interval95% CI 5.6 to 10.7n822Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2024SourceGeneral Social Survey, an interval that overlaps the one for 2011-2018 and contains the 8.18.1 percentage points (mean over institutions of the absolute Democratic-minus-Republican gap)Estimandmean over the institutions that pass the suppression floor of the absolute Democratic-minus-Republican gap; in-person interviews only; seeded respondent bootstrap interval (2,000 replicates)Interval95% CI 6.7 to 10.1n1,640Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2018SourceGeneral Social Survey of 2018, and among web respondents it was 12.012.0 percentage points (mean over institutions of the absolute Democratic-minus-Republican gap)Estimandmean over the institutions that pass the suppression floor of the absolute Democratic-minus-Republican gap; web interviews only; seeded respondent bootstrap interval (2,000 replicates)Interval95% CI 10.0 to 14.8n1,256Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2024SourceGeneral Social Survey in 2024. The 2024 in-person average covers twelve institutions because organized labor falls below the suppression floor. The rise after 2018 therefore rests on the in-person interviews of 2022 and on the web interviews, and the in-person interviews of 2024 do not establish it.

    Figure 1

    The mean absolute party gap rose from 4.8 points in 1973-1980 to 6.7 in 2011-2018 and was 9.4 in person in 2022-2024

    Mean over thirteen institutions of the absolute Democratic-minus-Republican gap in the share with a great deal of confidence, percentage points, by survey year (line) and by pooled period (steps), with 95% intervals; in-person, all-mode and web views

    Show data
    Data for The mean absolute party gap rose from 4.8 points in 1973-1980 to 6.7 in 2011-2018 and was 9.4 in person in 2022-2024
    conf_highconf_lowmodenseriesvalueyear
    8.304.90all1,288Mean absolute gap6.101973
    n/an/aalln/aSampling-noise floor2.301973
    9.306.10all1,258Mean absolute gap7.301974
    n/an/aalln/aSampling-noise floor2.301974
    8.305.20all1,262Mean absolute gap6.501975
    n/an/aalln/aSampling-noise floor2.301975
    9.105.80all1,250Mean absolute gap7.301976
    n/an/aalln/aSampling-noise floor2.301976
    7.404.40all1,338Mean absolute gap5.501977
    n/an/aalln/aSampling-noise floor2.301977
    6.603.80all1,296Mean absolute gap4.601978
    n/an/aalln/aSampling-noise floor2.501978
    7.403.30all1,210Mean absolute gap4.601980
    n/an/aalln/aSampling-noise floor2.501980
    9.905.90all1,612Mean absolute gap7.301982
    n/an/aalln/aSampling-noise floor2.401982
    6.904.30all1,388Mean absolute gap4.901983
    n/an/aalln/aSampling-noise floor2.301983
    9.704.40all1,280Mean absolute gap61984
    n/an/aalln/aSampling-noise floor31984
    6.103.30all1,271Mean absolute gap3.801986
    n/an/aalln/aSampling-noise floor2.201986
    74.50all1,595Mean absolute gap5.101987
    n/an/aalln/aSampling-noise floor2.201987
    8.204.80all1,290Mean absolute gap61988
    n/an/aalln/aSampling-noise floor2.601988
    8.605.10all1,327Mean absolute gap6.501989
    n/an/aalln/aSampling-noise floor2.601989
    8.203.50all1,201Mean absolute gap5.201990
    n/an/aalln/aSampling-noise floor2.701990
    10.805.80all1,309Mean absolute gap7.901991
    n/an/aalln/aSampling-noise floor2.601991
    7.104.20all1,375Mean absolute gap51993
    n/an/aalln/aSampling-noise floor2.301993
    5.503.30all2,530Mean absolute gap3.901994
    n/an/aalln/aSampling-noise floor1.801994
    6.103.90all2,398Mean absolute gap4.601996
    n/an/aalln/aSampling-noise floor1.801996
    6.303.80all2,283Mean absolute gap4.501998
    n/an/aalln/aSampling-noise floor1.901998
    6.404.10all2,191Mean absolute gap4.702000
    n/an/aalln/aSampling-noise floor22000
    9.305.90all2,153Mean absolute gap6.602002
    n/an/aalln/aSampling-noise floor2.902002
    12.207.20all2,300Mean absolute gap92004
    n/an/aalln/aSampling-noise floor2.902004
    11.307.30all3,422Mean absolute gap8.902006
    n/an/aalln/aSampling-noise floor22006
    7.404.70all1,650Mean absolute gap5.802008
    n/an/aalln/aSampling-noise floor2.302008
    8.705.60all1,619Mean absolute gap6.402010
    n/an/aalln/aSampling-noise floor2.402010
    9.906.90all1,533Mean absolute gap7.902012
    n/an/aalln/aSampling-noise floor2.402012
    8.705.90all1,948Mean absolute gap7.102014
    n/an/aalln/aSampling-noise floor22014
    8.406all2,290Mean absolute gap72016
    n/an/aalln/aSampling-noise floor1.902016
    106.90all1,824Mean absolute gap8.102018
    n/an/aalln/aSampling-noise floor2.402018
    12.209.90all3,069Mean absolute gap112021
    n/an/aalln/aSampling-noise floor1.702021
    13.209.80all2,568Mean absolute gap11.502022
    n/an/aalln/aSampling-noise floor2.202022
    12.308.60all2,356Mean absolute gap10.102024
    n/an/aalln/aSampling-noise floor2.202024
    5.504.30all10,281Period estimate4.801973
    5.504.30all10,281Period estimate4.801980
    5.804.50all12,330Period estimate5.101981
    5.804.50all12,330Period estimate5.101990
    4.803.60all12,086Period estimate4.101991
    4.803.60all12,086Period estimate4.102000
    7.305.60all11,144Period estimate6.302001
    7.305.60all11,144Period estimate6.302010
    7.506all7,595Period estimate6.702011
    7.506all7,595Period estimate6.702018
    11.509.80all7,993Period estimate10.602021
    11.509.80all7,993Period estimate10.602024
    8.305inperson1,288Mean absolute gap6.101973
    n/an/ainpersonn/aSampling-noise floor2.301973
    9.206.10inperson1,258Mean absolute gap7.301974
    n/an/ainpersonn/aSampling-noise floor2.301974
    8.305.10inperson1,262Mean absolute gap6.501975
    n/an/ainpersonn/aSampling-noise floor2.301975
    9.105.90inperson1,250Mean absolute gap7.301976
    n/an/ainpersonn/aSampling-noise floor2.301976
    7.404.40inperson1,338Mean absolute gap5.501977
    n/an/ainpersonn/aSampling-noise floor2.301977
    6.603.90inperson1,296Mean absolute gap4.601978
    n/an/ainpersonn/aSampling-noise floor2.501978
    7.403.30inperson1,210Mean absolute gap4.601980
    n/an/ainpersonn/aSampling-noise floor2.501980
    9.805.90inperson1,612Mean absolute gap7.301982
    n/an/ainpersonn/aSampling-noise floor2.401982
    74.30inperson1,388Mean absolute gap4.901983
    n/an/ainpersonn/aSampling-noise floor2.301983
    9.704.50inperson1,280Mean absolute gap61984
    n/an/ainpersonn/aSampling-noise floor31984
    6.103.30inperson1,271Mean absolute gap3.801986
    n/an/ainpersonn/aSampling-noise floor2.201986
    74.50inperson1,595Mean absolute gap5.101987
    n/an/ainpersonn/aSampling-noise floor2.201987
    8.304.90inperson1,290Mean absolute gap61988
    n/an/ainpersonn/aSampling-noise floor2.601988
    8.705.20inperson1,327Mean absolute gap6.501989
    n/an/ainpersonn/aSampling-noise floor2.601989
    8.203.50inperson1,201Mean absolute gap5.201990
    n/an/ainpersonn/aSampling-noise floor2.701990
    10.706inperson1,309Mean absolute gap7.901991
    n/an/ainpersonn/aSampling-noise floor2.601991
    7.204.10inperson1,375Mean absolute gap51993
    n/an/ainpersonn/aSampling-noise floor2.301993
    5.503.30inperson2,530Mean absolute gap3.901994
    n/an/ainpersonn/aSampling-noise floor1.801994
    6.204inperson2,398Mean absolute gap4.601996
    n/an/ainpersonn/aSampling-noise floor1.801996
    6.203.80inperson2,283Mean absolute gap4.501998
    n/an/ainpersonn/aSampling-noise floor1.901998
    6.404.10inperson2,191Mean absolute gap4.702000
    n/an/ainpersonn/aSampling-noise floor22000
    9.405.90inperson2,153Mean absolute gap6.602002
    n/an/ainpersonn/aSampling-noise floor2.902002
    12.607.20inperson1,796Mean absolute gap9.202004
    n/an/ainpersonn/aSampling-noise floor3.102004
    11.207.30inperson3,190Mean absolute gap8.702006
    n/an/ainpersonn/aSampling-noise floor2.102006
    7.404.40inperson1,456Mean absolute gap5.502008
    n/an/ainpersonn/aSampling-noise floor2.402008
    95.80inperson1,440Mean absolute gap6.902010
    n/an/ainpersonn/aSampling-noise floor2.602010
    10.406.90inperson1,326Mean absolute gap7.902012
    n/an/ainpersonn/aSampling-noise floor2.702012
    96inperson1,700Mean absolute gap7.302014
    n/an/ainpersonn/aSampling-noise floor2.102014
    8.606.10inperson2,125Mean absolute gap7.102016
    n/an/ainpersonn/aSampling-noise floor22016
    10.106.70inperson1,640Mean absolute gap8.102018
    n/an/ainpersonn/aSampling-noise floor2.502018
    14.209.70inperson1,158Mean absolute gap11.302022
    n/an/ainpersonn/aSampling-noise floor3.302022
    10.705.60inperson822Mean absolute gap7.302024
    n/an/ainpersonn/aSampling-noise floor3.702024
    5.504.30inperson10,281Period estimate4.801973
    5.504.30inperson10,281Period estimate4.801980
    5.804.50inperson12,330Period estimate5.101981
    5.804.50inperson12,330Period estimate5.101990
    4.803.60inperson12,086Period estimate4.101991
    4.803.60inperson12,086Period estimate4.102000
    7.205.50inperson10,035Period estimate6.202001
    7.205.50inperson10,035Period estimate6.202010
    7.506inperson6,791Period estimate6.702011
    7.506inperson6,791Period estimate6.702018
    11.808inperson1,980Period estimate9.402022
    11.808inperson1,980Period estimate9.402024
    12.4010web2,677Mean absolute gap11.202021
    n/an/awebn/aSampling-noise floor1.802021
    15.8010.80web1,156Mean absolute gap13.402022
    n/an/awebn/aSampling-noise floor2.902022
    14.8010web1,256Mean absolute gap122024
    n/an/awebn/aSampling-noise floor2.702024
    13.6010.60web2,412Period estimate122022
    13.6010.60web2,412Period estimate122024
    Annual values sit above the period estimates because sampling error inflates absolute gaps, and in person the 2022 survey is above the 2011-2018 interval while the 2024 survey overlaps it. The line is the mean of the thirteen absolute gaps with a bootstrap interval, found by resampling respondents many times, the steps are pooled period estimates with 95% intervals, and the dashed line is the average that sampling noise alone would produce if every true gap were zero.Note: Banks and financial institutions were first asked in 1975, so the averages for 1973 and 1974 cover twelve institutions. Organized labor is left out of the 2024 in-person average because its party cells fall below the suppression floor. Every other average covers thirteen institutions.Note: The mean of absolute values is pushed up by sampling error when true gaps are small, which is why the noise floor is shown.Note: The bootstrap resamples respondents and ignores the survey's sampling design, and a design-based version of these intervals has not been computed.Note: Lines are broken between 2018 and the next survey year in each view, because there is no 2020 survey and the 2021 survey had no in-person interviews.Source: General Social Survey, 1973-2024; weighted (weight_use); Democrats and Republicans include leaners; bootstrap intervals with 2,000 replicates and a fixed seed

    An average of absolute values is inflated by sampling error, because a true gap of zero still produces a small absolute gap in a sample. The dashed line in the figure is the average that noise alone would produce. It is between 0.80.8 percentage points (mean over institutions of the absolute Democratic-minus-Republican gap)Estimandsmallest across the periods before 2022 of the mean absolute gap expected from sampling noise alone if every true gap were zero (sqrt(2/pi) times the mean standard error of the gaps); in-person interviews onlyWeightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period1973-2018, by periodSourceGeneral Social Survey and 1.21.2 percentage points (mean over institutions of the absolute Democratic-minus-Republican gap)Estimandlargest across the periods before 2022 of the mean absolute gap expected from sampling noise alone if every true gap were zero; in-person interviews onlyWeightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period1973-2018, by periodSourceGeneral Social Survey points in the pooled periods before 2022 and 2.62.6 percentage points (mean over institutions of the absolute Democratic-minus-Republican gap)Estimandmean absolute gap expected from sampling noise alone if every true gap were zero; in-person interviews onlyWeightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022-2024SourceGeneral Social Survey points for the smaller in-person sample of 2022-2024, and every period sits above its floor. Annual values before 2022 rest on one survey each, have a higher floor and run above the period estimates.

    Leaners do not drive the pattern. Counting only strong and weak identifiers, the mean absolute gap was 5.4 (95% CI 4.8 to 6.1)5.4 percentage points (mean over institutions of the absolute Democratic-minus-Republican gap)Estimandmean over thirteen institutions of the absolute gap between strong and weak Democrats and strong and weak Republicans; in-person interviews (all surveys before 2004 were in person); seeded respondent bootstrapInterval95% CI 4.8 to 6.1n7,750Weightsweight_use (wtssnrps from 2004, wtssps before)Populationstrong and weak party identifiers only (Democrats: partyid 0-1, Republicans: partyid 5-6)Period1973-1980SourceGeneral Social Survey points in 1973-1980 and 8.7 (95% CI 7.2 to 11.2)8.7 percentage points (mean over institutions of the absolute Democratic-minus-Republican gap)Estimandmean over thirteen institutions of the absolute gap between strong and weak Democrats and strong and weak Republicans; in-person interviews only; seeded respondent bootstrapInterval95% CI 7.2 to 11.2n1,434Weightsweight_use (wtssnrps from 2004, wtssps before)Populationstrong and weak party identifiers only (Democrats: partyid 0-1, Republicans: partyid 5-6)Period2022-2024SourceGeneral Social Survey in 2022-2024 among in-person respondents.

    The number of institutions with an estimated gap of 1010 percentage pointsEstimandcutoff on the absolute gap (constant)Weightsnone (count)Populationthreshold for a large gapPeriodall periodsSourceGeneral Social Survey points or more went from 11 institutionsEstimandcount of institutions whose estimated absolute Democratic-minus-Republican gap (the point estimate, not its interval) is 10 points or more; in-person interviewsWeightsweight_use (wtssnrps from 2004, wtssps before)Populationthe thirteen institutionsPeriod1973-1980SourceGeneral Social Survey in 1973-1980 to 22 institutionsEstimandcount of institutions whose estimated absolute Democratic-minus-Republican gap (the point estimate, not its interval) is 10 points or more; in-person interviewsWeightsweight_use (wtssnrps from 2004, wtssps before)Populationthe thirteen institutionsPeriod2011-2018SourceGeneral Social Survey in 2011-2018 and 77 institutionsEstimandcount of institutions whose estimated absolute Democratic-minus-Republican gap (the point estimate, not its interval) is 10 points or more; in-person interviewsWeightsweight_use (wtssnrps from 2004, wtssps before)Populationthe thirteen institutionsPeriod2022-2024SourceGeneral Social Survey among in-person respondents in 2022-2024.Counting by interval, the gap excludes zero on the Democratic side for 55 institutionsEstimandcount of institutions whose 95% interval for the Democratic-minus-Republican gap excludes zero on the Democratic side; in-person interviewsWeightsweight_use (wtssnrps from 2004, wtssps before)Populationthe thirteen institutionsPeriod1973-1980SourceGeneral Social Survey of the thirteen institutions in 1973-1980 and 77 institutionsEstimandcount of institutions whose 95% interval for the Democratic-minus-Republican gap excludes zero on the Democratic side; in-person interviewsWeightsweight_use (wtssnrps from 2004, wtssps before)Populationthe thirteen institutionsPeriod2022-2024SourceGeneral Social Survey in 2022-2024 in person, and on the Republican side for 66 institutionsEstimandcount of institutions whose 95% interval for the Democratic-minus-Republican gap excludes zero on the Republican side; in-person interviewsWeightsweight_use (wtssnrps from 2004, wtssps before)Populationthe thirteen institutionsPeriod1973-1980SourceGeneral Social Survey and 33 institutionsEstimandcount of institutions whose 95% interval for the Democratic-minus-Republican gap excludes zero on the Republican side; in-person interviewsWeightsweight_use (wtssnrps from 2004, wtssps before)Populationthe thirteen institutionsPeriod2022-2024SourceGeneral Social Survey. Those counts use point estimates, and medicine, at +10.4+10.4 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in medicine; in-person interviews only; normal interval on the Kish effective nInterval95% CI 2.5 to 18.3n1,309Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022-2024SourceGeneral Social Survey, sits just above the cutoff.

    Parties differ most on science and the Supreme Court

    Among respondents interviewed in person in 2022-2024, the science gap was +19.8 (95% CI 11.8 to 27.8)+19.8 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in the scientific community; in-person interviews only; normal interval on the Kish effective nInterval95% CI 11.8 to 27.8n1,259Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022-2024SourceGeneral Social Survey points and the Supreme Court gap −16.2 (95% CI −22.8 to −9.5)−16.2 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in supreme court; in-person interviews only; normal interval on the Kish effective nInterval95% CI −22.8 to −9.5n1,289Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022-2024SourceGeneral Social Survey (Figure 2). The estimated gap is 1010 percentage pointsEstimandcutoff on the absolute gap (constant)Weightsnone (count)Populationthreshold for a large gapPeriodall periodsSourceGeneral Social Survey points or more for 77 institutionsEstimandcount of institutions whose estimated absolute Democratic-minus-Republican gap (the point estimate, not its interval) is 10 points or more; in-person interviewsWeightsweight_use (wtssnrps from 2004, wtssps before)Populationthe thirteen institutionsPeriod2022-2024SourceGeneral Social Survey institutions: science, the Supreme Court, the executive branch, education, the military, organized religion and medicine.

    The intervals for major companies, banks and organized labor include zero, but each reaches past the Congress gap of +5.0+5.0 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in congress; in-person interviews only; normal interval on the Kish effective nInterval95% CI 1.3 to 8.8n1,296Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022-2024SourceGeneral Social Survey points, so a gap of that size is not ruled out for any of the three. Web respondents give different results for companies and labor, reported below.

    Figure 2

    In person in 2022-2024 the estimated party gap is 10 points or more for 7 of thirteen institutions

    Share with a great deal of confidence in the people running each institution, percent, Democrats and Republicans including leaners, in-person interviews, 2022 and 2024 pooled, 95% intervals

    • Republicans
    • Democrats
    Show data
    Data for In person in 2022-2024 the estimated party gap is 10 points or more for 7 of thirteen institutions
    InstitutionRepublican share (%)Republican share, lower endRepublican share, upper endDemocratic share (%)Democratic share, lower endDemocratic share, upper endGap (points)Republican respondentsDemocratic respondents
    Scientific community26.420.832.046.240.551.9+19.8581678
    Executive branch5.12.37.819.715.224.2+14.6604691
    Education15.410.819.929.724.634.8+14.4607698
    Medicine28.522.934.239.033.544.4+10.4607702
    Press2.60.64.510.46.913.8+7.8604693
    Television4.92.27.612.08.315.6+7.1603688
    Congress3.00.95.28.05.011.1+5.0602694
    Organized labor17.912.123.718.713.523.9+0.8444520
    Banks and finance19.414.424.419.715.324.2+0.3607700
    Major companies16.611.821.315.711.619.9−0.8587678
    Organized religion25.720.131.214.010.118.0−11.6585679
    Military59.253.165.445.539.951.1−13.7609693
    Supreme Court26.921.332.610.87.314.3−16.2596693
    The two dots are farthest apart for science and the Supreme Court, and the dots and intervals overlap for banks, major companies and organized labor. Each row joins the Republican share (ring) to the Democratic share (solid dot), and rows run from the largest Democratic advantage to the largest Republican advantage. The right-hand column, headed Change, is the Democratic share minus the Republican share, the gap in percentage points.Note: The dots pool the in-person interviews of 2022 and 2024, and there is no in-person survey in 2021. The number of respondents behind each dot is in the data table.Note: Intervals assume a simple random sample of the weighted respondents (the Kish effective sample size); they do not use the survey's sampling design.Source: General Social Survey, 1973-2024; weighted (weight_use); Democrats and Republicans include leaners (in-person interviews, 2022 and 2024 pooled); intervals on the Kish effective sample size

    Congress, television and the press have gaps of +5.0+5.0 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in congress; in-person interviews only; normal interval on the Kish effective nInterval95% CI 1.3 to 8.8n1,296Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022-2024SourceGeneral Social Survey, +7.1+7.1 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in television; in-person interviews only; normal interval on the Kish effective nInterval95% CI 2.5 to 11.6n1,291Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022-2024SourceGeneral Social Survey and +7.8+7.8 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in press; in-person interviews only; normal interval on the Kish effective nInterval95% CI 3.8 to 11.8n1,297Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022-2024SourceGeneral Social Survey points at low levels of confidence for both parties. In Congress, 8.0%8.0% percent of respondents answering a great deal of confidenceEstimandweighted share saying a great deal of confidence in congress; in-person interviews onlyInterval95% CI 5.0% to 11.1%n694Weightsweight_use (wtssnrps from 2004, wtssps before)PopulationDemocrats who identify with or lean toward the partyPeriod2022-2024SourceGeneral Social Survey of Democrats and 3.0%3.0% percent of respondents answering a great deal of confidenceEstimandweighted share saying a great deal of confidence in congress; in-person interviews onlyInterval95% CI 0.9% to 5.2%n602Weightsweight_use (wtssnrps from 2004, wtssps before)PopulationRepublicans who identify with or lean toward the partyPeriod2022-2024SourceGeneral Social Survey of Republicans had a great deal of confidence, and the press shares are 10.4%10.4% percent of respondents answering a great deal of confidenceEstimandweighted share saying a great deal of confidence in press; in-person interviews onlyInterval95% CI 6.9% to 13.8%n693Weightsweight_use (wtssnrps from 2004, wtssps before)PopulationDemocrats who identify with or lean toward the partyPeriod2022-2024SourceGeneral Social Survey and 2.6%2.6% percent of respondents answering a great deal of confidenceEstimandweighted share saying a great deal of confidence in press; in-person interviews onlyInterval95% CI 0.6% to 4.5%n604Weightsweight_use (wtssnrps from 2004, wtssps before)PopulationRepublicans who identify with or lean toward the partyPeriod2022-2024SourceGeneral Social Survey, so a press gap of that size is large next to them. Gaps are compared in points, so a gap of a given size means more at a low level of confidence than at a high one, and the averages above do not adjust for the level.

    The levels behind the gaps are lower than in the 1970s for most institutions (Figure 3). From 1973-1980 to 2022-2024 among in-person respondents, the share with a great deal of confidence fell for both parties in 88 institutionsEstimandcount of institutions where the 95% interval of the share with a great deal of confidence in 2022-2024 lies below the interval for 1973-1980 for both Democrats and Republicans; in-person interviewsWeightsweight_use (wtssnrps from 2004, wtssps before)Populationthe thirteen institutionsPeriod1973-1980 to 2022-2024SourceGeneral Social Survey of the thirteen institutions, with intervals that do not overlap, and rose for both parties only in the military.

    Figure 3

    Between 1973-1980 and 2022-2024 the share with a great deal of confidence fell for both parties in 8 of thirteen institutions

    Share with "a great deal" of confidence in the people running each institution, percent, by survey year, Democrats (solid line) and Republicans (dashed line) including leaners, 95% intervals; in-person and all-mode views

    Share with a great deal of confidence (%)

    • Democrats
    • Republicans
    Scientific communityExecutive branchEducationMedicinePressTelevisionCongressOrganized laborBanks and financeMajor companiesOrganized religionMilitarySupreme Court
    Show data
    Data for Between 1973-1980 and 2022-2024 the share with a great deal of confidence fell for both parties in 8 of thirteen institutions
    conf_highconf_lowinstitutionmodenpartyshareweb_pctyear
    43.9036.20Scientific communityall740Democrats40.1001973
    53.1045.20Scientific communityall745Democrats49.2001974
    44.8036.90Scientific communityall730Democrats40.9001975
    49.4041.40Scientific communityall739Democrats45.4001976
    45.9038.30Scientific communityall810Democrats42.1001977
    47.3038.30Scientific communityall739Democrats42.8001978
    46.6037.40Scientific communityall683Democrats4201980
    42.2034.30Scientific communityall993Democrats38.3001982
    49.4040.50Scientific communityall789Democrats44.9001983
    5442.80Scientific communityall463Democrats48.4001984
    44.2036.20Scientific communityall697Democrats40.2001986
    50.2042.20Scientific communityall951Democrats46.2001987
    50.5040.20Scientific communityall443Democrats45.4001988
    48.2037.50Scientific communityall433Democrats42.9001989
    45.3034Scientific communityall379Democrats39.6001990
    42.9032.50Scientific communityall422Democrats37.7001991
    45.4035.10Scientific communityall451Democrats40.3001993
    45.2037.90Scientific communityall872Democrats41.5001994
    47.7040Scientific communityall807Democrats43.9001996
    4840.10Scientific communityall795Democrats44.1001998
    51.3043.20Scientific communityall754Democrats47.2002000
    48.7037Scientific communityall385Democrats42.9002002
    48.9036.80Scientific communityall387Democrats42.9002004
    50.2041.90Scientific communityall809Democrats4602006
    4838.90Scientific communityall639Democrats43.4002008
    51.5041.40Scientific communityall640Democrats46.5002010
    51.8041.90Scientific communityall612Democrats46.9002012
    50.8042.60Scientific communityall747Democrats46.7002014
    53.8045.80Scientific communityall904Democrats49.8002016
    60.3050.90Scientific communityall673Democrats55.6002018
    67.2060.20Scientific communityall1,194Democrats63.70882021
    5848.10Scientific communityall934Democrats53.10472022
    55.6045.70Scientific communityall812Democrats50.60542024
    49.7039.50Scientific communityall429Republicans44.6001973
    60.9050.20Scientific communityall384Republicans55.6001974
    50.5039.90Scientific communityall403Republicans45.2001975
    60.7049.80Scientific communityall379Republicans55.3001976
    56.4045.90Scientific communityall424Republicans51.2001977
    46.3034.70Scientific communityall452Republicans40.5001978
    56.6045.30Scientific communityall416Republicans50.9001980
    57.7046.30Scientific communityall460Republicans5201982
    53.3042.70Scientific communityall516Republicans4801983
    60.3047.10Scientific communityall336Republicans53.7001984
    5242.50Scientific communityall504Republicans47.2001986
    5747.10Scientific communityall535Republicans5201987
    49.4038.10Scientific communityall367Republicans43.7001988
    55.9045Scientific communityall385Republicans50.5001989
    50.4038.80Scientific communityall347Republicans44.6001990
    55.8045Scientific communityall401Republicans50.4001991
    47.8036.80Scientific communityall381Republicans42.3001993
    47.8039.80Scientific communityall743Republicans43.8001994
    49.9041.50Scientific communityall648Republicans45.7001996
    50.1041.60Scientific communityall635Republicans45.9001998
    50.9042.20Scientific communityall623Republicans46.5002000
    48.2035.30Scientific communityall290Republicans41.7002002
    51.3038.90Scientific communityall330Republicans45.1002004
    5242.80Scientific communityall640Republicans47.4002006
    41.3030.60Scientific communityall432Republicans3602008
    46.2034.80Scientific communityall416Republicans40.5002010
    45.7033.50Scientific communityall384Republicans39.6002012
    44.1034.30Scientific communityall499Republicans39.2002014
    41.6032.60Scientific communityall600Republicans37.1002016
    49.4037.80Scientific communityall516Republicans43.6002018
    37.5029.70Scientific communityall822Republicans33.60852021
    27.4018Scientific communityall761Republicans22.70462022
    29.3020.50Scientific communityall704Republicans24.90572024
    26.1019.70Executive branchall785Democrats22.9001973
    13.708.90Executive branchall809Democrats11.3001974
    13.909Executive branchall799Democrats11.4001975
    13.208.50Executive branchall812Democrats10.8001976
    34.2027.20Executive branchall852Democrats30.7001977
    18.7012.30Executive branchall780Democrats15.5001978
    19.6013Executive branchall730Democrats16.3001980
    16.8011.20Executive branchall1,074Democrats1401982
    126.90Executive branchall821Democrats9.4001983
    15.208Executive branchall486Democrats11.6001984
    20.3014.20Executive branchall724Democrats17.3001986
    16.3011Executive branchall996Democrats13.7001987
    14.508.10Executive branchall469Democrats11.3001988
    20.7012.90Executive branchall458Democrats16.8001989
    24.5015.60Executive branchall404Democrats20.1001990
    24.8016.40Executive branchall445Democrats20.6001991
    19.8012.40Executive branchall476Democrats16.1001993
    18.5013.20Executive branchall905Democrats15.9001994
    17.9012.40Executive branchall847Democrats15.2001996
    21.2015.20Executive branchall840Democrats18.2001998
    22.4016.20Executive branchall791Democrats19.3002000
    23.4014.40Executive branchall403Democrats18.9002002
    15.607.80Executive branchall395Democrats11.7002004
    10.205.70Executive branchall831Democrats802006
    11.506.30Executive branchall660Democrats8.9002008
    29.6020.90Executive branchall653Democrats25.3002010
    22.9015.20Executive branchall629Democrats1902012
    20.9014.60Executive branchall754Democrats17.7002014
    2215.80Executive branchall917Democrats18.9002016
    10.405.50Executive branchall683Democrats7.9002018
    21.7016Executive branchall1,191Democrats18.90882021
    22.5014.80Executive branchall936Democrats18.60472022
    25.2017.20Executive branchall820Democrats21.20542024
    50.6040.90Executive branchall475Republicans45.8001973
    24.7016.30Executive branchall424Republicans20.5001974
    24.3016.10Executive branchall441Republicans20.2001975
    22.8014.50Executive branchall406Republicans18.7001976
    28.8020Executive branchall444Republicans24.4001977
    11.605.30Executive branchall478Republicans8.5001978
    15.108.10Executive branchall445Republicans11.6001980
    38.5028.10Executive branchall496Republicans33.3001982
    2718.30Executive branchall534Republicans22.7001983
    37.3025.30Executive branchall348Republicans31.3001984
    36.5027.70Executive branchall524Republicans32.1001986
    33.7024.80Executive branchall550Republicans29.2001987
    32.3022.40Executive branchall383Republicans27.3001988
    34.5024.60Executive branchall398Republicans29.5001989
    36.1025.50Executive branchall365Republicans30.8001990
    42.5032.30Executive branchall419Republicans37.4001991
    104.40Executive branchall401Republicans7.2001993
    9.205.10Executive branchall766Republicans7.2001994
    9.505.20Executive branchall677Republicans7.3001996
    127.10Executive branchall660Republicans9.5001998
    15.409.70Executive branchall640Republicans12.5002000
    51.7038.90Executive branchall300Republicans45.3002002
    43.7031.60Executive branchall331Republicans37.7002004
    32.5024.30Executive branchall654Republicans28.4002006
    18.8011Executive branchall438Republicans14.9002008
    12.505.90Executive branchall427Republicans9.2002010
    10.504.10Executive branchall398Republicans7.3002012
    51.40Executive branchall512Republicans3.2002014
    105.10Executive branchall616Republicans7.6002016
    25.7016.40Executive branchall528Republicans2102018
    10.405.90Executive branchall826Republicans8.10852021
    5.401.30Executive branchall777Republicans3.30452022
    7.102.70Executive branchall718Republicans4.90562024
    42.8035.40Educationall790Democrats39.1001973
    5850.50Educationall813Democrats54.2001974
    36.2029Educationall804Democrats32.6001975
    44.4036.90Educationall817Democrats40.6001976
    45.5038.10Educationall866Democrats41.8001977
    36.1027.90Educationall790Democrats3201978
    3728.60Educationall741Democrats32.8001980
    37.5030Educationall1,079Democrats33.7001982
    34.5026.50Educationall829Democrats30.5001983
    34.6024.50Educationall490Democrats29.6001984
    33.2025.90Educationall731Democrats29.6001986
    4133.50Educationall1,010Democrats37.2001987
    33.4024.30Educationall473Democrats28.9001988
    37.3027.50Educationall470Democrats32.4001989
    31.3021.60Educationall413Democrats26.4001990
    3929.10Educationall447Democrats3401991
    27.9019.30Educationall480Democrats23.6001993
    30.3023.80Educationall913Democrats27.1001994
    3023.40Educationall872Democrats26.7001996
    32.8025.80Educationall852Democrats29.3001998
    34.4027.20Educationall820Democrats30.8002000
    33.8023.40Educationall409Democrats28.6002002
    34.4023.60Educationall401Democrats2902004
    31.4024.10Educationall846Democrats27.7002006
    36.4027.90Educationall666Democrats32.1002008
    32.7023.70Educationall661Democrats28.2002010
    34.5025.60Educationall640Democrats30.1002012
    33.2025.70Educationall763Democrats29.5002014
    33.2026Educationall931Democrats29.6002016
    32.6024.20Educationall686Democrats28.4002018
    22.5016.70Educationall1,196Democrats19.60882021
    27.8019.40Educationall938Democrats23.60472022
    29.5021.10Educationall828Democrats25.30532024
    42.2032.80Educationall476Republicans37.5001973
    47.4037.20Educationall427Republicans42.3001974
    34.9025.50Educationall441Republicans30.2001975
    41.3031.10Educationall408Republicans36.2001976
    43.7033.90Educationall457Republicans38.8001977
    30.7020.70Educationall485Republicans25.7001978
    33.4023.50Educationall446Republicans28.5001980
    40.6030.20Educationall500Republicans35.4001982
    36.3026.60Educationall534Republicans31.5001983
    34.4022.60Educationall344Republicans28.5001984
    31.9023.50Educationall524Republicans27.7001986
    37.1028.10Educationall561Republicans32.6001987
    37.7027.30Educationall386Republicans32.5001988
    35.6025.70Educationall407Republicans30.6001989
    33.9023.60Educationall364Republicans28.7001990
    31.8022.40Educationall422Republicans27.1001991
    25.9017Educationall405Republicans21.5001993
    29.4022.40Educationall765Republicans25.9001994
    25.4018.60Educationall687Republicans2201996
    27.2020.20Educationall669Republicans23.7001998
    27.4020.10Educationall647Republicans23.8002000
    2918.10Educationall300Republicans23.5002002
    34.2023Educationall335Republicans28.6002004
    32.9024.70Educationall661Republicans28.8002006
    34.9024.80Educationall444Republicans29.8002008
    32.4022.30Educationall430Republicans27.4002010
    22.5013.10Educationall401Republicans17.8002012
    21.9014.20Educationall516Republicans1802014
    2517.40Educationall620Republicans21.2002016
    23.8014.80Educationall530Republicans19.3002018
    17.9012Educationall823Republicans14.90852021
    13.907Educationall777Republicans10.50452022
    16.809.80Educationall715Republicans13.30552024
    56.8049.20Medicineall798Democrats5301973
    64.5057.20Medicineall814Democrats60.8001974
    54.1046.40Medicineall803Democrats50.3001975
    58.3050.70Medicineall822Democrats54.5001976
    57.1049.60Medicineall867Democrats53.3001977
    50.7042Medicineall794Democrats46.4001978
    59.5050.70Medicineall746Democrats55.1001980
    49.7041.90Medicineall1,094Democrats45.8001982
    57.3048.50Medicineall837Democrats52.9001983
    55.4044.40Medicineall492Democrats49.9001984
    50.4042.30Medicineall730Democrats46.3001986
    56.3048.50Medicineall1,014Democrats52.4001987
    54.9044.80Medicineall477Democrats49.8001988
    51.7041.30Medicineall470Democrats46.5001989
    49.9039Medicineall416Democrats44.4001990
    47.9037.70Medicineall449Democrats42.8001991
    44.8034.90Medicineall488Democrats39.9001993
    45.6038.40Medicineall918Democrats4201994
    47.8040.40Medicineall869Democrats44.1001996
    48.6040.90Medicineall850Democrats44.7001998
    50.4042.60Medicineall815Democrats46.5002000
    45.3034.10Medicineall409Democrats39.7002002
    41.8030.30Medicineall400Democrats36.1002004
    44.2036.20Medicineall847Democrats40.2002006
    43.5034.60Medicineall668Democrats39.1002008
    47.8037.90Medicineall660Democrats42.9002010
    46.4036.90Medicineall640Democrats41.7002012
    44.3036.30Medicineall763Democrats40.3002014
    41.2033.60Medicineall928Democrats37.4002016
    42.8033.70Medicineall687Democrats38.2002018
    48.5041.30Medicineall1,195Democrats44.90882021
    46.4036.70Medicineall945Democrats41.50462022
    39.3030Medicineall826Democrats34.60532024
    62.2052.60Medicineall476Republicans57.4001973
    69.7059.90Medicineall426Republicans64.8001974
    63.4053.40Medicineall442Republicans58.4001975
    62.8052.30Medicineall410Republicans57.5001976
    62.1052.10Medicineall458Republicans57.1001977
    55.7044.40Medicineall488Republicans50.1001978
    60.5049.60Medicineall449Republicans5501980
    5443.10Medicineall502Republicans48.5001982
    58.6048.30Medicineall535Republicans53.5001983
    61.7048.80Medicineall351Republicans55.2001984
    56.3046.90Medicineall524Republicans51.6001986
    60.6050.90Medicineall558Republicans55.8001987
    6554.20Medicineall390Republicans59.6001988
    58.2047.50Medicineall407Republicans52.9001989
    54.3042.90Medicineall366Republicans48.6001990
    60.2049.70Medicineall423Republicans54.9001991
    48.4037.70Medicineall402Republicans43.1001993
    50.5042.60Medicineall769Republicans46.5001994
    51.9043.70Medicineall686Republicans47.8001996
    51.9043.60Medicineall669Republicans47.8001998
    51.1042.70Medicineall649Republicans46.9002000
    46.1033.60Medicineall302Republicans39.9002002
    47.2035Medicineall334Republicans41.1002004
    48.7039.60Medicineall658Republicans44.1002006
    47.9037Medicineall447Republicans42.4002008
    50.2038.90Medicineall430Republicans44.6002010
    48.9036.70Medicineall400Republicans42.8002012
    44.6034.90Medicineall517Republicans39.8002014
    42.1033.20Medicineall621Republicans37.7002016
    44.8033.70Medicineall532Republicans39.2002018
    3830.20Medicineall824Republicans34.10852021
    31.4021.50Medicineall778Republicans26.40452022
    24.6016.40Medicineall716Republicans20.50562024
    29.6022.90Pressall795Democrats26.2001973
    33.6026.60Pressall816Democrats30.1001974
    32.5025.50Pressall794Democrats2901975
    36.9029.70Pressall815Democrats33.3001976
    31.3024.50Pressall861Democrats27.9001977
    25.6018.40Pressall792Democrats2201978
    28.8021Pressall736Democrats24.9001980
    2518.40Pressall1,074Democrats21.7001982
    17.5011.30Pressall830Democrats14.4001983
    22.1013.60Pressall487Democrats17.9001984
    24.1017.50Pressall727Democrats20.8001986
    25.9019.30Pressall995Democrats22.6001987
    26.8018.40Pressall471Democrats22.6001988
    27.5018.70Pressall462Democrats23.1001989
    19.6011.50Pressall403Democrats15.6001990
    18.2010.90Pressall445Democrats14.6001991
    15.909.20Pressall484Democrats12.6001993
    14.9010.10Pressall909Democrats12.5001994
    15.8010.70Pressall856Democrats13.2001996
    11.306.90Pressall844Democrats9.1001998
    12.607.80Pressall803Democrats10.2002000
    18.1010.10Pressall407Democrats14.1002002
    16.108.30Pressall401Democrats12.2002004
    14.909.50Pressall844Democrats12.2002006
    14.608.70Pressall664Democrats11.6002008
    17.5010.50Pressall654Democrats1402010
    15.909.40Pressall632Democrats12.7002012
    13.508.30Pressall753Democrats10.9002014
    12.307.60Pressall929Democrats1002016
    24.1016.60Pressall689Democrats20.3002018
    21.2015.50Pressall1,189Democrats18.40882021
    147.90Pressall939Democrats11472022
    18.2011.30Pressall822Democrats14.70542024
    24.8016.90Pressall474Republicans20.8001973
    2113.30Pressall426Republicans17.1001974
    23.2015.10Pressall438Republicans19.1001975
    25.9017.20Pressall412Republicans21.5001976
    26.9018.40Pressall452Republicans22.7001977
    25.2016Pressall484Republicans20.6001978
    2717.80Pressall449Republicans22.4001980
    17.209.70Pressall496Republicans13.4001982
    14.407.90Pressall535Republicans11.2001983
    2615.50Pressall351Republicans20.8001984
    20.4013.40Pressall522Republicans16.9001986
    16.9010.30Pressall560Republicans13.6001987
    2011.80Pressall382Republicans15.9001988
    16.809.60Pressall404Republicans13.2001989
    1810Pressall361Republicans1401990
    21.9013.80Pressall419Republicans17.8001991
    12.105.80Pressall400Republicans8.9001993
    116.50Pressall765Republicans8.8001994
    12.307.40Pressall678Republicans9.9001996
    11.406.60Pressall662Republicans901998
    12.707.50Pressall648Republicans10.1002000
    9.903.50Pressall301Republicans6.7002002
    8.402.70Pressall335Republicans5.6002004
    9.504.90Pressall659Republicans7.2002006
    7.102.40Pressall445Republicans4.8002008
    10.904.80Pressall430Republicans7.9002010
    7.302.10Pressall396Republicans4.7002012
    5.201.60Pressall517Republicans3.4002014
    6.302.50Pressall616Republicans4.4002016
    93.50Pressall531Republicans6.2002018
    4.201.50Pressall825Republicans2.80852021
    4.901.10Pressall776Republicans3452022
    4.601.20Pressall715Republicans2.90562024
    23.2017.10Televisionall798Democrats20.1001973
    30.6023.90Televisionall816Democrats27.3001974
    2417.70Televisionall794Democrats20.8001975
    23.7017.60Televisionall815Democrats20.7001976
    21.9016Televisionall865Democrats1901977
    18.7012.40Televisionall788Democrats15.5001978
    22.1015.20Televisionall744Democrats18.7001980
    20.5014.40Televisionall1,083Democrats17.4001982
    17.8011.50Televisionall835Democrats14.7001983
    16.308.90Televisionall490Democrats12.6001984
    17.9012.10Televisionall726Democrats1501986
    14.609.50Televisionall1,007Democrats12.1001987
    20.7013.20Televisionall475Democrats16.9001988
    20.3012.60Televisionall466Democrats16.5001989
    19.6011.60Televisionall415Democrats15.6001990
    19.6012.10Televisionall452Democrats15.9001991
    17.5010.50Televisionall488Democrats1401993
    14.9010Televisionall914Democrats12.4001994
    15.6010.50Televisionall866Democrats13.1001996
    12.607.90Televisionall850Democrats10.2001998
    11.807.20Televisionall808Democrats9.5002000
    15.207.90Televisionall405Democrats11.5002002
    16.108.30Televisionall401Democrats12.2002004
    12.807.80Televisionall846Democrats10.3002006
    12.907.40Televisionall660Democrats10.1002008
    15.909.30Televisionall659Democrats12.6002010
    13.507.50Televisionall631Democrats10.5002012
    15.309.90Televisionall758Democrats12.6002014
    12.707.90Televisionall924Democrats10.3002016
    13.107.40Televisionall682Democrats10.2002018
    14.109.40Televisionall1,197Democrats11.80882021
    137Televisionall931Democrats10472022
    126.40Televisionall823Democrats9.20542024
    21.1013.70Televisionall473Republicans17.4001973
    19.6012.10Televisionall427Republicans15.9001974
    20.7013Televisionall439Republicans16.8001975
    17.6010.20Televisionall409Republicans13.9001976
    21.7013.90Televisionall451Republicans17.8001977
    21.5012.90Televisionall483Republicans17.2001978
    18.9011Televisionall448Republicans1501980
    15.408.30Televisionall498Republicans11.9001982
    136.80Televisionall534Republicans9.9001983
    18.109.20Televisionall351Republicans13.7001984
    1811.30Televisionall520Republicans14.6001986
    14.608.40Televisionall560Republicans11.5001987
    16.709.30Televisionall387Republicans1301988
    15.408.40Televisionall403Republicans11.9001989
    17.409.60Televisionall366Republicans13.5001990
    18.5011.10Televisionall423Republicans14.8001991
    11.705.60Televisionall404Republicans8.6001993
    9.805.60Televisionall765Republicans7.7001994
    12.207.30Televisionall677Republicans9.7001996
    11.306.60Televisionall669Republicans8.9001998
    12.507.40Televisionall647Republicans9.9002000
    10.603.90Televisionall302Republicans7.2002002
    11.104.40Televisionall334Republicans7.8002004
    7.703.60Televisionall654Republicans5.6002006
    10.804.90Televisionall446Republicans7.9002008
    17.409.60Televisionall430Republicans13.5002010
    12.205.20Televisionall393Republicans8.7002012
    104.80Televisionall513Republicans7.4002014
    10.905.80Televisionall618Republicans8.3002016
    10.304.40Televisionall532Republicans7.3002018
    8.204.20Televisionall824Republicans6.20852021
    5.401.30Televisionall775Republicans3.40452022
    6.902.60Televisionall719Republicans4.70562024
    30.2023.40Congressall787Democrats26.8001973
    22.5016.50Congressall803Democrats19.5001974
    18.6013Congressall794Democrats15.8001975
    1812.50Congressall813Democrats15.3001976
    24.4018.30Congressall859Democrats21.4001977
    18.3012Congressall783Democrats15.1001978
    13.507.90Congressall727Democrats10.7001980
    17.7012Congressall1,082Democrats14.9001982
    17.3011.10Congressall822Democrats14.2001983
    19.4011.40Congressall488Democrats15.4001984
    20.7014.50Congressall723Democrats17.6001986
    22.3016.10Congressall992Democrats19.2001987
    23.1015.10Congressall468Democrats19.1001988
    23.3015.10Congressall464Democrats19.2001989
    18.1010.30Congressall402Democrats14.2001990
    20.4012.60Congressall437Democrats16.5001991
    115.50Congressall484Democrats8.3001993
    12.207.80Congressall899Democrats1001994
    9.505.50Congressall854Democrats7.5001996
    14.809.70Congressall832Democrats12.2001998
    16.5011Congressall791Democrats13.7002000
    20.1011.70Congressall404Democrats15.9002002
    20.4011.70Congressall399Democrats16.1002004
    12.507.50Congressall833Democrats1002006
    14.608.80Congressall661Democrats11.7002008
    17.5010.60Congressall653Democrats1402010
    10.205Congressall628Democrats7.6002012
    9.905.50Congressall753Democrats7.7002014
    7.203.60Congressall923Democrats5.4002016
    7.803.40Congressall677Democrats5.6002018
    11.407.20Congressall1,193Democrats9.30892021
    8.603.80Congressall938Democrats6.20472022
    9.704.70Congressall825Democrats7.20532024
    29.5021Congressall469Republicans25.3001973
    18.9011.50Congressall420Republicans15.2001974
    16.209.30Congressall437Republicans12.7001975
    16.309.20Congressall410Republicans12.8001976
    20.2012.60Congressall445Republicans16.4001977
    12.906.20Congressall478Republicans9.6001978
    9.704.10Congressall446Republicans6.9001980
    19.5011.60Congressall494Republicans15.6001982
    10.905.20Congressall532Republicans8.1001983
    19.9010.60Congressall349Republicans15.2001984
    20.4013.40Congressall524Republicans16.9001986
    17.7010.90Congressall555Republicans14.3001987
    18.5010.70Congressall386Republicans14.6001988
    2213.80Congressall402Republicans17.9001989
    23.6014.60Congressall367Republicans19.1001990
    25.5016.90Congressall417Republicans21.2001991
    8.703.50Congressall403Republicans6.1001993
    95Congressall762Republicans701994
    11.706.90Congressall682Republicans9.3001996
    137.90Congressall663Republicans10.4001998
    18.4012.20Congressall642Republicans15.3002000
    19.1010.10Congressall301Republicans14.6002002
    23.5013.80Congressall331Republicans18.6002004
    18.4011.90Congressall655Republicans15.1002006
    12.606.20Congressall441Republicans9.4002008
    10.104.20Congressall429Republicans7.2002010
    10.704.20Congressall397Republicans7.4002012
    4.801.30Congressall509Republicans3.1002014
    7.903.60Congressall618Republicans5.8002016
    9.703.90Congressall529Republicans6.8002018
    6.202.80Congressall827Republicans4.50852021
    4.500.900Congressall775Republicans2.70452022
    4.901.40Congressall716Republicans3.10562024
    22.1016.10Organized laborall770Democrats19.1001973
    26.4020Organized laborall803Democrats23.2001974
    15.5010.30Organized laborall765Democrats12.9001975
    18.5012.80Organized laborall770Democrats15.7001976
    21.1015.30Organized laborall847Democrats18.2001977
    17.6011.30Organized laborall770Democrats14.5001978
    21.8014.70Organized laborall714Democrats18.2001980
    20.4014.30Organized laborall1,056Democrats17.3001982
    13.908.30Organized laborall817Democrats11.1001983
    136.50Organized laborall488Democrats9.7001984
    148.80Organized laborall704Democrats11.4001986
    15.6010.20Organized laborall967Democrats12.9001987
    15.709Organized laborall461Democrats12.3001988
    18.5011.10Organized laborall448Democrats14.8001989
    17.309.60Organized laborall400Democrats13.4001990
    16.509.40Organized laborall428Democrats12.9001991
    13.106.90Organized laborall463Democrats1001993
    17.6012.30Organized laborall870Democrats14.9001994
    19.7013.90Organized laborall799Democrats16.8001996
    17.6011.90Organized laborall791Democrats14.7001998
    19.4013.40Organized laborall746Democrats16.4002000
    17.209.30Organized laborall396Democrats13.2002002
    19.9011.10Organized laborall384Democrats15.5002004
    16.3010.60Organized laborall816Democrats13.4002006
    19.5012.80Organized laborall643Democrats16.1002008
    18.3011Organized laborall634Democrats14.7002010
    20.1012.80Organized laborall619Democrats16.5002012
    1812Organized laborall734Democrats1502014
    19.5013.60Organized laborall908Democrats16.6002016
    16.6010.20Organized laborall667Democrats13.4002018
    17.7012.50Organized laborall1,187Democrats15.10892021
    20.5012.90Organized laborall911Democrats16.70482022
    25.5014.10Organized laborall409Democrats19.80582024
    16.8010.10Organized laborall463Republicans13.4001973
    18.6011.20Organized laborall419Republicans14.9001974
    104.60Organized laborall425Republicans7.3001975
    9.904.30Organized laborall390Republicans7.1001976
    13.607.30Organized laborall446Republicans10.4001977

    Showing the first 500 of 1680 rows.

    Comparing pooled periods, both parties' shares are lower in 2022-2024 than in 1973-1980 in 8 of the thirteen institutions, with intervals that do not overlap, and higher for both parties only in the military. Each panel is one institution, sorted from the largest Democratic advantage to the largest Republican advantage in 2022-2024. Shading marks 2021-2024, when web interviews began; this view plots only the in-person interviews of 2022 and 2024.Note: Web interviews were 87.3% of all 2021 respondents, 46.1% of 2022 respondents and 53.2% of 2024 respondents. Before 2021 the survey was in person, with some telephone interviews from 2004. The data table gives the web share of each cell.Note: Banks and financial institutions were first asked in 1975. The confidence items were not fielded in 1972 or 1985. Organized labor has no in-person 2024 cell because both party cells fall below the suppression floor.Note: Lines are broken between 2018 and the next survey year in each view, because there is no 2020 survey and the 2021 survey had no in-person interviews.Note: Intervals assume a simple random sample of the weighted respondents (the Kish effective sample size); they do not use the survey's sampling design.Source: General Social Survey, 1973-2024; weighted (weight_use); Democrats and Republicans include leaners; intervals on the Kish effective sample size

    The gaps have lasted since different periods

    A gap has lasted since the earliest period from which its interval stays on the same side of zero in that period and in every later period through 2022-2024, counting in-person interviews only. Figure 4 shows the gap for each institution in six periods, and Figure 5 gives the 2022-2024 gap with its interval and the period from which each gap has lasted.

    Television is the only gap that has lasted since the first period, from +4.0 (95% CI 2.1 to 5.9)+4.0 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in television; in-person interviews only; normal interval on the Kish effective nInterval95% CI 2.1 to 5.9n8,750Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period1973-1980SourceGeneral Social Survey in 1973-1980 to +7.1 (95% CI 2.5 to 11.6)+7.1 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in television; in-person interviews only; normal interval on the Kish effective nInterval95% CI 2.5 to 11.6n1,291Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022-2024SourceGeneral Social Survey in 2022-2024, and because the survey began in 1973 it may have started earlier. The Republican advantage on the military has lasted since 1981-1990 and was −22.3 (95% CI −25.6 to −18.9)−22.3 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in military; in-person interviews only; normal interval on the Kish effective nInterval95% CI −25.6 to −18.9n4,635Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2001-2010SourceGeneral Social Survey points in 2001-2010, the largest Republican advantage in any period, and −13.7−13.7 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in military; in-person interviews only; normal interval on the Kish effective nInterval95% CI −22.1 to −5.4n1,302Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022-2024SourceGeneral Social Survey in 2022-2024. Organized religion has favored Republicans since 1991-2000, the press has favored Democrats since 2001-2010, and the education and science gaps have lasted since 2011-2018. The press gap has favored Democrats in every period, although its interval includes zero in 1991-2000.

    Figure 4

    From 2011-2018 to 2022-2024 the Supreme Court gap moved from +1.1 to −16.2 points and the medicine gap from −1.8 to +10.4

    Democratic share minus Republican share with a great deal of confidence, percentage points, by institution and period; positive means Democrats are more confident; in-person and all-mode views

    −30.00.0+30.0RepublicanDemocratic
    A grid of thirteen institutions by six periods for in-person interviews, shaded blue where Democrats are more confident and red where Republicans are, with the gap printed in each cell. Science, education, medicine and the press are blue in the last period and the science cell is the deepest blue. The military, organized religion and the Supreme Court are red in the last period, and the military is deepest red in 2001-2010. Banks, companies and labor are pale in the last period.
    Institution1973-19801981-19901991-20002001-20102011-20182022-2024
    Scientific community−5.5, interval −8.1 to −2.9, n = 8,073−6.2, interval −8.8 to −3.6, n = 8,598−2.7, interval −5.2 to −0.1, n = 7,532+2.4, interval −1.1 to +5.9, n = 4,481+9.7, interval +6.1 to +13.4, n = 4,412+19.8, interval +11.8 to +27.8, n = 1,259
    Executive branch−4.5, interval −6.5 to −2.5, n = 8,680−15.6, interval −17.7 to −13.6, n = 9,030+5.1, interval +3.4 to +6.9, n = 7,867−12.3, interval −15.1 to −9.5, n = 4,600+5.9, interval +3.5 to +8.3, n = 4,508+14.6, interval +9.4 to +19.9, n = 1,295
    Education+5.2, interval +2.8 to +7.6, n = 8,761+0.5, interval −1.8 to +2.8, n = 9,115+4.4, interval +2.3 to +6.6, n = 7,979+1.3, interval −1.9 to +4.4, n = 4,653+9.7, interval +6.6 to +12.7, n = 4,550+14.4, interval +7.6 to +21.2, n = 1,305
    Medicine−3.6, interval −6.1 to −1.2, n = 8,793−4.3, interval −6.8 to −1.8, n = 9,163−4.1, interval −6.5 to −1.6, n = 7,987−2.6, interval −6.0 to +0.8, n = 4,653−1.8, interval −5.4 to +1.7, n = 4,554+10.4, interval +2.5 to +18.3, n = 1,309
    Press+7.1, interval +5.0 to +9.2, n = 8,744+5.3, interval +3.4 to +7.2, n = 9,060+1.4, interval −0.1 to +3.0, n = 7,913+5.6, interval +3.6 to +7.6, n = 4,642+8.3, interval +6.3 to +10.2, n = 4,533+7.8, interval +3.8 to +11.8, n = 1,297
    Television+4.0, interval +2.1 to +5.9, n = 8,750+2.7, interval +1.0 to +4.4, n = 9,116+2.4, interval +0.9 to +4.0, n = 7,963+2.7, interval +0.8 to +4.7, n = 4,639+3.3, interval +1.2 to +5.4, n = 4,519+7.1, interval +2.5 to +11.6, n = 1,291
    Congress+3.7, interval +1.9 to +5.5, n = 8,671+1.6, interval −0.2 to +3.5, n = 9,050−0.1, interval −1.7 to +1.5, n = 7,866+0.1, interval −2.3 to +2.4, n = 4,611+0.8, interval −1.0 to +2.5, n = 4,511+5.0, interval +1.3 to +8.8, n = 1,296
    Organized labor+6.9, interval +5.2 to +8.6, n = 8,486+5.2, interval +3.7 to +6.7, n = 8,865+5.1, interval +3.4 to +6.7, n = 7,513+2.5, interval +0.1 to +4.9, n = 4,481+4.6, interval +2.1 to +7.0, n = 4,405+0.8, interval −7.0 to +8.6, n = 964
    Banks and finance−6.0, interval −8.9 to −3.1, n = 6,227−2.2, interval −4.4 to 0.0, n = 9,081−2.0, interval −4.1 to +0.1, n = 7,925−6.1, interval −9.1 to −3.1, n = 4,629−5.2, interval −7.9 to −2.6, n = 4,551+0.3, interval −6.3 to +7.0, n = 1,307
    Major companies−12.2, interval −14.5 to −9.9, n = 8,493−12.2, interval −14.5 to −9.9, n = 8,892−12.2, interval −14.4 to −10.0, n = 7,783−7.7, interval −10.4 to −5.0, n = 4,591−11.4, interval −14.3 to −8.4, n = 4,494−0.8, interval −7.1 to +5.5, n = 1,265
    Organized religion−3.2, interval −5.7 to −0.8, n = 8,534−1.0, interval −3.3 to +1.3, n = 8,988−4.1, interval −6.4 to −1.9, n = 7,850−9.1, interval −12.1 to −6.1, n = 4,575−9.5, interval −12.6 to −6.5, n = 4,490−11.6, interval −18.4 to −4.8, n = 1,264
    Military−0.4, interval −2.8 to +2.0, n = 8,599−4.2, interval −6.6 to −1.8, n = 9,036−8.9, interval −11.4 to −6.5, n = 7,894−22.3, interval −25.6 to −18.9, n = 4,635−15.2, interval −18.7 to −11.7, n = 4,539−13.7, interval −22.1 to −5.4, n = 1,302
    Supreme Court0.0, interval −2.4 to +2.4, n = 8,541−5.3, interval −7.7 to −2.8, n = 8,942−0.6, interval −2.9 to +1.8, n = 7,737−6.5, interval −9.8 to −3.1, n = 4,571+1.1, interval −2.2 to +4.4, n = 4,497−16.2, interval −22.8 to −9.5, n = 1,289
    Show data
    Data for From 2011-2018 to 2022-2024 the Supreme Court gap moved from +1.1 to −16.2 points and the medicine gap from −1.8 to +10.4
    conf_highconf_loweragapinstitutionmoden
    −2.90−8.101973-1980−5.50Scientific communityall8,073
    −3.60−8.801981-1990−6.20Scientific communityall8,598
    −0.100−5.201991-2000−2.70Scientific communityall7,532
    5.40−1.302001-20102.10Scientific communityall4,968
    13.406.502011-20189.90Scientific communityall4,935
    33.3026.202021-202429.70Scientific communityall5,227
    −2.50−6.501973-1980−4.50Executive branchall8,680
    −13.60−17.701981-1990−15.60Executive branchall9,030
    6.903.401991-20005.10Executive branchall7,867
    −9−14.302001-2010−11.60Executive branchall5,092
    8.503.902011-20186.20Executive branchall5,037
    16.4011.602021-202414Executive branchall5,268
    7.602.801973-19805.20Educationall8,761
    2.80−1.801981-19900.500Educationall9,115
    6.602.301991-20004.40Educationall7,979
    4.10−1.802001-20101.20Educationall5,153
    137.302011-201810.10Educationall5,087
    12.406.702021-20249.50Educationall5,277
    −1.20−6.101973-1980−3.60Medicineall8,793
    −1.80−6.801981-1990−4.30Medicineall9,163
    −1.60−6.501991-2000−4.10Medicineall7,987
    0.300−6.202001-2010−2.90Medicineall5,155
    2.90−3.802011-2018−0.400Medicineall5,088
    17.4010.302021-202413.80Medicineall5,284
    9.2051973-19807.10Pressall8,744
    7.203.401981-19905.30Pressall9,060
    3−0.1001991-20001.40Pressall7,913
    8.104.302001-20106.20Pressall5,140
    10.306.602011-20188.50Pressall5,063
    14.3010.102021-202412.20Pressall5,266
    5.902.101973-19804Televisionall8,750
    4.4011981-19902.70Televisionall9,116
    40.9001991-20002.40Televisionall7,963
    51.102001-20103.10Televisionall5,137
    512011-20183Televisionall5,051
    7.703.802021-20245.80Televisionall5,269
    5.501.901973-19803.70Congressall8,671
    3.50−0.2001981-19901.60Congressall9,050
    1.50−1.701991-2000−0.100Congressall7,866
    2.30−2.202001-20100.100Congressall5,107
    2.40−0.9002011-20180.800Congressall5,034
    62.602021-20244.30Congressall5,274
    8.605.201973-19806.90Organized laborall8,486
    6.703.701981-19905.20Organized laborall8,865
    6.703.401991-20005.10Organized laborall7,513
    5.400.9002001-20103.20Organized laborall4,957
    7.102.502011-20184.80Organized laborall4,920
    7.701.902021-20244.80Organized laborall4,418
    −3.10−8.901973-1980−6Banks and financeall6,227
    0−4.401981-1990−2.20Banks and financeall9,081
    0.100−4.101991-2000−2Banks and financeall7,925
    −3−8.702001-2010−5.90Banks and financeall5,131
    −3.10−8.202011-2018−5.70Banks and financeall5,087
    1.70−4.302021-2024−1.30Banks and financeall5,291
    −9.90−14.501973-1980−12.20Major companiesall8,493
    −9.90−14.501981-1990−12.20Major companiesall8,892
    −10−14.401991-2000−12.20Major companiesall7,783
    −5.10−10.302001-2010−7.70Major companiesall5,081
    −8.10−13.802011-2018−10.90Major companiesall5,026
    −2.50−8.202021-2024−5.30Major companiesall5,240
    −0.800−5.701973-1980−3.20Organized religionall8,534
    1.30−3.301981-1990−1Organized religionall8,988
    −1.90−6.401991-2000−4.10Organized religionall7,850
    −6.30−122001-2010−9.20Organized religionall5,069
    −6−11.702011-2018−8.80Organized religionall5,012
    −7.20−132021-2024−10.10Organized religionall5,225
    2−2.801973-1980−0.400Militaryall8,599
    −1.80−6.601981-1990−4.20Militaryall9,036
    −6.50−11.401991-2000−8.90Militaryall7,894
    −18.60−25.102001-2010−21.90Militaryall5,132
    −12.10−18.702011-2018−15.40Militaryall5,069
    −10.20−17.802021-2024−14Militaryall5,278
    2.40−2.401973-19800Supreme Courtall8,541
    −2.80−7.701981-1990−5.30Supreme Courtall8,942
    1.80−2.901991-2000−0.600Supreme Courtall7,737
    −3.80−10.202001-2010−7Supreme Courtall5,063
    5.10−1.202011-20182Supreme Courtall5,022
    −9.70−16.102021-2024−12.90Supreme Courtall5,270
    −2.90−8.101973-1980−5.50Scientific communityinperson8,073
    −3.60−8.801981-1990−6.20Scientific communityinperson8,598
    −0.100−5.201991-2000−2.70Scientific communityinperson7,532
    5.90−1.102001-20102.40Scientific communityinperson4,481
    13.406.102011-20189.70Scientific communityinperson4,412
    27.8011.802022-202419.80Scientific communityinperson1,259
    −2.50−6.501973-1980−4.50Executive branchinperson8,680
    −13.60−17.701981-1990−15.60Executive branchinperson9,030
    6.903.401991-20005.10Executive branchinperson7,867
    −9.50−15.102001-2010−12.30Executive branchinperson4,600
    8.303.502011-20185.90Executive branchinperson4,508
    19.909.402022-202414.60Executive branchinperson1,295
    7.602.801973-19805.20Educationinperson8,761
    2.80−1.801981-19900.500Educationinperson9,115
    6.602.301991-20004.40Educationinperson7,979
    4.40−1.902001-20101.30Educationinperson4,653
    12.706.602011-20189.70Educationinperson4,550
    21.207.602022-202414.40Educationinperson1,305
    −1.20−6.101973-1980−3.60Medicineinperson8,793
    −1.80−6.801981-1990−4.30Medicineinperson9,163
    −1.60−6.501991-2000−4.10Medicineinperson7,987
    0.800−62001-2010−2.60Medicineinperson4,653
    1.70−5.402011-2018−1.80Medicineinperson4,554
    18.302.502022-202410.40Medicineinperson1,309
    9.2051973-19807.10Pressinperson8,744
    7.203.401981-19905.30Pressinperson9,060
    3−0.1001991-20001.40Pressinperson7,913
    7.603.602001-20105.60Pressinperson4,642
    10.206.302011-20188.30Pressinperson4,533
    11.803.802022-20247.80Pressinperson1,297
    5.902.101973-19804Televisioninperson8,750
    4.4011981-19902.70Televisioninperson9,116
    40.9001991-20002.40Televisioninperson7,963
    4.700.8002001-20102.70Televisioninperson4,639
    5.401.202011-20183.30Televisioninperson4,519
    11.602.502022-20247.10Televisioninperson1,291
    5.501.901973-19803.70Congressinperson8,671
    3.50−0.2001981-19901.60Congressinperson9,050
    1.50−1.701991-2000−0.100Congressinperson7,866
    2.40−2.302001-20100.100Congressinperson4,611
    2.50−12011-20180.800Congressinperson4,511
    8.801.302022-20245Congressinperson1,296
    8.605.201973-19806.90Organized laborinperson8,486
    6.703.701981-19905.20Organized laborinperson8,865
    6.703.401991-20005.10Organized laborinperson7,513
    4.900.1002001-20102.50Organized laborinperson4,481
    72.102011-20184.60Organized laborinperson4,405
    8.60−72022-20240.800Organized laborinperson964
    −3.10−8.901973-1980−6Banks and financeinperson6,227
    0−4.401981-1990−2.20Banks and financeinperson9,081
    0.100−4.101991-2000−2Banks and financeinperson7,925
    −3.10−9.102001-2010−6.10Banks and financeinperson4,629
    −2.60−7.902011-2018−5.20Banks and financeinperson4,551
    7−6.302022-20240.300Banks and financeinperson1,307
    −9.90−14.501973-1980−12.20Major companiesinperson8,493
    −9.90−14.501981-1990−12.20Major companiesinperson8,892
    −10−14.401991-2000−12.20Major companiesinperson7,783
    −5−10.402001-2010−7.70Major companiesinperson4,591
    −8.40−14.302011-2018−11.40Major companiesinperson4,494
    5.50−7.102022-2024−0.800Major companiesinperson1,265
    −0.800−5.701973-1980−3.20Organized religioninperson8,534
    1.30−3.301981-1990−1Organized religioninperson8,988
    −1.90−6.401991-2000−4.10Organized religioninperson7,850
    −6.10−12.102001-2010−9.10Organized religioninperson4,575
    −6.50−12.602011-2018−9.50Organized religioninperson4,490
    −4.80−18.402022-2024−11.60Organized religioninperson1,264
    2−2.801973-1980−0.400Militaryinperson8,599
    −1.80−6.601981-1990−4.20Militaryinperson9,036
    −6.50−11.401991-2000−8.90Militaryinperson7,894
    −18.90−25.602001-2010−22.30Militaryinperson4,635
    −11.70−18.702011-2018−15.20Militaryinperson4,539
    −5.40−22.102022-2024−13.70Militaryinperson1,302
    2.40−2.401973-19800Supreme Courtinperson8,541
    −2.80−7.701981-1990−5.30Supreme Courtinperson8,942
    1.80−2.901991-2000−0.600Supreme Courtinperson7,737
    −3.10−9.802001-2010−6.50Supreme Courtinperson4,571
    4.40−2.202011-20181.10Supreme Courtinperson4,497
    −9.50−22.802022-2024−16.20Supreme Courtinperson1,289
    Science, education, medicine and the press are blue in 2022-2024, the military, organized religion and the Supreme Court are red, and banks, companies and labor are pale. Blue cells are institutions where Democrats are more confident and red cells where Republicans are, and the printed number is the gap in points. Rows are sorted by the gap in 2022-2024, and this view uses in-person interviews, so the last column is the in-person interviews of 2022 and 2024. The table that follows gives the 2022-2024 intervals; hover a cell for its interval and sample size.Note: Cells combine the surveys of a period; the 1973-1980 period has twelve institutions in 1973 and 1974 and thirteen after, because banks were first asked in 1975.Note: Intervals assume a simple random sample of the weighted respondents (the Kish effective sample size); they do not use the survey's sampling design.Source: General Social Survey, 1973-2024; weighted (weight_use); Democrats and Republicans include leaners; intervals on the Kish effective sample size

    The science gap changed sign. Republicans were the more confident party in 1973-1980 (−5.5 (95% CI −8.1 to −2.9)−5.5 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in the scientific community; in-person interviews only; normal interval on the Kish effective nInterval95% CI −8.1 to −2.9n8,073Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period1973-1980SourceGeneral Social Survey), the interval for 2001-2010 includes zero (+2.4 (95% CI −1.1 to 5.9)+2.4 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in the scientific community; in-person interviews only; normal interval on the Kish effective nInterval95% CI −1.1 to 5.9n4,481Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2001-2010SourceGeneral Social Survey), and Democrats led in 2011-2018 (+9.7 (95% CI 6.1 to 13.4)+9.7 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in the scientific community; in-person interviews only; normal interval on the Kish effective nInterval95% CI 6.1 to 13.4n4,412Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2011-2018SourceGeneral Social Survey). In the annual data for all modes the point estimate has been positive in every survey year since 2008, when it was +7.5 (95% CI 0.4 to 14.6)+7.5 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in scientific community; all interview modes; normal interval on the Kish effective nInterval95% CI 0.4 to 14.6n1,071Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2008SourceGeneral Social Survey, and the interval excludes zero in 2008, in 2014 and in every survey year from 2016. An earlier study of the GSS found that self-identified conservatives went from the most to the least trusting of science between 1974 and 2010 [6], a reversal by ideology in the same direction as the reversal by party here.

    The medicine and Supreme Court gaps have lasted only since 2022-2024, and both institutions had earlier gaps. Medicine favored Republicans, with an interval below zero, in 1973-1980 and in the next two periods, and the Court favored Republicans in 1981-1990 and 2001-2010. Neither interval excluded zero in 2011-2018 (−1.8 (95% CI −5.4 to 1.7)−1.8 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in medicine; in-person interviews only; normal interval on the Kish effective nInterval95% CI −5.4 to 1.7n4,554Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2011-2018SourceGeneral Social Survey and +1.1 (95% CI −2.2 to 4.4)+1.1 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in supreme court; in-person interviews only; normal interval on the Kish effective nInterval95% CI −2.2 to 4.4n4,497Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2011-2018SourceGeneral Social Survey), and among in-person respondents in 2022-2024 they were +10.4 (95% CI 2.5 to 18.3)+10.4 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in medicine; in-person interviews only; normal interval on the Kish effective nInterval95% CI 2.5 to 18.3n1,309Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022-2024SourceGeneral Social Survey and −16.2 (95% CI −22.8 to −9.5)−16.2 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in supreme court; in-person interviews only; normal interval on the Kish effective nInterval95% CI −22.8 to −9.5n1,289Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022-2024SourceGeneral Social Survey, so the Court gap has the sign it had in earlier periods and the medicine gap has the opposite sign. The Congress gap has lasted since 2022-2024. It was also present in 1973-1980, and its intervals included zero from 1981-1990 to 2011-2018.

    Figure 5

    The television gap has lasted since 1973-1980 and the medicine, Congress and Supreme Court gaps only since 2022-2024

    Democratic share minus Republican share with a great deal of confidence, percentage points with 95% interval, in 2022-2024 among in-person respondents, and the period from which the gap has lasted

    InstitutionGap, 2022-2024, in person, points [95% interval]Lasting gap
    Scientific community+19.8 [+11.8, +27.8]Democrats more confident since 2011-2018
    Executive branch+14.6 [+9.4, +19.9]Reverses with the party in the White House
    Education+14.4 [+7.6, +21.2]Democrats more confident since 2011-2018
    Medicine+10.4 [+2.5, +18.3]Democrats more confident since 2022-2024
    Press+7.8 [+3.8, +11.8]Democrats more confident since 2001-2010
    Television+7.1 [+2.5, +11.6]Democrats more confident since 1973-1980
    Congress+5.0 [+1.3, +8.8]Democrats more confident since 2022-2024
    Organized labor+0.8 [−7.0, +8.6]None in 2022-2024 (interval includes zero)
    Banks and finance+0.3 [−6.3, +7.0]None in 2022-2024 (interval includes zero)
    Major companies−0.8 [−7.1, +5.5]None in 2022-2024 (interval includes zero)
    Organized religion−11.6 [−18.4, −4.8]Republicans more confident since 1991-2000
    Military−13.7 [−22.1, −5.4]Republicans more confident since 1981-1990
    Supreme Court−16.2 [−22.8, −9.5]Republicans more confident since 2022-2024
    Show data
    Data for The television gap has lasted since 1973-1980 and the medicine, Congress and Supreme Court gaps only since 2022-2024
    g3institutionlasting
    +19.8 [+11.8, +27.8]Scientific communityDemocrats more confident since 2011-2018
    +14.6 [+9.4, +19.9]Executive branchReverses with the party in the White House
    +14.4 [+7.6, +21.2]EducationDemocrats more confident since 2011-2018
    +10.4 [+2.5, +18.3]MedicineDemocrats more confident since 2022-2024
    +7.8 [+3.8, +11.8]PressDemocrats more confident since 2001-2010
    +7.1 [+2.5, +11.6]TelevisionDemocrats more confident since 1973-1980
    +5.0 [+1.3, +8.8]CongressDemocrats more confident since 2022-2024
    +0.8 [−7.0, +8.6]Organized laborNone in 2022-2024 (interval includes zero)
    +0.3 [−6.3, +7.0]Banks and financeNone in 2022-2024 (interval includes zero)
    −0.8 [−7.1, +5.5]Major companiesNone in 2022-2024 (interval includes zero)
    −11.6 [−18.4, −4.8]Organized religionRepublicans more confident since 1991-2000
    −13.7 [−22.1, −5.4]MilitaryRepublicans more confident since 1981-1990
    −16.2 [−22.8, −9.5]Supreme CourtRepublicans more confident since 2022-2024
    Positive gaps mean Democrats are more confident, and the last column names the period from which each gap has lasted: the earliest period from which its interval stays on the same side of zero in that period and in every later period through 2022-2024, counting in-person interviews only. The executive branch gap changes sign with the party in the White House, so no period is listed for it. Rows are sorted as in the heatmap.Note: The rule counts an institution as having no lasting gap when the 2022-2024 in-person interval includes zero, whatever the earlier periods show. The survey began in 1973, so a gap that has lasted since 1973-1980 may have started earlier.Source: General Social Survey, 1973-2024; weighted (weight_use); Democrats and Republicans include leaners; intervals on the Kish effective sample size

    Three gaps that existed in earlier periods no longer stand out among in-person respondents in 2022-2024. The major companies gap was −12.2 (95% CI −14.5 to −9.9)−12.2 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in major companies; in-person interviews only; normal interval on the Kish effective nInterval95% CI −14.5 to −9.9n8,493Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period1973-1980SourceGeneral Social Survey in 1973-1980 and −11.4 (95% CI −14.3 to −8.4)−11.4 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in major companies; in-person interviews only; normal interval on the Kish effective nInterval95% CI −14.3 to −8.4n4,494Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2011-2018SourceGeneral Social Survey in 2011-2018, and the organized labor gap was +6.9 (95% CI 5.2 to 8.6)+6.9 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in organized labor; in-person interviews only; normal interval on the Kish effective nInterval95% CI 5.2 to 8.6n8,486Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period1973-1980SourceGeneral Social Survey and +4.6 (95% CI 2.1 to 7.0)+4.6 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in organized labor; in-person interviews only; normal interval on the Kish effective nInterval95% CI 2.1 to 7.0n4,405Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2011-2018SourceGeneral Social Survey in the same periods, so Republicans were the more confident party about companies and Democrats about labor. Republicans also led on banks in 1973-1980. In 2022-2024 the in-person gaps for companies, labor and banks each have an interval that includes zero. Web respondents give −6.2 (95% CI −11.2 to −1.2)−6.2 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in major companies; web interviews only; normal interval on the Kish effective nInterval95% CI −11.2 to −1.2n1,629Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022-2024SourceGeneral Social Survey for major companies and +9.0 (95% CI 3.5 to 14.4)+9.0 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in organized labor; web interviews only; normal interval on the Kish effective nInterval95% CI 3.5 to 14.4n1,223Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022-2024SourceGeneral Social Survey for labor, so whether these two gaps have closed depends on the mode.

    The executive branch gap changes sign with the president

    The executive branch does not fit a pattern of a gap that opens and stays (Figure 6). Its gap is positive under Democratic presidents and negative under Republican presidents in 3333 survey yearsEstimandcount of years in which the sign of the Democratic-minus-Republican gap matches the party of the president (positive under a Democrat, negative under a Republican); all interview modesWeightsweight_use (wtssnrps from 2004, wtssps before)Populationsurvey years with the executive branch itemPeriod1973-2024SourceGeneral Social Survey of 3333 survey yearsEstimandcount of survey yearsWeightsnone (count)Populationsurvey years with the executive branch itemPeriod1973-2024SourceGeneral Social Survey survey years, and its interval excludes zero on the side of the president's party in 3232 survey yearsEstimandcount of years in which the 95% interval of the Democratic-minus-Republican gap excludes zero on the side of the president's party; all interview modesWeightsweight_use (wtssnrps from 2004, wtssps before)Populationsurvey years with the executive branch itemPeriod1973-2024SourceGeneral Social Survey of them. The exception is 1980, when the gap was +4.7 (95% CI −0.1 to 9.6)+4.7 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in executive branch; all interview modes; normal interval on the Kish effective nInterval95% CI −0.1 to 9.6n1,175Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period1980SourceGeneral Social Survey points. The gap was −20.4−20.4 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in executive branch; all interview modes; normal interval on the Kish effective nInterval95% CI −25.1 to −15.7n1,485Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2006SourceGeneral Social Survey points in 2006 under George W. Bush, +16.1+16.1 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in executive branch; all interview modes; normal interval on the Kish effective nInterval95% CI 10.6 to 21.6n1,080Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2010SourceGeneral Social Survey in 2010 under Barack Obama, −13.1−13.1 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in executive branch; all interview modes; normal interval on the Kish effective nInterval95% CI −18.4 to −7.8n1,211Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2018SourceGeneral Social Survey in 2018 under Donald Trump and +16.3+16.3 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in executive branch; all interview modes; normal interval on the Kish effective nInterval95% CI 11.7 to 20.9n1,538Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2024SourceGeneral Social Survey in 2024 under Joe Biden, each with an interval that excludes zero.

    The 2021 survey began interviewing in December 2020 [4], so 22.1%22.1% percent of respondents interviewed before the inaugurationEstimandshare interviewed from December 2020 to 19 January 2021, before the inauguration of 20 January 2021 (unweighted count)n1,793Weightsnone (count)Populationrespondents who identify with or lean toward a party and answered the executive branch item in 2021Period2021SourceGeneral Social Survey of its executive branch respondents were interviewed before the inauguration in January 2021. Their gap was −10.7 (95% CI −19.1 to −2.3)−10.7 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in the executive branch; all interview modes; 2021 interviews interviewed before the inauguration on 20 January 2021Interval95% CI −19.1 to −2.3n396Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6), interviewed before the inauguration on 20 January 2021Period2021SourceGeneral Social Survey points, against +16.0 (95% CI 11.7 to 20.4)+16.0 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in the executive branch; all interview modes; 2021 interviews interviewed from 20 January 2021Interval95% CI 11.7 to 20.4n1,397Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6), interviewed from 20 January 2021Period2021SourceGeneral Social Survey for the rest, so each part has the sign of the president in office when it was interviewed. An analysis of the GSS through 2006 described this party-in-power pattern for the executive branch [1], and the surveys since then continue it. Chapter 6.2, Trust in government (planned) asks whether trust in the federal government follows the same pattern in the American National Election Studies.

    Figure 6

    The executive branch gap has the sign of the president's party in 33 of 33 survey years

    Democratic minus Republican share with a great deal of confidence in the executive branch of the federal government, percentage points, by survey year, 1973-2024, 95% intervals

    Democratic minus Republican gap (points)

    Show data
    Data for The executive branch gap has the sign of the president's party in 33 of 33 survey years
    conf_highconf_lowgapmodenpresident_partyyear
    −17.10−28.70−22.90all1,260Republican1973
    −4.40−14−9.20all1,233Republican1974
    −3.90−13.50−8.70all1,240Republican1975
    −3−12.60−7.80all1,218Republican1976
    11.900.7006.30all1,296Democratic1977
    11.502.507all1,258Democratic1978
    9.60−0.1004.70all1,175Democratic1980
    −13.40−25.20−19.30all1,570Republican1982
    −8.20−18.30−13.30all1,355Republican1983
    −12.70−26.70−19.70all834Republican1984
    −9.50−20.20−14.80all1,248Republican1986
    −10.40−20.80−15.60all1,546Republican1987
    −10.10−21.90−16all852Republican1988
    −6.40−19−12.70all856Republican1989
    −3.80−17.70−10.70all769Republican1990
    −10.20−23.50−16.90all864Republican1991
    13.604.208.90all877Democratic1993
    125.308.70all1,671Democratic1994
    11.304.307.80all1,524Democratic1996
    12.504.808.60all1,500Democratic1998
    112.506.70all1,431Democratic2000
    −18.60−34.30−26.50all703Republican2002
    −18.80−33.20−26all726Republican2004
    −15.70−25.10−20.40all1,485Republican2006
    −1.40−10.70−6all1,098Republican2008
    21.6010.6016.10all1,080Democratic2010
    16.806.8011.80all1,027Democratic2012
    18.2010.9014.50all1,266Democratic2014
    15.307.4011.40all1,533Democratic2016
    −7.80−18.40−13.10all1,211Republican2018
    14.407.1010.70all2,017Democratic2021
    19.7010.9015.30all1,713Democratic2022
    20.9011.7016.30all1,538Democratic2024
    −17.10−28.70−22.90inperson1,260Republican1973
    −4.40−14−9.20inperson1,233Republican1974
    −3.90−13.50−8.70inperson1,240Republican1975
    −3−12.60−7.80inperson1,218Republican1976
    11.900.7006.30inperson1,296Democratic1977
    11.502.507inperson1,258Democratic1978
    9.60−0.1004.70inperson1,175Democratic1980
    −13.40−25.20−19.30inperson1,570Republican1982
    −8.20−18.30−13.30inperson1,355Republican1983
    −12.70−26.70−19.70inperson834Republican1984
    −9.50−20.20−14.80inperson1,248Republican1986
    −10.40−20.80−15.60inperson1,546Republican1987
    −10.10−21.90−16inperson852Republican1988
    −6.40−19−12.70inperson856Republican1989
    −3.80−17.70−10.70inperson769Republican1990
    −10.20−23.50−16.90inperson864Republican1991
    13.604.208.90inperson877Democratic1993
    125.308.70inperson1,671Democratic1994
    11.304.307.80inperson1,524Democratic1996
    12.504.808.60inperson1,500Democratic1998
    112.506.70inperson1,431Democratic2000
    −18.60−34.30−26.50inperson703Republican2002
    −18.20−33.60−25.90inperson579Republican2004
    −16.40−26.10−21.20inperson1,387Republican2006
    −0.700−10.60−5.70inperson974Republican2008
    20.508.7014.60inperson957Democratic2010
    17.907.2012.50inperson892Democratic2012
    1810.4014.20inperson1,098Democratic2014
    156.9010.90inperson1,424Democratic2016
    −8−19−13.50inperson1,094Republican2018
    23.209.7016.40inperson762Democratic2022
    19.303.2011.20inperson533Democratic2024
    14.206.6010.40web1,756Democratic2021
    20.80914.90web789Democratic2022
    251319web839Democratic2024
    The gap is above zero under Democratic presidents and below zero under Republican presidents, and it changes sign with each change of party in the White House. Blue shading marks the terms of Democratic presidents and red shading the terms of Republican presidents, and the dashed rule marks no gap. This view combines all interview modes and the line is broken between 2018 and 2021, when interviews moved to the web.Note: Presidential terms are assigned by the year of the inauguration; surveys fielded in an inauguration year are assigned to the new president. The shaded terms, from the left, are Nixon and Ford, Carter, Reagan and Bush, Clinton, Bush, Obama, Trump and Biden.Note: The interval excludes zero on the side of the president's party in 32 of the 33 years. In 1980 the gap was +4.7 points, an interval that includes zero.Note: The 2021 survey began interviewing in December 2020, and its point blends interviews before and after the inauguration of 20 January 2021 and is assigned to President Biden.Note: Intervals assume a simple random sample of the weighted respondents (the Kish effective sample size); they do not use the survey's sampling design.Source: General Social Survey, 1973-2024; weighted (weight_use); Democrats and Republicans include leaners; intervals on the Kish effective sample size

    The web science gap is larger before correcting for comparisons

    Figure 7 compares in-person and web respondents in 2022 and 2024 pooled. The point estimates have the same sign for 1313 institutionsEstimandcount of institutions where the in-person and web gaps have the same sign (point estimates)Weightsweight_use (wtssnrps from 2004, wtssps before)Populationthe thirteen institutionsPeriod2022-2024SourceGeneral Social Survey of the thirteen institutions, and the web gap is larger in absolute value for 88 institutionsEstimandcount of institutions where the absolute web gap exceeds the absolute in-person gapWeightsweight_use (wtssnrps from 2004, wtssps before)Populationthe thirteen institutionsPeriod2022-2024SourceGeneral Social Survey of them. Only the science difference has an interval that excludes zero: the gap is +34.3 (95% CI 27.8 to 40.7)+34.3 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in the scientific community; web interviews only; normal interval on the Kish effective nInterval95% CI 27.8 to 40.7n1,630Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022-2024SourceGeneral Social Survey points on the web and +19.8 (95% CI 11.8 to 27.8)+19.8 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in the scientific community; in-person interviews only; normal interval on the Kish effective nInterval95% CI 11.8 to 27.8n1,259Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022-2024SourceGeneral Social Survey in person, a difference of +14.5 (95% CI 4.2 to 24.7)+14.5 percentage points (web gap minus in-person gap)Estimanddifference between the web and in-person Democratic-minus-Republican gaps in confidence in the scientific community; standard errors of the two gaps combined in quadratureInterval95% CI 4.2 to 24.7n2,889Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022-2024SourceGeneral Social Survey points.

    Science is one of thirteen comparisons. Multiplying its p-value (the probability of a difference this large if the two modes gave the same gap) by thirteen gives about 0.070.07 p-value (the science difference in the gap, multiplied by thirteen for the thirteen comparisons)Estimandtwo-sided normal p-value of the web-minus-in-person difference in the science gap, multiplied by the thirteen institutions compared (Bonferroni adjustment) and capped at 1Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022-2024SourceGeneral Social Survey, so the difference is not established once the number of comparisons is counted. For the other institutions the interval of the difference includes zero, but the intervals are wide. The largest point difference is 8.68.6 percentage points (absolute web-minus-in-person difference in the gap)Estimandlargest absolute difference between the web and in-person gaps among the twelve institutions other than scienceWeightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022-2024SourceGeneral Social Survey points, for medicine, and its interval reaches 18.618.6 percentage points (end of a 95% interval for a web-minus-in-person difference)Estimandlargest absolute end of the 95% interval of the web-minus-in-person difference in the gap among the twelve institutions other than scienceWeightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022-2024SourceGeneral Social Survey points, so the data cannot rule out mode differences for other institutions that are as large as the science difference. The interval of each difference is in the data table under the figure. The pooled science gap across both modes, +28.2 (95% CI 23.4 to 32.9)+28.2 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in the scientific community; all interview modes; normal interval on the Kish effective nInterval95% CI 23.4 to 32.9n3,211Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022-2024SourceGeneral Social Survey points, is a blend of the two.

    Figure 7

    The web science gap differs from the in-person gap by +14.5 points, before correcting for thirteen comparisons

    Democratic minus Republican share with a great deal of confidence, percentage points with 95% intervals, respondents interviewed in person and on the web, 2022 and 2024 pooled

    • In person
    • Web
    Show data
    Data for The web science gap differs from the in-person gap by +14.5 points, before correcting for thirteen comparisons
    InstitutionIn-person gap (points)In-person gap, lower endIn-person gap, upper endWeb gap (points)Web gap, lower endWeb gap, upper endWeb minus in person (points)Difference, lower endDifference, upper endIn-person respondentsWeb respondents
    Scientific community+19.8+11.8+27.8+34.3+27.8+40.7+14.5+4.2+24.712591630
    Executive branch+14.6+9.4+19.9+17.2+13.0+21.5+2.6−4.2+9.412951628
    Education+14.4+7.6+21.2+12.6+8.0+17.2−1.8−10.1+6.413051622
    Medicine+10.4+2.5+18.3+19.0+12.8+25.2+8.6−1.4+18.613091626
    Press+7.8+3.8+11.8+11.8+8.0+15.6+4.0−1.5+9.512971628
    Television+7.1+2.5+11.6+4.1+1.0+7.2−3.0−8.5+2.512911628
    Congress+5.0+1.3+8.8+2.0−0.4+4.5−3.0−7.5+1.512961629
    Organized labor+0.8−7.0+8.6+9.0+3.5+14.4+8.2−1.3+17.69641223
    Banks and finance+0.3−6.3+7.0+1.0−4.0+6.0+0.6−7.7+9.013071632
    Major companies−0.8−7.1+5.5−6.2−11.2−1.2−5.4−13.5+2.712651629
    Organized religion−11.6−18.4−4.8−11.1−15.8−6.3+0.6−7.7+8.812641622
    Military−13.7−22.1−5.4−6.5−13.4+0.4+7.2−3.6+18.013021630
    Supreme Court−16.2−22.8−9.5−21.7−27.0−16.4−5.5−14.0+3.012891632
    The web gap is larger than the in-person gap for 8 of the thirteen institutions and the two point estimates have the same sign for 13, and before the correction for thirteen comparisons science is the only difference whose interval excludes zero. Each row joins the in-person gap (ring) to the web gap (solid dot), and the right-hand column, headed Change, is the web gap minus the in-person gap. Intervals are wide for most rows, so the chart cannot rule out mode differences as large as the science difference for the other institutions. Rows are sorted as in the heatmap.Note: Telephone and other-mode interviews of 2022 and 2024 are left out of this figure; they are included in the all-modes estimates elsewhere in the chapter. The interval of each web-minus-in-person difference is in the data table.Note: Intervals assume a simple random sample of the weighted respondents (the Kish effective sample size); they do not use the survey's sampling design.Source: General Social Survey, 1973-2024; weighted (weight_use); Democrats and Republicans include leaners (2022 and 2024 pooled); intervals on the Kish effective sample size

    Figure 8 follows four institutions year by year. In person, the science gap was +10.7 (95% CI 3.0 to 18.3)+10.7 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in scientific community; in-person interviews only; normal interval on the Kish effective nInterval95% CI 3.0 to 18.3n1,073Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2018SourceGeneral Social Survey points in 2018, +24.1 (95% CI 13.7 to 34.6)+24.1 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in scientific community; in-person interviews only; normal interval on the Kish effective nInterval95% CI 13.7 to 34.6n747Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022SourceGeneral Social Survey in 2022 and +11.8 (95% CI 0.1 to 23.5)+11.8 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in scientific community; in-person interviews only; normal interval on the Kish effective nInterval95% CI 0.1 to 23.5n512Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2024SourceGeneral Social Survey in 2024. On the web it was +38.0+38.0 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in scientific community; web interviews only; normal interval on the Kish effective nInterval95% CI 28.6 to 47.4n789Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022SourceGeneral Social Survey in 2022 and +31.4+31.4 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in scientific community; web interviews only; normal interval on the Kish effective nInterval95% CI 22.7 to 40.2n841Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2024SourceGeneral Social Survey in 2024. The web estimate is higher in each year that has both modes, by +13.9 (95% CI −0.2 to 28.0)+13.9 percentage points (later gap minus earlier gap, or web gap minus in-person gap)Estimandweb gap minus in-person gap in confidence in the scientific community, 2022; standard errors of the two gaps combined in quadratureInterval95% CI −0.2 to 28.0n1,536Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022SourceGeneral Social Survey points in 2022 and +19.6 (95% CI 5.0 to 34.3)+19.6 percentage points (later gap minus earlier gap, or web gap minus in-person gap)Estimandweb gap minus in-person gap in confidence in the scientific community, 2024; standard errors of the two gaps combined in quadratureInterval95% CI 5.0 to 34.3n1,353Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2024SourceGeneral Social Survey in 2024, so the difference is established for 2024 only.

    In person the gap changed by +13.5 (95% CI 0.5 to 26.4)+13.5 percentage points (later gap minus earlier gap, or web gap minus in-person gap)Estimandchange in the in-person Democratic-minus-Republican gap in confidence in the scientific community, 2018 to 2022; standard errors of the two gaps combined in quadratureInterval95% CI 0.5 to 26.4n1,820Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2018 to 2022SourceGeneral Social Survey points from 2018 to 2022 and by −12.3 (95% CI −28.0 to 3.4)−12.3 percentage points (later gap minus earlier gap, or web gap minus in-person gap)Estimandchange in the in-person Democratic-minus-Republican gap in confidence in the scientific community, 2022 to 2024; standard errors of the two gaps combined in quadratureInterval95% CI −28.0 to 3.4n1,259Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022 to 2024SourceGeneral Social Survey from 2022 to 2024. The second interval includes zero, so the in-person data do not show whether the 2024 gap is lower than the 2022 gap, and the pooled in-person figure averages the surveys of 2022 and 2024, whose point estimates differ. The 2024 interval is wide, with a lower end close to zero.

    The medicine gap has the same shape in person, with a wide interval in 2024. It was −3.4 (95% CI −10.7 to 4.0)−3.4 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in medicine; in-person interviews only; normal interval on the Kish effective nInterval95% CI −10.7 to 4.0n1,100Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2018SourceGeneral Social Survey in 2018, +15.2 (95% CI 4.6 to 25.7)+15.2 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in medicine; in-person interviews only; normal interval on the Kish effective nInterval95% CI 4.6 to 25.7n772Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022SourceGeneral Social Survey in 2022 and +1.8 (95% CI −9.1 to 12.8)+1.8 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in medicine; in-person interviews only; normal interval on the Kish effective nInterval95% CI −9.1 to 12.8n537Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2024SourceGeneral Social Survey in 2024, while the web gap in 2024 was +19.9 (95% CI 11.7 to 28.0)+19.9 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in medicine; web interviews only; normal interval on the Kish effective nInterval95% CI 11.7 to 28.0n838Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2024SourceGeneral Social Survey.

    The military gap narrowed by 2024

    In person the military gap went from −21.0 (95% CI −28.1 to −13.9)−21.0 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in military; in-person interviews only; normal interval on the Kish effective nInterval95% CI −28.1 to −13.9n1,100Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2018SourceGeneral Social Survey in 2018 to −17.5−17.5 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in military; in-person interviews only; normal interval on the Kish effective nInterval95% CI −28.6 to −6.4n766Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022SourceGeneral Social Survey in 2022 and −6.7 (95% CI −18.6 to 5.2)−6.7 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in military; in-person interviews only; normal interval on the Kish effective nInterval95% CI −18.6 to 5.2n536Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2024SourceGeneral Social Survey in 2024. The change from 2018 to 2024 is +14.3 (95% CI 0.4 to 28.1)+14.3 percentage points (later gap minus earlier gap, or web gap minus in-person gap)Estimandchange in the in-person Democratic-minus-Republican gap in confidence in the military, 2018 to 2024; standard errors of the two gaps combined in quadratureInterval95% CI 0.4 to 28.1n1,636Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2018 to 2024SourceGeneral Social Survey points, and the change from 2022 to 2024 is +10.8 (95% CI −5.4 to 27.1)+10.8 percentage points (later gap minus earlier gap, or web gap minus in-person gap)Estimandchange in the in-person Democratic-minus-Republican gap in confidence in the military, 2022 to 2024; standard errors of the two gaps combined in quadratureInterval95% CI −5.4 to 27.1n1,302Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022 to 2024SourceGeneral Social Survey, an interval that includes zero.

    On the web the gap went from −18.7−18.7 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in military; web interviews only; normal interval on the Kish effective nInterval95% CI −28.8 to −8.5n791Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022SourceGeneral Social Survey in 2022 to +2.2+2.2 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in military; web interviews only; normal interval on the Kish effective nInterval95% CI −7.1 to 11.5n839Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2024SourceGeneral Social Survey in 2024, a change of +20.9 (95% CI 7.1 to 34.6)+20.9 percentage points (later gap minus earlier gap, or web gap minus in-person gap)Estimandchange in the web Democratic-minus-Republican gap in confidence in the military, 2022 to 2024; standard errors of the two gaps combined in quadratureInterval95% CI 7.1 to 34.6n1,630Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022 to 2024SourceGeneral Social Survey, and across all modes the 2024 gap was +0.3 (95% CI −6.7 to 7.3)+0.3 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in military; all interview modes; normal interval on the Kish effective nInterval95% CI −6.7 to 7.3n1,542Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2024SourceGeneral Social Survey.

    The in-person narrowing since 2018 comes mostly from Republicans. Their share with a great deal of confidence in the military went from 73.5% (95% CI 68.4% to 78.6%)73.5% percent of respondents answering a great deal of confidenceEstimandweighted share saying a great deal of confidence in military; in-person interviews onlyInterval95% CI 68.4% to 78.6%n486Weightsweight_use (wtssnrps from 2004, wtssps before)PopulationRepublicans who identify with or lean toward the partyPeriod2018SourceGeneral Social Survey in 2018 to 54.8% (95% CI 46.2% to 63.4%)54.8% percent of respondents answering a great deal of confidenceEstimandweighted share saying a great deal of confidence in military; in-person interviews onlyInterval95% CI 46.2% to 63.4%n252Weightsweight_use (wtssnrps from 2004, wtssps before)PopulationRepublicans who identify with or lean toward the partyPeriod2024SourceGeneral Social Survey in 2024, while the Democratic share went from 52.5% (95% CI 47.6% to 57.4%)52.5% percent of respondents answering a great deal of confidenceEstimandweighted share saying a great deal of confidence in military; in-person interviews onlyInterval95% CI 47.6% to 57.4%n614Weightsweight_use (wtssnrps from 2004, wtssps before)PopulationDemocrats who identify with or lean toward the partyPeriod2018SourceGeneral Social Survey to 48.1% (95% CI 39.9% to 56.3%)48.1% percent of respondents answering a great deal of confidenceEstimandweighted share saying a great deal of confidence in military; in-person interviews onlyInterval95% CI 39.9% to 56.3%n284Weightsweight_use (wtssnrps from 2004, wtssps before)PopulationDemocrats who identify with or lean toward the partyPeriod2024SourceGeneral Social Survey.

    The Supreme Court gap opened by 2022

    In person the Court gap was −3.9 (95% CI −11.2 to 3.3)−3.9 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in supreme court; in-person interviews only; normal interval on the Kish effective nInterval95% CI −11.2 to 3.3n1,086Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2018SourceGeneral Social Survey in 2018 and −19.8 (95% CI −28.6 to −10.9)−19.8 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in supreme court; in-person interviews only; normal interval on the Kish effective nInterval95% CI −28.6 to −10.9n756Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022SourceGeneral Social Survey in 2022, and on the web it was −5.5 (95% CI −10.8 to −0.2)−5.5 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in supreme court; web interviews only; normal interval on the Kish effective nInterval95% CI −10.8 to −0.2n1,762Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2021SourceGeneral Social Survey in 2021 and −19.7 (95% CI −27.6 to −11.9)−19.7 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in supreme court; web interviews only; normal interval on the Kish effective nInterval95% CI −27.6 to −11.9n790Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2022SourceGeneral Social Survey in 2022. In person there is no survey between 2018 and 2022, so the timing rests on the web series. In 2024 the gap was −9.5−9.5 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in supreme court; in-person interviews only; normal interval on the Kish effective nInterval95% CI −18.9 to −0.2n533Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2024SourceGeneral Social Survey in person and −23.0−23.0 percentage points (Democratic share minus Republican share; positive means Democrats are more confident)Estimandweighted Democratic share minus weighted Republican share with a great deal of confidence in supreme court; web interviews only; normal interval on the Kish effective nInterval95% CI −30.2 to −15.8n842Weightsweight_use (wtssnrps from 2004, wtssps before)Populationrespondents who identify with or lean toward a party (Democrats: partyid 0-2, Republicans: partyid 4-6)Period2024SourceGeneral Social Survey on the web (Figure 8).

    Figure 8

    The Supreme Court gap was on the Republican side by 2022 in both interview modes

    Democratic minus Republican share with a great deal of confidence, percentage points, four institutions, in-person interviews (solid line) and web interviews (dashed line), 2014-2024, 95% intervals

    Democratic minus Republican gap (points)

    • In person
    • Web
    Scientific communityMedicineMilitarySupreme Court
    Show data
    Data for The Supreme Court gap was on the Republican side by 2022 in both interview modes
    conf_highconf_lowgapinstitutioninterviewnyear
    140.3007.10Scientific communityIn person1,0762014
    19.707.1013.40Scientific communityIn person1,3972016
    18.30310.70Scientific communityIn person1,0732018
    34.6013.7024.10Scientific communityIn person7472022
    23.500.10011.80Scientific communityIn person5122024
    5.50−8−1.30MedicineIn person1,1092014
    3.90−8.30−2.20MedicineIn person1,4412016
    4−10.70−3.40MedicineIn person1,1002018
    25.704.6015.20MedicineIn person7722022
    12.80−9.101.80MedicineIn person5372024
    −0.200−13.90−7.10MilitaryIn person1,1032014
    −7.30−19.70−13.50MilitaryIn person1,4392016
    −13.90−28.10−21MilitaryIn person1,1002018
    −6.40−28.60−17.50MilitaryIn person7662022
    5.20−18.60−6.70MilitaryIn person5362024
    11.50−0.4005.50Supreme CourtIn person1,0942014
    9.80−1.504.10Supreme CourtIn person1,4282016
    3.30−11.20−3.90Supreme CourtIn person1,0862018
    −10.90−28.60−19.80Supreme CourtIn person7562022
    −0.200−18.90−9.50Supreme CourtIn person5332024
    36.6025.4031Scientific communityWeb1,7552021
    47.4028.6038Scientific communityWeb7892022
    40.2022.7031.40Scientific communityWeb8412024
    15.804.4010.10MedicineWeb1,7592021
    27.108.2017.70MedicineWeb7882022
    2811.7019.90MedicineWeb8382024
    −21.60−32.90−27.30MilitaryWeb1,7572021
    −8.50−28.80−18.70MilitaryWeb7912022
    11.50−7.102.20MilitaryWeb8392024
    −0.200−10.80−5.50Supreme CourtWeb1,7622021
    −11.90−27.60−19.70Supreme CourtWeb7902022
    −15.80−30.20−23Supreme CourtWeb8422024
    The Court gap moves from near zero to the Republican side by 2022 in both modes. In 2024 the web gaps for science and medicine sit above the in-person gaps, the in-person science gap is close to its 2018 value and the in-person military gap is closer to zero than in 2018, with wide intervals. The dashed rule marks no gap.Note: There are no web interviews before 2021.Note: Lines are broken between 2018 and the next survey year in each view, because there is no 2020 survey and the 2021 survey had no in-person interviews.Note: Intervals assume a simple random sample of the weighted respondents (the Kish effective sample size); they do not use the survey's sampling design.Source: General Social Survey, 1973-2024; weighted (weight_use); Democrats and Republicans include leaners; intervals on the Kish effective sample size

    Most of the change comes from Democrats. Among in-person respondents their share with a great deal of confidence in the Court fell from 32.7%32.7% percent of respondents answering a great deal of confidenceEstimandweighted share saying a great deal of confidence in supreme court; in-person interviews onlyInterval95% CI 28.1% to 37.4%n605Weightsweight_use (wtssnrps from 2004, wtssps before)PopulationDemocrats who identify with or lean toward the partyPeriod2018SourceGeneral Social Survey in 2018 to 8.9%8.9% percent of respondents answering a great deal of confidenceEstimandweighted share saying a great deal of confidence in supreme court; in-person interviews onlyInterval95% CI 4.6% to 13.2%n405Weightsweight_use (wtssnrps from 2004, wtssps before)PopulationDemocrats who identify with or lean toward the partyPeriod2022SourceGeneral Social Survey in 2022, and among web respondents it fell from 25.5%25.5% percent of respondents answering a great deal of confidenceEstimandweighted share saying a great deal of confidence in supreme court; web interviews onlyInterval95% CI 22.1% to 28.9%n1,058Weightsweight_use (wtssnrps from 2004, wtssps before)PopulationDemocrats who identify with or lean toward the partyPeriod2021SourceGeneral Social Survey in 2021 to 7.1%7.1% percent of respondents answering a great deal of confidenceEstimandweighted share saying a great deal of confidence in supreme court; web interviews onlyInterval95% CI 3.5% to 10.6%n441Weightsweight_use (wtssnrps from 2004, wtssps before)PopulationDemocrats who identify with or lean toward the partyPeriod2022SourceGeneral Social Survey in 2022.

    The Court decided Dobbs v. Jackson Women's Health Organization in June 2022 [7], but the 2022 survey does not date the fall to that decision. Web interviewing in 2022 began in May, after a news report in early May 2022 published the Court's draft opinion [8], so an interview before the ruling was not an interview before the news. Web interviews of Democrats before the ruling already show a low share, and its interval is cut off at zero and overlaps the interval for interviews after the ruling.Among Democrats interviewed on the web in 2022, the share with a great deal of confidence in the Court was 4.6% (95% CI 0.0% to 9.2%)4.6% percent of respondents answering a great deal of confidenceEstimandweighted share of Democrats saying a great deal of confidence in the Supreme Court; web interviews only; normal interval on the Kish effective nInterval95% CI 0.0% to 9.2%n137Weightsweight_use (wtssnrps from 2004, wtssps before)PopulationDemocrats (including leaners), interviewed on the web before 25 June 2022, the day after the ruling of 24 June 2022Period2022SourceGeneral Social Survey before the ruling and 7.8% (95% CI 3.3% to 12.4%)7.8% percent of respondents answering a great deal of confidenceEstimandweighted share of Democrats saying a great deal of confidence in the Supreme Court; web interviews only; normal interval on the Kish effective nInterval95% CI 3.3% to 12.4%n304Weightsweight_use (wtssnrps from 2004, wtssps before)PopulationDemocrats (including leaners), interviewed on the web from 25 June 2022Period2022SourceGeneral Social Survey from the day after it. Too few in-person interviews were completed before the ruling to run the same check.

    What these data cannot show

    The GSS interviews a new sample each time and asks party identification and confidence in the same interview, so it cannot say whether individuals changed their confidence or whether the people who call themselves Democrats or Republicans changed. A gap between party identifiers can widen when the membership of the parties shifts along lines such as education or religion (party sortingParty sortingThe alignment of party with other traits, such as ideology, religion, race and place, so that people who share a party also tend to share those traits., taken up in Chapter 1.4, Cohort change in partisan sorting). The comparison of in-person and web respondents cannot separate an effect of the mode from a difference between the people who answered in each mode. In 2022 respondents were randomly assigned to a sequence of modes, face to face first or web first, and each sequence followed up nonrespondents in the other mode [5], so the two samples reflect who answered which invitation. The 2021 survey was designed as a web survey, and its documentation cautions that changes in 2021 may reflect the change in method as well as changes in opinion [4].

    The item asks about the people running an institution, with three answers, so a change in the share answering "a great deal" does not show where the other answers went. The executive branch item in particular follows the incumbent administration. The thirteen institutions are the ones the GSS asks about, and the survey has no items for others such as the police. Causes are outside the design. The rise after 2018 coincides with a change of president, a pandemic and a change of mode, and the data cannot say which of them, if any, moved the gaps. The same holds for the Court's ruling in 2022. The in-person cells of 2022 and 2024 are small, so the estimates for those years are less precise than those for earlier periods. The sample is national, and it supports no claims about states or counties.

    The intervals assume a simple random sample of the weighted respondents, but the GSS samples clusters of households within strata of areas. The intervals for the mean absolute gap come from resampling respondents and ignore the sampling design in the same way, and no design-based version of them has been run.A design-based standard error uses the survey's strata and clusters. Across 585585 gap estimatesEstimandcount of gap estimates comparedWeightsnone (count)Populationgap estimates: annual, period and pooled 2022-2024 by mode, from 1975 onPeriod1975-2024SourceGeneral Social Survey gap estimates, the median ratio of the design-based standard error to the one that assumes a simple random sample is 1.011.01 ratio of standard errors (using the survey design over assuming a simple random sample)Estimandmedian across gap estimates of the ratio of the Taylor-linearized standard error (GSS strata and PSUs) to the standard error that assumes a simple random sample of the weighted respondentsWeightsweight_use (wtssnrps from 2004, wtssps before)Populationgap estimates: annual, period and pooled 2022-2024 by mode, from 1975 onPeriod1975-2024SourceGeneral Social Survey, with most between 0.870.87 ratio of standard errors (using the survey design over assuming a simple random sample)Estimand5th percentile across gap estimates of the ratio of the Taylor-linearized standard error (GSS strata and PSUs) to the standard error that assumes a simple random sample of the weighted respondentsWeightsweight_use (wtssnrps from 2004, wtssps before)Populationgap estimates: annual, period and pooled 2022-2024 by mode, from 1975 onPeriod1975-2024SourceGeneral Social Survey and 1.171.17 ratio of standard errors (using the survey design over assuming a simple random sample)Estimand95th percentile across gap estimates of the ratio of the Taylor-linearized standard error (GSS strata and PSUs) to the standard error that assumes a simple random sample of the weighted respondentsWeightsweight_use (wtssnrps from 2004, wtssps before)Populationgap estimates: annual, period and pooled 2022-2024 by mode, from 1975 onPeriod1975-2024SourceGeneral Social Survey, and in 1515 gap estimatesEstimandcount of gap estimates whose 95% interval excludes zero under one standard error and not the otherWeightsweight_use (wtssnrps from 2004, wtssps before)Populationgap estimates: annual, period and pooled 2022-2024 by mode, from 1975 onPeriod1975-2024SourceGeneral Social Survey of them the interval excludes zero under one method and not the other.

    Chapter 6.5, Institutional trust outside the United States (planned) asks whether comparable gaps appear in other democracies, Chapter 6.4, Generalized and institutional trust (planned) asks whether trust in other people moved with confidence in institutions, and Chapter 6.6, Tolerance of speech turns to another GSS battery.

    Data sources

    • General Social SurveyGeneral Social Survey cumulative data, 1972-2024, confidence items fielded 1973-2024; interview mode and interview date from the raw fileYears: 1972-2024 (35 study years) · Weights: weight_use (wtssnrps from 2004, wtssps before) · License: See the provider's terms of use.

    All data sources

    Methods

    Sample: General Social Survey (GSS) respondents from the analysis subset (gss_sub2024) in the 33 survey years from 1973 to 2024 in which the confidence items were fielded (they were not asked in 1972 or 1985), who identify with or lean toward a party: 41,797 respondents with at least one confidence answer. Each confidence item was asked of a random part of the sample, so cell sizes vary by year. Items: confidence in the people running thirteen institutions (the scientific community, medicine, the military, the Supreme Court, education, the executive branch of the federal government, Congress, the press, television, major companies, banks and financial institutions, organized labor and organized religion), answered a great deal, only some or hardly any. Banks and financial institutions were added in 1975. Measure: a 0/1 indicator for a great deal. Party: Democrats are partyid 0 to 2 and Republicans are partyid 4 to 6, so leaners are counted with the party they lean toward; independents (3) and other-party identifiers (7) are dropped. The robustness cut uses strong and weak identifiers only (0 to 1 against 5 to 6). Gap: Democratic share minus Republican share, in percentage points, throughout. Weights: weight_use (wtssnrps from 2004, wtssps before). Mode: interview mode is joined from the raw GSS file by verified row position; in person means mode coded face to face from 2004 and every interview before 2004, which was in person by design; web means mode coded web (2021, 2022, 2024). There were no in-person interviews in 2021. Periods: 1973-1980, 1981-1990, 1991-2000, 2001-2010, 2011-2018 and 2022-2024, all among in-person interviews, which the text uses for every period estimate, and, in the all-modes views, 2021-2024 (2021, 2022 and 2024 pooled; 64 percent of these respondents were interviewed on the web). Intervals: 95% normal intervals on the Kish effective sample size for shares and, with standard errors combined in quadrature, for gaps, for changes in a gap between surveys and for the web-minus-in-person difference; these assume a simple random sample of the weighted respondents and ignore the survey's strata and clusters. A normal interval for a share near zero is cut off at zero, and a Wilson interval on the same effective sample size would be wider and asymmetric for shares under 10 percent. The mean absolute gap has a percentile bootstrap interval from resampling respondents (2,000 replicates, seed 12345), which also ignores strata and clusters; its noise floor is sqrt(2/pi) times the mean standard error of the gaps, the mean absolute gap expected if every true gap were zero, and an institution enters the average only when both of its party cells pass the suppression floor. The science web-minus-in-person difference is one of thirteen comparisons; its p-value is multiplied by thirteen (Bonferroni) in the text. Suppression: a party-by-year cell needs at least 150 respondents and an effective sample of 100; 2 party-by-year cells are suppressed (organized labor, democrats, 2024, in person; organized labor, republicans, 2024, in person); the smallest cell shown has 198 respondents, and the backstop of 30 respondents applies to every row. Design check: for every gap from 1975 on (annual, period and pooled 2022-2024 by mode; 585 estimates) a Taylor-linearized standard error using the GSS strata and PSUs (vstrat, vpsu; a PSU without domain respondents counts as zero) has a median ratio to the standard error that assumes a simple random sample of 1.01 (5th to 95th percentile 0.87 to 1.17); the 95% interval excludes zero under one method and not the other for 15 of them, and the lasting-gap classification of all 13 institutions is unchanged when the period intervals use the design-based standard errors. Because 1973 and 1974 carry no design variables, the 1973-1980 cells compare the design-based standard error with the standard error of the same 1975-1980 respondents. Lasting gap: the earliest period from which the 95% interval of the gap excludes zero on the same side in that period and every later period through 2022-2024 in person; an institution whose interval includes zero in 2022-2024 in person has none, and one whose significant gaps change sign at least twice is reported as reversing with the party in the White House. Presidential terms are assigned by inauguration year; the 2021 executive branch answers are also split at the inauguration of 20 January 2021 by interview date. Interview dates for the 2022 comparison are the raw dateintv values; interviews on or after 25 June 2022 are counted as after the ruling of 24 June 2022, and web interviewing in 2022 began in May, after the draft opinion was reported on 2 May 2022.

    References

    1. Smith, Tom W. (2008). Trends in Confidence in Institutions, 1973-2006. GSS Social Change Report No. 54, NORC at the University of Chicago. gss.norc.org/content/dam/gss/get-documentation/pdf/reports/social-change-reports/SC54%20Trends%20in%20Confidence%20in%20Institutions.pdf
    2. Kish, Leslie (1965). Survey Sampling. John Wiley and Sons.
    3. Lipset, Seymour Martin; Schneider, William (1983). The Confidence Gap: Business, Labor, and Government in the Public Mind. Free Press, New York.
    4. NORC at the University of Chicago (2021). 2021 GSS (Cross-section Study) Methodological Primer. General Social Survey documentation, NORC at the University of Chicago (November 2021). gss.norc.org/content/dam/gss/get-documentation/pdf/other/2021%20XSEC%20Methodological%20Primer.pdf
    5. Wells, Brian M.; Christian, Leah; Bautista, René; Lafia, Sara; Davern, Michael (2024). Mode-Sequencing Insights from the General Social Survey. NORC at the University of Chicago (September 2024). www.norc.org/research/library/mode-sequencing-insights-from-the-general-social-survey.html
    6. Gauchat, Gordon (2012). Politicization of Science in the Public Sphere: A Study of Public Trust in the United States, 1974 to 2010. American Sociological Review 77(2): 167-187. doi:10.1177/0003122412438225
    7. Supreme Court of the United States (2022). Dobbs v. Jackson Women's Health Organization. 597 U.S. 215, decided 24 June 2022. www.supremecourt.gov/opinions/21pdf/19-1392_6j37.pdf
    8. Gerstein, Josh; Ward, Alexander (2022). Supreme Court has voted to overturn abortion rights, draft opinion shows. Politico, 2 May 2022. www.politico.com/news/2022/05/02/supreme-court-abortion-draft-opinion-00029473

    How to cite

    @misc{singh2026confidenceininstitutions,
  author = {Singh, Vishal},
  title = {Confidence in thirteen institutions},
  year = {2026},
  url = {https://vishalsingh.org/politics/confidence-in-institutions},
  note = {In The Public Mind. Published 30 September 2026.}
}

    Reading paths