How Plato's spell still shapes our thinking on democracy
Democracy, fairness, and Popper
Threats to democracy?
Like many Trump critics, I often find myself worried about the president’s numerous attempts to curtail the constitution, skirt common decency, and undermine our democratic system. Trump is also friendly with some of the world’s most aggressively illiberal leaders like Vladimir Putin and Nayib Bukele, which is reasonably concerning to many in the West.
One name that’s become increasingly prominent in that conversation is Viktor Orbán.
After he and his party Fidesz won with a supermajority, Orbán rewrote Hungary’s constitution and amended it dozens of times. He packed the courts. Gerrymandered the maps. Extended voting rights to Hungarians abroad who overwhelmingly supported his rule. State-aligned entities acquired most of the media market. Public funds were directed toward loyal oligarchs.
As political scientists debated whether it still qualified as a democracy by any meaningful definition, Orbán and his party were voted out.
And in light of that event, Ross Douthat made a simple observation that mimicked the ancient excuse Republicans love to parrot whenever Trump's democratic record comes up: the institutions held!
People who think Trump is tyrannical tend to see Orbán as validation of their fears despite his ultimate departure, while those who think the threat is exaggerated use Orbán’s nonviolent removal as evidence that the framing around Trump is overblown.
Both sides are genuinely convinced they’re arguing about democratic legitimacy—and in a sense, they are.
But much too often, our debates around consolidation of power and unpopular governments stray into questions that sound like they’re about democracy but are really about something else entirely: Did the right people win? Are my preferences being represented? Is the system ensuring fair outcomes? Who deserves to be in charge?
These common inquiries are about whether the system is producing the right outcomes for the right people given a specific input—in other words, whether it's fair. Once fairness becomes the standard for preventing tyranny, any widely disliked government is already suspect and any attempt by it to govern looks like a democratic failure.
These must therefore be the wrong questions.
The wrong question
Since Ancient Greece, every civilization that has tried to organize politically has attempted to answer the question of who should rule.
Plato argued for philosopher-kings. Others responded with God and their anointed authorities. Eventually a more convincing answer formed: the will of the people. But which people? Karl Marx argued for the working class—a “dictatorship of the proletariat”. William Godwin’s theories centered on a tyranny of the individual. Hegel thought the question is resolved by history itself.
The cost contained in all these answers, and in any further attempts, is that if you provide one, you’ve made it permanent.
In The Open Society and Its Enemies, philosopher Karl Popper noticed the error: every attempt to answer who should rule becomes justification for someone to take power and refuse to give it up; whether through wisdom, divine right, historical necessity, or even the will of the majority.
“The question of who should rule contains buried inside it the seed of tyranny.”
—Karl Popper
As was Poppers’s style, he decided to throw the question out altogether.
What we should be asking isn’t who should rule—it’s how do we get rid of rulers without violence?
A ruler is only tolerated because there exists a mechanism to remove them peacefully. Elections from this perspective aren’t popularity contests but rather a “Day of Judgment”—when those who governed can be dismissed without bloodshed if the verdict goes against them. But a vote is only a single blunt instrument. And even then, it’s possible for people to remove those who weren’t actually responsible—leaving the mechanism misdirected.
Ensuring those in power are held responsible for bad outcomes is therefore essential to the effectiveness of peaceful removal.
With great power comes great responsibility
For any solution to be useful, it must actually address the problem it’s purporting to. It’s not enough that a populace can remove rulers without violence; the removal mechanisms must be aimed at those controlling the levers—those genuinely responsible for bad outcomes.
It necessitates structuring our system so voters can identify who to blame.
Responsibility should assigned so when a party wins, they govern alone: no coalition compromise and no diluted policies. What they do is fully theirs. Those policies are exposed to reality, and they’re responsible for the outcomes. When the removal mechanism fires, it hits those in control.
In my view, much of Western commentary treats this clarity of responsibility as a flaw rather than a feature. A system is often considered more legitimate the more it incorporates everyone’s preferences into its decisions—leading to arguments in favor of proportional representation, ranked-choice voting, and other variations. But this is just a reformulation of the question of who should rule followed by a response of "everyone, equally, in proportion to their numbers."
“An election does not play the same role in a rational society as consulting an oracle or a priest, or obeying orders from the king, did in earlier societies… The voters are not a fount of wisdom from which the right policies can be empirically derived. they are attempting, fallibly, to explain the world and thereby improve it.”
—David Deutsch, The Beginning of Infinity
But a rational analysis of democracy should concentrate not on questions of how inputs in decision-making can lead to a specific outcome but on how democratic institutions contribute to the assigning of responsibility, the removal of bad policies and rulers, and the creation of new options.
The ‘guardrails’—so to speak—are there to keep those in power exposed to the consequences of governing badly; not to ensure a preferable outcome.
How the guardrails hold
The removal mechanism operates on two levels, though Popper mainly focused on the first.
At the formal level, elections are held, leaders can lose, and the apparatus of displacement is present. Simple enough.
The second level deals with function: criticism can broadly accumulate and translate into distributable removal of rulers. And notably: we can satisfy the first condition while degrading the second—as in the case of Orbán and in partisan assessments of Trump.
Democracy doesn’t just happen. We set up structures that make certain actions possible, expected, and rational—and others not. These mechanisms depend on institutions; not physical buildings or bodies of experts, but shared expectations about what counts as legitimate. Think of them as accessible systems of rules, norms, and roles that structure how individuals understand and act within political life. They shape what a judge does, what a voter expects, what a peaceful transfer of power looks like.
And the reality of modern democracies is that removal processes are distributed through discrete competitive units (districts, seats, offices) each with its own threshold for displacement.
A wave of criticism must find purchase in enough of these units to produce an actual change in governing power. A system in which many units are genuinely competitive translates distributed criticism into displacement relatively cheaply. And so the system in which few are competitive requires criticism to be more concentrated, more intense, and more precisely targeted before it becomes effective.
The rate and fidelity of institutional learning—what we might call a system's dynamism—is the right measure of democratic health.
In a healthy system, even parties with large majorities remain vulnerable to being blamed and to being voted out. It’s the institutions that keep responsibility alive between elections; what makes blame assignable and dissatisfaction legible before one even arrives at the voting booth.
So rather than asking whether a system is democratic or tyrannical, we should be asking how dynamically is it functioning? Which specific institutions are making that easier or harder? And how can we improve them?
Hungary is the most recent test of the right questions.
The right question
Measured not by fairness but by the two institutional levels we’ve established, the case of Hungary is genuinely complicated.
Some of Orbán’s reforms actually sharpened accountability. Eliminating the second round in favor of single-round plurality meant whoever won governed alone, owned outcomes alone, and could be removed alone. Governing alone clarifies responsibility.
However, other moves by Fidesz cannot survive the same analysis. Many of their actions were genuine attempts to reduce vulnerability without any corresponding increase in the clarity of responsibility: a deliberate decoupling of democracy’s two core functions. And yet, Orbán and his party, after more than a decade of entrenching themselves in power, were indeed removed peacefully through the very institutional mechanisms they had spent years shaping.
The question was never who should rule Hungary or how to make the outcomes more fair; it was always how effectively its institutions could remove those in power. Those who point to peaceful removal as proof of democracy are right. Those who point to institutional degradation are also right.
Where both sides go wrong is in treating Hungary as a verdict in the domestic ideological conflict about the United States and Donald Trump.
Look: January 6th was bad. I feel very strongly about that, and writing this piece has given me a better explanation for why: it was an explicit attempt to curtail the removal mechanism itself by rejecting the institutional bounds of conduct and using violence and the threat of it to hold onto power rather than submit to it.
But I also don’t think January 6th-styled takeovers are encroaching on us each and every day Trump is in office.
Even if Trump declared himself king tomorrow, that wouldn’t mean we’d simply tolerate it. The same institutions that removed Orbán, that reconvened Congress on the night of January 6th, that refused to certify fraudulent results don’t disappear because one person (or even an immoral group) wills it.
We must throw away the question of who should rule for a dynamic system we control individually and collectively. We are each the guardrails. We retain the ability to criticize and remove those in power because we impose its constraints, we criticize abuses, and we alter what isn’t working.
The fact is our concern for proportionality and fairness in the United States is often an attempt to prevent an immoral majority from entrenching themselves. But a voting system can accomplish this.
What it can do is better enable the structures to make removal possible and effective while keeping responsibility clear. I don’t want to discredit the enlightened institutions we’ve built, the norms of tolerance and error-correction we’ve accumulated, or ignore the fact that we retain the ability to do something about it—all problems are soluble.
Winston Churchill said “Democracy is the worst form of government, except for all the others that have been tried” and he’s right. Democracy did what no other civilization or religion or king ever attempted: it abandoned the question of who should rule and replaced it with the ability to recognize mistakes, remove those responsible, and keep trying.
Interesting. The shift from who should rule to whether a system can actually remove those in power is pretty useful. I think Trump is definitely trying to degrade the removal aspects but the responsibility part is still very much intact.
We gotta make sure the institutions do a better job of holding but they have so far done an okay job!