The Machinery of Legitimation I

How a State Gets Made and Made Legitimate

Ask yourself this question. What is the legitimate basis of political and economic authority, and what happens when that authority is systematically used against the people it claims to represent?

Not Founded, Accreted

No state was founded. It accreted. The version most people carry from school — a great lawgiver, a founding council, a single document that brought order out of chaos — is a story about the moment a state announced itself, not about how it actually came to exist. By the time anyone is writing a constitution, the underlying machinery has usually been running for centuries.

The earliest order ran on people you were related to, not rules you agreed to. Anthropologists studying kinship-based societies — the !Kung of the Kalahari, Amazonian tribes, medieval Icelandic clans — find the same logic worldwide: small kin groups solve “how do we not kill each other” through reciprocity and revenge. You didn’t take from your cousin because your cousin’s whole extended family would come for you, and everyone knew it. No court, no police, no ruler required.

This wasn’t a single ladder everyone climbed in order. Recent anthropology (Graeber and Wengrow’s The Dawn of Everything is the best-known version of the argument) pushes back hard on the old idea that societies march in one direction — small and equal, then large and hierarchical, forever. Plenty of societies moved back and forth, hierarchical for the harvest season and egalitarian the rest of the year, by choice. What follows is the pattern that, this time, didn’t reverse.

That system has a hard ceiling, and hitting it produces the first leaders. Reciprocity scales badly. Once a group gets too large for everyone to track everyone else’s debts and grievances personally, the kin-based system fails exactly where it’s needed most — strangers, distant clans, disputes nobody remembers the origin of. What fills the gap isn’t a vote for a king. It’s whoever can already solve a coordination problem the group has: organising a hunt, leading a raid, mediating a dispute between clans who don’t trust each other. Authority starts as a job, not a title.

Bureaucracy is the technology that lets authority survive the person holding it — and accounting is a stronger candidate for the oldest profession than the one usually given that title. A war leader’s power dies with him unless something outlives him, and what outlives him is a record. Mesopotamian cities kept grain and tribute inventories on clay tablets centuries before anyone wrote down a law code. Egypt shows the same pattern from the other direction: the annual Nile flood made grain surplus predictable enough that someone had to count, store and redistribute it at scale — and that counting apparatus, not the pharaoh’s divinity, is the structure underneath the throne.

Shang China recorded harvests and tribute on oracle bones. The Indus Valley cities ran on standardised weights and measures centuries before their script has even been deciphered. The Inca ran an empire of several million people using knotted cords — quipu — with no writing system at all. The technology varies. The function doesn’t.

Kinship handles small-scale order and runs out of road. Leadership fills the gap at exactly the point reciprocity fails. Bureaucracy makes that leadership outlast the individual who first held it. None of this required a founding genius, a single date, or a plan — and it didn’t even require a single direction of travel. It required scale, a coordination problem, and a way to keep score.

The state’s arrival never erases what came before it, either. Kinship-based organisation hasn’t disappeared inside modern states — it’s taken one of a few different shapes. Sometimes it survives hollowed out, sold back as heritage rather than left as politics: Sri Lanka’s Vedda communities, resettled off their traditional land decades ago, now visited more as a tourist stop than encountered as a governing system.

Sometimes it survives absorbed as an administrative layer, ruling through existing local authority rather than replacing it — much of British colonial Africa ran this way, installing tribal leaders as the local arm of the state rather than dismantling them, a structure that in many places outlived colonial rule itself. Sometimes it runs genuinely in parallel, wherever the state’s own reach stays thin: jirga councils still settle disputes in parts of Afghanistan and Pakistan the formal courts barely touch. And sometimes it’s written into law as its own jurisdiction rather than merely tolerated: US tribal sovereignty, New Zealand’s Waitangi Tribunal and iwi structures, both genuine, ongoing exceptions to “the state” being one uniform thing. A state’s authority is rarely as total or as settled as it looks from the centre.

History Without Great Men

Most of the states named so far didn’t develop toward Rome. They didn’t develop toward much of anything, for a very long time. Egypt ran essentially the same basic model — pharaoh, temple, granary bureaucracy, a cosmology (maat) tying the ruler’s legitimacy to maintaining cosmic order — for close to three thousand years, longer than the entire span from Rome’s founding to now. Mesopotamia cycled through Akkadian, Babylonian and Assyrian empires while recycling the same cuneiform-bureaucracy-plus-divine-kingship toolkit for millennia.

The Indus Valley cities ran standardised weights, brick sizes and drainage across a vast area for centuries with no obvious palace or monarchy at all, then declined for reasons still debated, their script still undeciphered. China’s Shang and Zhou dynasties ran on the Mandate of Heaven — a cosmological legitimacy claim built to outlast any individual dynasty: when a dynasty fell, the explanation wasn’t that the system had failed, it was that heaven had withdrawn its mandate, which let the same underlying state technology restart under new management, repeatedly, for centuries. Mesoamerica and the Andes — Olmec through the Maya, later the Inca — ran their own multi-century versions of this same pattern, independently, with no contact with anything just described.

Greece gets an outsized share of the credit for “democracy,” against societies already running sophisticated bureaucratic states for two thousand years. It wasn’t more advanced. It’s just that successors and Western thought leaders decided it was. Rome later borrowed selectively from Greek political thought and constitutional practice, and centuries after that, Enlightenment-era European writers went looking for a usable origin story for their own new institutions and picked Athens for the part, after the fact.

What Athens actually did, on its own terms rather than as a prequel to anything else, is still striking: it ran large parts of government by lottery, not election. Most people assume voting is democracy’s defining feature; classical Athens used election mainly for a small number of high-trust military and financial posts, and filled most public offices — including most of the Council of 500 that set the Assembly’s agenda — by sortition, drawing citizens’ names at random for fixed terms. The Assembly itself, open to any citizen who turned up, voted directly on laws and war — no representatives standing in for anyone. Ostracism let citizens vote, once a year, to exile any individual for ten years if he looked to be building dangerous personal power, no trial or crime required.

“Citizen” excluded women, excluded the enslaved population that did much of the work the city ran on, and excluded resident foreigners — so the franchise this system is famous for reached perhaps one adult in five, in a city built substantially on the unfree labour of everyone else. Rome, when it later drew on Greek constitutional thinking, dropped direct citizen voting almost entirely and kept the checks-and-balances instinct instead — the mixed constitution, power split and balanced against itself, is Rome’s contribution and later theorists’, not a straight inheritance from Athens.

Rome shows what happens when bureaucracy’s logic is pushed as far as it can go. At its height it governed tens of millions of people across three continents with no printing press, no telegraph, nothing faster than a horse — through roads, a common currency, standardised law, and a professional administrative class (the cursus honorum, a genuine career ladder of public office) rather than through the emperor personally knowing anyone. Citizenship did the job kinship used to do — telling you who was in — extended eventually to conquered peoples far from Rome, without requiring anyone to actually be a relative.

In western Europe specifically, most of that machinery stopped working, and it isn’t a story about the whole world going dark. When the western Roman empire collapsed in the fifth century, its roads, currency, tax bureaucracy and standing legions broke down across the western territory within a couple of generations. Rome’s eastern half didn’t collapse at all: Byzantium ran a recognisably continuous, tax-collecting, law-codifying Roman state for another thousand years. Further east and south, the Islamic caliphates — Umayyad, then Abbasid — built some of the most sophisticated tax-literate bureaucratic states anywhere on earth across this same period, and it’s their scholars who preserved and extended much of the Greek philosophy and science western Europe had itself let lapse. “The Dark Ages” is a western European story, told from inside western Europe, about western Europe’s own institutions.

Not every part of western Europe took this route, and the exception matters for what comes later. Venice, Genoa and Florence kept commercial administration, urban bureaucracy and, in Venice’s case, an elected doge checked by councils and a senate, running continuously through the same centuries the rest of the region was fragmenting into lordship and patronage. The Hanseatic League ran a comparable trading-town network across northern Germany and the Baltic. These weren’t feudal states that happened to trade — they were a genuinely separate political form, urban and mercantile rather than land-and-loyalty based, and Florence’s version of it is exactly what the next section returns to.

What filled the gap in the west wasn’t Rome’s bureaucracy rebuilt. It was four much older, much simpler mechanisms standing in for it. Hereditary monarchy replaced the cursus honorum‘s career civil service with something far cruder — power passed by blood, because there was no functioning administrative apparatus left to select, train or promote anyone on merit. Patronage replaced Roman law’s impersonal claims with personal loyalty: a lord protected you, and armed you, in exchange for service and obedience sworn directly to him, not to an office. Manorialism replaced the tax bureaucracy with the manor itself — a mostly self-sufficient estate where the lord extracted labour and produce directly from the peasants tied to his land. Revenue-raising itself shrank from something a distant treasury administered to something a lord collected, by force, custom or consent, from the people physically nearest him.

In short feudalism. Though “feudalism” is doing some flattening here — historians increasingly avoid the word for that reason, since what actually varied enormously by region and century, but the underlying shift it points to is real: less central bureaucracy, more personal and local power.

Charlemagne’s Carolingian Empire, crowned in 800, looked briefly like Rome rebuilt — but it was one man’s personal military achievement, not an institution, and it fractured into pieces within a generation of his death precisely because it had no cursus honorum underneath it to outlast him, only patronage and hereditary claims among his grandsons. The Holy Roman Empire that followed centuries later is the purest case of all: an emperor with a grand Roman-sounding title presiding, in practice, over hundreds of effectively independent princes, bishops and cities — sovereignty in name, fragmentation in fact, for most of its eight hundred years. Alfred the Great points forward rather than back: his burhs and shire system were early, real steps back toward the kind of standing infrastructure — tax base, defended territory, administrative division — the next section shows fully arriving centuries later.

Go back to this piece’s own opening line — the story most people carry from school, a founding lawgiver bringing order out of chaos. That story doesn’t just get the history wrong. It teaches, quietly and very early, that order comes from betters — that hierarchy is what stands between us and chaos, and that the person on top has earned the standing to demand obedience by the fact of having imposed order at all. Feudalism’s parade of kings and queens is the simplest possible version of that lesson, learned in childhood long before anyone reads a word of political philosophy.

Two shocks then drag western Europe back toward centralised state capacity, and neither is a great man’s idea.The Black Death, from 1347, killed roughly a third of Europe’s population and broke the feudal labour system from below — with workers suddenly scarce, peasants could demand wages or walk, and the old personal bonds of obligation started to look optional. From above, gunpowder artillery made the feudal lord’s castle obsolete almost overnight, a shift historians call the military revolution: a cannon could knock down walls a private army couldn’t take, and only a ruler with tax revenue large enough to buy cannon, and gunners to fire them, could compete. Both pressures point the same way — power consolidating upward, away from thousands of local lords and back toward fewer centres able to tax, and therefore fight, at scale.

What comes out the other side, by the 1500s and 1600s, is recognisably the modern state’s ancestor: the fiscal-military state. A monarch who can levy taxes systematically, fund a standing professional army instead of borrowed feudal knights, and run a permanent civil administration to collect the money and keep the records — Tudor England’s Exchequer, Louis XIV’s France, are the standard examples. The printing press adds a further multiplier: for the first time a ruler’s laws, currency and propaganda could reach the whole realm in identical copies, rather than degrading with every hand-copied transcription.

Jean Bodin, writing in the 1570s amid France’s wars of religion, coined the concept modern states still run on: sovereignty — a single, indivisible, supreme authority within a territory, answerable to no one above it. Before anyone can ask why a subject obeys, there has to be a single who the obeying is owed to.

Inventing Consent

To recap: even Athens’s “democracy,” the model most people trace the word back to, was a contested and exclusionary category from the start — direct votes for perhaps one adult in five, women, slaves and foreigners excluded entirely. The word has never had one fixed, agreed meaning. It has always been a boundary people fight over.

Montesquieu, writing in 1748, gave the modern version of that fight its most durable design answer: executive, legislature, judiciary, split across separate hands, so that no single office — elected or not — can simply do what it likes and call it legitimate. It is not, on its own, a guarantee that the split is real rather than decorative.

Bodin gave the state a who. What it still needs is a why — and the why does not arrive out of nowhere; it takes about two centuries of intellectual upheaval to get built. Florence is a useful place to pick up the thread, because it was itself an unusual kind of state — a republic of merchants and guilds rather than a hereditary court, one of the exceptions the previous section already flagged. Florence’s scholars revived something Rome had and medieval Europe had largely lost: the idea that a citizen could reason about how a state ought to be organised, rather than simply inheriting an answer from scripture or a king. Out of that civic humanism — Machiavelli, writing from inside Florentine politics, is the sharpest example — comes a habit of mind that spreads across the Renaissance: questions about power, authority and human nature are things educated people argue about in print, rather than settled facts.

That habit of argument runs straight into a century of catastrophe, and the catastrophe forces the question from interesting to urgent. The late 1500s and early 1600s are one of the most intellectually fertile stretches in European history, and also the run-up to the Thirty Years’ War, 1618 to 1648, a conflict fought substantially over who has the right to rule and on what religious and political grounds, which killed somewhere between a quarter and a third of the population in parts of central Europe before it burned out. England, Scotland and Ireland fought their own version of the same collision in the 1640s — a conflict now often called the Wars of the Three Kingdoms, since it was never only an English civil war.

All of this — Florence, the Renaissance, the wars of religion — adds up to what is usually marked as the start of Western political philosophy as its own subject, distinct from theology. Political authority stops being something argued mainly by quoting scripture, and starts being something argued by reasoning from first principles about human nature and consent. Hobbes is where that subject gets its first fully worked example.

Thomas Hobbes wrote Leviathan in 1651, in the middle of the English Civil War, and it shows. His starting point, the state of nature, imagines life with no state at all: no shared authority to enforce agreements, so trust between strangers collapses and everyone is, in effect, a potential threat to everyone else. Life there, in his famous phrase, is solitary, poor, nasty, brutish, and short. His answer is stark: people rationally trade away almost all their freedom to an absolute sovereign, his Leviathan, in exchange for protection from that chaos. Obedience, for Hobbes, is the price of not being murdered by your neighbour.

John Locke, writing a generation later, in 1689, just after England’s own less bloody second revolution, takes the same starting point and reaches a much less absolute answer. People still leave the state of nature and consent to government, but only to protect rights — life, liberty, and property — they already had beforehand, not to hand away everything. Locke’s consent is conditional: a government that seriously breaches that trust can be resisted and replaced. This is the version that echoes loudest in the reader’s own civic inheritance — it sits substantially underneath the American Declaration of Independence’s “life, liberty, and the pursuit of happiness,” and its claim that a government which becomes destructive of these ends can be altered or abolished by the people.

Jean-Jacques Rousseau, writing on the eve of the French Revolution in 1762, goes further still and asks for something closer to genuine collective self-rule. His general will is not simply the sum of individual preferences or a monarch’s decree — it is what a community, reasoning together about its shared good, would will collectively, and legitimate authority for Rousseau has to track that general will rather than any one ruler’s or faction’s private interest.

None of this is really about a historical event, and David Hume said so directly, aimed squarely at all three men just named. His 1748 essay “Of the Original Contract” argued that almost no government in recorded history actually began with anything resembling consent — conquest, usurpation, one faction imposing itself on another, then the arrangement simply persisting long enough to feel natural. What sustains obedience, on his account, isn’t a remembered agreement. It’s habit and custom. No one alive has ever signed a social contract, and by Hume’s own telling, no one ever really did. What Hobbes, Locke and Rousseau built instead is a justifying story — a way of explaining why obedience to a state should feel different from obedience to a gang with better weapons. It echoes the “great man” schooling from the previous section: hierarchy teaching obedience to betters before anyone can question it. Social contract theory is the grown-up, philosophically respectable version of the same move.

Bodin gave the state a who. This section has now given it a family of whys. What’s still missing is how that sovereign gets recognised by anyone else, and what it actually does, day to day, to make the obedience stick.

Nobody Is a State Until Somebody Else Agrees

Consent answered why a subject might owe a sovereign obedience. It said nothing about why any other sovereign should treat that state as real. Hugo Grotius, a Dutch jurist writing On the Law of War and Peace in 1625, mid-Thirty Years’ War, asked the mirror question to Hobbes’s: not why a subject obeys a sovereign, but what rules bind sovereigns to each other, when by definition there is no higher sovereign above them to enforce anything. His answer — a body of natural law that states owe each other regardless of religion or ruler — is usually credited as the founding move of international law.

The Peace of Westphalia in 1648, ending the Thirty Years’ War, is where that kind of thinking stops being a jurist’s theory and becomes an operating system. A cluster of treaties settled the war on a simple organising principle: each state, however small, controls its own territory and its own internal affairs, and no other power gets to override that from outside. For a thousand years the western idea of a single Christian order, with the Pope and the Holy Roman Emperor sitting notionally above every king, had been at least the official theory of how Europe worked. Westphalia buries it. Sovereign states, formally equal to each other regardless of size, dealing with each other through negotiation rather than a shared overlord, is still, in its basic shape, the system the world runs on.

Westphalia wasn’t the last word. It began a pattern of exactly this kind of resettlement, recurring every time the existing order broke down and had to be renegotiated by whoever came out on top. Paris in 1783 recognised American independence. Vienna in 1815 redrew Europe after Napoleon. Versailles in 1919 redrew it again after the first World War, alongside the League of Nations’ attempt to formalise the whole system into something permanent. The post-1945 settlement — the United Nations, Bretton Woods, the division of Germany — did it once more. Each is called a negotiation. Each was also a settlement shaped substantially by whoever had just won, presented afterward as agreement among equals.

And that system means something specific: statehood is a status other states agree to grant, not a fact a country simply has. You can control territory, govern a population, and run every function of a state, and still not be treated as one if enough other states decline to recognise you. Taiwan runs its own government, currency, and military, and is not recognised as a state by the United Nations because Beijing’s objection carries enough diplomatic weight to block it almost everywhere. Somaliland has functioned as a genuinely self-governing territory since 1991 and is recognised by no one. Kosovo is recognised by roughly a hundred countries and not by others, including several EU members.

That vote, historically, is decided by force first and diplomacy second — the same pattern as above, running through how individual states are made real, not just through how the whole state system periodically remakes itself. Most existing states, including the ones a reader would consider most secure and legitimate, were made real through conquest, war, or violent secession before they were ever made real through recognition — the United States through revolution and, in reaching its current borders, wars and forced removals against Indigenous nations and Mexico; Germany unified through a deliberate sequence of Prussian-led wars in the 1860s and 1870s; virtually every African and Middle Eastern state’s borders were drawn by European powers with a ruler and no local consultation whatsoever, then simply handed independence along those same lines.

Recent history shows every possible outcome. South Sudan seceded from Sudan in 2011 after a referendum and was recognised almost universally within days. Czechoslovakia’s “Velvet Divorce” the same decade split into the Czech Republic and Slovakia with no war, no referendum even, both recognised immediately — sitting deliberately next to Yugoslavia’s violent breakup in the same era, same fall of Soviet-bloc pressure, entirely different outcome, down to whether the parties involved chose to fight about it. The post-Soviet “frozen conflicts” — Abkhazia and South Ossetia, recognised only by Russia and a handful of its allies; Transnistria, recognised by no one but running its own currency, army and border for over three decades — are functioning, self-governing territories the world has simply declined to call countries. Nagorno-Karabakh took it one step further: it functioned as an unrecognised state for three decades, then in 2023, after Azerbaijan’s military campaign, it didn’t just lose a war, it ceased to exist. No recognition, and no powerful backer willing to make recognition stick by force, meant there was nothing to fall back on. Palestine is recognised by roughly three-quarters of UN member states but not by the US, the UK, or most of western Europe, a live and contested case rather than settled history.

The Islamic State’s “caliphate,” 2014 to 2019, is the sharpest test of the whole system. It met more of the functional tests of statehood than most of the cases above — it controlled a defined territory with millions of people in it, ran courts, collected taxes, issued its own currency, administered schools and utilities. For years it had something close to a working monopoly of force within its borders. It was recognised by no state on earth, treated universally as a criminal, terrorist entity, and eventually destroyed by an international coalition rather than negotiated with. Recognition was never really about whether the functional boxes get ticked. It is a judgment about legitimacy, made by existing states, and they can withhold it completely even from something that otherwise looks like a state.

The Kurds connect hardest to what comes at the end of this piece. Somewhere around thirty to forty million people, the largest stateless nation on earth, spread across Turkey, Iraq, Syria and Iran, with genuinely functioning autonomous governments in parts of northern Iraq and northern Syria, and no realistic path to a state of their own, because doing so would require four different existing states to simultaneously agree to give up territory, which none of them will.

Everything above has a mirror image: the failed state. Every case above is a place with the practical substance of a state and no recognition. A failed state is the reverse — full, unquestioned legal recognition, a UN seat, a flag, and almost none of the actual internal capacity that is supposed to come with it. Libya since the 2011 uprising and NATO intervention has kept its UN seat and its flag while splitting into rival governments in Tripoli and Benghazi, neither commanding a real monopoly of force across the whole territory. International relations scholars call this distinction juridical sovereignty — the recognition, the paperwork — versus empirical sovereignty — the actual capacity to govern. Recognition can be entirely real and entirely hollow at the same time.

A border was always kinship’s question — who’s in, who’s out — answered at continental scale and enforced with the state’s own monopoly of force rather than a clan’s collective memory of who’s owed what. The mechanism hasn’t changed since the very first section of this piece. Only the size of the group doing the drawing, and the weapons — military, economic, ideological or otherwise — doing the enforcing, have.

The Monopoly That Makes It Real

Everything so far has been building toward one sentence. The German sociologist Max Weber, in a 1919 lecture, defined the state as the human community that successfully claims the monopoly on the legitimate use of physical force within a given territory. Nearly every word in that definition does real work.

Monopoly means exactly what it sounds like: everyone else’s violence gets defined as illegitimate by comparison.When an individual hits someone, it’s assault. When the state imprisons that same person for it, that’s law enforcement. The physical act can look similar; what separates them is that only one of the two parties is allowed to use force and call it something other than a crime. Every other actor with the capacity for violence — a private militia, a criminal gang, a corporation’s security detail — is, by definition, either operating with the state’s specific permission or is something the state is actively trying to suppress.

Legitimate is the word carrying the most weight, and the one that connects back to everything already covered. A monopoly on force alone just makes the state the biggest gang — Hobbes’s nightmare with better weapons. What makes it a state rather than an occupying army is that enough of the population, most of the time, without being forced to at gunpoint, accepts that the state’s use of force is rightful rather than merely powerful. Obedience that runs on genuine acceptance looks identical, from the outside, to obedience that runs on simple fear, right up until the moment it’s tested.

Territory is the previous section’s border, doing double duty. The monopoly only has to be total inside the line already drawn — which is why recognition and borders had to come before this, not after it.

Most days, that monopoly is invisible precisely because it’s working. Law is the routine, low-intensity face of a threat that almost never has to be carried out. A speeding fine, a contract dispute, a planning permission refusal: none of these involve anyone pointing a weapon at anyone else, and that’s the point. Law is what a monopoly on violence looks like when it doesn’t need to use itself — a written threat credible enough that compliance, most of the time, makes the actual force unnecessary.

Army and police are the same monopoly, split by direction. The army exists to defend, or extend, the state’s monopoly against other states’ monopolies — territory, defended. The police exist to maintain that same monopoly against threats from inside the state’s own population. Most democracies keep this split rigid — a soldier is not supposed to police their own citizens. Hold that thought. We will come back to this.

Threat, internal or external, is the justification the state reaches for whenever it wants to expand what the monopoly is allowed to do — and that’s true of functioning democracies as much as anything more sinister. An external threat, war or terrorism, justifies a bigger army, more surveillance, tighter borders. An internal threat, crime or unrest, justifies a bigger police presence, broader emergency powers, more intrusive law. Both are frequently genuine. Both are also, historically, the single most reliable lever for a state to acquire powers it wouldn’t otherwise be granted.

Weber’s definition is best tested against the cases the previous section already introduced. A state can lose the actual capacity Weber describes while keeping every formal trapping of it, the way a failed state does. Or it can hold the capacity with none of the formal recognition — the Islamic State’s brief, real monopoly of force over territory and people, backed by no recognition at all. Weber’s monopoly and Grotius and Westphalia’s recognition are two separate achievements. A state, in the fullest sense, needs both.

With the monopoly defined, the obvious next question is who gets to hold it, and how they’re chosen.

An Election Is Not a Democracy

“Democracy” is a word regimes with almost nothing else in common have all reached for. The German Democratic Republic — East Germany — held elections. So did the Soviet Union, for its Supreme Soviet. North Korea’s Supreme People’s Assembly is elected too, with results regularly reported above 99% turnout and near-unanimous approval. The word sits in the formal name of states that look nothing like each other in practice. That isn’t an abuse of language. It’s the same efficiency this piece has been describing since its first section. David Hume put the underlying claim plainly in 1741: all government, however despotic, rests ultimately on opinion, because no ruler in history has ever had enough soldiers to hold down an unwilling population by force alone. Coercion is expensive — a state that can only make people obey by watching them do it needs eyes everywhere, all the time, which no state has ever actually had. A population that believes its own obedience is owed, rather than merely extracted, polices itself, and polices its neighbours too, multiplying the state’s real reach without multiplying its actual force by a single soldier.

That’s the practical payoff underneath everything already shown. Kinship needing a shared sense of who belongs to whom. The schooling that teaches hierarchy as natural before anyone can question it. The philosophers building a story that made obedience feel chosen rather than imposed. Weber’s own insistence that legitimate authority has to be believed in, not merely feared. Every one of those was the same need, arriving in a different century’s clothing. Democracy, however twisted a given regime’s version of it, is history’s fullest working answer to that need — the most complete mechanism yet built for manufacturing belief in rightful rule at scale. The election is simply its visible proof, the signifier standing in for something built out of far older and far deeper material.

An election, held on its own terms, is a performance of consent — and a performance can be staged as easily as it can be genuine. A regime that holds one, however tightly controlled the candidate list or predetermined the outcome, gets to say afterward that the people chose this. That claim persuades even when the choice on offer was never real, because “we won an election” sounds, from outside, indistinguishable from “we won a free and fair one.”

Montesquieu’s split, already named, isn’t one design. It’s a family of them, and the differences between versions matter more than the shared vocabulary suggests. England fuses two of the three: the prime minister sits inside the legislature, drawn from and answerable to Parliament directly, with no separate presidential election at all — the Westminster model, later exported to Australia and dozens of former colonies. The American version does the opposite on purpose: a president elected entirely separately from Congress, deliberately able to be from a different party than the legislature controls, with the judiciary appointed for life specifically to insulate it from either. Argentina runs a presidential system too, close on paper to the American template — and its repeated economic crises show what happens when a strong, separately-elected executive collides with a legislature and judiciary too weak, or too crisis-battered, to actually check it.

So the real question was never whether a state holds elections. It is whether the election actually constrains the people it elects. There is a serious research framework built for exactly that gap, rather than picking examples by feel. The V-Dem Institute, at the University of Gothenburg, has tracked roughly 180 countries for two decades using a large panel of country experts rather than any government’s own account of itself, and splits the results into four bands. Closed autocracy: no meaningful multiparty elections at all. Electoral autocracy: elections happen, multiple parties technically appear on the ballot, but the press freedom, freedom of association, or fair competition that would make the choice meaningful are missing or suppressed. Electoral democracy: elections are genuinely free, fair and contested. Liberal democracy: electoral democracy plus enforced individual rights, equality before the law, and an executive actually constrained day to day by the legislature and courts, not just on paper.

Band	What defines it	Anchor examples
Closed autocracy	No meaningful multiparty elections at all	North Korea, Afghanistan, Eritrea, Turkmenistan
Electoral autocracy	Elections exist on paper; press, opposition, or fair competition are suppressed enough to make the result not genuinely contestable	Russia, Belarus, Venezuela, Cambodia, Egypt, Pakistan
Electoral democracy	Elections genuinely free, fair and contested; rights protections or real executive constraint fall short	India, Brazil, Indonesia, Argentina, South Africa, the United States, as of the most recent assessment
Liberal democracy	Electoral democracy plus enforced rights and an executive genuinely constrained in practice	Canada, Switzerland, Botswana, Australia, New Zealand, the Nordic states, Uruguay, Japan, Germany

Two things matter more than any single row in the table. Movement between bands runs in both directions and never stops — the most recent global data shows considerably more states sliding down than climbing up in recent years. And the top band is genuinely small: liberal democracies, by this measure, now cover well under a tenth of the world’s population.

Switzerland and Botswana. Switzerland runs on a seven-person collegial executive with a rotating ceremonial presidency, real power held by its cantons rather than the federal centre, and the heaviest use of binding citizen referendums anywhere in the world — a genuinely different shape of democracy from the single-executive model most of the examples above share. Botswana has held regular, genuinely contested multiparty elections without interruption since independence in 1966, and has managed, even with recent challenges, its diamond wealth through relatively transparent, accountable institutions rather than the resource-curse pattern that has destabilised many mineral-rich states elsewhere on the continent. Neither is an outlier picked to make a point. They’re two working examples of the small category the point is about, sitting alongside the Nordic states, Uruguay, Canada, Australia and New Zealand, Japan and Germany.

This is not a story about safely distant countries. The same research institute’s most recent report reclassified the United States itself, for the first time in over fifty years, out of the liberal democracy band and down into electoral democracy: elections still genuinely free and contested, but the executive-constraint and rights-protection criteria no longer fully met by the researchers’ own measure. Not a partisan claim. A research institute’s own methodology, applied identically to all 180 countries, moving one particular country down a tier. The United States now sits in the same band as Brazil, Argentina and India — three large, genuinely contested democracies, each pulled into that middle tier by a different mechanism entirely, worth returning to properly elsewhere rather than compressed into a single sentence here. The United Kingdom remains classified a liberal democracy but is one of the countries the same report newly flags as moving in the wrong direction, alongside the United States, Italy, Slovakia, Slovenia and Croatia — still in the top band, trending out of it.

Movement down the table, one plausible-sounding step at a time, without anyone ever announcing that democracy has ended, is exactly what comes next.

Methodology note: this piece was written collaboratively between a human and an AI — the human providing the instincts, provocations, editorial judgement and voice; the AI providing research synthesis, intellectual scaffolding and drafting. Full method at [athomehefeelslikeatourist.blog].