Democracies Don't Warn You Before They Stop Being One

There is no moment when a democracy becomes something else. No siren. No formal declaration. The constitution usually stays in place. The elections keep happening. The flags still hang outside the same buildings. What changes is subtler and, for that reason, much harder to argue about in real time: who the rules protect, how consistently they are enforced, whether institutions act as genuine constraints on power or as theater for the appearance of constraint. By the time the change is obvious to everyone, the machinery that would have stopped it has already been hollowed out.

In 2025, two of the most widely used datasets in comparative politics quietly registered that hollowing-out process as underway in the United States. The Polity data series — maintained by political scientists and used extensively by researchers, international organizations, and security analysts — moved the U.S. score in a direction that, in other countries, has historically preceded a qualitative shift in the kind of government being studied. Bright Line Watch^[3], a monitoring project run by political scientists who specialize in democratic erosion, published survey data showing significant movement on several of their benchmark indicators. Neither organization announced a coup. Neither declared the United States a dictatorship. That is precisely what makes their findings worth reading carefully.

Comparative politics scholars have spent roughly two decades studying what they call democratic backsliding — the gradual, largely legal process by which elected governments weaken the institutions meant to check them. The case studies are well documented: Hungary under Viktor Orbán beginning in 2010, Turkey under Recep Tayyip Erdoğan beginning around 2013, Venezuela under Hugo Chávez and then Nicolás Maduro across the 2000s and 2010s. In each case, researchers have identified a recognizable sequence. In each case, that sequence looked, at each individual step, like ordinary politics. The accumulation is what matters, and you can only see it clearly if you know what you are looking at.

The goal here is not to tell you how to feel about American politics in 2025. It is to explain the framework that political scientists use to measure democratic health, why the indicators they track are specifically those indicators and not others, what the historical comparisons actually show, and why the methodology matters more than any single data point. If you understand the measurement, you can read the news differently. That is the only useful thing this article can do.

What the Scorecards Are Actually Measuring

Polity has been measuring regime types since the 1970s. Its scores run from negative ten (fully autocratic) to positive ten (fully democratic) and are built from observable institutional characteristics: how executive power is recruited and constrained, how competitive political participation is structured, and how much institutional authority limits what leaders can do unilaterally. It is not a measure of policy outcomes or cultural values. It is a structural assessment — are the mechanisms of accountability present and functional? The United States has scored a ten for most of its measured history. Scores do not usually move on this scale by accident.

Bright Line Watch operates differently but with the same underlying logic. Its founding researchers — drawn primarily from American political science departments with expertise in democratic erosion — identified a set of specific institutional behaviors that, in other countries, marked the transition from competitive democracy toward competitive authoritarianism. The term competitive authoritarianism, developed by Steven Levitsky and Lucan Way in their 2002 research and later expanded in book form, describes a system where elections still occur and are nominally contested, but the playing field has been tilted sufficiently that meaningful democratic competition is undermined. Bright Line Watch surveys both political scientists and members of the public on whether specific conditions hold in the United States, tracking movement on each indicator over time.

“A score doesn't move by accident. It moves because something structural moved first.”

What both instruments share is a rejection of the dramatic. They are not measuring whether a leader sounds authoritarian. They are not tracking rhetoric, personality, or the temperature of political discourse. They are measuring institutional behavior: whether courts remain independent in practice, whether the civil service can function without political coercion, whether the press faces legal harassment from state power, whether electoral administration is insulated from partisan manipulation, whether prosecutorial authority is being used against political opponents. These are behaviors that either occur or do not occur. They are measurable precisely because they happen in the real world and leave evidence.

The Sequence That Appears Again and Again

Across the major case studies, scholars have mapped a rough sequence of how backsliding unfolds. It rarely begins with a grab for total control. It begins with what Levitsky and Ziblatt — writing in their 2018 work — call capturing the referees. Before changing the rules of the game, an aspiring authoritarian government typically moves to control the institutions that enforce the rules: courts, prosecutors, inspectors general, regulatory bodies, election commissions. These moves are almost always legally defensible in isolation. Each one can be described as a reform, a correction, a rebalancing, a return to constitutional intent. The problem is the pattern and the direction.

In Hungary, Orbán's government passed a new constitution in 2011, restructured the Constitutional Court to limit its jurisdiction over budget matters^[2], packed the court with loyalists, expanded the Supreme Court and then replaced its chief justice, and created a new judicial administrative body run by an Orbán ally. None of these steps, taken individually, was a coup. The court still met. Judges still wore robes. Rulings still issued. But the institution no longer operated as a genuine constraint. It became a ratifying body for decisions already made. Turkey followed a similar sequence with the judiciary, the military, and the press over roughly the same period, accelerating significantly after the 2016 coup attempt gave Erdoğan's government emergency justification for mass purges of civil institutions. Venezuela's trajectory involved more economic coercion and more outright violence, but the foundational move was the same: control the institutions that could push back before they do.

The second phase, in each case study, involves what scholars call using incumbency advantages to tilt the playing field. This means leveraging state resources, regulatory authority, and legal exposure to disadvantage opponents. It means targeting media organizations, donors, and civil society groups that support the opposition. It means using the justice system selectively — aggressively against critics, permissively toward allies. It does not require making opposition illegal. It only requires making opposition expensive, risky, and exhausting. That asymmetry compounds over time in ways that surveys and vote tallies can fail to capture until the underlying structure has already shifted.

“You don't have to make opposition illegal. You only have to make it expensive enough that it stops.”

Why Legality Is Not the Same as Democratic Health

One of the most important things comparative politics research has established is that legal does not mean democratic. Every tool used in the backsliding sequences in Hungary, Turkey, and Venezuela was formally legal under the systems in which it was applied. Laws were passed by legislatures. Court rulings were issued by courts. Appointments were made through established procedures. The governments in question consistently described their actions as constitutionally proper, as anti-corruption measures, as fixing broken systems — and in many cases, those arguments were difficult to defeat on purely procedural grounds, because the existing rules did not explicitly prohibit what was being done.

This is why political scientists focus on norms alongside rules. Democratic systems depend on what Levitsky and Ziblatt call mutual toleration and institutional forbearance^[4] — the informal agreements between political competitors that certain things, though legal, will not be done, because doing them would damage the system everyone depends on for their own future power. When those informal restraints erode, the written rules are rarely sufficient to replace them. Courts can be legally packed. Civil servants can be legally fired. Prosecutorial discretion can be legally directed. Each action can survive a legal challenge while still representing a fundamental change in how power behaves.

What the 2025 downgrade signals, in this framework, is not that the United States has crossed some definitive threshold into authoritarianism. What it signals is that several indicators that have historically preceded that crossing — in other countries, in well-documented sequences — have moved in the same direction at once. The methodological point is not alarmism. It is pattern recognition. Researchers who have spent careers studying what the early stages of backsliding look like are saying that what they are observing in the United States resembles what they have observed before. That observation should be treated as evidence, evaluated carefully, and neither dismissed because it sounds extreme nor accepted uncritically because it confirms a prior political feeling.

The American Case Is Both Typical and Unusual

Every backsliding case has features that look unique from the inside. Citizens in each country believed, at critical junctures, that their particular institutions were too strong, their history too democratic, their culture too resistant for the sequence to complete. Hungarian liberals noted that Hungary had NATO membership, EU structural funding requirements, and an opposition that had governed within living memory. Turkish secularists trusted the military as a backstop. Venezuelan civil society initially believed that Chávez's broad popularity would force policy moderation. These are not stupid beliefs. They are the normal beliefs of people who live inside systems that have mostly worked, reasoning about a future they have not yet seen.

The United States has features that make backsliding harder than in most cases. The constitutional system is genuinely federal, which means institutional resistance can be organized at the state level even if federal institutions are captured. The press, though financially stressed and ideologically sorted, retains significant independent reach and legal protection. Civil society is large, legally experienced, and well-funded enough to mount sustained litigation. The opposition party controls significant state governments. These are real structural buffers. Researchers who study backsliding take them seriously. They are also the same researchers who note that Orbán, Erdoğan, and their counterparts operated against institutional resistance for years before that resistance broke or was neutralized — and that the resistance itself, by remaining energized, sometimes obscures from its own participants how much ground has been conceded.

What to Watch Instead of What to Feel

The practical use of comparative backsliding research is not that it tells you how outraged to be. It tells you which events to monitor as structurally significant versus which are noise. Rhetoric is noise. Rhetoric is a signal of direction and intent, but on its own it does not move institutional indicators. What moves them is behavior: whether courts rule independently or in alignment with executive preference across a pattern of cases; whether the civil service can resist politically motivated dismissals or whether those dismissals succeed and establish a precedent; whether press organizations face a pattern of legal, regulatory, and financial pressure that demonstrably changes their coverage; whether prosecutorial decisions follow evidence or follow the interests of the incumbent; whether electoral administration becomes an arena of partisan contestation rather than technocratic management.

These are all observable. They produce evidence. They can be compared to historical baselines and to the documented sequences in other countries. Researchers are doing that comparison now, in real time, which is both more sophisticated and more useful than either reflexive panic or reflexive reassurance. When Polity moves a score and Bright Line Watch registers movement on multiple indicators in the same period, the appropriate response is to understand what those instruments measure, examine whether the underlying behavioral evidence supports the movement, and use that framework to track what happens next. The downgrade is not a verdict. It is a read of current conditions from people who built the instruments to catch exactly this kind of development before the public vocabulary catches up with it.

“The instruments were built to catch this before the public vocabulary catches up with it.”

The One Thing the Comparisons Cannot Tell You

No case study from Hungary, Turkey, or Venezuela can tell you how the American sequence ends. The comparative literature identifies patterns, sequences, and mechanisms. It does not produce deterministic predictions. Backsliding processes have been reversed. They have stalled at intermediate stages. They have been accelerated by crises or slowed by unexpected institutional resistance. The value of the framework is not prophecy. It is diagnosis — knowing what is happening clearly enough to respond to it rather than to the simplified version that partisan media on either side will offer instead.

Democratic erosion is hard to argue about in real time because each individual step is defensible and because the people implementing it always have an explanation that sounds like good governance. That is not a coincidence. It is the method. The countries that held the line, or recovered, did so not primarily because their citizens felt the right emotions, but because enough institutional actors, legal challengers, journalists, civil society organizations, and ordinary voters understood the mechanism clearly enough to resist it at the right moments. Understanding the mechanism is not a guarantee. It is a precondition. The downgrade is a warning light, not a prognosis — and the difference between those two things is whether enough people know how to read the instrument.

References

1. Project MUSE -- Verification required! (muse.jhu.edu)
2. Wrong Direction on Rights (hrw.org)
   Documents Hungary's 2011 constitutional changes and judicial restructuring that weakened legal checks on executive authority.
3. Bright Line Watch (brightlinewatch.org)
   Monitors democratic practices and institutional threats in the United States through surveys of political scientists and the public.
4. The Crisis of American Democracy (aft.org)
   Defines mutual toleration and institutional forbearance as informal agreements between political competitors essential to democratic systems.