Chapter 37 Key Takeaways

The chapter is the book's most ideologically sensitive. Its takeaways are stated as analytical commitments and institutional facts, not as policy positions. Each survives across the chapter's four steel-manned positions on the contested question of where, if anywhere, American democracy is currently eroding.

On the analytic distinction

  • Erosion is not collapse. Collapse is sudden and often violent. Erosion is gradual, operates through formally legal moves, and produces a regime in which elections still occur but the playing field is no longer level. The comparative-democracy literature distinguishing the two has been developed since the 1990s and now routinely turns its instruments on the United States itself, with mixed and contested findings.

  • Erosion is not a synonym for "policies I oppose." A government can pursue policies you strongly disagree with without eroding democracy in the institutional sense, and vice versa.

On measurement

  • V-Dem, Freedom House, EIU, and Bertelsmann all show a coordinated decline in U.S. composite scores from approximately 2015–16 onward. V-Dem's Liberal Democracy Index for the United States dropped from approximately 0.88 in 2015 to approximately 0.73 in 2024 — the largest decline among advanced industrial democracies in the period. Freedom House dropped the U.S. from 94/100 in 2010 to 83/100 in 2024. The U.S. remains classified as Free.

  • Read the numbers carefully. They are imperfect instruments, but their convergence across multiple independent frameworks is harder to dismiss as artifact than any single index would be on its own.

On norms versus rules

  • The U.S. Constitution does not specify how the federal government operates day-to-day. The gaps are filled by informal norms — patterns of behavior political actors observe even when not legally required to. Once broken, norms are slow to restore: the rational political actor does not return to a norm unilaterally if they fear non-reciprocation.

On constitutional hardball

  • Constitutional hardball — legal but norm-violating exercises of formal power — is bipartisan in pattern and escalating in trajectory. The chain documented in Case Study 2 includes the 1987 Bork rejection, the 2013 Reid nuclear option, the 2016 Garland blockade, the 2017 McConnell extension, the 2020 Barrett confirmation, the 2020–21 Democratic filibuster-reform and court-expansion proposals, and the 2025 Trump-2 executive-branch actions. Whether the post-2016 acceleration has been more pronounced on one side (the asymmetric-hardball debate, Fishkin/Pozen versus Lee/Schickler/Bernstein) is contested empirical territory the chapter does not adjudicate.

On the institutional facts of 2020–25

  • The 2020 election was won by Joseph R. Biden Jr. This is not contested at the level of the data. Federal courts heard and rejected more than 60 post-election lawsuits; CISA called the 2020 election "the most secure in American history"; Attorney General Barr told the AP the DOJ had not seen fraud sufficient to have changed the outcome. A meaningful share of the public continues to believe otherwise; that gap is itself a flagged pattern in the comparative-democracy literature.

  • January 6, 2021, is treated in the chapter as institutional fact. The Capitol was breached; certification was delayed approximately six hours; approximately 140 officers were injured; five people died on or near the day; ~1,500 individuals were prosecuted; the Senate impeachment trial returned a 57–43 vote to convict, the largest cross-party vote against any President in any impeachment trial in American history but short of the 67 required. On January 20, 2025, nearly all approximately 1,500 January 6 defendants received executive grants of clemency, including those convicted of violent offenses and seditious conspiracy. What these facts mean is the reader's task to assess.

  • The 2024 election was won by Donald J. Trump in both the popular vote and the Electoral College (312–226). The election was administered, certified, and the transfer of power was peaceful. The outcome is not contested at the level of the data.

  • The Trump-2 administration since January 20, 2025, has taken executive-branch actions including Schedule F revival, the firing of 17 Inspectors General without statutory 30-day notice, mass pardons of January 6 defendants, personnel actions at DOJ and FBI, National Guard federalizations, public calls for prosecution of named opponents, IEEPA-based tariff actions, and executive orders restructuring agency authorities. Whether these represent a difference of degree or of kind from prior administrations' executive overreach is the central analytical question.

On the four steel-manned positions

The chapter steel-mans four positions: (1) the asymmetric-erosion-is-real position (Levitsky/Ziblatt; Cheney; Kristol; Niskanen); (2) the institutional-system-is-responding-correctly position (Federalist Society; National Review tradition); (3) the real-erosion-is-on-the-left position (New Right; Vermeule; Caldwell; Rufo); and (4) the institutional-credibility position (Yuval Levin; Jonathan Rauch; Gurri). The chapter does not adjudicate. The analytical equipment to engage all four is the chapter's offer.

On the cross-partisan civic obligation

  • The Madisonian framework relies on ambition counteracting ambition. That counteraction operates only if mass-public actors hold both parties accountable for institutional behavior. Same-party criticism — costlier and less expected than opposition-party criticism — is the most informative norm signal.

  • Power flows to those who show up. Theme 3 of this textbook applies. The work is voting in primaries, holding one's own party's leaders accountable, supporting civil society, defending free expression including expression one disagrees with, recognizing cross-coalition allies, and engaging in peaceful protest. The list is incomplete; the work is what citizenship is.

What the chapter does not do

The chapter does not predict the future of American democracy. It does not declare American democracy in crisis or robust. It does not declare any administration's actions decisive. The trajectory is uncertain; the interpretation is contested; the bet the Founders made — that ordinary people, given decent institutions and a serious civic education, are capable of self-government — has not yet been resolved. The reader is one of the actors who will resolve it.