Skip to content
Contents
The Public Mind

    Type to search. Planned chapters match by title only.

    Chapter 2.1

    Cohort change in partisan sorting

    Among ANES partisans who place themselves on the ideology scale, conservative Democrats and liberal Republicans fell from 19.6% in 1972 to 6.3% in 2024.

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

    This draft describes birth cohorts and cannot follow individuals. It does not yet use a panel study that re-interviews the same people.

    Conservative Democrats and liberal Republicans together made up 19.6%19.6% percent of placed partisansEstimandweighted share who are conservative Democrats or liberal Republicans (95% Wilson interval)Interval95% CI 17.6% to 21.8%n1,371Weightsweight (VCF0009z)Populationplaced partisans (Democrats and Republicans, leaners included, who place themselves on the seven-point ideology scale)Period1972SourceANES Cumulative Data File of the partisans who placed themselves on the seven-point liberal-conservative scale of the American National Election Studies (ANES) in 1972 and 6.3%6.3% percent of placed partisansEstimandweighted share who are conservative Democrats or liberal Republicans (95% Wilson interval)Interval95% CI 5.4% to 7.4%n4,482Weightsweight (VCF0009z)Populationplaced partisans (Democrats and Republicans, leaners included, who place themselves on the seven-point ideology scale)Period2024SourceANES Cumulative Data File in 2024. This chapter calls them cross-pressured partisans, people whose party and ideological label point in different directions.The word has another meaning in Planned, Partisans in the other party's territory (planned), which is about partisans who live where their party loses. Chapter 1.1, Party and ideology measures the same link as a correlation, and Chapter 4.7, Sort dates by attitude dates the sort attitude by attitude. The national shares show the group shrinking and leave open who changed. The ANES draws a new sample at each election, so it cannot follow any one conservative Democrat and cannot show that any one of them changed party or label. It can follow people born in the same years, a birth cohortCohortPeople 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., from one decade to the next, and ask whether the share fell inside each cohort or only as cohorts came and went.

    The fall came on both sides. Among placed Democrats the conservative share went from 24.6%24.6% percent of placed DemocratsEstimandweighted share of placed Democrats who are conservative (95% Wilson interval)Interval95% CI 21.7% to 27.8%n776Weightsweight (VCF0009z)Populationplaced Democrats (leaners included)Period1972SourceANES Cumulative Data File in 1972 to 8.2%8.2% percent of placed DemocratsEstimandweighted share of placed Democrats who are conservative (95% Wilson interval)Interval95% CI 6.7% to 10.0%n2,310Weightsweight (VCF0009z)Populationplaced Democrats (leaners included)Period2024SourceANES Cumulative Data File in 2024, and among placed Republicans the liberal share went from 13.1%13.1% percent of placed RepublicansEstimandweighted share of placed Republicans who are liberal (95% Wilson interval)Interval95% CI 10.6% to 16.1%n595Weightsweight (VCF0009z)Populationplaced Republicans (leaners included)Period1972SourceANES Cumulative Data File to 4.4%4.4% percent of placed RepublicansEstimandweighted share of placed Republicans who are liberal (95% Wilson interval)Interval95% CI 3.3% to 5.8%n2,172Weightsweight (VCF0009z)Populationplaced Republicans (leaners included)Period2024SourceANES Cumulative Data File. Democrats and Republicans include independents who say they lean toward a party, the 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., throughout.

    Figure 1

    The conservative share of placed Democrats fell from 24.6% to 8.2%, and the liberal share of placed Republicans from 13.1% to 4.4%

    Percent of placed partisans who are conservative Democrats or liberal Republicans, ANES 1972-2024, with 95% intervals

    Cross-pressured share within party (%)

    Show data
    Data for Cross-pressured partisans fell from 19.6% of placed partisans in 1972 to 6.3% in 2024
    MeasureYearShare (%)Interval lowInterval highRespondents in the base
    Cross-pressured share of placed partisans197219.617.621.81,371
    Cross-pressured share of placed partisans197418.415.721.3977
    Cross-pressured share of placed partisans197617.815.820.01,306
    Cross-pressured share of placed partisans197818.016.120.01,430
    Cross-pressured share of placed partisans198020.317.823.1892
    Cross-pressured share of placed partisans198218.115.620.9806
    Cross-pressured share of placed partisans198417.015.119.01,401
    Cross-pressured share of placed partisans198618.516.520.51,441
    Cross-pressured share of placed partisans198818.616.620.81,286
    Cross-pressured share of placed partisans199019.717.522.01,220
    Cross-pressured share of placed partisans199216.915.218.81,630
    Cross-pressured share of placed partisans199413.611.715.81,279
    Cross-pressured share of placed partisans199614.512.516.91,233
    Cross-pressured share of placed partisans199815.613.218.3922
    Cross-pressured share of placed partisans200016.013.019.5603
    Cross-pressured share of placed partisans200214.912.617.51,171
    Cross-pressured share of placed partisans200411.59.414.0849
    Cross-pressured share of placed partisans200811.39.413.41,448
    Cross-pressured share of placed partisans20129.88.711.04,613
    Cross-pressured share of placed partisans20167.26.28.42,949
    Cross-pressured share of placed partisans20205.24.55.96,353
    Cross-pressured share of placed partisans20246.35.47.44,482
    Conservative share of placed Democrats197224.621.727.8776
    Conservative share of placed Democrats197423.019.227.1573
    Conservative share of placed Democrats197624.121.027.4725
    Conservative share of placed Democrats197823.120.426.0867
    Conservative share of placed Democrats198028.024.332.0515
    Conservative share of placed Democrats198225.521.829.6479
    Conservative share of placed Democrats198422.319.425.4719
    Conservative share of placed Democrats198624.621.727.7793
    Conservative share of placed Democrats198827.424.131.0638
    Conservative share of placed Democrats199025.922.729.3673
    Conservative share of placed Democrats199223.420.726.3871
    Conservative share of placed Democrats199422.619.226.5622
    Conservative share of placed Democrats199621.818.525.4667
    Conservative share of placed Democrats199821.517.925.6515
    Conservative share of placed Democrats200022.317.627.7319
    Conservative share of placed Democrats200221.117.325.6568
    Conservative share of placed Democrats200416.112.720.2438
    Conservative share of placed Democrats200816.613.719.9894
    Conservative share of placed Democrats201214.512.716.42,709
    Conservative share of placed Democrats201610.38.612.31,467
    Conservative share of placed Democrats20206.65.57.83,277
    Conservative share of placed Democrats20248.26.710.02,310
    Liberal share of placed Republicans197213.110.616.1595
    Liberal share of placed Republicans197411.48.315.4404
    Liberal share of placed Republicans19769.87.612.5581
    Liberal share of placed Republicans197810.17.912.9563
    Liberal share of placed Republicans19809.87.213.2377
    Liberal share of placed Republicans19827.35.010.7327
    Liberal share of placed Republicans198411.49.314.0682
    Liberal share of placed Republicans198611.08.813.6648
    Liberal share of placed Republicans19889.97.812.4648
    Liberal share of placed Republicans199012.19.615.1547
    Liberal share of placed Republicans19929.57.611.8759
    Liberal share of placed Republicans19945.13.57.3657
    Liberal share of placed Republicans19966.14.38.7566
    Liberal share of placed Republicans19988.35.911.7407
    Liberal share of placed Republicans20008.75.813.0284
    Liberal share of placed Republicans20028.46.111.4603
    Liberal share of placed Republicans20046.64.49.7411
    Liberal share of placed Republicans20085.23.57.8554
    Liberal share of placed Republicans20124.63.55.91,904
    Liberal share of placed Republicans20164.03.05.41,482
    Liberal share of placed Republicans20203.73.04.73,076
    Liberal share of placed Republicans20244.43.35.82,172
    Fully sorted share of placed partisans197243.641.046.31,371
    Fully sorted share of placed partisans197446.042.449.6977
    Fully sorted share of placed partisans197646.944.249.71,306
    Fully sorted share of placed partisans197846.744.149.31,430
    Fully sorted share of placed partisans198050.246.953.5892
    Fully sorted share of placed partisans198247.544.151.0806
    Fully sorted share of placed partisans198451.649.054.21,401
    Fully sorted share of placed partisans198646.143.548.71,441
    Fully sorted share of placed partisans198851.849.154.51,286
    Fully sorted share of placed partisans199044.842.147.61,220
    Fully sorted share of placed partisans199253.250.755.61,630
    Fully sorted share of placed partisans199453.250.256.21,279
    Fully sorted share of placed partisans199655.051.958.01,233
    Fully sorted share of placed partisans199850.246.753.7922
    Fully sorted share of placed partisans200054.249.758.5603
    Fully sorted share of placed partisans200257.754.361.11,171
    Fully sorted share of placed partisans200457.053.360.6849
    Fully sorted share of placed partisans200863.160.066.11,448
    Fully sorted share of placed partisans201260.658.762.54,613
    Fully sorted share of placed partisans201668.866.870.72,949
    Fully sorted share of placed partisans202071.770.273.16,353
    Fully sorted share of placed partisans202469.367.371.24,482
    Moderate share of placed partisans197236.834.239.31,371
    Moderate share of placed partisans197435.632.239.2977
    Moderate share of placed partisans197635.332.738.01,306
    Moderate share of placed partisans197835.332.937.81,430
    Moderate share of placed partisans198029.526.632.6892
    Moderate share of placed partisans198234.431.237.7806
    Moderate share of placed partisans198431.429.033.91,401
    Moderate share of placed partisans198635.533.038.01,441
    Moderate share of placed partisans198829.627.232.21,286
    Moderate share of placed partisans199035.532.938.21,220
    Moderate share of placed partisans199229.927.832.21,630
    Moderate share of placed partisans199433.130.436.01,279
    Moderate share of placed partisans199630.527.733.41,233
    Moderate share of placed partisans199834.231.037.6922
    Moderate share of placed partisans200029.926.034.1603
    Moderate share of placed partisans200227.424.430.61,171
    Moderate share of placed partisans200431.528.235.0849
    Moderate share of placed partisans200825.622.928.51,448
    Moderate share of placed partisans201229.627.931.44,613
    Moderate share of placed partisans201624.022.225.92,949
    Moderate share of placed partisans202023.221.824.66,353
    Moderate share of placed partisans202424.422.626.24,482
    Cross-pressured share of placed partisans197219.617.621.81,371
    Cross-pressured share of placed partisans197418.415.721.3977
    Cross-pressured share of placed partisans197617.815.820.01,306
    Cross-pressured share of placed partisans197818.016.120.01,430
    Cross-pressured share of placed partisans198020.317.823.1892
    Cross-pressured share of placed partisans198218.115.620.9806
    Cross-pressured share of placed partisans198417.015.119.01,401
    Cross-pressured share of placed partisans198618.516.520.51,441
    Cross-pressured share of placed partisans198818.616.620.81,286
    Cross-pressured share of placed partisans199019.717.522.01,220
    Cross-pressured share of placed partisans199216.915.218.81,630
    Cross-pressured share of placed partisans199413.611.715.81,279
    Cross-pressured share of placed partisans199614.512.516.91,233
    Cross-pressured share of placed partisans199815.613.218.3922
    Cross-pressured share of placed partisans200016.013.019.5603
    Cross-pressured share of placed partisans200214.912.617.51,171
    Cross-pressured share of placed partisans200411.59.414.0849
    Cross-pressured share of placed partisans200811.39.413.41,448
    Cross-pressured share of placed partisans20129.88.711.04,613
    Cross-pressured share of placed partisans20167.26.28.42,949
    Cross-pressured share of placed partisans20205.24.55.96,353
    Cross-pressured share of placed partisans20246.35.47.44,482

    Both lines fall. The upper line is the conservative share of placed Democrats and the lower line the liberal share of placed Republicans, so the two do not add up to a single share. Together the two groups were 19.6%19.6% percent of placed partisansEstimandweighted share who are conservative Democrats or liberal Republicans (95% Wilson interval)Interval95% CI 17.6% to 21.8%n1,371Weightsweight (VCF0009z)Populationplaced partisans (Democrats and Republicans, leaners included, who place themselves on the seven-point ideology scale)Period1972SourceANES Cumulative Data File of placed partisans in 1972 and 6.3%6.3% percent of placed partisansEstimandweighted share who are conservative Democrats or liberal Republicans (95% Wilson interval)Interval95% CI 5.4% to 7.4%n4,482Weightsweight (VCF0009z)Populationplaced partisans (Democrats and Republicans, leaners included, who place themselves on the seven-point ideology scale)Period2024SourceANES Cumulative Data File in 2024, the line in the first tab. The rule at 2012 marks the year the ANES added an online sample, and the bands are 95% intervals. Values in Table 1.

    Note: There was no ANES study in 2006, 2010 or 2014, so the line joins the neighboring study years.Note: The 2000 point rests on 603 placed partisans, the fewest in any year. The seven-point item went to a subsample of respondents that year: 52.1% of respondents have no recorded answer, and 35.6% of partisans placed themselves.Note: The ANES adds an online sample in 2012, and 2020 has no separate weights by interview mode. In 2024 the share is 7.2% in the face-to-face sample, 4.6% in the web sample and 7.1% among the 2,049 panel and paper cases (45.7% of the placed partisans in 2024), which have no separate weights by interview mode.Source: ANES Cumulative Data File, 1972-2024; weighted (VCF0009z); placed partisans, leaners included

    The fall appears inside the cohorts born 1927 to 1974

    Two accounts of the fall differ in where it happens. Under replacement, each cohort keeps roughly the rate it had when it entered the electorate. The national share then falls as the older cohorts that held mixed positions leave the sample and younger cohorts with lower rates arrive. Under within-cohort change, the rate falls inside each cohort as the years pass.

    Figure 2 gives one panel per birth cohort, with the ANES study years pooled by decade so that each cell holds enough respondents. Among placed partisans born 1943 to 1958, the cross-pressured share was 17.6%17.6% percent of placed partisansEstimandweighted share cross-pressured within the birth cohort and pooled period (95% Wilson interval)Interval95% CI 15.7% to 19.6%n1,702Weightsweight (VCF0009z)Populationplaced partisans born 1943 to 1958Period1972-1978 pooledSourceANES Cumulative Data File in the 1970s and 5.0%5.0% percent of placed partisansEstimandweighted share cross-pressured within the birth cohort and pooled period (95% Wilson interval)Interval95% CI 4.0% to 6.1%n2,910Weightsweight (VCF0009z)Populationplaced partisans born 1943 to 1958Period2020-2024 pooledSourceANES Cumulative Data File in the 2020s. The cohorts born 1927 to 1942 and 1959 to 1974 fell as well, and in all three the fall comes mostly after 1990. The cohort born 1991 or later also ends lower than it starts, between the 2010s and the 2020s. The two oldest cohorts are observed only through the 1980s and the 2000s, and their changes have intervals that include zero, as does the change for the cohort born 1975 to 1990.

    Figure 2

    Every cohort born 1927 to 1974 fell, mostly after 1990

    Cross-pressured share of placed partisans by birth cohort (solid) and for all placed partisans (dashed), ANES survey years pooled by decade, 1972-2024, with 95% intervals

    Share of placed partisans (%)

    • This cohort
    • All placed partisans
    Born 1895-1910Born 1911-1926Born 1927-1942Born 1943-1958Born 1959-1974Born 1975-1990Born 1991 or later

    Survey years pooled by decade (plotted at the midpoint)

    Show data
    Data for Every cohort born 1927 to 1974 fell, mostly after 1990
    PanelSeriesSurvey years pooledShare (%)Interval lowInterval highPlaced partisans
    Born 1895-1910This cohort1972-197819.716.623.3644
    Born 1895-1910This cohort1980-198816.512.820.9328
    Born 1895-1910All placed partisans1972-197818.417.319.65,084
    Born 1895-1910All placed partisans1980-198818.417.419.45,826
    Born 1895-1910All placed partisans1990-199816.215.217.16,284
    Born 1895-1910All placed partisans2000-200812.911.814.24,071
    Born 1895-1910All placed partisans2012-20168.88.09.77,562
    Born 1895-1910All placed partisans2020-20245.75.16.310,835
    Born 1911-1926This cohort1972-197818.115.920.51,262
    Born 1911-1926This cohort1980-198818.516.321.01,053
    Born 1911-1926This cohort1990-199819.817.122.7835
    Born 1911-1926This cohort2000-200816.611.523.4201
    Born 1911-1926All placed partisans1972-197818.417.319.65,084
    Born 1911-1926All placed partisans1980-198818.417.419.45,826
    Born 1911-1926All placed partisans1990-199816.215.217.16,284
    Born 1911-1926All placed partisans2000-200812.911.814.24,071
    Born 1911-1926All placed partisans2012-20168.88.09.77,562
    Born 1911-1926All placed partisans2020-20245.75.16.310,835
    Born 1927-1942This cohort1972-197818.716.521.01,324
    Born 1927-1942This cohort1980-198818.316.220.51,276
    Born 1927-1942This cohort1990-199816.614.519.01,174
    Born 1927-1942This cohort2000-200812.29.615.5678
    Born 1927-1942This cohort2012-20167.95.810.7797
    Born 1927-1942This cohort2020-202410.07.313.7625
    Born 1927-1942All placed partisans1972-197818.417.319.65,084
    Born 1927-1942All placed partisans1980-198818.417.419.45,826
    Born 1927-1942All placed partisans1990-199816.215.217.16,284
    Born 1927-1942All placed partisans2000-200812.911.814.24,071
    Born 1927-1942All placed partisans2012-20168.88.09.77,562
    Born 1927-1942All placed partisans2020-20245.75.16.310,835
    Born 1943-1958This cohort1972-197817.615.719.61,702
    Born 1943-1958This cohort1980-198817.516.019.12,279
    Born 1943-1958This cohort1990-199815.113.516.72,120
    Born 1943-1958This cohort2000-200813.511.415.81,263
    Born 1943-1958This cohort2012-20168.97.510.52,392
    Born 1943-1958This cohort2020-20245.04.06.12,910
    Born 1943-1958All placed partisans1972-197818.417.319.65,084
    Born 1943-1958All placed partisans1980-198818.417.419.45,826
    Born 1943-1958All placed partisans1990-199816.215.217.16,284
    Born 1943-1958All placed partisans2000-200812.911.814.24,071
    Born 1943-1958All placed partisans2012-20168.88.09.77,562
    Born 1943-1958All placed partisans2020-20245.75.16.310,835
    Born 1959-1974This cohort1980-198821.418.824.3863
    Born 1959-1974This cohort1990-199816.514.818.31,881
    Born 1959-1974This cohort2000-200812.710.715.01,248
    Born 1959-1974This cohort2012-20169.47.911.12,037
    Born 1959-1974This cohort2020-20246.55.47.92,785
    Born 1959-1974All placed partisans1972-197818.417.319.65,084
    Born 1959-1974All placed partisans1980-198818.417.419.45,826
    Born 1959-1974All placed partisans1990-199816.215.217.16,284
    Born 1959-1974All placed partisans2000-200812.911.814.24,071
    Born 1959-1974All placed partisans2012-20168.88.09.77,562
    Born 1959-1974All placed partisans2020-20245.75.16.310,835
    Born 1975-1990This cohort1990-199810.46.217.1153
    Born 1975-1990This cohort2000-200812.39.615.8641
    Born 1975-1990This cohort2012-20168.46.910.31,758
    Born 1975-1990This cohort2020-20245.54.56.82,789
    Born 1975-1990All placed partisans1972-197818.417.319.65,084
    Born 1975-1990All placed partisans1980-198818.417.419.45,826
    Born 1975-1990All placed partisans1990-199816.215.217.16,284
    Born 1975-1990All placed partisans2000-200812.911.814.24,071
    Born 1975-1990All placed partisans2012-20168.88.09.77,562
    Born 1975-1990All placed partisans2020-20245.75.16.310,835
    Born 1991 or laterThis cohort2012-20168.96.013.0407
    Born 1991 or laterThis cohort2020-20244.02.85.71,363
    Born 1991 or laterAll placed partisans1972-197818.417.319.65,084
    Born 1991 or laterAll placed partisans1980-198818.417.419.45,826
    Born 1991 or laterAll placed partisans1990-199816.215.217.16,284
    Born 1991 or laterAll placed partisans2000-200812.911.814.24,071
    Born 1991 or laterAll placed partisans2012-20168.88.09.77,562
    Born 1991 or laterAll placed partisans2020-20245.75.16.310,835

    In each panel the solid line is one birth cohort and the grey dashed line is all placed partisans. Look first at the panel for people born 1943 to 1958, where the solid line falls mostly after 1990. The panels for 1927 to 1942 and 1959 to 1974 fall as well, mostly after 1990. The two panels for the cohorts born before 1927 are short, and their changes have intervals that include zero. The bands are 95% intervals, and a line starts and ends where the cohort has enough respondents. Values in Table 2.

    Note: A cell is plotted only when it has 100 or more placed partisans and an effective sample (the sample size after weighting) of 100 or more, the floor of this chapter, so a line starts and ends where the cohort is well observed. The two oldest cohorts are plotted for only two and four decades.Note: The pooled 2020s combine 2020 and 2024, and 553 placed partisans with no birth cohort in the file are excluded from the cohort lines. The dashed line uses all placed partisans, including those with no birth cohort.Note: Two cells below the floor are not plotted: the 1895-1910 cohort in the 1990s (13.1%, 105 respondents) and the 1911-1926 cohort in the 2010s (12.8%, 60 respondents).Note: The 1927-1942 cohort is at 7.9% in the 2010s and 10.0% in the 2020s, a difference of +2.1 points that is within sampling error.Source: ANES Cumulative Data File, 1972-2024; weighted (VCF0009z); placed partisans, leaners included

    A new sample at every election

    The ANES interviews a fresh sample of Americans in each study year and asks each respondent for a party and for a place on the scale. The chapter uses every study year from 1972 to 2024 that carries the seven-point item, which gives 39,66239,662 respondentsEstimandcount of placed partisans over all study yearsWeightsnone (count or constant)Populationplaced partisans (Democrats and Republicans, leaners included, who place themselves on the seven-point ideology scale)Period1972-2024SourceANES Cumulative Data File placed partisans. Every share is weightedSurvey weightsNumbers that scale each respondent so the sample matches the population on known traits such as age, race, education and region. A weighted average counts a respondent with a large weight for more than one with a small weight. so that each year's sample matches the population on known traits. The qualification that changes the meaning is the design itself. A repeated cross-section follows a cohort's share and not its members, so a fall inside a cohort shows that the cohort's share fell and does not show that the same people changed their minds. The intervals treat each weighted sample as a simple random sample and ignore the clustering and stratification of the ANES design, which can raise or lower the variance. The report gives every estimate with its interval and sample size.

    What freezing each cohort's rate would have predicted

    To give the replacement account a number, apply fixed cohort rates to the mix of birth cohorts actually present among placed partisans in each decade. The mix changed over the decades, as the start and end of each line in Figure 2 show. The first version freezes every cohort at its 1970s rate and gives the cohorts without enough respondents in the 1970s the average 1970s rate. Under it the predicted share stays near its 1970s start, and the mix alone changes the share by +0.1 (95% CI −0.2 to 0.3)+0.1 percentage pointsEstimandchange in the share from the cohort mix alone, cohorts frozen at 1970s rates; bootstrap 95% percentile interval (1,000 replicates over respondents within pooled periods, seed 12345)Interval95% CI −0.2 to 0.3n10,489Weightsweight (VCF0009z)Populationplaced partisans with a known birth cohortPeriod1972-1978 to 2020-2024SourceANES Cumulative Data File pointsPercentage pointThe unit for the difference between two percentages. A change from one share to another is stated in points, not in percent, so that absolute and relative change are not confused., an interval that includes zero. That result is close to built in, because the assumption gives later cohorts the rates of the older ones.

    The second version freezes every cohort at its entry rate, the rate it had in the first decade in which it has enough respondents, so each later cohort keeps the rate it arrived with. Under it the mix accounts for 25%25.2% percent of the fallEstimandcohort-mix change with cohorts frozen at entry rates as a share of the observed fall; bootstrap 95% percentile interval (1,000 replicates over respondents within pooled periods, seed 12345)Interval95% CI 10.8% to 38.0%n10,489Weightsweight (VCF0009z)Populationplaced partisans with a known birth cohortPeriod1972-1978 to 2020-2024SourceANES Cumulative Data File of the fall from the 1970s to the 2020s, and the rest of the fall occurs inside cohorts. This version credits replacement with everything that distinguishes a cohort when it arrives, including the part that comes from the years in which it arrived. The chapter calls it the more generous of the two and not an upper bound.

    Figure 3

    Freezing cohort rates predicts a 2020s cross-pressured share of 15.1% to 18.3%, against 5.6% observed

    Cross-pressured share of placed partisans, observed and under two frozen-rate counterfactuals, ANES survey years pooled by decade, 1972-2024, with 95% bootstrap intervals

    Show data
    Data for Freezing cohort rates predicts a 2020s cross-pressured share of 15.1% to 18.3%, against 5.6% observed
    SeriesSurvey years pooledShare (%)Interval lowInterval highPlaced partisans
    Observed1972-197818.517.319.65,059
    Observed1980-198818.417.419.35,808
    Observed1990-199816.215.217.16,268
    Observed2000-200812.911.814.24,034
    Observed2012-20168.88.09.77,451
    Observed2020-20245.65.06.310,489
    1970s rates1972-197818.317.119.55,059
    1970s rates1980-198818.216.919.45,808
    1970s rates1990-199818.217.019.46,268
    1970s rates2000-200818.317.019.54,034
    1970s rates2012-201618.317.019.47,451
    1970s rates2020-202418.317.119.510,489
    Entry rates1972-197818.317.119.55,059
    Entry rates1980-198818.617.519.75,808
    Entry rates1990-199818.817.620.06,268
    Entry rates2000-200817.516.119.14,034
    Entry rates2012-201616.214.617.97,451
    Entry rates2020-202415.113.516.710,489

    The solid line is the observed share, level through the 1980s and then falling. The dashed line frozen at 1970s rates stays near its start. The dashed line frozen at entry rates ends at 15.1%15.1% percent of placed partisansEstimandcross-pressured share with every cohort held at its entry rate, bootstrap 95% percentile interval (1,000 replicates over respondents within pooled periods, seed 12345)Interval95% CI 13.5% to 16.7%n10,489Weightsweight (VCF0009z)Populationplaced partisans with a known birth cohortPeriod2020-2024 pooledSourceANES Cumulative Data File in the 2020s, above the observed line. Bands are 95% intervals from a bootstrap over respondents, drawn for the observed and entry-rate lines. Values in Table 4.

    Note: Placed partisans with no birth cohort in the file are left out of all three series so that the accounting is exact.Note: Cohorts not observed with 100 or more placed partisans in the 1970s take the mean 1970s cohort rate of 18.5% in the first counterfactual.Note: The within-cohort part of the fall is the observed fall minus the mix part, so it includes a baseline residual: the frozen rates applied to the 1970s mix predict a 1970s share 0.2 points (1970s rates) and 0.2 points (entry rates) below the observed one.Note: The bootstrap re-draws respondents with replacement within each pooled decade many times and re-estimates every quantity each time; the intervals are the 2.5th and 97.5th percentiles.Note: Bands are drawn for the observed and entry-rate lines; the interval of the 1970s-rate line is in the data table. Differences are computed from unrounded values, so a printed difference can differ in the last digit from the difference of the printed shares.Source: ANES Cumulative Data File, 1972-2024; weighted (VCF0009z); placed partisans, leaners included; intervals from a bootstrap over respondents within pooled decades (1,000 replicates, seed 12345)

    The answer depends on the rate given to a cohort before it is observed. Table 5 gives the mix effect under five assumptions. Two are the frozen-rate versions, two vary the rule for a well-observed cell, and the last is a limiting case that holds the cohorts born after 1958 at their own 2020s rates. The limiting case counts the fall inside those cohorts as a fixed cohort trait, so it is no longer a replacement counterfactual. The share of the fall credited to the mix runs from the first version, within sampling error of zero, to 74%74.0% percent of the fallEstimandcohort-mix change as a share of the observed fall, with the cohorts born after 1958 held at their own 2020s rates instead of their entry rates; bootstrap 95% percentile interval (1,000 replicates over respondents within pooled periods, seed 12345)Interval95% CI 70.6% to 77.8%n10,489Weightsweight (VCF0009z)Populationplaced partisans with a known birth cohortPeriod1972-1978 to 2020-2024SourceANES Cumulative Data File in the limiting case, and the data do not choose among the assumptions. Every version is an accounting statement that holds cohort rates fixed by assumption and does not say why the rates fell.

    A cohort line mixes change with loss and entry

    A cohort line follows a group of people born in the same years, and the group changes as the years pass. It mixes individuals who changed party or ideology with three other things. Some members die or stop answering surveys, the youngest members reach adulthood and join the sample, and the share who are partisans and who place themselves on the scale changes. A simple bound shows what loss alone would require. In the cohort born 1943 to 1958 the odds of being cross-pressured in the 2020s are 0.240.24 ratio of survival oddsEstimandodds ratio of the cross-pressured share (2020s over 1970s), read as the survival of the cross-pressured relative to other partisans if the fall were all differential lossInterval95% CI 0.19 to 0.32n4,612Weightsweight (VCF0009z)Populationplaced partisans born 1943 to 1958Period1972-1978 versus 2020-2024SourceANES Cumulative Data File times the odds in the 1970s. If loss alone produced that fall, the cross-pressured would have had to survive, and stay in the survey samples, at only that fraction of the rate of other partisans over the whole period. The bound covers one channel and estimates nobody's mortality.

    Studies that re-interview the same people find movement in both directions. In a panel of the 1992 to 1996 elections, people who saw party differences on an issue they cared about changed party to fit their views. People who saw differences on an issue they did not care about changed their views to fit their party [1]. A panel that followed one cohort from 1973 to 1997 found that partisanship shaped issue orientations early and that issue orientations shaped partisanship later [2]. Either move takes a conservative Democrat out of the cross-pressured group, and the ANES share counts the two the same way. What remains is the pattern of 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., liberals becoming Democrats and conservatives becoming Republicans, that Levendusky describes [3]. These data hold no measure of what party leaders did, so the chapter tests none of his account of why.

    What repeated cross-sections cannot tell you

    The data are repeated cross-sections with no panel, so the chapter cannot show that any person switched and cannot say which identity moved. Age, period and cohort are not separated, because a cohort ages a decade at each step and the three cannot be told apart without further assumptions [4]. Chapter 3.7, Age and generation takes up what repeated surveys can and cannot separate. The shares cover only partisans who place themselves on the scale, and many Americans do not describe their views with ideological labels [5]; with the unplaced in the base the fall is smaller. The intervals ignore the clustering of the sample, and the report says which one verdict a larger variance would change.

    Go deeper

    • The report: every estimate with its interval and sample size.
    • Tables: every figure of the report as a table.

    Same question at another grain:

    Reading paths that include this chapter:

    Data sources

    • ANES Cumulative Data FileCumulative Data File, study years 1972-2024 that carry the seven-point ideology itemYears: 1948-2024 (33 study years; even years only after 2004; no 2006, 2010 or 2014 study) · Weights: weight (VCF0009z) · License: Free download after accepting the ANES terms of use; cite the release.
    • ANES 2024 Time Series2024 Time Series, Census region of the sample location (V243007) for the 2024 panel and paper casesYears: 2024 (pre-election and post-election interviews) · Weights: none (a respondent attribute) · License: Free download after accepting the ANES terms of use; cite the release.
    • Cooperative Election StudyCumulative file, even-year waves 2006-2024, party identification and five-point ideology, for the second-survey comparisonYears: 2006-2024 (every year; self-reported turnout is available in the 10 even years) · Weights: weight of each wave (the cumulative weight gives the same within-year shares) · License: See the provider's terms of use.

    All data sources

    References

    1. 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
    2. Highton, Benjamin; Kam, Cindy D. (2011). The Long-Term Dynamics of Partisanship and Issue Orientations. Journal of Politics 73(1): 202-215.
    3. Levendusky, Matthew (2009). The Partisan Sort: How Liberals Became Democrats and Conservatives Became Republicans. University of Chicago Press.
    4. Glenn, Norval D. (1976). Cohort Analysts' Futile Quest: Statistical Attempts to Separate Age, Period and Cohort Effects. American Sociological Review 41(5): 900-904. doi:10.2307/2094738
    5. Kinder, Donald R.; Kalmoe, Nathan P. (2017). Neither Liberal nor Conservative: Ideological Innocence in the American Public. University of Chicago Press. press.uchicago.edu/ucp/books/book/chicago/N/bo25841664.html

    How to cite

    @misc{singh2026howthesorthappened,
  author = {Singh, Vishal},
  title = {Cohort change in partisan sorting},
  year = {2026},
  url = {https://vishalsingh.org/politics/how-the-sort-happened},
  note = {In The Public Mind. Published 30 September 2026.}
}