Knowledge & Expertise Institutions: The Education Divide

Science: the flip and the floor

The lead chart of this chapter is the time series for confidence in the scientific community, by party. For the first thirty years of the GSS, the two lines tracked closely — sometimes with Republicans modestly more confident, sometimes Democrats, but always within a handful of points. Around 2000 the lines began to separate. After 2018 they collapsed apart.

Three things in Figure 1 are worth slowing down on.

The first is the direction of the original 1973 partisan order. The contemporary American conversation about science and politics often assumes that science has been a Democratic-coalition object for a long time — that the Republican coalition's distance from federal scientific agencies is in some sense definitional. The GSS does not support that. For roughly the first quarter-century of the series, Republicans were the slightly more pro-science party. The realignment is recent.

The second is the asymmetry of the recent movement. Between 2010 and 2024, the Democratic line on confidence in science rose modestly (47% → 51%, a 4-point gain). Over the same window, the Republican line fell sharply (41% → 25%, a 16-point loss). The 2024 partisan gap is not a story of two coalitions moving in opposite directions. It is a story of one coalition moving while the other holds.

The third is the speed of the post-2018 leg. Republican confidence in science fell from 44% in 2018 to 34% in 2021 to 23% in 2022 to 25% in 2024 — a 19-point drop in six years. Some of that drop is mode-shift; some of it is COVID-era controversies over the CDC, the NIH, and federal public-health authority; some of it is post-Trump-era institutional realignment. The data cannot adjudicate among these. The data show only that something in the post-2018 period broke the Republican coalition's relationship to the institution it had been mildly more pro than the Democratic coalition was for the first thirty years of the question.

The ideological view sharpens the partisan one.

The ideological version of the chart is even cleaner than the partisan version. In 1974, Conservatives reported higher confidence in the scientific community than Liberals did — the post-Sputnik, defense-laboratories, NASA-program coalition. In 2024, Conservatives report less than half the confidence of Liberals. The 35-point ideological gap on confidence in science is the largest in the history of the question and is approached by very few items elsewhere in this book.

The diploma divide, and what it doesn't explain

The proposal that became this book described the science-confidence story as an example of the diploma divide reinforcing the partisan divide. The reading is partly right and partly wrong. The college-educated population is more confident in the scientific community than the non-college population, and has been for the entire GSS series. But the recent partisan widening is not primarily a college-vs-non-college story.

Two of this book's repeated findings show up in Figure 3. The first is that the diploma divide on a knowledge institution is real and persistent — about 15–25 points across the entire fifty-year series. The second is that the diploma divide is not the master variable on every cultural question. Between 2018 and 2024, college-educated confidence in science fell almost as much in absolute terms as non-college confidence did. The educated segment did not stay where it was; the educated segment came down too, though to a still-higher level. Whatever happened to American confidence in science in the post-2018 period is not primarily a story about the difference between people with and without a bachelor's degree. It is a story about the difference between Republicans and Democrats, and on that axis the gap is now twice the size of the diploma gap.

Medicine: the COVID footprint

Of the three institutions in this chapter, medicine has the cleanest single-event footprint. The 2018 GSS wave showed the two parties within one point of each other on confidence in medicine (D=38, R=39); the 2021 GSS wave showed Democrats 11 points ahead; the 2024 wave shows Democrats 14 points ahead. The widening sits directly on top of the COVID-19 pandemic.

The pattern is similar in shape to the science series, but the relevant timing is shorter and more clearly identifiable. The relationship between the Republican coalition and the federal public-health apparatus changed during a specific, short, observable period. The institutions involved — the CDC, the FDA, state and local health departments, the doctors and hospitals associated with vaccine and treatment recommendations — were the same institutions before, during, and after COVID. The change is in the public's relationship to them.

The topline shows the magnitude of what was lost.

Half a century, half the confidence. Medicine in 1973 was an institution that more than half of Americans expressed a great deal of confidence in. Medicine in 2024 is an institution that one in four Americans expresses that level of confidence in. The 2024 Republican reading (21%) is the lowest in the entire series for any partisan group. The 2024 Democratic reading (35%) is below where Republicans were as recently as 2010.

Education: the institution everyone trusts less, but unevenly

The third item in this chapter — confidence in the people running "education" — adds one more axis of polarization that the other two do not. Confidence in education has fallen across the board (37% in 1973, 20% in 2024). But the partisan dynamics within that fall are recent and specific.

What is happening on education is not the same as what is happening on science or medicine. The Democratic line on education has been roughly flat at 25-30% for the entire post-2000 era; it is not the case that Democratic confidence in education has risen. What has happened is that Republican confidence in education has fallen sharply. The 2024 Republican reading of 13% is the lowest in the series and is below the Republican reading on any of the three institutions in this chapter. Whatever Republican-coalition voters mean when they answer this question in 2024, they are reporting that the people running American education are not earning their confidence.

The college-education view contains a finding that runs against the standard "diploma divide" framing of the other items in this chapter.

Figure 7 is an important caveat to the diploma-divide story. On science and medicine, college-educated respondents are systematically more confident in the institution. On education, that pattern does not hold. Throughout the series, non-college respondents have been somewhat more confident in the people running education than college-educated respondents have been. The 1973 gap was 10 points in the non-college direction; the 2024 gap is roughly 2 points in the same direction. If you believe the survey, the population most distant from the K-through-PhD educational system is not the population least confident in it.

One possible reading: college-educated Americans have direct exposure to the institution and report on it from that exposure, while non-college Americans report on it from a more distant — and apparently more favorable — vantage point. Another possible reading: the contemporary "education" question is being heard by college-educated respondents as a question about higher education, with all the culture-war content that carries, while non-college respondents are hearing it as a question about K–12 and report on it more positively. The GSS data cannot adjudicate; both readings are consistent with the pattern.

Three institutions, one pattern

What is striking about the three knowledge-and-expertise institutions covered in this chapter is that they have moved together, in the same direction, over the same period. On all three, the post-2018 leg shows:

• A sharp Republican-coalition decline in confidence (science −19pts, medicine −19pts, education −6pts).
• A much smaller Democratic-coalition movement (science +1pt, medicine −2pts, education −3pts).
• A widening partisan gap of 10–30 points by 2024.

This pattern is the strongest argument in this book for what political scientists call coalition realignment — a structural change in which kinds of voters identify with which coalition, in which "the Republican coalition is now the coalition of people skeptical of federal scientific authority" is a recent fact, not a longstanding one. The 1973 GSS showed Republicans the slightly more pro-science party. The 2024 GSS shows them more than 25 points behind on the same question.

How much of this is COVID, how much is Trump-era institutional disruption, how much is the cumulative effect of conservative-media frames around the EPA, the CDC, the NIH, and university administration — the data cannot say. The data say only that the fact of the change is real. The Republican coalition in 2024 is, on this measure, no longer the coalition of trusted federal scientists, doctors, and educators. That coalition has migrated.

The next chapter turns from knowledge institutions to the media institutions — the press and television — where the same partisan-flip story is even sharper, and where the post-2010 readings are extreme enough that the GSS series almost runs out of resolution at the bottom of its scale.