Government Institutions: The Partisan Flip

Executive: the flip, fifty years deep

The most famous chart in the institutional-confidence literature — Robert Putnam's drew it; every political scientist who has worked with the GSS has redrawn it — is the partisan-flip on confidence in the executive branch.

The pattern is so reliable that it functions as a sanity check on the GSS itself. If the executive series did not flip with the presidency, something would be wrong with the data. Confidence in the executive branch is, in 2024, essentially a question about which party the respondent's party belongs to and which party currently holds the White House.

Three observations about the flip are worth slowing down on.

The first is that the baseline — what each party reports about the executive when the other side holds the office — has fallen over time. In 1973, with Nixon in the White House, Democrats expressed 23% high confidence in the executive. In 2024, with Biden in the White House, Republicans expressed 5% high confidence. The out-party floor has dropped 18 points over half a century. The phenomenon is not "Democrats and Republicans switch sides each election"; it is "Democrats and Republicans switch sides each election, and the side they're on no longer extends any benefit of the doubt to the other side."

The second is that the swing amplitude has grown. The 1973-vs-1977 Nixon-to-Carter shift was 23 points on the Republican line (46% → 24%) and 8 points on the Democratic line (23% → 31%). The 2009-vs-2017 Obama-to-Trump shift involved a much larger flip: Republicans went from single digits under Obama to majority-confidence under Trump, while Democrats did the reverse. The signal is the same; the noise around it has compressed and the amplitude has grown.

The third is that the topline — the unconditional share of Americans saying they have high confidence in the executive — has fallen on both sides of the flip.

The combination of those three observations is what makes the executive series different from the institutional-confidence items elsewhere in this chapter. The Congress series shows a bipartisan collapse — both parties losing confidence together, at roughly the same rate, with the partisan signal small and stable. The Court series shows a recent partisan event — the Dobbs decision triggered a sharp Democratic collapse with no Republican movement. The executive series shows neither of those things. It shows partisanship interacting with the presidency itself: confidence has become a function of which party is on which side of the White House door, and the function has grown sharper.

Congress: the institution everyone has stopped trusting

If the executive series is about partisan reading of a partisan institution, the Congress series is the opposite. Almost nobody trusts Congress, in either party, in any administration, and almost nobody has for a long time.

The Congress collapse is one of the most-cited findings in the institutional-confidence literature, and the GSS has been documenting it since 1973. Half the country expressed at least some confidence in Congress in the mid-1970s; one in twenty does today. The 2024 reading of 5.6% is the lowest in the series and represents a small but real continuation of a downward trend that has not, in fifty years, ever reversed for more than a wave or two.

The partisan structure underneath this topline is the third pattern worth seeing.

Congress is the cleanest example in this book of a genuinely bipartisan loss of confidence. It does not matter which party controls the chamber; it does not matter which party the respondent supports. The American public has lost confidence in Congress more or less uniformly, decade after decade. The 1973 partisan gap was 1.5 points; the 1990 partisan gap was about 5 points (with Republicans briefly more trusting during the post-1994 era); the 2024 partisan gap is 4 points. None of these is a story; the level is the story.

The cohort breakdown on Congress is similarly featureless. Younger Americans have always been somewhat less confident in Congress than older Americans, but the gap is small and stable and does not carry a generational-replacement story. Congress has earned its low confidence across every age bracket, in every wave, in nearly every demographic the GSS tracks. It is the institution against which the rest of the chapter's data should be interpreted: when an item moves, that is the finding; the Congress baseline is what doesn't move.

Supreme Court: the Dobbs cliff

The Supreme Court series is the strangest of the three. For the first four decades the GSS observed it, confidence in the Supreme Court behaved roughly like confidence in Congress — declining slowly, with small and shifting partisan gaps, no sharp inflections. Then 2022 happened.

The topline collapse is real, and it is recent. For nearly fifty years the Court hovered in the 28–40% range, with no clear directional trend. Between 2018 and 2022, it dropped 15 points to the series low. The 2024 reading at 18% has not recovered. The Court has lost more confidence in six years than it had lost in the prior forty-five.

That topline collapse, however, is almost entirely a one-coalition event.

The Dobbs decision was issued in June 2022, between GSS waves. The 2022 GSS wave was fielded across the months that followed. The 24-point Democratic drop on Court confidence — from 32% in 2018 to 8% in 2022 — is one of the largest single-coalition shifts on any institution in this book. The Republican line over the same window moved from 36% to 27% — a 9-point drop, real but much smaller. By 2024, Democrats sit at 10% and Republicans at 29%. The institution has been functionally renamed in Democratic-coalition surveys; Republican-coalition surveys see roughly the same Court they saw four years earlier.

The ideology view sharpens this further.

The post-Dobbs Court is, on the GSS's evidence, a partisan-coded institution in a way it was not before 2022. Conservatives report roughly the same confidence in the Court they reported under the Rehnquist years. Liberals report less confidence in the Supreme Court than any partisan or ideological group has reported in any institution in this entire chapter at any time in the series — including Congress in its 2024 trough.

The college-education view tells a smaller version of the same story.

On the Court specifically, party and ideology dominate; education is a secondary axis. This is in contrast to several institutions covered later in the book — the press (§2.4), the scientific community (§2.3) — where the diploma divide is the dominant story. The Court is sorted by which side of Dobbs the respondent fell on, not by which side of the diploma the respondent went to.

Three institutions, three different futures

The three institutions in this chapter are supposed to share a single feature: legitimacy that does not depend on partisan outcomes. The Constitution allocates power between them precisely so that any single one's partisan misuse can be checked by the others. The GSS data show that the public, in 2024, no longer extends to any of the three the kind of trust that the constitutional design assumes.

Executive confidence has become a function of which party holds the office. The series is vivid in 2024 — partisan gaps of 16-30 points within each administration — but it carries an unsettling implication: there is no longer a shared baseline of confidence in the office itself. Whichever side wins the next presidential election, more than 90% of the losing side's voters will report that they have less than a great deal of confidence in the executive branch.

Congressional confidence has collapsed across the board. The institution is at 5.6% in 2024, the lowest reading on any institution in this chapter, and the partisan structure is too small to be a story. Whatever Congress is doing, the American public is not impressed, and the parties more or less agree about that. This is the closest the chapter gets to a "bipartisan finding."

Supreme Court confidence held steady for half a century and then cracked in 2022. The crack is large, recent, and one-coalition — a Democratic-line collapse from 32% to 8% in two years, with the Republican line barely moving. The institution that was supposed to be insulated from partisan politics — the institution where the constitutional design works hardest — is now, by some margin, the most partisan-coded of the three. Whether the post-Dobbs Court can recover its bipartisan legitimacy is, on this evidence, an open question. There is no fifty-year analog in the series.

The chapter ahead, on knowledge and expertise institutions (the scientific community, medicine, education), takes the same three-institution framing to a different set of bodies whose legitimacy is also being contested — and where the diploma divide that was a secondary axis here becomes the dominant one.