Skip to content
Contents
The Public Mind

    Type to search. Planned chapters match by title only.

    Chapter 1.1

    Party and ideology

    The correlation between party identification and liberal-conservative self-placement rose from +0.32 in 1972 to +0.72 in 2024 in the American National Election Studies, ended higher on two other surveys, and rose in every cohort, education and region group compared.

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

    Intervals are model based and do not follow the sampling design. A design-based jackknife was computed only for the GSS cells behind the interview-mode comparison; it has not been computed for the ANES or the CES.

    The correlationCorrelationA number from minus 1 to plus 1 for how closely two measures move together in a straight-line pattern. Zero means no straight-line pattern; a correlation does not say that one measure causes the other. between party identification and liberal-conservative self-placement in the American National Election Studies (ANES) rose from +0.32 (95% CI 0.27 to 0.36)+0.32 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.27 to 0.36n1,545Weightsweight (VCF0009z)PopulationANES respondents with a party identification (leaners kept) and a liberal-conservative self-placementPeriod1972SourceANES Cumulative Data File in 1972 to +0.72 (95% CI 0.70 to 0.74)+0.72 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.70 to 0.74n4,716Weightsweight (VCF0009z)PopulationANES respondents with a party identification (leaners kept) and a liberal-conservative self-placementPeriod2024SourceANES Cumulative Data File in 2024. The correlation is zero when a person's party says nothing about their placement and approaches one when the party nearly fixes it. The General Social Survey (GSS) goes from +0.17+0.17 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.12 to 0.23n1,345Weightsweight_usePopulationGSS respondents with a party identification (leaners kept, other party excluded) and a liberal-conservative self-placementPeriod1974SourceGeneral Social Survey in 1974 to +0.63+0.63 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.60 to 0.66n3,062Weightsweight_usePopulationGSS respondents with a party identification (leaners kept, other party excluded) and a liberal-conservative self-placementPeriod2024SourceGeneral Social Survey in 2024 and the Cooperative Election Study (CES), which starts in 2006, from +0.66+0.66 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.65 to 0.67n33,755Weightsweight (single wave); weight_cumulative (pooled years)PopulationCES respondents with a party identification (leaners kept, not sure excluded) and an ideology answer (not sure excluded)Period2006SourceCooperative Election Study to +0.72+0.72 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.71 to 0.73n55,835Weightsweight (single wave); weight_cumulative (pooled years)PopulationCES respondents with a party identification (leaners kept, not sure excluded) and an ideology answer (not sure excluded)Period2024SourceCooperative Election Study, so all three surveys end higher than they start, at different levels.

    Party identification is a seven-point scale from strong Democrat to strong Republican, and 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. keep their own points on it. Self-placement runs from extremely liberal to extremely conservative on a seven-point scale in the ANES and the GSS and a five-point scale in the CES. Every correlation, difference and share is weighted (survey weightsSurvey 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.) and carries 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., shown in parentheses or in the popover on the number.

    Researchers call the movement of liberals into the Democratic Party and of conservatives into the Republican Party the partisan sort [1], and the wider alignment of party with traits such as religion, race and place 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.. Early survey research found that most Americans' answers to separate political questions were only loosely tied together [2], and work on the ANES of the 1970s to 1990s found party identification realigning along ideological lines [3]. This chapter measures how far the two labels merged, whether the three surveys agree on the size of the change, and whether the change appears inside every group. Chapter 1.4, Cohort change in partisan sorting describes the change by birth cohort, which cannot show whether individuals changed, and Chapter 1.3, Sort dates by attitude dates the point at which particular attitudes became divided by party.

    The correlation over time on three surveys

    Figure 1 plots the correlation for each survey year.Each interval is computed on the Fisher z transformation of the correlation, which makes its sampling error close to normal [4], with 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. [5] in place of the number of respondents. It treats the weighted sample as a simple random sample, so it does not follow the sampling design, which selects respondents in clusters within sampled areas; the section on interview mode reports the design-based check that was possible. A difference between two surveys combines their two standard errors, because the surveys share no respondents. In the ANES it stayed between +0.32+0.32 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.27 to 0.36n1,545Weightsweight (VCF0009z)PopulationANES respondents with a party identification (leaners kept) and a liberal-conservative self-placementPeriod1972SourceANES Cumulative Data File and +0.41+0.41 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.36 to 0.47n891Weightsweight (VCF0009z)PopulationANES respondents with a party identification (leaners kept) and a liberal-conservative self-placementPeriod1982SourceANES Cumulative Data File in every survey year from 1972 through 1990, and every ANES estimate from 1992 on is higher than the 1982 estimate, the highest before 1992. The highest GSS estimate before 1993 is +0.31+0.31 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.26 to 0.36n1,390Weightsweight_usePopulationGSS respondents with a party identification (leaners kept, other party excluded) and a liberal-conservative self-placementPeriod1986SourceGeneral Social Survey in 1986, and every GSS estimate from 1993 on is higher.

    Both series also rose earlier, and the ANES estimate for 1992 and the GSS estimate for 1993 do not differ detectably from the earlier peaks, so neither series dates the step to a single year.The ANES correlation went from +0.32+0.32 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.27 to 0.36n1,545Weightsweight (VCF0009z)PopulationANES respondents with a party identification (leaners kept) and a liberal-conservative self-placementPeriod1972SourceANES Cumulative Data File in 1972 to +0.41+0.41 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.36 to 0.47n891Weightsweight (VCF0009z)PopulationANES respondents with a party identification (leaners kept) and a liberal-conservative self-placementPeriod1982SourceANES Cumulative Data File in 1982, a change of +0.09 (95% CI 0.02 to 0.16)+0.09 difference in correlation coefficientsEstimand1982 minus 1972 ANES correlation; standard errors of independent estimates combined in quadratureInterval95% CI 0.02 to 0.16n2,436Weightsweight (VCF0009z)PopulationANES respondents with a party identification (leaners kept) and a liberal-conservative self-placementPeriod1972 and 1982SourceANES Cumulative Data File, and the GSS correlation went from +0.14+0.14 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.08 to 0.21n1,418Weightsweight_usePopulationGSS respondents with a party identification (leaners kept, other party excluded) and a liberal-conservative self-placementPeriod1980SourceGeneral Social Survey in 1980 to +0.31+0.31 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.26 to 0.36n1,390Weightsweight_usePopulationGSS respondents with a party identification (leaners kept, other party excluded) and a liberal-conservative self-placementPeriod1986SourceGeneral Social Survey in 1986, a change of +0.17 (95% CI 0.08 to 0.25)+0.17 difference in correlation coefficientsEstimand1986 minus 1980 GSS correlation; standard errors of independent estimates combined in quadratureInterval95% CI 0.08 to 0.25n2,808Weightsweight_usePopulationGSS respondents with a party identification (leaners kept, other party excluded) and a liberal-conservative self-placementPeriod1980 and 1986SourceGeneral Social Survey. The 1992 ANES estimate minus the 1982 estimate is +0.02 (95% CI −0.04 to 0.09)+0.02 difference in correlation coefficientsEstimand1992 minus 1982 ANES correlation (1982 is the highest estimate before 1992); standard errors of independent estimates combined in quadratureInterval95% CI −0.04 to 0.09n2,710Weightsweight (VCF0009z)PopulationANES respondents with a party identification (leaners kept) and a liberal-conservative self-placementPeriod1982 and 1992SourceANES Cumulative Data File, and the 1993 GSS estimate minus the 1986 estimate is +0.02 (95% CI −0.05 to 0.10)+0.02 difference in correlation coefficientsEstimand1993 minus 1986 GSS correlation (1986 is the highest estimate before 1993); standard errors of independent estimates combined in quadratureInterval95% CI −0.05 to 0.10n2,918Weightsweight_usePopulationGSS respondents with a party identification (leaners kept, other party excluded) and a liberal-conservative self-placementPeriod1986 and 1993SourceGeneral Social Survey.

    The chapter compares two windows, 1980 to 1988 and 2020 to 2024, each estimated from all respondents in its survey years together (pooled). In the ANES the pooled correlation is +0.39 (95% CI 0.37 to 0.41)+0.39 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.37 to 0.41n6,482Weightsweight (VCF0009z)PopulationANES respondents with a party identification (leaners kept) and a liberal-conservative self-placementPeriod1980-1988, waves pooledSourceANES Cumulative Data File for 1980 to 1988 and +0.73 (95% CI 0.72 to 0.74)+0.73 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.72 to 0.74n11,757Weightsweight (VCF0009z)PopulationANES respondents with a party identification (leaners kept) and a liberal-conservative self-placementPeriod2020 and 2024, waves pooledSourceANES Cumulative Data File for 2020 to 2024, a change of +0.34 (95% CI 0.31 to 0.36)+0.34 difference in correlation coefficientsEstimandchange in the pooled ANES correlation between the two windows; standard errors of independent estimates combined in quadratureInterval95% CI 0.31 to 0.36n18,239Weightsweight (VCF0009z)PopulationANES respondents with a party identification (leaners kept) and a liberal-conservative self-placementPeriod1980-1988 to 2020-2024SourceANES Cumulative Data File. The 2020 estimate is +0.74 (95% CI 0.72 to 0.75)+0.74 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.72 to 0.75n7,041Weightsweight (VCF0009z)PopulationANES respondents with a party identification (leaners kept) and a liberal-conservative self-placementPeriod2020SourceANES Cumulative Data File, and the 2024 minus 2020 difference is −0.01 (95% CI −0.04 to 0.01)−0.01 difference in correlation coefficientsEstimand2024 minus 2020 ANES correlation; standard errors of independent estimates combined in quadratureInterval95% CI −0.04 to 0.01n11,757Weightsweight (VCF0009z)PopulationANES respondents with a party identification (leaners kept) and a liberal-conservative self-placementPeriod2020 and 2024SourceANES Cumulative Data File, an interval that includes zero, so the ANES shows no further rise between 2020 and 2024. The GSS goes from +0.27 (95% CI 0.24 to 0.29)+0.27 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.24 to 0.29n11,201Weightsweight_usePopulationGSS respondents with a party identification (leaners kept, other party excluded) and a liberal-conservative self-placementPeriod1980-1988, waves pooledSourceGeneral Social Survey pooled over 1980 to 1988 to +0.63 (95% CI 0.62 to 0.65)+0.63 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.62 to 0.65n10,198Weightsweight_usePopulationGSS respondents with a party identification (leaners kept, other party excluded) and a liberal-conservative self-placementPeriod2021, 2022 and 2024, waves pooled; web and in-person interviewsSourceGeneral Social Survey pooled over 2021, 2022 and 2024, a change of +0.37 (95% CI 0.34 to 0.40)+0.37 difference in correlation coefficientsEstimandchange in the pooled GSS correlation between the two windows; standard errors of independent estimates combined in quadratureInterval95% CI 0.34 to 0.40n21,399Weightsweight_usePopulationGSS respondents with a party identification (leaners kept, other party excluded) and a liberal-conservative self-placementPeriod1980-1988 to 2020-2024SourceGeneral Social SurveyNoteThe 2020-2024 window includes web interviews..

    The CES has no reading before 2006, when its correlation of +0.66+0.66 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.65 to 0.67n33,755Weightsweight (single wave); weight_cumulative (pooled years)PopulationCES respondents with a party identification (leaners kept, not sure excluded) and an ideology answer (not sure excluded)Period2006SourceCooperative Election Study was already above the ANES and GSS readings of that period, and it is +0.69+0.69 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.69 to 0.69n211,673Weightsweight_cumulativePopulationCES respondents with a party identification (leaners kept, not sure excluded) and an ideology answer (not sure excluded)Period2020-2024, waves pooledSourceCooperative Election Study pooled over 2020 to 2024. Its 2024 minus 2006 difference is +0.06 (95% CI 0.05 to 0.07)+0.06 difference in correlation coefficientsEstimand2024 minus 2006 CES correlation; standard errors of independent estimates combined in quadratureInterval95% CI 0.05 to 0.07n89,590WeightsweightPopulationCES respondents with a party identification (leaners kept, not sure excluded) and an ideology answer (not sure excluded)Period2006 and 2024SourceCooperative Election Study, but the path in between is not steady. The CES correlation falls from +0.68+0.68 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.67 to 0.68n29,939Weightsweight (single wave); weight_cumulative (pooled years)PopulationCES respondents with a party identification (leaners kept, not sure excluded) and an ideology answer (not sure excluded)Period2008SourceCooperative Election Study in 2008 to +0.59+0.59 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.58 to 0.60n58,788Weightsweight (single wave); weight_cumulative (pooled years)PopulationCES respondents with a party identification (leaners kept, not sure excluded) and an ideology answer (not sure excluded)Period2016SourceCooperative Election Study in 2016, a change of −0.09 (95% CI −0.10 to −0.08)−0.09 difference in correlation coefficientsEstimand2016 minus 2008 CES correlation; standard errors of independent estimates combined in quadratureInterval95% CI −0.10 to −0.08n88,727WeightsweightPopulationCES respondents with a party identification (leaners kept, not sure excluded) and an ideology answer (not sure excluded)Period2008 and 2016SourceCooperative Election Study, while the ANES 2016 minus 2008 difference is +0.08 (95% CI 0.04 to 0.13)+0.08 difference in correlation coefficientsEstimand2016 minus 2008 ANES correlation; standard errors of independent estimates combined in quadratureInterval95% CI 0.04 to 0.13n4,905Weightsweight (VCF0009z)PopulationANES respondents with a party identification (leaners kept) and a liberal-conservative self-placementPeriod2008 and 2016SourceANES Cumulative Data File.

    Figure 1

    The party-ideology correlation ended higher than it began on all three surveys

    Weighted correlation between party identification (seven-point scale, leaners kept) and liberal-conservative self-placement, respondents who answered both, ANES 1972 to 2024, GSS 1974 to 2024, CES 2006 to 2024, with 95% intervals; the differences view plots the ANES value minus each other survey's value in the same year

    Show data
    Data for The party-ideology correlation ended higher than it began on all three surveys
    SurveyRespondents includedYearEstimateInterval lowInterval highRespondentsEffective n
    ANESAll interviews1972+0.32+0.27+0.361,5451,545
    ANESAll interviews1974+0.33+0.27+0.391,140857
    ANESAll interviews1976+0.38+0.33+0.421,4971,441
    ANESAll interviews1978+0.35+0.31+0.391,6581,658
    ANESAll interviews1980+0.39+0.33+0.441,0041,004
    ANESAll interviews1982+0.41+0.36+0.47891891
    ANESAll interviews1984+0.41+0.37+0.451,5461,546
    ANESAll interviews1986+0.36+0.32+0.411,6221,622
    ANESAll interviews1988+0.40+0.36+0.451,4191,419
    ANESAll interviews1990+0.34+0.29+0.391,3151,315
    ANESAll interviews1992+0.44+0.40+0.471,8191,812
    ANESAll interviews1994+0.51+0.46+0.551,3981,144
    ANESAll interviews1996+0.54+0.50+0.581,3261,070
    ANESAll interviews1998+0.44+0.39+0.491,012858
    ANESAll interviews2000+0.47+0.40+0.53669546
    ANESAll interviews2002+0.52+0.47+0.571,223828
    ANESAll interviews2004+0.58+0.53+0.63913766
    ANESAll interviews2008+0.61+0.57+0.651,6071,053
    ANESAll interviews2012+0.61+0.59+0.635,2882,970
    ANESAll interviews2016+0.70+0.68+0.723,2982,332
    ANESAll interviews2020+0.74+0.72+0.757,0413,929
    ANESAll interviews2024+0.72+0.70+0.744,7162,228
    GSSAll interviews1974+0.17+0.12+0.231,3451,122
    GSSAll interviews1975+0.23+0.18+0.291,3841,136
    GSSAll interviews1976+0.22+0.16+0.271,3931,137
    GSSAll interviews1977+0.22+0.16+0.271,4401,160
    GSSAll interviews1978+0.16+0.09+0.221,422885
    GSSAll interviews1980+0.14+0.08+0.211,418941
    GSSAll interviews1982+0.24+0.19+0.301,7231,032
    GSSAll interviews1983+0.28+0.20+0.36763482
    GSSAll interviews1984+0.26+0.20+0.321,386889
    GSSAll interviews1985+0.28+0.22+0.341,444919
    GSSAll interviews1986+0.31+0.26+0.361,3901,132
    GSSAll interviews1987+0.31+0.25+0.361,6661,126
    GSSAll interviews1988+0.30+0.24+0.351,4111,139
    GSSAll interviews1989+0.27+0.21+0.321,4291,127
    GSSAll interviews1990+0.29+0.23+0.351,3011,023
    GSSAll interviews1991+0.26+0.20+0.311,4431,105
    GSSAll interviews1993+0.34+0.28+0.381,5281,196
    GSSAll interviews1994+0.37+0.34+0.412,8062,204
    GSSAll interviews1996+0.38+0.35+0.422,7022,168
    GSSAll interviews1998+0.36+0.32+0.402,6222,033
    GSSAll interviews2000+0.34+0.30+0.382,5932,038
    GSSAll interviews2002+0.37+0.31+0.421,299944
    GSSAll interviews2004+0.43+0.38+0.481,289937
    GSSAll interviews2006+0.47+0.44+0.504,2532,952
    GSSAll interviews2008+0.48+0.44+0.521,8901,324
    GSSAll interviews2010+0.50+0.46+0.541,9141,231
    GSSAll interviews2012+0.47+0.42+0.511,8151,153
    GSSAll interviews2014+0.50+0.46+0.532,3681,734
    GSSAll interviews2016+0.50+0.46+0.532,6681,782
    GSSAll interviews2018+0.56+0.52+0.602,1491,285
    GSSAll interviews2021+0.64+0.62+0.673,8332,501
    GSSAll interviews2022+0.63+0.59+0.663,3031,304
    GSSAll interviews2024+0.63+0.60+0.663,0621,557
    CESAll interviews2006+0.66+0.65+0.6733,75526,295
    CESAll interviews2007+0.68+0.67+0.708,7874,593
    CESAll interviews2008+0.68+0.67+0.6829,93920,888
    CESAll interviews2009+0.66+0.64+0.6712,2299,223
    CESAll interviews2010+0.66+0.65+0.6651,57420,318
    CESAll interviews2011+0.60+0.58+0.6118,3298,715
    CESAll interviews2012+0.63+0.62+0.6448,97221,136
    CESAll interviews2013+0.60+0.58+0.6114,7127,302
    CESAll interviews2014+0.62+0.61+0.6349,87621,741
    CESAll interviews2015+0.63+0.61+0.6412,8076,907
    CESAll interviews2016+0.59+0.58+0.6058,78830,529
    CESAll interviews2017+0.63+0.62+0.6516,3567,800
    CESAll interviews2018+0.68+0.68+0.6954,06325,503
    CESAll interviews2019+0.69+0.68+0.7016,4789,481
    CESAll interviews2020+0.69+0.68+0.6955,39527,431
    CESAll interviews2021+0.69+0.68+0.7023,45013,577
    CESAll interviews2022+0.70+0.69+0.7054,96424,917
    CESAll interviews2023+0.65+0.64+0.6622,02912,323
    CESAll interviews2024+0.72+0.71+0.7355,83518,536
    ANESAll respondents (mode not recorded before 2008)1972+0.32+0.27+0.361,5451,545
    ANESAll respondents (mode not recorded before 2008)1974+0.33+0.27+0.391,140857
    ANESAll respondents (mode not recorded before 2008)1976+0.38+0.33+0.421,4971,441
    ANESAll respondents (mode not recorded before 2008)1978+0.35+0.31+0.391,6581,658
    ANESAll respondents (mode not recorded before 2008)1980+0.39+0.33+0.441,0041,004
    ANESAll respondents (mode not recorded before 2008)1982+0.41+0.36+0.47891891
    ANESAll respondents (mode not recorded before 2008)1984+0.41+0.37+0.451,5461,546
    ANESAll respondents (mode not recorded before 2008)1986+0.36+0.32+0.411,6221,622
    ANESAll respondents (mode not recorded before 2008)1988+0.40+0.36+0.451,4191,419
    ANESAll respondents (mode not recorded before 2008)1990+0.34+0.29+0.391,3151,315
    ANESAll respondents (mode not recorded before 2008)1992+0.44+0.40+0.471,8191,812
    ANESAll respondents (mode not recorded before 2008)1994+0.51+0.46+0.551,3981,144
    ANESAll respondents (mode not recorded before 2008)1996+0.54+0.50+0.581,3261,070
    ANESLeft out: telephone or mixed-mode study1998n/an/an/an/an/a
    ANESLeft out: telephone or mixed-mode study2000n/an/an/an/an/a
    ANESLeft out: telephone or mixed-mode study2002n/an/an/an/an/a
    ANESAll respondents (mode not recorded before 2008)2004+0.58+0.53+0.63913766
    ANESFace-to-face respondents2008+0.61+0.57+0.651,6071,053
    ANESFace-to-face respondents2012+0.63+0.59+0.671,454794
    ANESFace-to-face respondents2016+0.67+0.62+0.71903598
    ANESFace-to-face respondents2024+0.72+0.68+0.77868442
    GSSAll respondents (no mode variable before 2004)1974+0.17+0.12+0.231,3451,122
    GSSAll respondents (no mode variable before 2004)1975+0.23+0.18+0.291,3841,136
    GSSAll respondents (no mode variable before 2004)1976+0.22+0.16+0.271,3931,137
    GSSAll respondents (no mode variable before 2004)1977+0.22+0.16+0.271,4401,160
    GSSAll respondents (no mode variable before 2004)1978+0.16+0.09+0.221,422885
    GSSAll respondents (no mode variable before 2004)1980+0.14+0.08+0.211,418941
    GSSAll respondents (no mode variable before 2004)1982+0.24+0.19+0.301,7231,032
    GSSAll respondents (no mode variable before 2004)1983+0.28+0.20+0.36763482
    GSSAll respondents (no mode variable before 2004)1984+0.26+0.20+0.321,386889
    GSSAll respondents (no mode variable before 2004)1985+0.28+0.22+0.341,444919
    GSSAll respondents (no mode variable before 2004)1986+0.31+0.26+0.361,3901,132
    GSSAll respondents (no mode variable before 2004)1987+0.31+0.25+0.361,6661,126
    GSSAll respondents (no mode variable before 2004)1988+0.30+0.24+0.351,4111,139
    GSSAll respondents (no mode variable before 2004)1989+0.27+0.21+0.321,4291,127
    GSSAll respondents (no mode variable before 2004)1990+0.29+0.23+0.351,3011,023
    GSSAll respondents (no mode variable before 2004)1991+0.26+0.20+0.311,4431,105
    GSSAll respondents (no mode variable before 2004)1993+0.34+0.28+0.381,5281,196
    GSSAll respondents (no mode variable before 2004)1994+0.37+0.34+0.412,8062,204
    GSSAll respondents (no mode variable before 2004)1996+0.38+0.35+0.422,7022,168
    GSSAll respondents (no mode variable before 2004)1998+0.36+0.32+0.402,6222,033
    GSSAll respondents (no mode variable before 2004)2000+0.34+0.30+0.382,5932,038
    GSSAll respondents (no mode variable before 2004)2002+0.37+0.31+0.421,299944
    GSSIn-person respondents2004+0.44+0.38+0.501,025761
    GSSIn-person respondents2006+0.47+0.44+0.493,9722,812
    GSSIn-person respondents2008+0.46+0.42+0.511,6671,180
    GSSIn-person respondents2010+0.49+0.45+0.541,7061,100
    GSSIn-person respondents2012+0.47+0.42+0.521,5791,016
    GSSIn-person respondents2014+0.48+0.44+0.522,0541,503
    GSSIn-person respondents2016+0.51+0.47+0.542,4871,651
    GSSIn-person respondents2018+0.55+0.51+0.591,9201,177
    GSSIn-person respondents2022+0.59+0.53+0.641,424566
    GSSIn-person respondents2024+0.56+0.50+0.611,029536
    ANES minus GSSAll interviews in both surveys1974+0.15+0.07+0.242,485n/a
    ANES minus GSSAll interviews in both surveys1976+0.16+0.09+0.232,890n/a
    ANES minus GSSAll interviews in both surveys1978+0.19+0.12+0.273,080n/a
    ANES minus GSSAll interviews in both surveys1980+0.24+0.16+0.332,422n/a
    ANES minus GSSAll interviews in both surveys1982+0.17+0.09+0.252,614n/a
    ANES minus GSSAll interviews in both surveys1984+0.15+0.07+0.222,932n/a
    ANES minus GSSAll interviews in both surveys1986+0.05−0.01+0.123,012n/a
    ANES minus GSSAll interviews in both surveys1988+0.11+0.04+0.182,830n/a
    ANES minus GSSAll interviews in both surveys1990+0.05−0.02+0.122,616n/a
    ANES minus GSSAll interviews in both surveys1994+0.13+0.08+0.194,204n/a
    ANES minus GSSAll interviews in both surveys1996+0.16+0.10+0.214,028n/a
    ANES minus GSSAll interviews in both surveys1998+0.08+0.01+0.153,634n/a
    ANES minus GSSAll interviews in both surveys2000+0.12+0.05+0.203,262n/a
    ANES minus GSSAll interviews in both surveys2002+0.15+0.07+0.222,522n/a
    ANES minus GSSAll interviews in both surveys2004+0.15+0.08+0.222,202n/a
    ANES minus GSSAll interviews in both surveys2008+0.13+0.07+0.193,497n/a
    ANES minus GSSAll interviews in both surveys2012+0.14+0.09+0.197,103n/a
    ANES minus GSSAll interviews in both surveys2016+0.20+0.16+0.245,966n/a
    ANES minus GSSAll interviews in both surveys2024+0.09+0.06+0.137,778n/a
    ANES minus GSSANES face-to-face and GSS in-person respondents2008+0.15+0.09+0.213,274n/a
    ANES minus GSSANES face-to-face and GSS in-person respondents2012+0.16+0.10+0.223,033n/a
    ANES minus GSSANES face-to-face and GSS in-person respondents2016+0.16+0.10+0.223,390n/a
    ANES minus GSSANES face-to-face and GSS in-person respondents2024+0.16+0.09+0.241,897n/a
    ANES minus CESAll interviews in both surveys2008−0.06−0.10−0.0331,546n/a
    ANES minus CESAll interviews in both surveys2012−0.02−0.04+0.0154,260n/a
    ANES minus CESAll interviews in both surveys2016+0.11+0.09+0.1362,086n/a
    ANES minus CESAll interviews in both surveys2020+0.05+0.04+0.0762,436n/a
    ANES minus CESAll interviews in both surveys20240.00−0.02+0.0260,551n/a
    Each line ends higher than it starts, and the CES line falls between 2008 and 2016. Each point is one survey year and the band is its 95% interval. Grey shading marks 1980 to 1988 and 2020 to 2024, the two windows compared later in the chapter.Note: The in-person view keeps all respondents in the years before the file records interview mode; the methods note gives its rules.Note: In the all-interviews view, rings mark ANES 2000 and GSS 1983, 2002 and 2004. The label is the share of respondents with a party answer who placed themselves on the ideology scale, fewer than in the years around them, so each ringed point rests on fewer respondents.Note: A year is drawn only when it has at least 150 respondents who answered both questions and passes a second size condition given in the methods note. A difference needs both surveys to pass.Source: ANES Cumulative Data File, 1972-2024 (weight VCF0009z; weight_ftf for the in-person view); GSS, 1974-2024 (weight_use); CES cumulative file, 2006-2024 (weight); all estimates weighted

    The surveys sit at different levels

    The three lines end higher than they start and do not coincide. The differences view of Figure 1 subtracts the GSS correlation from the ANES correlation in each of the 1919 survey yearsEstimandcount of matched survey yearsWeightscount of yearly estimates, each weightedPopulationyears in which both the ANES and the GSS asked both questionsPeriod1974-2024SourceANES Cumulative Data File years in which both surveys ran. The ANES correlation is the higher of the two in all 1919 survey yearsEstimandcount of matched years in which the ANES correlation exceeds the GSS correlationWeightscount of yearly estimates, each weightedPopulationmatched ANES and GSS survey yearsPeriod1974-2024SourceANES Cumulative Data File of them, and the interval of the difference is above zero in 1717 survey yearsEstimandcount of matched years in which the interval of the ANES minus GSS difference lies above zeroWeightscount of yearly estimates, each weightedPopulationmatched ANES and GSS survey yearsPeriod1974-2024SourceANES Cumulative Data File and includes zero in the other 22 survey yearsEstimandcount of matched years in which the interval of the ANES minus GSS difference includes zeroWeightscount of yearly estimates, each weightedPopulationmatched ANES and GSS survey yearsPeriod1974-2024SourceANES Cumulative Data File. The mean of the yearly differences is +0.14 (95% CI 0.12 to 0.15)+0.14 difference in correlation coefficientsEstimandmean over matched years of the ANES minus GSS correlation; standard errors of independent estimates combined in quadratureInterval95% CI 0.12 to 0.15n69,077Weightsweight (VCF0009z)PopulationANES respondents with a party identification (leaners kept) and a liberal-conservative self-placementPeriod1974-2024, 19 matched survey yearsSourceANES Cumulative Data FileNoteMean of yearly differences between two independent surveys., and the difference ranges from +0.05 (95% CI −0.02 to 0.12)+0.05 difference in correlation coefficientsEstimandANES minus GSS correlation in the year of the smallest difference; standard errors of independent estimates combined in quadratureInterval95% CI −0.02 to 0.12n2,616Weightsweight (VCF0009z)PopulationANES respondents with a party identification (leaners kept) and a liberal-conservative self-placementPeriod1990SourceANES Cumulative Data FileNoteYear 1990. in 1990 to +0.24 (95% CI 0.16 to 0.33)+0.24 difference in correlation coefficientsEstimandANES minus GSS correlation in the year of the largest difference; standard errors of independent estimates combined in quadratureInterval95% CI 0.16 to 0.33n2,422Weightsweight (VCF0009z)PopulationANES respondents with a party identification (leaners kept) and a liberal-conservative self-placementPeriod1980SourceANES Cumulative Data FileNoteYear 1980. in 1980. It is present in both windows: +0.13 (95% CI 0.10 to 0.16)+0.13 difference in correlation coefficientsEstimandANES minus GSS pooled correlation; standard errors of independent estimates combined in quadratureInterval95% CI 0.10 to 0.16n17,683Weightsweight (VCF0009z)PopulationANES respondents with a party identification (leaners kept) and a liberal-conservative self-placementPeriod1980-1988SourceANES Cumulative Data File pooled over 1980 to 1988 and +0.10 (95% CI 0.08 to 0.12)+0.10 difference in correlation coefficientsEstimandANES minus GSS pooled correlation; standard errors of independent estimates combined in quadratureInterval95% CI 0.08 to 0.12n21,955Weightsweight (VCF0009z)PopulationANES respondents with a party identification (leaners kept) and a liberal-conservative self-placementPeriod2020-2024SourceANES Cumulative Data File pooled over 2020 to 2024. The ANES figure for the 1980s therefore describes the ANES and does not stand in for the GSS.

    The CES does not sit on the ANES path either. The ANES minus CES difference is −0.06 (95% CI −0.10 to −0.03)−0.06 difference in correlation coefficientsEstimandANES minus CES correlation in the year of the smallest difference; standard errors of independent estimates combined in quadratureInterval95% CI −0.10 to −0.03n31,546Weightsweight (VCF0009z)PopulationANES respondents with a party identification (leaners kept) and a liberal-conservative self-placementPeriod2008SourceANES Cumulative Data File in 2008 and +0.11 (95% CI 0.09 to 0.13)+0.11 difference in correlation coefficientsEstimandANES minus CES correlation in the year of the largest difference; standard errors of independent estimates combined in quadratureInterval95% CI 0.09 to 0.13n62,086Weightsweight (VCF0009z)PopulationANES respondents with a party identification (leaners kept) and a liberal-conservative self-placementPeriod2016SourceANES Cumulative Data File in 2016, and it averages +0.02 (95% CI 0.01 to 0.03)+0.02 difference in correlation coefficientsEstimandmean over matched years of the ANES minus CES correlation; standard errors of independent estimates combined in quadratureInterval95% CI 0.01 to 0.03n270,879Weightsweight (VCF0009z)PopulationANES respondents with a party identification (leaners kept) and a liberal-conservative self-placementPeriod2008, 2012, 2016, 2020 and 2024SourceANES Cumulative Data FileNoteMean of yearly differences between two independent surveys. over the years in which both surveys ran. In the pooled 2020 to 2024 window the ANES minus CES difference is +0.04 (95% CI 0.03 to 0.05)+0.04 difference in correlation coefficientsEstimandANES minus CES pooled correlation; standard errors of independent estimates combined in quadratureInterval95% CI 0.03 to 0.05n223,430Weightsweight (VCF0009z)PopulationANES respondents with a party identification (leaners kept) and a liberal-conservative self-placementPeriod2020-2024SourceANES Cumulative Data File and the CES minus GSS difference is +0.05 (95% CI 0.04 to 0.07)+0.05 difference in correlation coefficientsEstimandCES minus GSS pooled correlation; standard errors of independent estimates combined in quadratureInterval95% CI 0.04 to 0.07n221,871Weightsweight_cumulative; weight_usePopulationCES respondents with a party identification (leaners kept, not sure excluded) and an ideology answer (not sure excluded)Period2020-2024SourceCooperative Election Study, so no two of the three surveys give the same level in that window.

    The surveys differ in sampling frame, interview mode, the number of points on the ideology scale and the way a respondent who declines to place themselves is recorded. These data do not identify which of those differences produces the level differences. The chapter therefore reports each survey at its own level and compares changes within each survey.

    Interview mode and the level difference

    Both the ANES and the GSS added web interviews, and 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. can move a correlation without any change in the public. The files identify mode for some years only: the ANES mode weights cover 2008, 2012, 2016 and 2024, and the GSS records mode from 2004. Figure 2 splits the GSS estimates by mode. The in-person view of Figure 1 uses the recorded mode where there is one. For earlier years it keeps all respondents, on the ground that those studies were mainly in person, and it leaves out the ANES studies of 1998, 2000 and 2002, which used telephone interviews wholly or in part.

    Figure 2

    In the GSS, web respondents show a higher correlation than in-person respondents

    Weighted party-ideology correlation by interview mode among GSS respondents who answered both questions, 2004 to 2024; bands are 95% intervals for the in-person and web lines

    Show data
    Data for In the GSS, web respondents show a higher correlation than in-person respondents
    Interview modeYearEstimateInterval lowInterval highRespondentsEffective n
    All interviews20040.430.380.481,289937
    All interviews20060.470.440.504,2532,952
    All interviews20080.480.440.521,8901,324
    All interviews20100.500.460.541,9141,231
    All interviews20120.470.420.511,8151,153
    All interviews20140.500.460.532,3681,734
    All interviews20160.500.460.532,6681,782
    All interviews20180.560.520.602,1491,285
    All interviews20210.640.620.673,8332,501
    All interviews20220.630.590.663,3031,304
    All interviews20240.630.600.663,0621,557
    In person20040.440.380.501,025761
    In person20060.470.440.493,9722,812
    In person20080.460.420.511,6671,180
    In person20100.490.450.541,7061,100
    In person20120.470.420.521,5791,016
    In person20140.480.440.522,0541,503
    In person20160.510.470.542,4871,651
    In person20180.550.510.591,9201,177
    In person20220.590.530.641,424566
    In person20240.560.500.611,029536
    Web20210.650.620.673,3702,198
    Web20220.690.650.731,561650
    Web20240.700.670.731,683843
    The web minus in-person difference is +0.11 in 2022 and +0.14 in 2024. The in-person estimate is +0.55 in 2018 and +0.56 in 2024, while the all-interviews estimate goes from +0.56 to +0.63. 2021 has a web estimate of +0.65 and no in-person estimate. In the ANES, the interval of the web minus face-to-face difference includes zero in 2012, 2016 and 2024.Note: Telephone and other GSS interviews are in the all-interviews line and in neither of the other two. The all-interviews line has no band; its intervals are in the data table.Source: GSS, 2004-2024 (weight_use, interview mode from the raw file); all estimates weighted

    In the GSS the web minus in-person difference is +0.11 (95% CI 0.04 to 0.17)+0.11 difference in correlation coefficientsEstimandweb minus in-person GSS correlation; standard errors of independent estimates combined in quadratureInterval95% CI 0.04 to 0.17n2,985Weightsweight_usePopulationGSS respondents with a party identification (leaners kept, other party excluded) and a liberal-conservative self-placementPeriod2022SourceGeneral Social Survey in 2022 and +0.14 (95% CI 0.07 to 0.21)+0.14 difference in correlation coefficientsEstimandweb minus in-person GSS correlation; standard errors of independent estimates combined in quadratureInterval95% CI 0.07 to 0.21n2,712Weightsweight_usePopulationGSS respondents with a party identification (leaners kept, other party excluded) and a liberal-conservative self-placementPeriod2024SourceGeneral Social Survey in 2024. The 2021 GSS was mostly web and has no in-person estimate. Among in-person respondents the GSS correlation is +0.55 (95% CI 0.51 to 0.59)+0.55 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.51 to 0.59n1,920Weightsweight_usePopulationGSS respondents with a party identification (leaners kept, other party excluded) and a liberal-conservative self-placementPeriod2018, in-person interviewsSourceGeneral Social Survey in 2018 and +0.56 (95% CI 0.50 to 0.61)+0.56 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.50 to 0.61n1,029Weightsweight_usePopulationGSS respondents with a party identification (leaners kept, other party excluded) and a liberal-conservative self-placementPeriod2024, in-person interviewsSourceGeneral Social Survey in 2024, two readings whose intervals overlap, while the estimate over all interviews goes from +0.56+0.56 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.52 to 0.60n2,149Weightsweight_usePopulationGSS respondents with a party identification (leaners kept, other party excluded) and a liberal-conservative self-placementPeriod2018SourceGeneral Social Survey to +0.63+0.63 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.60 to 0.66n3,062Weightsweight_usePopulationGSS respondents with a party identification (leaners kept, other party excluded) and a liberal-conservative self-placementPeriod2024SourceGeneral Social Survey. Pooled over 2022 and 2024, the in-person correlation is +0.58 (95% CI 0.53 to 0.62)+0.58 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.53 to 0.62n2,453Weightsweight_usePopulationGSS respondents with a party identification (leaners kept, other party excluded) and a liberal-conservative self-placementPeriod2022 and 2024, waves pooled; in-person interviews onlySourceGeneral Social Survey, against +0.63+0.63 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.62 to 0.65n10,198Weightsweight_usePopulationGSS respondents with a party identification (leaners kept, other party excluded) and a liberal-conservative self-placementPeriod2021, 2022 and 2024, waves pooled; web and in-person interviewsSourceGeneral Social Survey for all interviews pooled over 2021, 2022 and 2024. The GSS rise of the 2020s therefore combines any change among Americans with a change in how the GSS interviewed them, and these data cannot separate the two.

    Standard errors that follow the sampling design leave these results in place. The web minus in-person difference is +0.11 (95% CI 0.03 to 0.18)+0.11 difference in correlation coefficientsEstimandweb minus in-person GSS correlation; jackknife standard error over primary sampling units within strata, normal intervalInterval95% CI 0.03 to 0.18n2,985Weightsweight_usePopulationGSS respondents with a party identification (leaners kept, other party excluded) and a liberal-conservative self-placementPeriod2022SourceGeneral Social Survey in 2022 and +0.14 (95% CI 0.04 to 0.24)+0.14 difference in correlation coefficientsEstimandweb minus in-person GSS correlation; jackknife standard error over primary sampling units within strata, normal intervalInterval95% CI 0.04 to 0.24n2,712Weightsweight_usePopulationGSS respondents with a party identification (leaners kept, other party excluded) and a liberal-conservative self-placementPeriod2024SourceGeneral Social Survey in 2024, with intervals above zero, and the in-person estimate for 2024 minus the one for 2018 is 0.00 (95% CI −0.09 to 0.10)0.00 difference in correlation coefficientsEstimand2024 minus 2018 GSS correlation among in-person respondents; jackknife standard error over primary sampling units within strata, normal intervalInterval95% CI −0.09 to 0.10n2,949Weightsweight_usePopulationGSS respondents with a party identification (leaners kept, other party excluded) and a liberal-conservative self-placementPeriod2018 and 2024SourceGeneral Social Survey, an interval that includes zero.The GSS files carry the stratum and the primary sampling unit of each respondent. A jackknife that leaves out one primary sampling unit at a time within its stratum gives design-based standard errors for the 77 year-and-mode cellsEstimandcount of GSS year and interview-mode cells whose correlation was re-estimated with a design-based standard errorWeightscount of yearly estimates, each weightedPopulationGSS cells behind the interview-mode comparison (2018, 2022, 2024)Period2018-2024SourceGeneral Social Survey GSS year-and-mode cells quoted in this section. They are 1.131.13 ratio of standard errorsEstimandsmallest ratio of the jackknife standard error over primary sampling units within strata to the model-based standard error of the correlationWeightsweight_usePopulationGSS respondents with both answers in the 7 year-and-mode cells behind the interview-mode comparisonPeriod2018, 2022 and 2024SourceGeneral Social Survey to 1.681.68 ratio of standard errorsEstimandlargest ratio of the jackknife standard error over primary sampling units within strata to the model-based standard error of the correlationWeightsweight_usePopulationGSS respondents with both answers in the 7 year-and-mode cells behind the interview-mode comparisonPeriod2018, 2022 and 2024SourceGeneral Social Survey times the model-based standard errors, which treat the weighted sample as a simple random sample. The same check was not made for the ANES or the CES.

    In the ANES, the web minus face-to-face difference is −0.02 (95% CI −0.07 to 0.03)−0.02 difference in correlation coefficientsEstimandweb minus face-to-face ANES correlation; standard errors of independent estimates combined in quadratureInterval95% CI −0.07 to 0.03n5,288Weightsweight_web; weight_ftfPopulationANES respondents with a party identification (leaners kept) and a liberal-conservative self-placementPeriod2012SourceANES Cumulative Data File in 2012, +0.04 (95% CI −0.01 to 0.09)+0.04 difference in correlation coefficientsEstimandweb minus face-to-face ANES correlation; standard errors of independent estimates combined in quadratureInterval95% CI −0.01 to 0.09n3,298Weightsweight_web; weight_ftfPopulationANES respondents with a party identification (leaners kept) and a liberal-conservative self-placementPeriod2016SourceANES Cumulative Data File in 2016 and +0.03 (95% CI −0.02 to 0.08)+0.03 difference in correlation coefficientsEstimandweb minus face-to-face ANES correlation; standard errors of independent estimates combined in quadratureInterval95% CI −0.02 to 0.08n2,587Weightsweight_web; weight_ftfPopulationANES respondents with a party identification (leaners kept) and a liberal-conservative self-placementPeriod2024SourceANES Cumulative Data File in 2024, and each interval includes zero. The face-to-face ANES estimate for 2024 is +0.72 (95% CI 0.68 to 0.77)+0.72 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.68 to 0.77n868Weightsweight_ftfPopulationANES respondents with a party identification (leaners kept) and a liberal-conservative self-placementPeriod2024, face-to-face interviewsSourceANES Cumulative Data File, against +0.72+0.72 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.70 to 0.74n4,716Weightsweight (VCF0009z)PopulationANES respondents with a party identification (leaners kept) and a liberal-conservative self-placementPeriod2024SourceANES Cumulative Data File for all interviews.

    Mode does not remove the level difference between the ANES and the GSS. The in-person line of the differences view of Figure 1 compares face-to-face ANES respondents with in-person GSS respondents. The mean of the ANES minus GSS differences is +0.16 (95% CI 0.13 to 0.19)+0.16 difference in correlation coefficientsEstimandmean over matched years of the ANES face-to-face minus GSS in-person correlation; standard errors of independent estimates combined in quadratureInterval95% CI 0.13 to 0.19n11,594Weightsweight_ftf; weight_usePopulationANES respondents with a party identification (leaners kept) and a liberal-conservative self-placementPeriod2008, 2012, 2016 and 2024SourceANES Cumulative Data File over the 44 survey yearsEstimandcount of matched years with mode information in both surveysWeightscount of yearly estimates, each weightedPopulationyears in which both the ANES and the GSS identify in-person interviewsPeriod2008-2024SourceANES Cumulative Data File years in which both files identify in-person interviews, and the 2024 difference is +0.16 (95% CI 0.09 to 0.24)+0.16 difference in correlation coefficientsEstimandANES face-to-face minus GSS in-person correlation; standard errors of independent estimates combined in quadratureInterval95% CI 0.09 to 0.24n1,897Weightsweight_ftf; weight_usePopulationANES respondents with a party identification (leaners kept) and a liberal-conservative self-placementPeriod2024SourceANES Cumulative Data File. Among web respondents in 2024 the ANES minus GSS difference is +0.05 (95% CI 0.01 to 0.09)+0.05 difference in correlation coefficientsEstimandANES web minus GSS web correlation; standard errors of independent estimates combined in quadratureInterval95% CI 0.01 to 0.09n3,402Weightsweight_web; weight_usePopulationANES respondents with a party identification (leaners kept) and a liberal-conservative self-placementPeriod2024SourceANES Cumulative Data File.

    Distance between the party averages

    The correlation measures how well one answer predicts the other. A second measure asks how far apart the parties' supporters sit on the ideology scale. For each survey year, take the mean placement of Republican identifiers, leaners included, and subtract the mean placement of Democratic identifiers. In the ANES that difference was 0.9 (95% CI 0.7 to 1.0)0.9 points on the seven-point liberal-conservative scaleEstimandweighted mean self-placement of Republican identifiers minus that of Democratic identifiersInterval95% CI 0.7 to 1.0n1,371Weightsweight (VCF0009z)PopulationANES identifiers (leaners included; pure independents excluded) who placed themselvesPeriod1972SourceANES Cumulative Data File points on the seven-point scale in 1972 and 2.4 (95% CI 2.3 to 2.5)2.4 points on the seven-point liberal-conservative scaleEstimandweighted mean self-placement of Republican identifiers minus that of Democratic identifiersInterval95% CI 2.3 to 2.5n4,482Weightsweight (VCF0009z)PopulationANES identifiers (leaners included; pure independents excluded) who placed themselvesPeriod2024SourceANES Cumulative Data File points in 2024.

    Figure 3 divides the difference by a standard deviation of placements, so that surveys with different scales can be compared. Dividing by the ANES standard deviation across all years, 1.481.48 points on the seven-point liberal-conservative scaleEstimandweighted standard deviation of liberal-conservative self-placementWeightsweight (VCF0009z)PopulationANES respondents who placed themselves, all survey years 1972-2024 pooledPeriod1972-2024SourceANES Cumulative Data File scale points, gives 0.58 (95% CI 0.50 to 0.67)0.58 standard deviations of self-placement (survey-wide, all years pooled)Estimandweighted mean ideological self-placement of Republican identifiers minus that of Democratic identifiers, divided by the survey-wide standard deviation of placementsInterval95% CI 0.50 to 0.67n1,371Weightsweight (VCF0009z)PopulationANES identifiers (leaners included; pure independents excluded) who placed themselvesPeriod1972SourceANES Cumulative Data File in 1972 and 1.60 (95% CI 1.53 to 1.67)1.60 standard deviations of self-placement (survey-wide, all years pooled)Estimandweighted mean ideological self-placement of Republican identifiers minus that of Democratic identifiers, divided by the survey-wide standard deviation of placementsInterval95% CI 1.53 to 1.67n4,482Weightsweight (VCF0009z)PopulationANES identifiers (leaners included; pure independents excluded) who placed themselvesPeriod2024SourceANES Cumulative Data File in 2024. The 2024 value means that the average Republican identifier places themselves 1.601.60 standard deviations of self-placement (survey-wide, all years pooled)Estimandweighted mean ideological self-placement of Republican identifiers minus that of Democratic identifiers, divided by the survey-wide standard deviation of placementsInterval95% CI 1.53 to 1.67n4,482Weightsweight (VCF0009z)PopulationANES identifiers (leaners included; pure independents excluded) who placed themselvesPeriod2024SourceANES Cumulative Data File standard deviations to the right of the average Democratic identifier, where a standard deviation is the ANES figure above. The unit is fixed across years, so a change in the value is a change in the distance between two averages. The value says nothing about how spread out the electorate is on either side of the middle.

    Dividing instead by the standard deviation of the same year gives 0.68 (95% CI 0.58 to 0.78)0.68 standard deviations of self-placement in the same yearEstimandweighted mean ideological self-placement of Republican identifiers minus that of Democratic identifiers, divided by the same-year standard deviation of placementsInterval95% CI 0.58 to 0.78n1,371Weightsweight (VCF0009z)PopulationANES identifiers (leaners included; pure independents excluded) who placed themselvesPeriod1972SourceANES Cumulative Data File in 1972 and 1.43 (95% CI 1.37 to 1.50)1.43 standard deviations of self-placement in the same yearEstimandweighted mean ideological self-placement of Republican identifiers minus that of Democratic identifiers, divided by the same-year standard deviation of placementsInterval95% CI 1.37 to 1.50n4,482Weightsweight (VCF0009z)PopulationANES identifiers (leaners included; pure independents excluded) who placed themselvesPeriod2024SourceANES Cumulative Data File in 2024. The rise is smaller under this scaling because the spread of ANES placements within a year also grew, from 1.281.28 points on the seven-point liberal-conservative scaleEstimandweighted standard deviation of liberal-conservative self-placement within the yearWeightsweight (VCF0009z)PopulationANES respondents who placed themselves that yearPeriod1972SourceANES Cumulative Data File scale points in 1972 to 1.661.66 points on the seven-point liberal-conservative scaleEstimandweighted standard deviation of liberal-conservative self-placement within the yearWeightsweight (VCF0009z)PopulationANES respondents who placed themselves that yearPeriod2024SourceANES Cumulative Data File in 2024.

    The GSS series goes from 0.39 (95% CI 0.27 to 0.50)0.39 standard deviations of self-placement (survey-wide, all years pooled)Estimandweighted mean ideological self-placement of Republican identifiers minus that of Democratic identifiers, divided by the survey-wide standard deviation of placementsInterval95% CI 0.27 to 0.50n1,208Weightsweight_usePopulationGSS identifiers (leaners included; pure independents excluded) who placed themselvesPeriod1974SourceGeneral Social Survey in 1974 to 1.50 (95% CI 1.39 to 1.60)1.50 standard deviations of self-placement (survey-wide, all years pooled)Estimandweighted mean ideological self-placement of Republican identifiers minus that of Democratic identifiers, divided by the survey-wide standard deviation of placementsInterval95% CI 1.39 to 1.60n2,304Weightsweight_usePopulationGSS identifiers (leaners included; pure independents excluded) who placed themselvesPeriod2024SourceGeneral Social Survey in 2024 and the CES series from 1.32 (95% CI 1.30 to 1.34)1.32 standard deviations of self-placement (survey-wide, all years pooled)Estimandweighted mean ideological self-placement of Republican identifiers minus that of Democratic identifiers, divided by the survey-wide standard deviation of placementsInterval95% CI 1.30 to 1.34n30,222Weightsweight (single wave); weight_cumulative (pooled years)PopulationCES identifiers (leaners included; pure independents excluded) who placed themselvesPeriod2006SourceCooperative Election Study in 2006 to 1.60 (95% CI 1.58 to 1.62)1.60 standard deviations of self-placement (survey-wide, all years pooled)Estimandweighted mean ideological self-placement of Republican identifiers minus that of Democratic identifiers, divided by the survey-wide standard deviation of placementsInterval95% CI 1.58 to 1.62n49,320Weightsweight (single wave); weight_cumulative (pooled years)PopulationCES identifiers (leaners included; pure independents excluded) who placed themselvesPeriod2024SourceCooperative Election Study in 2024. The ANES series is higher than the GSS series in 1919 survey yearsEstimandcount of matched years in which the ANES distance between the party centers exceeds the GSS distance (each in its own standard deviations)Weightscount of yearly estimates, each weightedPopulationyears in which both the ANES and the GSS asked both questionsPeriod1974-2024SourceANES Cumulative Data File of the 1919 survey yearsEstimandcount of matched survey yearsWeightscount of yearly estimates, each weightedPopulationyears in which both the ANES and the GSS asked both questionsPeriod1974-2024SourceANES Cumulative Data File years that both ran, with each survey in its own standard deviations, and the interval of the difference lies above zero in 1515 survey yearsEstimandcount of matched years in which the interval of the ANES minus GSS distance lies above zero (each in its own standard deviations)Weightscount of yearly estimates, each weightedPopulationyears in which both the ANES and the GSS asked both questionsPeriod1974-2024SourceANES Cumulative Data File of them. In 2024 the ANES minus GSS difference is +0.10 (95% CI −0.02 to 0.23)+0.10 difference in standard deviations of self-placement (each survey in its own survey-wide standard deviation)EstimandANES minus GSS distance between the party centers; standard errors of independent estimates combined in quadratureInterval95% CI −0.02 to 0.23n6,786Weightsweight (VCF0009z); weight_usePopulationANES and GSS identifiers (leaners included; pure independents excluded) who placed themselvesPeriod2024SourceANES Cumulative Data File, an interval that includes zero. The CES series begins above the GSS series.

    Figure 3

    The two parties' average placements moved further apart on all three surveys

    Mean liberal-conservative self-placement of Republican identifiers minus that of Democratic identifiers (leaners included, pure independents left out), in standard deviations (SD) of placement, with 95% intervals

    Republican minus Democratic average (SD)

    Show data
    Data for The two parties' average placements moved further apart on all three surveys
    SurveyYearEstimateInterval lowInterval highRespondentsEffective n
    ANES19720.580.500.671,371595
    ANES19740.690.570.82977303
    ANES19760.710.620.811,306558
    ANES19780.740.650.831,430563
    ANES19800.780.660.89892377
    ANES19820.770.660.88806327
    ANES19840.790.700.881,401682
    ANES19860.640.560.721,441648
    ANES19880.750.660.851,286638
    ANES19900.600.510.701,220547
    ANES19920.890.800.971,630756
    ANES19940.920.821.021,279500
    ANES19960.990.891.081,233461
    ANES19980.800.680.91922345
    ANES20000.940.791.09603240
    ANES20021.060.931.181,171372
    ANES20041.131.011.25849351
    ANES20081.311.211.411,448409
    ANES20121.261.191.324,6131,161
    ANES20161.541.471.612,9491,021
    ANES20201.711.651.766,3531,745
    ANES20241.601.531.674,4821,035
    GSS19740.390.270.501,208360
    GSS19750.530.410.651,196360
    GSS19760.510.390.631,195336
    GSS19770.470.350.581,290374
    GSS19780.360.230.491,225288
    GSS19800.300.170.421,189315
    GSS19820.540.420.671,528315
    GSS19830.540.380.70678174
    GSS19840.490.370.621,244322
    GSS19850.510.380.631,316394
    GSS19860.570.460.681,221427
    GSS19870.600.480.711,502397
    GSS19880.600.480.711,248450
    GSS19890.580.460.701,278497
    GSS19900.540.410.661,162455
    GSS19910.530.420.651,279510
    GSS19930.640.530.751,347492
    GSS19940.760.680.852,471869
    GSS19960.750.670.832,312870
    GSS19980.770.680.862,211766
    GSS20000.750.660.842,113756
    GSS20020.790.670.921,061356
    GSS20040.870.741.001,092377
    GSS20061.030.951.103,3651,044
    GSS20081.050.951.151,611471
    GSS20101.121.001.231,583445
    GSS20121.060.941.181,487360
    GSS20141.060.971.161,922590
    GSS20161.151.061.252,245652
    GSS20181.221.101.331,790435
    GSS20211.531.461.613,046846
    GSS20221.521.411.632,522422
    GSS20241.501.391.602,304580
    CES20061.321.301.3430,22211,342
    CES20071.401.351.448,0011,983
    CES20081.471.441.4927,3549,277
    CES20091.361.331.4010,8593,973
    CES20101.271.251.2946,4308,965
    CES20111.081.051.1116,0203,710
    CES20121.221.201.2443,8478,888
    CES20131.201.171.2412,2502,683
    CES20141.181.161.2042,9018,687
    CES20151.261.231.3011,0252,725
    CES20161.141.121.1650,08211,665
    CES20171.291.261.3313,7242,859
    CES20181.601.581.6247,24510,926
    CES20191.711.671.7414,2764,090
    CES20201.591.571.6148,07111,264
    CES20211.641.611.6719,8635,244
    CES20221.611.591.6347,7419,659
    CES20231.511.481.5518,9415,124
    CES20241.601.581.6249,3207,984
    Each line ends higher than it starts, and the CES line falls from 1.47 in 2008 to 1.14 in 2016. One unit is the standard deviation of placements in all years of the same survey, so it means the same in every year of that survey. Each survey uses its own standard deviation, so levels are not comparable across surveys.Note: Standard deviations are of the seven-point placement in the ANES and GSS and of the five-point placement in the CES, so one unit is not the same number of scale points across surveys.Source: ANES Cumulative Data File, 1972-2024 (weight VCF0009z); GSS, 1974-2024 (weight_use); CES cumulative file, 2006-2024 (weight_cumulative); all estimates weighted

    The rise inside every group

    Figure 4 re-estimates the ANES correlation by year inside three birth cohortsCohortPeople born in the same period, such as the same decade. Comparing cohorts shows whether a difference comes from when people were born, as distinct from their age or the year they were surveyed., three education groups and two regions. In each of the 88 groupsEstimandcount of groups with a yearly correlation lineWeightscount of yearly estimates, each weightedPopulationANES birth-cohort, education and region groupsPeriod1972-2024SourceANES Cumulative Data File groups the line ends above where it starts. Figure 5 compares the pooled 1980 to 1988 window with the pooled 2020 to 2024 window. Of the 77 groupsEstimandcount of groups compared across the two windowsWeightscount of yearly estimates, each weightedPopulationANES cohort, education and region groups with a reading in both windowsPeriod1980-1988 and 2020-2024SourceANES Cumulative Data File groups with a reading in both windows, which is every group except the youngest cohort, 77 groupsEstimandcount of groups whose change between the windows has an interval above zeroWeightscount of yearly estimates, each weightedPopulationANES cohort, education and region groups with a reading in both windowsPeriod1980-1988 and 2020-2024SourceANES Cumulative Data File rose by an amount whose interval lies above zero.

    Figure 4

    The correlation rose inside every cohort, education and region group in the ANES

    Weighted party-ideology correlation by year within each group (solid lines) and among all ANES respondents (dashed line), 1972 to 2024

    Show data
    Data for The correlation rose inside every cohort, education and region group in the ANES
    GroupingLineYearEstimateInterval lowInterval highRespondentsEffective n
    cohortBorn before 194319720.320.260.371,0711,071
    cohortBorn before 194319740.330.250.39818616
    cohortBorn before 194319760.390.330.45920883
    cohortBorn before 194319780.370.310.42913913
    cohortBorn before 194319800.440.380.51554554
    cohortBorn before 194319820.400.320.47436436
    cohortBorn before 194319840.400.340.47675675
    cohortBorn before 194319860.370.300.43649649
    cohortBorn before 194319880.460.390.52565565
    cohortBorn before 194319900.330.250.41467467
    cohortBorn before 194319920.420.350.48625622
    cohortBorn before 194319940.430.350.50485407
    cohortBorn before 194319960.460.370.53432369
    cohortBorn before 194319980.360.240.47265217
    cohortBorn before 194320000.550.410.65163136
    cohortBorn before 194320020.510.410.60328228
    cohortBorn before 194320040.590.480.68195163
    cohortBorn before 194320080.640.540.72249173
    cohortBorn before 194320120.630.570.69641377
    cohortBorn before 194320160.730.660.79297196
    cohortBorn before 194320200.700.630.76467251
    cohortBorn before 194320240.610.480.72217105
    cohortBorn 1943 to 197419720.320.240.40469469
    cohortBorn 1943 to 197419740.260.140.37312250
    cohortBorn 1943 to 197419760.340.270.41570555
    cohortBorn 1943 to 197419780.340.270.40738738
    cohortBorn 1943 to 197419800.310.220.39448448
    cohortBorn 1943 to 197419820.440.360.51454454
    cohortBorn 1943 to 197419840.420.360.47858858
    cohortBorn 1943 to 197419860.370.310.42972972
    cohortBorn 1943 to 197419880.380.320.43852852
    cohortBorn 1943 to 197419900.350.290.41848848
    cohortBorn 1943 to 197419920.460.410.501,1881,184
    cohortBorn 1943 to 197419940.550.500.60883731
    cohortBorn 1943 to 197419960.580.530.63853698
    cohortBorn 1943 to 197419980.490.420.55636548
    cohortBorn 1943 to 197420000.480.390.55451380
    cohortBorn 1943 to 197420020.540.480.60804614
    cohortBorn 1943 to 197420040.580.520.64543462
    cohortBorn 1943 to 197420080.600.550.65928614
    cohortBorn 1943 to 197420120.620.590.653,2411,859
    cohortBorn 1943 to 197420160.720.700.751,7901,306
    cohortBorn 1943 to 197420200.750.730.773,6572,172
    cohortBorn 1943 to 197420240.720.690.742,4361,163
    cohortBorn 1975 or later20040.540.410.65175144
    cohortBorn 1975 or later20080.600.510.67404254
    cohortBorn 1975 or later20120.600.550.641,364782
    cohortBorn 1975 or later20160.650.600.681,121778
    cohortBorn 1975 or later20200.730.710.762,6401,428
    cohortBorn 1975 or later20240.750.720.771,937961
    cohortAll respondents19720.320.270.361,5451,545
    cohortAll respondents19740.330.270.391,140857
    cohortAll respondents19760.380.330.421,4971,441
    cohortAll respondents19780.350.310.391,6581,658
    cohortAll respondents19800.390.330.441,0041,004
    cohortAll respondents19820.410.360.47891891
    cohortAll respondents19840.410.370.451,5461,546
    cohortAll respondents19860.360.320.411,6221,622
    cohortAll respondents19880.400.360.451,4191,419
    cohortAll respondents19900.340.290.391,3151,315
    cohortAll respondents19920.440.400.471,8191,812
    cohortAll respondents19940.510.460.551,3981,144
    cohortAll respondents19960.540.500.581,3261,070
    cohortAll respondents19980.440.390.491,012858
    cohortAll respondents20000.470.400.53669546
    cohortAll respondents20020.520.470.571,223828
    cohortAll respondents20040.580.530.63913766
    cohortAll respondents20080.610.570.651,6071,053
    cohortAll respondents20120.610.590.635,2882,970
    cohortAll respondents20160.700.680.723,2982,332
    cohortAll respondents20200.740.720.757,0413,929
    cohortAll respondents20240.720.700.744,7162,228
    educationHigh school or less19720.260.200.32950950
    educationHigh school or less19740.230.150.31689517
    educationHigh school or less19760.310.250.37822791
    educationHigh school or less19780.270.210.32926926
    educationHigh school or less19800.290.210.37522522
    educationHigh school or less19820.260.170.35401401
    educationHigh school or less19840.260.200.33744744
    educationHigh school or less19860.240.170.31785785
    educationHigh school or less19880.280.210.35654654
    educationHigh school or less19900.170.090.24633633
    educationHigh school or less19920.280.220.35769765
    educationHigh school or less19940.320.240.40552461
    educationHigh school or less19960.380.300.47490394
    educationHigh school or less19980.230.120.33363313
    educationHigh school or less20000.330.170.47179146
    educationHigh school or less20020.410.300.51342250
    educationHigh school or less20040.450.340.55290242
    educationHigh school or less20080.440.350.51634398
    educationHigh school or less20120.470.420.521,708987
    educationHigh school or less20160.590.530.65648480
    educationHigh school or less20200.630.590.681,209722
    educationHigh school or less20240.600.540.65878522
    educationSome college19720.360.270.46320320
    educationSome college19740.450.320.56240181
    educationSome college19760.390.290.47348336
    educationSome college19780.430.350.51392392
    educationSome college19800.410.300.51255255
    educationSome college19820.480.370.57256256
    educationSome college19840.460.390.53455455
    educationSome college19860.370.280.45419419
    educationSome college19880.410.320.49376376
    educationSome college19900.390.290.48308308
    educationSome college19920.490.420.56460459
    educationSome college19940.550.470.62390317
    educationSome college19960.570.490.64391319
    educationSome college19980.500.400.58307267
    educationSome college20000.440.320.55227188
    educationSome college20020.490.400.58385264
    educationSome college20040.530.440.61292260
    educationSome college20080.610.540.67522351
    educationSome college20120.630.590.661,746961
    educationSome college20160.690.650.721,132842
    educationSome college20200.740.710.762,3201,349
    educationSome college20240.720.690.761,461667
    educationCollege graduate19720.500.410.59274274
    educationCollege graduate19740.530.410.64202153
    educationCollege graduate19760.560.480.63322310
    educationCollege graduate19780.540.460.61338338
    educationCollege graduate19800.580.490.66225225
    educationCollege graduate19820.630.550.70233233
    educationCollege graduate19840.630.560.69341341
    educationCollege graduate19860.600.530.66403403
    educationCollege graduate19880.610.540.67358358
    educationCollege graduate19900.610.540.67349349
    educationCollege graduate19920.650.600.70540539
    educationCollege graduate19940.770.730.81422358
    educationCollege graduate19960.740.690.78443385
    educationCollege graduate19980.650.580.71337292
    educationCollege graduate20000.660.580.72262232
    educationCollege graduate20020.700.650.75487390
    educationCollege graduate20040.760.710.80331288
    educationCollege graduate20080.780.730.82441313
    educationCollege graduate20120.770.740.791,7711,019
    educationCollege graduate20160.790.770.811,4901,169
    educationCollege graduate20200.800.780.813,4172,240
    educationCollege graduate20240.790.770.812,3421,111
    educationAll respondents19720.320.270.361,5451,545
    educationAll respondents19740.330.270.391,140857
    educationAll respondents19760.380.330.421,4971,441
    educationAll respondents19780.350.310.391,6581,658
    educationAll respondents19800.390.330.441,0041,004
    educationAll respondents19820.410.360.47891891
    educationAll respondents19840.410.370.451,5461,546
    educationAll respondents19860.360.320.411,6221,622
    educationAll respondents19880.400.360.451,4191,419
    educationAll respondents19900.340.290.391,3151,315
    educationAll respondents19920.440.400.471,8191,812
    educationAll respondents19940.510.460.551,3981,144
    educationAll respondents19960.540.500.581,3261,070
    educationAll respondents19980.440.390.491,012858
    educationAll respondents20000.470.400.53669546
    educationAll respondents20020.520.470.571,223828
    educationAll respondents20040.580.530.63913766
    educationAll respondents20080.610.570.651,6071,053
    educationAll respondents20120.610.590.635,2882,970
    educationAll respondents20160.700.680.723,2982,332
    educationAll respondents20200.740.720.757,0413,929
    educationAll respondents20240.720.700.744,7162,228
    regionSouth19720.260.170.34473473
    regionSouth19740.250.130.35363273
    regionSouth19760.350.260.43414398
    regionSouth19780.310.230.38574574
    regionSouth19800.360.260.45336336
    regionSouth19820.330.220.42290290
    regionSouth19840.320.230.40447447
    regionSouth19860.260.180.34523523
    regionSouth19880.310.220.39431431
    regionSouth19900.240.150.34376376
    regionSouth19920.380.310.45561559
    regionSouth19940.470.390.54475387
    regionSouth19960.560.480.62458361
    regionSouth19980.370.270.47348298
    regionSouth20000.400.270.51226181
    regionSouth20020.460.370.55424290
    regionSouth20040.490.390.58295246
    regionSouth20080.590.530.65746471
    regionSouth20120.570.530.612,0121,102
    regionSouth20160.670.630.711,215887
    regionSouth20200.700.670.732,5671,461
    regionSouth20240.720.680.76958482
    regionOutside the South19720.360.310.411,0721,072
    regionOutside the South19740.380.310.45777584
    regionOutside the South19760.400.350.451,0831,043
    regionOutside the South19780.380.330.431,0841,084
    regionOutside the South19800.410.340.47668668
    regionOutside the South19820.470.400.53601601
    regionOutside the South19840.460.410.501,0991,099
    regionOutside the South19860.410.360.461,0991,099
    regionOutside the South19880.450.400.50988988
    regionOutside the South19900.390.330.44939939
    regionOutside the South19920.460.420.501,2581,254
    regionOutside the South19940.530.470.58923758
    regionOutside the South19960.530.480.58868709
    regionOutside the South19980.480.420.54664560
    regionOutside the South20000.500.420.57441363
    regionOutside the South20020.560.500.61799539
    regionOutside the South20040.630.570.68618520
    regionOutside the South20080.620.560.66861583
    regionOutside the South20120.640.610.663,2761,868
    regionOutside the South20160.710.680.732,0831,450
    regionOutside the South20200.760.740.774,4742,468
    regionOutside the South20240.730.700.761,629801
    regionAll respondents19720.320.270.361,5451,545
    regionAll respondents19740.330.270.391,140857
    regionAll respondents19760.380.330.421,4971,441
    regionAll respondents19780.350.310.391,6581,658
    regionAll respondents19800.390.330.441,0041,004
    regionAll respondents19820.410.360.47891891
    regionAll respondents19840.410.370.451,5461,546
    regionAll respondents19860.360.320.411,6221,622
    regionAll respondents19880.400.360.451,4191,419
    regionAll respondents19900.340.290.391,3151,315
    regionAll respondents19920.440.400.471,8191,812
    regionAll respondents19940.510.460.551,3981,144
    regionAll respondents19960.540.500.581,3261,070
    regionAll respondents19980.440.390.491,012858
    regionAll respondents20000.470.400.53669546
    regionAll respondents20020.520.470.571,223828
    regionAll respondents20040.580.530.63913766
    regionAll respondents20080.610.570.651,6071,053
    regionAll respondents20120.610.590.635,2882,970
    regionAll respondents20160.700.680.723,2982,332
    regionAll respondents20200.740.720.757,0413,929
    regionAll respondents20240.720.700.744,7162,228
    All three cohorts end above where they start. The oldest cohort falls from +0.73 in 2016 to +0.61 in 2024, and its 2024 estimate rests on an effective sample of 104 respondents. The youngest cohort has readings only from 2004.Note: A year is drawn only when it has at least 150 respondents who answered both questions and passes a second size condition given in the methods note.Note: The chart draws no interval bands, so that the lines stay readable; every yearly estimate has a 95% interval in the data table.Source: ANES Cumulative Data File, 1972-2024 (weight VCF0009z); all estimates weighted

    The groups started at different levels. In 1980 to 1988 the correlation was +0.26+0.26 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.23 to 0.30n3,106Weightsweight (VCF0009z)PopulationANES respondents in the group 'High school or less' with both answersPeriod1980-1988, waves pooledSourceANES Cumulative Data File among people with a high school education or less and +0.61+0.61 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.58 to 0.64n1,560Weightsweight (VCF0009z)PopulationANES respondents in the group 'College graduate' with both answersPeriod1980-1988, waves pooledSourceANES Cumulative Data File among college graduates, and it was +0.31+0.31 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.27 to 0.35n2,027Weightsweight (VCF0009z)PopulationANES respondents in the group 'South' with both answersPeriod1980-1988, waves pooledSourceANES Cumulative Data File in the South against +0.44+0.44 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.41 to 0.46n4,455Weightsweight (VCF0009z)PopulationANES respondents in the group 'Outside the South' with both answersPeriod1980-1988, waves pooledSourceANES Cumulative Data File in the rest of the country. College graduates rose less than either of the other two education groups. The change among college graduates minus the change among people with a high school education or less is −0.17 (95% CI −0.23 to −0.11)−0.17 difference in changes of correlation coefficientsEstimandchange in the pooled ANES correlation among college graduates minus the change among respondents with high school or less; standard errors of independent estimates combined in quadratureInterval95% CI −0.23 to −0.11n12,512Weightsweight (VCF0009z)PopulationANES college graduates and respondents with high school or less (disjoint education groups)Period1980-1988 to 2020-2024SourceANES Cumulative Data File, and minus the change among people with some college it is −0.12 (95% CI −0.18 to −0.07)−0.12 difference in changes of correlation coefficientsEstimandchange in the pooled ANES correlation among college graduates minus the change among respondents with some college; standard errors of independent estimates combined in quadratureInterval95% CI −0.18 to −0.07n12,861Weightsweight (VCF0009z)PopulationANES college graduates and respondents with some college (disjoint education groups)Period1980-1988 to 2020-2024SourceANES Cumulative Data File. A correlation cannot pass one, so a group that starts at +0.61+0.61 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.58 to 0.64n1,560Weightsweight (VCF0009z)PopulationANES respondents in the group 'College graduate' with both answersPeriod1980-1988, waves pooledSourceANES Cumulative Data File has less room to rise than a group that starts at +0.26+0.26 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.23 to 0.30n3,106Weightsweight (VCF0009z)PopulationANES respondents in the group 'High school or less' with both answersPeriod1980-1988, waves pooledSourceANES Cumulative Data File.

    The differences between groups are smaller in 2020 to 2024. The three education groups stand at +0.62+0.62 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.58 to 0.65n2,087Weightsweight (VCF0009z)PopulationANES respondents in the group 'High school or less' with both answersPeriod2020 and 2024, waves pooledSourceANES Cumulative Data File to +0.80+0.80 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.78 to 0.81n5,759Weightsweight (VCF0009z)PopulationANES respondents in the group 'College graduate' with both answersPeriod2020 and 2024, waves pooledSourceANES Cumulative Data File, against +0.26+0.26 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.23 to 0.30n3,106Weightsweight (VCF0009z)PopulationANES respondents in the group 'High school or less' with both answersPeriod1980-1988, waves pooledSourceANES Cumulative Data File to +0.61+0.61 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.58 to 0.64n1,560Weightsweight (VCF0009z)PopulationANES respondents in the group 'College graduate' with both answersPeriod1980-1988, waves pooledSourceANES Cumulative Data File in 1980 to 1988, and the South, at +0.71+0.71 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.69 to 0.73n3,525Weightsweight (VCF0009z)PopulationANES respondents in the group 'South' with both answersPeriod2020 and 2024, waves pooledSourceANES Cumulative Data File, is still below the rest of the country, at +0.75+0.75 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.73 to 0.76n6,103Weightsweight (VCF0009z)PopulationANES respondents in the group 'Outside the South' with both answersPeriod2020 and 2024, waves pooledSourceANES Cumulative Data File, by less than in 1980 to 1988.

    Figure 5

    The correlation rose in every group, and less among college graduates than among other education groups

    Weighted party-ideology correlation estimated from all respondents in 1980 to 1988 together and in 2020 to 2024 together, by group, with 95% intervals; rows run by classification and, within each, by the size of the change

    • 1980-88
    • 2020-24
    Show data
    Data for The correlation rose in every group, and less among college graduates than among other education groups
    SurveyGroup1980 to 1988LowHighRespondents2020 to 2024LowHighRespondentsChange
    ANESEveryone0.390.370.416,4820.730.720.7411,757+0.34
    ANESBorn 1943 to 19740.380.350.413,5840.740.720.756,093+0.35
    ANESBorn before 19430.410.380.442,8790.670.600.72684+0.25
    ANESHigh school or less0.260.230.303,1060.620.580.652,087+0.35
    ANESSome college0.420.380.461,7610.730.710.753,781+0.31
    ANESCollege graduate0.610.580.641,5600.800.780.815,759+0.19
    ANESSouth0.310.270.352,0270.710.690.733,525+0.40
    ANESOutside the South0.440.410.464,4550.750.730.766,103+0.31
    GSSEveryone0.270.240.2911,2010.580.530.622,453+0.31
    GSSBorn 1943 to 19740.270.240.295,8600.590.530.641,153+0.32
    GSSNo bachelor's degree0.200.180.239,2200.490.440.551,751+0.29
    GSSBachelor's degree or more0.550.520.591,9610.760.710.81702+0.21
    College graduates start highest and rise less than the other education groups, and every group moves to the right. Open dot: 1980 to 1988. Filled dot: 2020 to 2024. The GSS view uses in-person interviews only in 2020 to 2024 and has no group for people born before 1943, who are too few in that window.Note: Americans born 1975 or later have no 1980 to 1988 reading, so they are not in the comparison. Their 2020 to 2024 correlation is +0.74.Source: ANES Cumulative Data File, 1980-1988 and 2020-2024 (weight VCF0009z); GSS, 1980-1988 and 2020-2024 in-person interviews (weight_use); all estimates weighted

    Respondents born before 1943 went from +0.41+0.41 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.38 to 0.44n2,879Weightsweight (VCF0009z)PopulationANES respondents in the group 'Born before 1943' with both answersPeriod1980-1988, waves pooledSourceANES Cumulative Data File to +0.67+0.67 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.60 to 0.72n684Weightsweight (VCF0009z)PopulationANES respondents in the group 'Born before 1943' with both answersPeriod2020 and 2024, waves pooledSourceANES Cumulative Data File, a change of +0.25 (95% CI 0.19 to 0.32)+0.25 difference in correlation coefficientsEstimandchange in the pooled ANES correlation among respondents born before 1943; standard errors of independent estimates combined in quadratureInterval95% CI 0.19 to 0.32n3,563Weightsweight (VCF0009z)PopulationANES respondents born before 1943 with both answersPeriod1980-1988 to 2020-2024SourceANES Cumulative Data File. Respondents born in 1975 or later, who were too young for the earlier window, stand at +0.74+0.74 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.72 to 0.76n4,577Weightsweight (VCF0009z)PopulationANES respondents in the group 'Born 1975 or later' with both answersPeriod2020 and 2024, waves pooledSourceANES Cumulative Data File in the 2020s. The two smallest changes of the 77 groupsEstimandcount of groups compared across the two windowsWeightscount of yearly estimates, each weightedPopulationANES cohort, education and region groups with a reading in both windowsPeriod1980-1988 and 2020-2024SourceANES Cumulative Data File groups belong to college graduates and to people born before 1943. The chapter does not compare those two changes with an interval, because a college graduate can also be born before 1943 and the two groups overlap; changes are compared only between groups of one classification.

    The GSS gives the same result for the groups it can define: among in-person respondents in 2020 to 2024, 33 groupsEstimandcount of groups whose change between the windows has an interval above zeroWeightscount of yearly estimates, each weightedPopulationGSS cohort and education groups with a reading in both windows (in-person interviews)Period1980-1988 and 2020-2024SourceGeneral Social Survey of the 33 groupsEstimandcount of groups compared across the two windowsWeightscount of yearly estimates, each weightedPopulationGSS cohort and education groups with a reading in both windows (in-person interviews)Period1980-1988 and 2020-2024SourceGeneral Social Survey groups with a reading in both windows rose by an amount whose interval lies above zero.

    A cohort is not a fixed set of people. The group born before 1943 in the 2020s consists of its oldest surviving members who agreed to be interviewed, so a rise inside a cohort does not show that the same people changed. Chapter 1.4, Cohort change in partisan sorting follows the cohorts and states what that comparison cannot show.

    Partisans and their camps

    Figure 6 counts the party's identifiers, leaners included, who place themselves left of the middle, at the middle or right of the middle. On the seven-point scales, left and right of the middle include slightly liberal and slightly conservative. In the ANES the liberal share of Democrats who placed themselves rose from 36% (95% CI 33% to 40%)36% percent of the party's identifiers who placed themselvesEstimandweighted share in the liberal camp, Wilson intervalInterval95% CI 33% to 40%n776Weightsweight (VCF0009z)PopulationANES Democrat identifiers (leaners included) who placed themselves on the liberal-conservative scalePeriod1972SourceANES Cumulative Data File in 1972 to 62% (95% CI 59% to 65%)62% percent of the party's identifiers who placed themselvesEstimandweighted share in the liberal camp, Wilson intervalInterval95% CI 59% to 65%n2,310Weightsweight (VCF0009z)PopulationANES Democrat identifiers (leaners included) who placed themselves on the liberal-conservative scalePeriod2024SourceANES Cumulative Data File in 2024, and the conservative share of Republicans rose from 53% (95% CI 49% to 57%)53% percent of the party's identifiers who placed themselvesEstimandweighted share in the conservative camp, Wilson intervalInterval95% CI 49% to 57%n595Weightsweight (VCF0009z)PopulationANES Republican identifiers (leaners included) who placed themselves on the liberal-conservative scalePeriod1972SourceANES Cumulative Data File to 77% (95% CI 75% to 80%)77% percent of the party's identifiers who placed themselvesEstimandweighted share in the conservative camp, Wilson intervalInterval95% CI 75% to 80%n2,172Weightsweight (VCF0009z)PopulationANES Republican identifiers (leaners included) who placed themselves on the liberal-conservative scalePeriod2024SourceANES Cumulative Data File. The opposite camps shrank: conservative Democrats went from 25%25% percent of the party's identifiers who placed themselvesEstimandweighted share in the conservative camp, Wilson intervalInterval95% CI 22% to 28%n776Weightsweight (VCF0009z)PopulationANES Democrat identifiers (leaners included) who placed themselves on the liberal-conservative scalePeriod1972SourceANES Cumulative Data File to 8%8% percent of the party's identifiers who placed themselvesEstimandweighted share in the conservative camp, Wilson intervalInterval95% CI 7% to 10%n2,310Weightsweight (VCF0009z)PopulationANES Democrat identifiers (leaners included) who placed themselves on the liberal-conservative scalePeriod2024SourceANES Cumulative Data File, and liberal Republicans from 13%13% percent of the party's identifiers who placed themselvesEstimandweighted share in the liberal camp, Wilson intervalInterval95% CI 11% to 16%n595Weightsweight (VCF0009z)PopulationANES Republican identifiers (leaners included) who placed themselves on the liberal-conservative scalePeriod1972SourceANES Cumulative Data File to 4%4% percent of the party's identifiers who placed themselvesEstimandweighted share in the liberal camp, Wilson intervalInterval95% CI 3% to 6%n2,172Weightsweight (VCF0009z)PopulationANES Republican identifiers (leaners included) who placed themselves on the liberal-conservative scalePeriod2024SourceANES Cumulative Data File. The GSS shows the same direction, with liberal Democrats going from 36%36% percent of the party's identifiers who placed themselvesEstimandweighted share in the liberal camp, Wilson intervalInterval95% CI 32% to 40%n789Weightsweight_usePopulationGSS Democrat identifiers (leaners included) who placed themselves on the liberal-conservative scalePeriod1974SourceGeneral Social Survey in 1974 to 59%59% percent of the party's identifiers who placed themselvesEstimandweighted share in the liberal camp, Wilson intervalInterval95% CI 55% to 63%n1,215Weightsweight_usePopulationGSS Democrat identifiers (leaners included) who placed themselves on the liberal-conservative scalePeriod2024SourceGeneral Social Survey in 2024 and conservative Republicans from 42%42% percent of the party's identifiers who placed themselvesEstimandweighted share in the conservative camp, Wilson intervalInterval95% CI 37% to 47%n419Weightsweight_usePopulationGSS Republican identifiers (leaners included) who placed themselves on the liberal-conservative scalePeriod1974SourceGeneral Social Survey to 71%71% percent of the party's identifiers who placed themselvesEstimandweighted share in the conservative camp, Wilson intervalInterval95% CI 67% to 74%n1,089Weightsweight_usePopulationGSS Republican identifiers (leaners included) who placed themselves on the liberal-conservative scalePeriod2024SourceGeneral Social Survey. In the CES, which starts in 2006, the liberal share of Democrats rose and the conservative share of Republicans did not change detectably.

    Figure 6

    Larger shares of Democrats are liberal and of Republicans conservative than in the 1970s

    Share of each party's identifiers (leaners included) who place themselves as liberal, moderate or conservative, among identifiers who placed themselves, with 95% intervals; liberal and conservative include slightly liberal and slightly conservative on the seven-point scales of the ANES and GSS

    Share of the party's identifiers (%)

    DemocratsRepublicans
    Show data
    Data for Larger shares of Democrats are liberal and of Republicans conservative than in the 1970s
    SurveyPartyCampYearShareInterval lowInterval highRespondentsEffective n
    ANESDemocratsLiberal197236.1%32.8%39.5%776776
    ANESDemocratsLiberal197439.3%34.8%44.0%573431
    ANESDemocratsLiberal197636.6%33.1%40.2%725699
    ANESDemocratsLiberal197837.1%34.0%40.4%867867
    ANESDemocratsLiberal198037.9%33.8%42.1%515515
    ANESDemocratsLiberal198234.2%30.1%38.6%479479
    ANESDemocratsLiberal198440.5%36.9%44.1%719719
    ANESDemocratsLiberal198635.7%32.4%39.1%793793
    ANESDemocratsLiberal198838.1%34.4%41.9%638638
    ANESDemocratsLiberal199036.7%33.1%40.4%673673
    ANESDemocratsLiberal199244.1%40.8%47.4%871868
    ANESDemocratsLiberal199435.5%31.4%39.8%622500
    ANESDemocratsLiberal199640.9%36.8%45.1%667537
    ANESDemocratsLiberal199838.1%33.7%42.7%515437
    ANESDemocratsLiberal200044.1%38.1%50.2%319255
    ANESDemocratsLiberal200246.5%41.5%51.6%568372
    ANESDemocratsLiberal200444.2%39.2%49.4%438363
    ANESDemocratsLiberal200851.7%47.5%55.9%894547
    ANESDemocratsLiberal201249.0%46.4%51.6%2,7091,417
    ANESDemocratsLiberal201661.2%58.2%64.2%1,4671,021
    ANESDemocratsLiberal202065.1%62.9%67.3%3,2771,745
    ANESDemocratsLiberal202461.7%58.8%64.5%2,3101,087
    ANESDemocratsModerate197239.3%35.9%42.8%776776
    ANESDemocratsModerate197437.7%33.3%42.4%573431
    ANESDemocratsModerate197639.4%35.8%43.0%725699
    ANESDemocratsModerate197839.8%36.6%43.1%867867
    ANESDemocratsModerate198034.2%30.2%38.4%515515
    ANESDemocratsModerate198240.3%36.0%44.7%479479
    ANESDemocratsModerate198437.3%33.8%40.9%719719
    ANESDemocratsModerate198639.7%36.4%43.2%793793
    ANESDemocratsModerate198834.5%30.9%38.3%638638
    ANESDemocratsModerate199037.4%33.9%41.2%673673
    ANESDemocratsModerate199232.6%29.5%35.8%871868
    ANESDemocratsModerate199441.9%37.6%46.2%622500
    ANESDemocratsModerate199637.3%33.3%41.5%667537
    ANESDemocratsModerate199840.4%35.9%45.1%515437
    ANESDemocratsModerate200033.7%28.2%39.7%319255
    ANESDemocratsModerate200232.4%27.8%37.3%568372
    ANESDemocratsModerate200439.7%34.8%44.8%438363
    ANESDemocratsModerate200831.7%28.0%35.8%894547
    ANESDemocratsModerate201236.6%34.1%39.1%2,7091,417
    ANESDemocratsModerate201628.5%25.8%31.3%1,4671,021
    ANESDemocratsModerate202028.3%26.2%30.5%3,2771,745
    ANESDemocratsModerate202430.1%27.5%32.9%2,3101,087
    ANESDemocratsConservative197224.6%21.7%27.8%776776
    ANESDemocratsConservative197423.0%19.2%27.1%573431
    ANESDemocratsConservative197624.1%21.0%27.4%725699
    ANESDemocratsConservative197823.1%20.4%26.0%867867
    ANESDemocratsConservative198028.0%24.3%32.0%515515
    ANESDemocratsConservative198225.5%21.8%29.6%479479
    ANESDemocratsConservative198422.3%19.4%25.4%719719
    ANESDemocratsConservative198624.6%21.7%27.7%793793
    ANESDemocratsConservative198827.4%24.1%31.0%638638
    ANESDemocratsConservative199025.9%22.7%29.3%673673
    ANESDemocratsConservative199223.4%20.7%26.3%871868
    ANESDemocratsConservative199422.6%19.2%26.5%622500
    ANESDemocratsConservative199621.8%18.5%25.4%667537
    ANESDemocratsConservative199821.5%17.9%25.6%515437
    ANESDemocratsConservative200022.3%17.6%27.7%319255
    ANESDemocratsConservative200221.1%17.3%25.6%568372
    ANESDemocratsConservative200416.1%12.7%20.2%438363
    ANESDemocratsConservative200816.6%13.7%19.9%894547
    ANESDemocratsConservative201214.5%12.7%16.4%2,7091,417
    ANESDemocratsConservative201610.3%8.6%12.3%1,4671,021
    ANESDemocratsConservative20206.6%5.5%7.8%3,2771,745
    ANESDemocratsConservative20248.2%6.7%10.0%2,3101,087
    ANESRepublicansLiberal197213.1%10.6%16.1%595595
    ANESRepublicansLiberal197411.4%8.3%15.4%404303
    ANESRepublicansLiberal19769.8%7.6%12.5%581558
    ANESRepublicansLiberal197810.1%7.9%12.9%563563
    ANESRepublicansLiberal19809.8%7.2%13.2%377377
    ANESRepublicansLiberal19827.3%5.0%10.7%327327
    ANESRepublicansLiberal198411.4%9.3%14.0%682682
    ANESRepublicansLiberal198611.0%8.8%13.6%648648
    ANESRepublicansLiberal19889.9%7.8%12.4%648648
    ANESRepublicansLiberal199012.1%9.6%15.1%547547
    ANESRepublicansLiberal19929.5%7.6%11.8%759756
    ANESRepublicansLiberal19945.1%3.5%7.3%657547
    ANESRepublicansLiberal19966.1%4.3%8.7%566461
    ANESRepublicansLiberal19988.3%5.9%11.7%407345
    ANESRepublicansLiberal20008.7%5.8%13.0%284240
    ANESRepublicansLiberal20028.4%6.1%11.4%603427
    ANESRepublicansLiberal20046.6%4.4%9.7%411351
    ANESRepublicansLiberal20085.2%3.5%7.8%554409
    ANESRepublicansLiberal20124.6%3.5%5.9%1,9041,161
    ANESRepublicansLiberal20164.0%3.0%5.4%1,4821,063
    ANESRepublicansLiberal20203.7%3.0%4.7%3,0761,816
    ANESRepublicansLiberal20244.4%3.3%5.8%2,1721,035
    ANESRepublicansModerate197233.4%29.8%37.3%595595
    ANESRepublicansModerate197432.4%27.4%37.9%404303
    ANESRepublicansModerate197630.0%26.4%34.0%581558
    ANESRepublicansModerate197828.4%24.8%32.3%563563
    ANESRepublicansModerate198023.1%19.1%27.6%377377
    ANESRepublicansModerate198225.7%21.3%30.7%327327
    ANESRepublicansModerate198425.2%22.1%28.6%682682
    ANESRepublicansModerate198630.2%26.8%33.9%648648
    ANESRepublicansModerate198824.8%21.7%28.3%648648
    ANESRepublicansModerate199033.1%29.3%37.1%547547
    ANESRepublicansModerate199226.9%23.9%30.2%759756
    ANESRepublicansModerate199424.9%21.4%28.6%657547
    ANESRepublicansModerate199622.6%19.0%26.6%566461
    ANESRepublicansModerate199826.5%22.2%31.4%407345
    ANESRepublicansModerate200025.5%20.4%31.4%284240
    ANESRepublicansModerate200222.3%18.6%26.5%603427
    ANESRepublicansModerate200422.8%18.7%27.5%411351
    ANESRepublicansModerate200818.7%15.2%22.7%554409
    ANESRepublicansModerate201221.7%19.4%24.2%1,9041,161
    ANESRepublicansModerate201619.5%17.2%22.0%1,4821,063
    ANESRepublicansModerate202018.0%16.3%19.8%3,0761,816
    ANESRepublicansModerate202418.3%16.0%20.7%2,1721,035
    ANESRepublicansConservative197253.4%49.4%57.4%595595
    ANESRepublicansConservative197456.2%50.6%61.7%404303
    ANESRepublicansConservative197660.2%56.0%64.1%581558
    ANESRepublicansConservative197861.5%57.4%65.4%563563
    ANESRepublicansConservative198067.1%62.2%71.7%377377
    ANESRepublicansConservative198267.0%61.7%71.8%327327
    ANESRepublicansConservative198463.3%59.7%66.9%682682
    ANESRepublicansConservative198658.8%55.0%62.5%648648
    ANESRepublicansConservative198865.3%61.5%68.8%648648
    ANESRepublicansConservative199054.8%50.7%59.0%547547
    ANESRepublicansConservative199263.6%60.1%67.0%759756
    ANESRepublicansConservative199470.0%66.1%73.7%657547
    ANESRepublicansConservative199671.3%67.0%75.2%566461
    ANESRepublicansConservative199865.1%60.0%70.0%407345
    ANESRepublicansConservative200065.7%59.5%71.5%284240
    ANESRepublicansConservative200269.3%64.8%73.5%603427
    ANESRepublicansConservative200470.6%65.7%75.2%411351
    ANESRepublicansConservative200876.1%71.7%80.0%554409
    ANESRepublicansConservative201273.7%71.1%76.2%1,9041,161
    ANESRepublicansConservative201676.5%73.8%78.9%1,4821,063
    ANESRepublicansConservative202078.3%76.4%80.1%3,0761,816
    ANESRepublicansConservative202477.4%74.7%79.8%2,1721,035
    GSSDemocratsLiberal197436.0%32.4%39.7%789653
    GSSDemocratsLiberal197538.2%34.5%42.1%766623
    GSSDemocratsLiberal197633.4%29.8%37.1%791641
    GSSDemocratsLiberal197735.9%32.3%39.6%841664
    GSSDemocratsLiberal197833.8%29.7%38.2%756479
    GSSDemocratsLiberal198028.2%24.4%32.3%741486
    GSSDemocratsLiberal198234.1%30.4%38.0%1,041594
    GSSDemocratsLiberal198330.0%24.8%35.7%420269
    GSSDemocratsLiberal198431.8%27.7%36.1%724466
    GSSDemocratsLiberal198534.8%30.6%39.4%719448
    GSSDemocratsLiberal198631.4%27.7%35.3%703570
    GSSDemocratsLiberal198738.4%34.6%42.3%961615
    GSSDemocratsLiberal198839.4%35.4%43.5%697559
    GSSDemocratsLiberal198941.1%36.9%45.5%666504
    GSSDemocratsLiberal199038.1%33.8%42.6%590461
    GSSDemocratsLiberal199139.1%34.9%43.4%655510
    GSSDemocratsLiberal199340.8%36.8%44.9%725567
    GSSDemocratsLiberal199439.2%36.4%42.2%1,3701,074
    GSSDemocratsLiberal199636.8%33.8%39.8%1,272997
    GSSDemocratsLiberal199838.8%35.8%41.8%1,274986
    GSSDemocratsLiberal200036.0%32.9%39.2%1,189914
    GSSDemocratsLiberal200237.4%32.9%42.1%596425
    GSSDemocratsLiberal200439.8%35.2%44.6%588411
    GSSDemocratsLiberal200642.7%40.0%45.4%1,9181,312
    GSSDemocratsLiberal200841.7%38.0%45.5%954653
    GSSDemocratsLiberal201046.9%42.9%51.0%932569
    GSSDemocratsLiberal201243.2%39.3%47.2%902605
    GSSDemocratsLiberal201441.5%38.2%44.9%1,143818
    GSSDemocratsLiberal201648.7%45.4%52.1%1,334865
    GSSDemocratsLiberal201852.2%48.3%56.0%1,013636
    GSSDemocratsLiberal202158.3%55.4%61.1%1,8181,138
    GSSDemocratsLiberal202259.8%55.8%63.7%1,416592
    GSSDemocratsLiberal202458.9%54.9%62.7%1,215603
    GSSDemocratsModerate197439.5%35.8%43.3%789653
    GSSDemocratsModerate197540.6%36.8%44.5%766623
    GSSDemocratsModerate197642.4%38.7%46.3%791641
    GSSDemocratsModerate197740.8%37.1%44.6%841664
    GSSDemocratsModerate197839.4%35.1%43.9%756479
    GSSDemocratsModerate198045.0%40.6%49.4%741486
    GSSDemocratsModerate198240.9%37.0%44.9%1,041594
    GSSDemocratsModerate198345.0%39.2%51.0%420269
    GSSDemocratsModerate198440.9%36.5%45.4%724466
    GSSDemocratsModerate198540.3%35.9%44.9%719448
    GSSDemocratsModerate198643.3%39.3%47.4%703570
    GSSDemocratsModerate198740.3%36.5%44.2%961615
    GSSDemocratsModerate198836.7%32.8%40.8%697559
    GSSDemocratsModerate198938.4%34.2%42.7%666504
    GSSDemocratsModerate199037.1%32.8%41.6%590461
    GSSDemocratsModerate199138.5%34.4%42.8%655510
    GSSDemocratsModerate199335.0%31.2%39.1%725567
    GSSDemocratsModerate199438.0%35.1%40.9%1,3701,074
    GSSDemocratsModerate199638.7%35.8%41.8%1,272997
    GSSDemocratsModerate199838.9%35.9%41.9%1,274986
    GSSDemocratsModerate200043.6%40.4%46.8%1,189914
    GSSDemocratsModerate200243.2%38.5%47.9%596425
    GSSDemocratsModerate200439.0%34.4%43.8%588411
    GSSDemocratsModerate200639.6%37.0%42.3%1,9181,312
    GSSDemocratsModerate200840.9%37.2%44.7%954653
    GSSDemocratsModerate201037.2%33.3%41.3%932569
    GSSDemocratsModerate201239.4%35.6%43.3%902605
    GSSDemocratsModerate201442.0%38.7%45.4%1,143818
    GSSDemocratsModerate201635.2%32.1%38.5%1,334865
    GSSDemocratsModerate201835.5%31.9%39.3%1,013636
    GSSDemocratsModerate202132.4%29.7%35.2%1,8181,138
    GSSDemocratsModerate202230.1%26.5%33.9%1,416592
    GSSDemocratsModerate202429.7%26.2%33.5%1,215603
    GSSDemocratsConservative197424.6%21.4%28.0%789653
    GSSDemocratsConservative197521.2%18.2%24.6%766623
    GSSDemocratsConservative197624.2%21.1%27.7%791641
    GSSDemocratsConservative197723.3%20.3%26.7%841664
    GSSDemocratsConservative197826.8%23.0%30.9%756479
    GSSDemocratsConservative198026.8%23.1%30.9%741486
    GSSDemocratsConservative198225.0%21.7%28.7%1,041594
    GSSDemocratsConservative198325.0%20.2%30.5%420269
    GSSDemocratsConservative198427.4%23.5%31.6%724466
    GSSDemocratsConservative198524.8%21.0%29.0%719448
    GSSDemocratsConservative198625.3%21.9%29.0%703570
    GSSDemocratsConservative198721.3%18.2%24.7%961615
    GSSDemocratsConservative198823.9%20.6%27.6%697559
    GSSDemocratsConservative198920.5%17.2%24.2%666504
    GSSDemocratsConservative199024.8%21.1%29.0%590461
    GSSDemocratsConservative199122.4%19.0%26.3%655510
    GSSDemocratsConservative199324.2%20.9%27.9%725567
    GSSDemocratsConservative199422.8%20.4%25.4%1,3701,074
    GSSDemocratsConservative199624.5%21.9%27.3%1,272997
    GSSDemocratsConservative199822.4%19.9%25.1%1,274986
    GSSDemocratsConservative200020.4%18.0%23.2%1,189914
    GSSDemocratsConservative200219.4%16.0%23.5%596425
    GSSDemocratsConservative200421.2%17.5%25.4%588411
    GSSDemocratsConservative200617.7%15.7%19.9%1,9181,312
    GSSDemocratsConservative200817.4%14.7%20.5%954653
    GSSDemocratsConservative201015.9%13.1%19.1%932569
    GSSDemocratsConservative201217.4%14.6%20.6%902605
    GSSDemocratsConservative201416.5%14.1%19.2%1,143818
    GSSDemocratsConservative201616.0%13.8%18.6%1,334865
    GSSDemocratsConservative201812.4%10.0%15.2%1,013636
    GSSDemocratsConservative20219.3%7.8%11.1%1,8181,138
    GSSDemocratsConservative202210.1%8.0%12.8%1,416592
    GSSDemocratsConservative202411.4%9.1%14.2%1,215603
    GSSRepublicansLiberal197419.0%15.3%23.4%419360
    GSSRepublicansLiberal197518.5%14.8%22.8%430360
    GSSRepublicansLiberal197618.7%14.9%23.3%404336
    GSSRepublicansLiberal197721.2%17.3%25.6%449374
    GSSRepublicansLiberal197820.2%16.0%25.2%469288
    GSSRepublicansLiberal198020.7%16.6%25.5%448315
    GSSRepublicansLiberal198214.1%10.7%18.4%487315
    GSSRepublicansLiberal198314.5%10.1%20.5%258174
    GSSRepublicansLiberal198412.5%9.3%16.6%520322
    GSSRepublicansLiberal198514.6%11.5%18.5%597394
    GSSRepublicansLiberal198614.5%11.4%18.1%518427
    GSSRepublicansLiberal198717.1%13.7%21.1%541397
    GSSRepublicansLiberal198816.0%12.9%19.6%551450
    GSSRepublicansLiberal198916.8%13.8%20.4%612497
    GSSRepublicansLiberal199019.0%15.6%22.8%572455
    GSSRepublicansLiberal199118.1%15.0%21.6%624513
    GSSRepublicansLiberal199313.4%10.7%16.7%622492
    GSSRepublicansLiberal199414.8%12.6%17.3%1,101869
    GSSRepublicansLiberal199611.3%9.3%13.6%1,040870
    GSSRepublicansLiberal199813.0%10.8%15.6%937766
    GSSRepublicansLiberal200013.0%10.8%15.6%924756
    GSSRepublicansLiberal200210.6%7.8%14.3%465356
    GSSRepublicansLiberal200410.3%7.6%13.7%504377
    GSSRepublicansLiberal20069.0%7.4%10.9%1,4471,044
    GSSRepublicansLiberal20086.0%4.2%8.6%657471
    GSSRepublicansLiberal201010.3%7.8%13.4%651445
    GSSRepublicansLiberal20128.2%5.8%11.5%585360
    GSSRepublicansLiberal20147.4%5.5%9.8%779590
    GSSRepublicansLiberal20166.8%5.1%8.9%911652
    GSSRepublicansLiberal20188.7%6.4%11.7%777435
    GSSRepublicansLiberal20213.7%2.6%5.2%1,228846
    GSSRepublicansLiberal20224.2%2.7%6.6%1,106422
    GSSRepublicansLiberal20245.7%4.1%7.9%1,089580
    GSSRepublicansModerate197439.1%34.2%44.2%419360
    GSSRepublicansModerate197537.0%32.2%42.1%430360
    GSSRepublicansModerate197632.9%28.1%38.1%404336
    GSSRepublicansModerate197733.0%28.5%37.9%449374
    GSSRepublicansModerate197832.4%27.2%38.0%469288
    GSSRepublicansModerate198032.2%27.3%37.6%448315
    GSSRepublicansModerate198236.8%31.6%42.2%487315
    GSSRepublicansModerate198331.2%24.8%38.4%258174
    GSSRepublicansModerate198435.9%30.8%41.3%520322
    GSSRepublicansModerate198537.4%32.8%42.3%597394
    GSSRepublicansModerate198633.3%29.0%37.9%518427
    GSSRepublicansModerate198734.6%30.1%39.4%541397
    GSSRepublicansModerate198832.7%28.6%37.2%551450
    GSSRepublicansModerate198936.3%32.1%40.6%612497
    GSSRepublicansModerate199031.0%26.9%35.4%572455
    GSSRepublicansModerate199137.0%32.9%41.3%624513
    GSSRepublicansModerate199334.1%30.1%38.4%622492
    GSSRepublicansModerate199428.6%25.7%31.7%1,101869
    GSSRepublicansModerate199632.6%29.6%35.8%1,040870
    GSSRepublicansModerate199829.7%26.6%33.1%937766
    GSSRepublicansModerate200030.4%27.2%33.8%924756
    GSSRepublicansModerate200229.3%24.8%34.3%465356
    GSSRepublicansModerate200431.6%27.1%36.4%504377
    GSSRepublicansModerate200628.7%26.0%31.5%1,4471,044
    GSSRepublicansModerate200829.1%25.2%33.3%657471
    GSSRepublicansModerate201027.2%23.3%31.5%651445
    GSSRepublicansModerate201226.2%21.9%31.0%585360
    GSSRepublicansModerate201428.6%25.1%32.4%779590
    GSSRepublicansModerate201629.0%25.7%32.6%911652
    GSSRepublicansModerate201831.3%27.1%35.8%777435
    GSSRepublicansModerate202124.7%21.9%27.7%1,228846
    GSSRepublicansModerate202225.4%21.4%29.7%1,106422
    GSSRepublicansModerate202423.4%20.1%27.0%1,089580
    GSSRepublicansConservative197441.9%36.9%47.0%419360
    GSSRepublicansConservative197544.5%39.5%49.7%430360
    GSSRepublicansConservative197648.4%43.1%53.7%404336
    GSSRepublicansConservative197745.8%40.8%50.9%449374
    GSSRepublicansConservative197847.4%41.7%53.2%469288
    GSSRepublicansConservative198047.1%41.6%52.6%448315
    GSSRepublicansConservative198249.1%43.7%54.6%487315
    GSSRepublicansConservative198354.3%46.9%61.5%258174
    GSSRepublicansConservative198451.6%46.1%57.0%520322
    GSSRepublicansConservative198547.9%43.0%52.9%597394
    GSSRepublicansConservative198652.3%47.5%57.0%518427
    GSSRepublicansConservative198748.3%43.4%53.2%541397
    GSSRepublicansConservative198851.3%46.7%55.9%551450
    GSSRepublicansConservative198946.9%42.6%51.3%612497
    GSSRepublicansConservative199050.0%45.5%54.6%572455
    GSSRepublicansConservative199144.9%40.7%49.3%624513
    GSSRepublicansConservative199352.5%48.1%56.9%622492
    GSSRepublicansConservative199456.6%53.3%59.9%1,101869
    GSSRepublicansConservative199656.1%52.8%59.4%1,040870
    GSSRepublicansConservative199857.3%53.7%60.7%937766
    GSSRepublicansConservative200056.6%53.0%60.1%924756
    GSSRepublicansConservative200260.0%54.9%65.0%465356
    GSSRepublicansConservative200458.2%53.1%63.0%504377
    GSSRepublicansConservative200662.3%59.3%65.2%1,4471,044
    GSSRepublicansConservative200864.9%60.5%69.1%657471
    GSSRepublicansConservative201062.6%58.0%66.9%651445
    GSSRepublicansConservative201265.6%60.6%70.3%585360
    GSSRepublicansConservative201464.0%60.0%67.8%779590
    GSSRepublicansConservative201664.2%60.5%67.8%911652
    GSSRepublicansConservative201860.1%55.4%64.5%777435
    GSSRepublicansConservative202171.6%68.5%74.5%1,228846
    GSSRepublicansConservative202270.4%65.9%74.6%1,106422
    GSSRepublicansConservative202470.9%67.0%74.4%1,089580
    CESDemocratsLiberal200647.1%46.2%48.0%15,56012,187
    CESDemocratsLiberal200751.3%49.2%53.4%4,2812,226
    CESDemocratsLiberal200850.9%49.9%52.0%13,7939,277
    CESDemocratsLiberal200949.7%48.2%51.3%5,5813,973
    CESDemocratsLiberal201048.9%47.8%49.9%23,3529,329
    CESDemocratsLiberal201136.1%34.7%37.6%8,8704,108
    CESDemocratsLiberal201246.7%45.7%47.7%23,7739,916
    CESDemocratsLiberal201347.9%46.3%49.6%7,1033,435
    CESDemocratsLiberal201444.8%43.9%45.8%24,36710,280
    CESDemocratsLiberal201547.8%46.1%49.5%6,2993,272
    CESDemocratsLiberal201645.1%44.3%45.9%29,78414,600
    CESDemocratsLiberal201749.3%47.7%50.9%8,1053,795
    CESDemocratsLiberal201859.6%58.7%60.5%26,26812,043
    CESDemocratsLiberal201962.4%60.9%63.8%7,9634,354
    CESDemocratsLiberal202058.6%57.7%59.4%28,30313,342
    CESDemocratsLiberal202160.4%59.2%61.6%11,9876,476
    CESDemocratsLiberal202260.2%59.3%61.0%28,10212,736
    CESDemocratsLiberal202356.6%55.3%57.9%10,3915,748
    CESDemocratsLiberal202459.9%58.9%60.9%28,5598,993
    CESDemocratsModerate200646.9%46.0%47.8%15,56012,187
    CESDemocratsModerate200741.5%39.4%43.5%4,2812,226
    CESDemocratsModerate200841.4%40.4%42.4%13,7939,277
    CESDemocratsModerate200942.2%40.7%43.8%5,5813,973
    CESDemocratsModerate201042.0%41.0%43.0%23,3529,329
    CESDemocratsModerate201151.7%50.2%53.2%8,8704,108
    CESDemocratsModerate201242.7%41.8%43.7%23,7739,916
    CESDemocratsModerate201340.0%38.4%41.7%7,1033,435
    CESDemocratsModerate201442.9%41.9%43.8%24,36710,280
    CESDemocratsModerate201541.0%39.3%42.7%6,2993,272
    CESDemocratsModerate201641.6%40.8%42.4%29,78414,600
    CESDemocratsModerate201740.0%38.5%41.6%8,1053,795
    CESDemocratsModerate201831.8%31.0%32.6%26,26812,043
    CESDemocratsModerate201930.6%29.3%32.0%7,9634,354
    CESDemocratsModerate202034.2%33.4%35.0%28,30313,342
    CESDemocratsModerate202133.0%31.9%34.2%11,9876,476
    CESDemocratsModerate202233.7%32.9%34.6%28,10212,736
    CESDemocratsModerate202336.0%34.7%37.2%10,3915,748
    CESDemocratsModerate202435.1%34.2%36.1%28,5598,993
    CESDemocratsConservative20066.0%5.6%6.4%15,56012,187
    CESDemocratsConservative20077.2%6.2%8.4%4,2812,226
    CESDemocratsConservative20087.7%7.2%8.3%13,7939,277
    CESDemocratsConservative20098.0%7.2%8.9%5,5813,973
    CESDemocratsConservative20109.2%8.6%9.8%23,3529,329
    CESDemocratsConservative201112.2%11.2%13.2%8,8704,108
    CESDemocratsConservative201210.6%10.0%11.2%23,7739,916
    CESDemocratsConservative201312.0%11.0%13.2%7,1033,435
    CESDemocratsConservative201412.3%11.7%12.9%24,36710,280
    CESDemocratsConservative201511.2%10.2%12.3%6,2993,272
    CESDemocratsConservative201613.2%12.7%13.8%29,78414,600
    CESDemocratsConservative201710.7%9.7%11.7%8,1053,795
    CESDemocratsConservative20188.6%8.1%9.1%26,26812,043
    CESDemocratsConservative20197.0%6.3%7.8%7,9634,354
    CESDemocratsConservative20207.2%6.8%7.7%28,30313,342
    CESDemocratsConservative20216.6%6.0%7.2%11,9876,476
    CESDemocratsConservative20226.1%5.7%6.5%28,10212,736
    CESDemocratsConservative20237.4%6.8%8.2%10,3915,748
    CESDemocratsConservative20245.0%4.5%5.4%28,5598,993
    CESRepublicansLiberal20062.4%2.1%2.7%14,66211,342
    CESRepublicansLiberal20072.0%1.5%2.7%3,7201,983
    CESRepublicansLiberal20082.8%2.5%3.1%13,56110,315
    CESRepublicansLiberal20092.2%1.8%2.7%5,2784,289
    CESRepublicansLiberal20102.3%2.0%2.7%23,0788,965
    CESRepublicansLiberal20111.9%1.5%2.4%7,1503,710
    CESRepublicansLiberal20122.2%1.9%2.5%20,0748,888
    CESRepublicansLiberal20132.5%2.0%3.1%5,1472,683
    CESRepublicansLiberal20142.2%1.9%2.6%18,5348,687
    CESRepublicansLiberal20152.3%1.8%3.0%4,7262,725
    CESRepublicansLiberal20162.5%2.2%2.8%20,29811,665
    CESRepublicansLiberal20171.9%1.4%2.4%5,6192,859
    CESRepublicansLiberal20182.9%2.6%3.3%20,97710,926
    CESRepublicansLiberal20193.8%3.2%4.4%6,3134,090
    CESRepublicansLiberal20203.3%3.0%3.6%19,76811,264
    CESRepublicansLiberal20212.7%2.3%3.2%7,8765,244
    CESRepublicansLiberal20222.7%2.4%3.0%19,6399,659
    CESRepublicansLiberal20233.4%3.0%4.0%8,5505,124
    CESRepublicansLiberal20242.3%2.0%2.7%20,7617,984
    CESRepublicansModerate200624.8%24.0%25.6%14,66211,342
    CESRepublicansModerate200722.5%20.7%24.4%3,7201,983
    CESRepublicansModerate200821.2%20.5%22.0%13,56110,315
    CESRepublicansModerate200921.2%20.0%22.5%5,2784,289
    CESRepublicansModerate201019.9%19.1%20.8%23,0788,965
    CESRepublicansModerate201122.6%21.3%24.0%7,1503,710
    CESRepublicansModerate201220.3%19.5%21.2%20,0748,888
    CESRepublicansModerate201322.1%20.6%23.7%5,1472,683
    CESRepublicansModerate201422.8%22.0%23.7%18,5348,687
    CESRepublicansModerate201523.9%22.3%25.5%4,7262,725
    CESRepublicansModerate201624.8%24.0%25.6%20,29811,665
    CESRepublicansModerate201722.0%20.5%23.6%5,6192,859
    CESRepublicansModerate201821.0%20.2%21.8%20,97710,926
    CESRepublicansModerate201920.2%19.0%21.5%6,3134,090
    CESRepublicansModerate202022.3%21.6%23.1%19,76811,264
    CESRepublicansModerate202122.1%21.0%23.3%7,8765,244
    CESRepublicansModerate202223.3%22.5%24.2%19,6399,659
    CESRepublicansModerate202325.4%24.2%26.6%8,5505,124
    CESRepublicansModerate202424.5%23.5%25.4%20,7617,984
    CESRepublicansConservative200672.8%72.0%73.7%14,66211,342
    CESRepublicansConservative200775.4%73.5%77.3%3,7201,983
    CESRepublicansConservative200876.0%75.1%76.8%13,56110,315
    CESRepublicansConservative200976.6%75.3%77.8%5,2784,289
    CESRepublicansConservative201077.7%76.8%78.6%23,0788,965
    CESRepublicansConservative201175.4%74.0%76.8%7,1503,710
    CESRepublicansConservative201277.5%76.6%78.3%20,0748,888
    CESRepublicansConservative201375.4%73.7%77.0%5,1472,683
    CESRepublicansConservative201474.9%74.0%75.8%18,5348,687
    CESRepublicansConservative201573.8%72.1%75.4%4,7262,725
    CESRepublicansConservative201672.7%71.9%73.5%20,29811,665
    CESRepublicansConservative201776.1%74.5%77.6%5,6192,859
    CESRepublicansConservative201876.1%75.3%76.9%20,97710,926
    CESRepublicansConservative201976.0%74.6%77.3%6,3134,090
    CESRepublicansConservative202074.4%73.6%75.2%19,76811,264
    CESRepublicansConservative202175.1%73.9%76.3%7,8765,244
    CESRepublicansConservative202274.0%73.1%74.9%19,6399,659
    CESRepublicansConservative202371.2%69.9%72.4%8,5505,124
    CESRepublicansConservative202473.2%72.2%74.2%20,7617,984
    In the ANES and GSS views, the liberal share of Democrats and the conservative share of Republicans are larger at the end than at the start; in the ANES the opposite camps are smaller. Each line is the share of one party's identifiers in one camp. Liberal and conservative are the three points on each side of the middle in the ANES and GSS scales and the two points on each side in the CES scale.Note: Respondents who did not place themselves are left out of every share, and the share who placed themselves is reported in the text.Source: ANES Cumulative Data File, 1972-2024 (weight VCF0009z); GSS, 1974-2024 (weight_use); CES cumulative file, 2006-2024 (weight); all estimates weighted

    These shares describe people who placed themselves. In the ANES the share of respondents with a party answer who placed themselves was 57% (95% CI 55% to 59%)57% percent of respondents with a party identificationEstimandweighted share who placed themselves on the liberal-conservative scale, Wilson intervalInterval95% CI 55% to 59%n2,695Weightsweight (VCF0009z)PopulationANES respondents with a party identification (other party and not sure excluded)Period1972SourceANES Cumulative Data File in 1972 and 84% (95% CI 82% to 85%)84% percent of respondents with a party identificationEstimandweighted share who placed themselves on the liberal-conservative scale, Wilson intervalInterval95% CI 82% to 85%n5,483Weightsweight (VCF0009z)PopulationANES respondents with a party identification (other party and not sure excluded)Period2024SourceANES Cumulative Data File in 2024, so the 1972 correlation describes a smaller share of respondents than the 2024 correlation.In 2000 only 34%34% percent of respondents with a party identificationEstimandweighted share who placed themselves on the liberal-conservative scale, Wilson intervalInterval95% CI 32% to 37%n1,792Weightsweight (VCF0009z)PopulationANES respondents with a party identification (other party and not sure excluded)Period2000SourceANES Cumulative Data File of ANES respondents with a party answer have a placement, and in 2002 only 48%48% percent of respondents with a party identificationEstimandweighted share who placed themselves on the liberal-conservative scale, Wilson intervalInterval95% CI 46% to 51%n2,681Weightsweight_usePopulationGSS respondents with a party identification (other party and not sure excluded)Period2002SourceGeneral Social Survey of GSS respondents do; the GSS pattern repeats in 1983 and 2004. Those years carry smaller samples, and the placement rates quoted here come from other years. The GSS placement rate was 96%96% percent of respondents with a party identificationEstimandweighted share who placed themselves on the liberal-conservative scale, Wilson intervalInterval95% CI 95% to 97%n1,402Weightsweight_usePopulationGSS respondents with a party identification (other party and not sure excluded)Period1974SourceGeneral Social Survey in 1974 and 96%96% percent of respondents with a party identificationEstimandweighted share who placed themselves on the liberal-conservative scale, Wilson intervalInterval95% CI 94% to 96%n3,185Weightsweight_usePopulationGSS respondents with a party identification (other party and not sure excluded)Period2024SourceGeneral Social Survey in 2024, and its correlation rose over that span from +0.17+0.17 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.12 to 0.23n1,345Weightsweight_usePopulationGSS respondents with a party identification (leaners kept, other party excluded) and a liberal-conservative self-placementPeriod1974SourceGeneral Social Survey to +0.63+0.63 correlation coefficient, minus 1 to plus 1Estimandweighted Pearson correlation between party identification and liberal-conservative self-placementInterval95% CI 0.60 to 0.66n3,062Weightsweight_usePopulationGSS respondents with a party identification (leaners kept, other party excluded) and a liberal-conservative self-placementPeriod2024SourceGeneral Social Survey, so a change in who places themselves cannot be the only source of the rise.

    Agreement on the ideology label is not agreement on issues. Chapter 1.5, Constraint among issue positions (planned) asks whether issue positions are more tightly bundled than they were in the 1980s, and Chapter 7.2, Economic and cultural dimensions asks whether economic and cultural conservatism form one dimension or two.

    What these data cannot show

    The data are repeated cross-sections, so they show that party and self-placement became more tightly linked and do not show which of the two changed. A panel study that followed the same people found movement in both directions: some changed party identification to match their issue positions, and others changed issue positions to match their party [6]. A study of issue attitudes over 1972 to 2004 found that attitudes came to line up with party more than they lined up with each other [7]. The rise in the correlation here is a rise in the link between two labels and is silent on which label moved first.

    Self-placement is a label. Its meaning for respondents draws partly on their feelings toward the groups and symbols tied to each label [8], and the public is more likely to call itself conservative than liberal while favoring more government action on many problems [9]. A rise in the correlation between party and label is not a rise in the correlation between party and any set of policy positions. The correlation also says nothing about extremity, and the distance between party averages in scale points is a separate measure.

    The three surveys give different levels, and these data cannot say which is closer to the level in the population. The GSS reading of the 2020s mixes web and in-person interviews that give different correlations, and the ANES mode weights are missing for 2020, so mode is identified in only some years. Every estimate describes respondents who placed themselves, and the share who did so differs between ANES years. The intervals are model based: they treat each weighted sample as a simple random sample and do not follow the sampling design. Where the design could be checked, in the GSS cells behind the mode comparison, the design-based standard errors were 1.131.13 ratio of standard errorsEstimandsmallest ratio of the jackknife standard error over primary sampling units within strata to the model-based standard error of the correlationWeightsweight_usePopulationGSS respondents with both answers in the 7 year-and-mode cells behind the interview-mode comparisonPeriod2018, 2022 and 2024SourceGeneral Social Survey to 1.681.68 ratio of standard errorsEstimandlargest ratio of the jackknife standard error over primary sampling units within strata to the model-based standard error of the correlationWeightsweight_usePopulationGSS respondents with both answers in the 7 year-and-mode cells behind the interview-mode comparisonPeriod2018, 2022 and 2024SourceGeneral Social Survey times the model-based ones; the ANES and CES intervals were not checked that way. The CES draws its respondents from an online panel of volunteers and weights them to match known population totals, so its intervals rest on that weighting.

    The data hold no measure of what party leaders did. They cannot say whether the electorate moved toward parties that had moved apart or the parties moved apart because the electorate had. Chapter 1.2, Polarization in Congress (planned) takes up how far apart the parties are in Congress.

    Data sources

    • ANES Cumulative Data FileANES Time Series Cumulative Data File, 1972-2024: seven-point party identification and seven-point liberal-conservative scale; interview mode from the mode weights (2008, 2012, 2016, 2024)Years: 1948-2024 (33 study years; even years only after 2004; no 2006, 2010 or 2014 study) · Weights: weight (VCF0009z); weight_ftf and weight_web for interview mode · License: Free download after accepting the ANES terms of use; cite the release.
    • General Social SurveyGeneral Social Survey, 1974-2024: partyid and polviews; interview mode joined from the raw file (from 2004)Years: 1972-2024 (35 study years) · Weights: weight_use · License: See the provider's terms of use.
    • Cooperative Election StudyCooperative Election Study cumulative file, 2006-2024: pid7 and ideo5Years: 2006-2024 (every year; self-reported turnout is available in the 10 even years) · Weights: weight (single wave); weight_cumulative (pooled waves) · License: See the provider's terms of use.

    All data sources

    Methods

    Sample: respondents who answered both the party-identification question and the liberal-conservative self-placement question: 43,947 ANES respondents in 22 survey years from 1972 to 2024, 64,452 GSS respondents in 33 survey years from 1974 to 2024 and 638,338 CES respondents in 19 survey years from 2006 to 2024. The three surveys share no respondents, and every estimate is computed inside one survey; surveys are compared, never pooled. Measures: party identification is the seven-point scale (ANES pid_7pt; GSS partyid 0 to 6 with the other-party code dropped; CES pid7 with not sure dropped), which gives leaners their own points. Ideology is the seven-point liberal-conservative scale in the ANES (lib_con_7pt) and the GSS (polviews) and the five-point scale in the CES (ideo5, not sure dropped). Every scale runs from Democratic or liberal to Republican or conservative by its codebook, so a positive correlation means that conservatives sit with Republicans; the sign is never chosen from a result. Respondents who do not place themselves on the ideology scale are missing on that item and are left out, so every estimate describes people who placed themselves. Estimators: the survey-weighted Pearson correlation of the party scale and the ideology scale within a survey and year; for pooled windows (1980 to 1988 and 2020 to 2024) the respondents of all survey years of one survey in the window are estimated together with its weight. The distance between the party centers is the weighted mean placement of Republican identifiers (leaners included) minus that of Democratic identifiers (leaners included), divided by the weighted standard deviation of placements among all placers in the survey over all years (first view) or in the same year (second view); pure independents are left out of the distance. Shares of ideological camps are weighted shares among identifiers who placed themselves. Weights: ANES weight (VCF0009z); interview-mode estimates use weight_ftf for face-to-face respondents and weight_web for web respondents, which exist for 2008, 2012, 2016 and 2024 only; GSS weight_use, with interview mode joined from the raw file (recorded from 2004); CES weight within a survey year and weight_cumulative for pooled windows. In-person view of Figure 1: it uses the recorded mode where the file has one (ANES face-to-face from 2008, GSS in person from 2004) and all respondents in the earlier years, which were interviewed mainly in person, except that it leaves out the ANES studies of 1998, 2000 and 2002, which used telephone interviews wholly or in part. It leaves out the CES, which is online only, and draws no line across a year with no in-person estimate. In the interview-mode figure, telephone and other GSS interviews are in the all-interviews line and in neither of the other two; in the ANES the 2024 face-to-face and web estimates leave out the panel and paper respondents. Intervals: 95% intervals from Fisher z on the Kish effective sample size for correlations, a normal interval on the effective sample size for means, and a Wilson interval for shares. A difference between two independent estimates combines their standard errors in quadrature, and the standard error of a correlation r is taken as (1 minus r squared) divided by the square root of the effective sample size minus 3. The mean of yearly differences treats years as independent samples. Differences in changes are formed only between disjoint groups. All intervals are model based: they treat the weighted sample as a simple random sample and do not follow the sampling design (stratification and clustering of respondents within sampled areas). For the seven GSS year-and-interview-mode cells behind the interview-mode comparison (2018, 2022 and 2024) the chapter also reports a design-based check: the stratified delete-one-primary-sampling-unit jackknife on the GSS variables vstrat and vpsu (two primary sampling units in every stratum of those years), with the same estimator and weights. Its standard errors are 1.13 to 1.68 times the model-based ones, and the web minus in-person differences and the 2018 to 2024 in-person difference use it. Design-based intervals were not computed for the ANES (its variance-estimation variables take only one or two values a year from 1972 to 1990) or the CES (an online panel of volunteers, weighted to match population totals). Suppression: a survey-year or group-year cell is drawn only with at least 150 respondents and a Kish effective sample of at least 100; a pooled group cell needs at least 100 respondents and an effective sample of at least 100 in each window. Groups: birth cohort is the ANES cohort code grouped into born before 1943, 1943 to 1974 and 1975 or later (GSS: year minus age); education groups grade school with high school (GSS: no bachelor's degree against bachelor's degree or more, because the GSS degree item cannot separate some college from a high school diploma); region is the census South against the other census regions. The GSS comparison of the two windows uses in-person interviews only in 2020 to 2024. Coverage: among ANES respondents with a party answer, 34 percent have an ideology placement in 2000, against 57 to 91 percent in the other years (1972 to 2024); among GSS respondents, 48, 48 and 46 percent have one in 1983, 2002 and 2004, against 93 percent or more in the other years. The share who placed themselves is reported only from years outside those exceptions. The ANES 2020 wave has no mode weights. Item order, wording and response options differ across the three surveys, and this chapter does not adjust for those differences.

    References

    1. Levendusky, Matthew (2009). The Partisan Sort: How Liberals Became Democrats and Conservatives Became Republicans. University of Chicago Press.
    2. Converse, Philip E. (1964). The Nature of Belief Systems in Mass Publics. In David E. Apter (ed.), Ideology and Discontent. New York: Free Press, pp. 206-261.
    3. Abramowitz, Alan I.; Saunders, Kyle L. (1998). Ideological Realignment in the U.S. Electorate. Journal of Politics 60(3): 634-652. doi:10.2307/2647642
    4. Fisher, Ronald A. (1921). On the Probable Error of a Coefficient of Correlation Deduced from a Small Sample. Metron 1(4): 3-32.
    5. Kish, Leslie (1965). Survey Sampling. John Wiley and Sons.
    6. Carsey, Thomas M.; Layman, Geoffrey C. (2006). Changing Sides or Changing Minds? Party Identification and Policy Preferences in the American Electorate. American Journal of Political Science 50(2): 464-477. doi:10.1111/j.1540-5907.2006.00196.x
    7. Baldassarri, Delia; Gelman, Andrew (2008). Partisans without Constraint: Political Polarization and Trends in American Public Opinion. American Journal of Sociology 114(2): 408-446. doi:10.1086/590649
    8. Conover, Pamela Johnston; Feldman, Stanley (1981). The Origins and Meaning of Liberal/Conservative Self-Identifications. American Journal of Political Science 25(4): 617-645. www.jstor.org/stable/2110756
    9. Ellis, Christopher; Stimson, James A. (2012). Ideology in America. Cambridge University Press.

    How to cite

    @misc{singh2026partyandideology,
  author = {Singh, Vishal},
  title = {Party and ideology},
  year = {2026},
  url = {https://vishalsingh.org/politics/party-and-ideology},
  note = {In The Public Mind. Published 30 September 2026.}
}

    Reading paths