In 2018, Harvard political scientists Steven Levitsky and Daniel Ziblatt published How Democracies Die, seven years before any of what follows in this piece happened. Their argument: modern democracies rarely fall to a coup. They fall to their own elected leaders, using the institutions already in place — gradually, subtly, and even legally. Building on a framework the political scientist Juan Linz had proposed decades earlier, they named four specific warning signs any citizen could check a leader against: rejection of, or weak commitment to, democratic rules of the game; denial of the legitimacy of political opponents; toleration or encouragement of violence; and readiness to curtail the civil liberties of opponents, including the media.^[1]

The book's own argument has a two-thousand-year-old precedent it names directly: the Roman Republic. Historian Edward Watts, in Mortal Republic: How Rome Fell into Tyranny (2018), argues the Republic did not fall to a single law or a single conquest — it fell because Roman politicians stopped honoring its unwritten norms one at a time, each violation individually defensible, until nothing was left to restrain the next one.^[2] The first break came in 133 BCE: the tribune Tiberius Gracchus, campaigning for land reform, was clubbed to death by a mob led by a sitting senator during his re-election campaign, and roughly three hundred of his own supporters were killed alongside him — the first time in Roman political history that an election was settled by violence instead of a vote.^[3] Twelve years later, in 121 BCE, his brother Gaius Gracchus was hunted by a Senate-authorized mob and took his own life rather than be captured.^[3] Neither killing broke a written law. Both broke a norm that had held for centuries, and by the time Augustus became the first emperor decades later, there was no norm left standing to break.

This site has already traced the specific American version of the pattern that framework describes in the abstract. Congress has not formally declared war since June 4, 1942. The Supreme Court has narrowed independent-agency protections case by case over five years — Seila Law v. CFPB (2020), Collins v. Yellen (2021), Trump v. Slaughter (2025) — consolidating power toward a single point. The same Court, in Rucho v. Common Cause (2019), ruled that federal courts have no authority to police partisan gerrymandering at all. None of these is a coup, and none happened in a single dramatic moment — which is exactly the "gradually, subtly, even legally" pattern named seven years earlier.

Separately from that book and separately from each other, three organizations with no coordination between them have spent years building the actual measurement infrastructure for this question, for reasons that have nothing to do with any single argument or any single year. Freedom House has published Freedom in the World since the 1970s, scoring every country on a 0-100 scale since 2002. The V-Dem Institute, based at the University of Gothenburg, codes hundreds of democracy indicators across roughly 200 countries using teams of country-level experts. Bright Line Watch has regularly surveyed several hundred American political scientists specifically on the health of U.S. democracy since 2017. None of the three was built to answer a question about this particular year.

In 2026, all three published what they found. Freedom House gave the United States 81 out of 100 — the sharpest single-year decline of any country still rated "free," the first time the U.S. has fallen outside the global top 20, and its lowest score since the 0-100 scale began. The report names specific reasons: weakened anticorruption safeguards, the president's family business ventures creating documented conflicts of interest, a pattern of clemency benefiting people with ties to the president's campaign or family enterprises, and "escalating legislative dysfunction and executive dominance over Congress's traditional powers."^[4]

V-Dem went further, formally downgrading the United States from "liberal democracy" to "electoral democracy" — a category change, not just a lower number — and described the pace of that shift as unprecedented in modern times, pulling U.S. democracy back to roughly its 1965 level by the index's own measure. The same report found freedom of expression in the U.S. at its lowest level since the end of World War II, citing the "suppression and intimidation of media and dissenting voices."^[5]

Bright Line Watch's roughly 600 surveyed political scientists put the United States at 57 out of 100 as of their most recent survey — closer to their own hypothetical benchmark for an "illiberal democracy" (47) than for a "strong democracy" (93). The honest complication, stated because it is true and not because it softens anything: that 57 is actually up from a low of 53 in April 2025, an improvement the surveyed scholars specifically attributed to two checks that visibly worked — a Supreme Court ruling against the administration on tariffs, and the end of domestic deployments of the National Guard.^[6] The system is not moving in only one direction at every moment. It is measurably worse than it was, and it remains capable of correcting itself when a check actually gets used.

Three organizations that do not coordinate with each other, and a framework published seven years before any of this happened, are not describing three different things. They are describing the same mechanism from three different distances — one naming how this particular kind of loss happens in general, the other two measuring, in real time, that it is happening here.