Who Should Rule?

Two Questions

Around 330 BCE, in his Politics, Aristotle sorted the constitutions of the Greek world by two questions: how many ruled, one, a few or the many, and for whose good, the common good or the rulers' own. The answers gave six forms. Rule by one was kingship or, corrupted, tyranny. Rule by the few was aristocracy when the best governed for all and oligarchy when the rich governed for themselves. Rule by the many was polity, which blended the claims of the rich and the poor, or, corrupted, democracy, the rule of the poor in their own interest. He then complicated the scheme: what made a constitution oligarchic was not the number of its rulers but the principle on which they held power, so that wherever the rich ruled because they were rich there was an oligarchy, even if they were many.

The questions were older than Aristotle. Herodotus staged a debate among the Persian nobles who had overthrown a usurper in 522 BCE, in which Otanes spoke for the many, Megabyzus for a council of the best men, and Darius, who won the argument and then the throne, for monarchy. Nearly two centuries after Aristotle, Polybius, a Greek hostage in Rome, set the six forms in motion: each good form decays into its corruption, and each corruption provokes the next form, in a cycle that begins again. Only a mixed constitution could slow the decay, and Polybius found one in the Roman Republic, where the consuls, the Senate and the assemblies stood for monarchy, aristocracy and democracy and checked one another.

The scheme outlived the cities that produced it, from Cicero and Thomas Aquinas, who commended mixed government, to Montesquieu and the authors of The Federalist. It survived because every state must settle who makes its laws and how its rulers can be restrained or replaced. This essay follows his three answers and the forms they do not easily fit, then sets them beside the record of 5,468 states in the Institute's register.

Rule by One

Through most of recorded history the commonest answer to Aristotle's first question was one, and writers on monarchy asked less whether one should rule than how to tell lawful rule from its counterfeit. They applied four tests: purpose, title, law and, in Rome, time.

Purpose was the hardest to judge, since almost every ruler claimed to serve the common good; in China the doctrine of the Mandate of Heaven left the judgement to Heaven, which withdrew the right to rule from a wicked or incompetent ruler. Title was easier. The Greek word tyranny first described how power had been won, not how it was used: Cypselus became tyrant of Corinth by overthrowing the Bacchiad clan, traditionally in 657 BCE, and some tyrants, such as Peisistratus at Athens, were remembered for good government. The philosophers nonetheless made the tyrant the type of the worst ruler, and in the 1350s the jurist Bartolus of Sassoferrato separated a tyrant by defect of title, who seized power he had no right to, from a tyrant by conduct, a lawful ruler who governed unjustly.

Law divided early modern Europe. Jean Bodin, writing during the French Wars of Religion, argued in 1576 that sovereignty, the absolute and perpetual power of a commonwealth, must be undivided, and Thomas Hobbes drew the same lesson from the English Civil War. The absolute monarchy they defended reached its classic form under Louis XIV and its most explicit statement in Denmark–Norway, whose King's Law of 1665 set the king “above all human laws”. Its defenders still denied that absolute meant arbitrary: Bishop Bossuet distinguished absolute government, which no subject might lawfully resist, from arbitrary government, which disregarded law and property. John Locke answered that “wherever law ends, tyranny begins”, and Montesquieu that a monarchy which destroyed the intermediate powers, such as the nobility and the parlements, would become a despotism.

England answered with a limited crown. The Bill of Rights 1689, enacted after the Glorious Revolution, forbade the monarch to suspend laws, levy taxes or keep a standing army in peacetime without Parliament's consent, and constitutional monarchy came to mean a crown that reigned while ministers answerable to Parliament governed, keeping, in Walter Bagehot's words, “the right to be consulted, the right to encourage, the right to warn”. Its critics attacked heredity itself: Thomas Paine thought hereditary legislators as absurd as a hereditary mathematician. An older alternative to heredity was elective monarchy, from the seven prince-electors of the Holy Roman Empire, fixed by the Golden Bull of 1356, to the nobles of the Polish–Lithuanian Commonwealth, who elected their kings from 1573 to 1764 and bound each one by the Henrician Articles. Election constrained rulers but invited bribery and disputed successions.

Rome added the test of time. The dictator of the Roman Republic was a lawful magistrate, named by a consul for a single task and for at most six months; according to Livy, Cincinnatus laid down the office in 458 BCE after sixteen days, and Machiavelli defended it in his Discourses on Livy because it was conferred for a limited time and a limited purpose. Sulla revived it in 82 BCE to rewrite the laws, and Julius Caesar was made dictator for life shortly before his murder in 44 BCE. The modern word kept the Roman name and dropped the limits. Unlike purpose, title, law and time could be written into institutions: rules of succession, courts and fixed terms of office.

Rule by the Few

Aristotle's line between aristocracy and oligarchy was easier to draw in theory than in practice, because the few were nearly always a hereditary class whose claim to excellence rested on land and wealth. The Bacchiadae of Corinth, a clan that married only among themselves, are counted among the first aristocracies, yet Herodotus called their regime an oligarchy, and the historian Ronald Syme held that whatever the form of government “an oligarchy lurks behind the façade”.

The longest-lived oligarchy was the Republic of Venice, whose Great Council the Serrata of 1297 began to close to new families. The patriciate guarded the state against its own members: from 1268 the doge was chosen in ten rounds that alternated drawing lots with voting to defeat bribery, and in 1355 the Council of Ten had Doge Marino Faliero beheaded for plotting to make himself lord of the city. No revolution overthrew the patriciate before the Great Council voted the Republic out of existence under Napoleon's ultimatum in 1797, although most of its people had no political voice.

Defenders of the few argued that property and leisure fitted men to govern and made them a check on kings and crowds, and even Thomas Jefferson granted that there was a “natural aristocracy” grounded in “virtue and talents”, though he wanted it chosen by election rather than by birth. Critics pointed to the costs of privilege, as in Poland–Lithuania, where from 1652 the liberum veto let a single deputy break up the Sejm.

Theocracy claims to place rule beyond human hands. Josephus coined the word in the 90s CE for a constitution “placing all sovereignty and authority in the hands of God”, but God's law needs interpreters, and theocracies have been governed by a few: the prelates of the Papal States from 756 to 1870, the government of the Dalai Lamas in Tibet from 1642 to 1959, and, under Iran's constitution of 1979, a Supreme Leader chosen by an assembly of Islamic jurists. Defenders held that laws given by God did not change with the whims of rulers; critics, that a ruler who speaks for God cannot be voted out.

Two forms of rule by the few belong above all to the twentieth century. A military junta governs as a committee of officers, usually after a coup. The first clear case came in the Korean kingdom of Goryeo, whose officers seized power in 1170, and after 1945 juntas governed much of Latin America, Africa and Asia. Samuel P. Huntington argued that officers could act as reformers where landed oligarchies held power; Argentina's junta of 1976–1983 left almost 9,000 disappearances documented in the Nunca Más report, and up to 30,000 by the estimate of human rights groups. In a one-party state the few are the leadership of a party, a model that Lenin's Bolsheviks built in Russia between 1918 and 1921 and that Julius Nyerere defended for new nations, where, he argued, rival parties would only foster faction, while Milovan Djilas, once a leader of communist Yugoslavia, charged in 1957 that the party bureaucracy had become a new privileged class. Barbara Geddes, who divides dictatorships into personalist, military and one-party regimes, found the one-party kind the most durable and the military the shortest-lived.

Behind all these forms lay a harder question: whether rule by the few is one type of government or a condition of all of them. In 1911 Robert Michels, studying Germany's Social Democratic Party, proposed an “iron law of oligarchy”, by which any large organisation falls to its leaders: “Who says organization, says oligarchy.” His critics answered with exceptions, from a printers' union that Seymour Martin Lipset and his colleagues found still democratic in 1956 to classical Athens, which Josiah Ober has argued disproves the law.

Rule by the Many

The first well-documented direct democracy was that of Athens, from the reforms of Cleisthenes in 508–507 BCE until 322 BCE, when the Macedonian regent Antipater limited citizenship to men owning at least 2,000 drachmas. An Assembly open to all adult male citizens decided war and peace by show of hands, and the Council of Five Hundred, the juries and most magistrates were chosen by lot, which Aristotle reported was commonly thought democratic, as election was thought oligarchic. The citizens were a minority: in the fourth century BCE perhaps 30,000 adult male citizens lived beside some 10,000 resident foreigners and many thousands of slaves, and women had no political rights.

In the funeral speech that Thucydides gives to Pericles, the Athenians regard a man who takes no part in public affairs not as quiet but as useless, and Aristotle granted that the many may judge better together than a few experts, as a feast to which many contribute is better than one provided by a single host. Plato, whose teacher Socrates was condemned to death by an Athenian jury in 399 BCE, compared the democratic city to a ship whose crew ignore the true navigator, and Thucydides blamed Athens's defeat by Sparta in part on the Assembly's enthusiasm for the Sicilian Expedition.

Rome offered another model. The republic, res publica, the “public thing” of its citizens, had no hereditary monarch and was run by magistrates elected for a year and checked by colleagues and tribunes. It was not the same as democracy: Montesquieu called a republic aristocratic when only part of the people held supreme power, and most republics before the nineteenth century, from Rome to Venice and the Dutch Republic, were aristocratic in his sense. Montesquieu and Rousseau thought republics workable only in small states, and Rousseau held that sovereignty could not be represented. The founders of the United States answered with representation: James Madison argued in The Federalist No. 10 that a large representative republic would control faction better than a small democracy, and Bernard Manin has argued that the founders valued election, unlike the Athenian lot, as a way of choosing a distinguished few.

Madison's republic was also a compound one. The Constitution that took effect in 1789 made the United States the first modern federation, dividing power between a federal government and the states, neither able to abolish the other, and so giving, in Madison's words, “a double security” to the rights of the people. Federations now hold about 40 per cent of the world's population. Their critics point to secession, from the American Civil War to Biafra, and to the shelter that the rights of states have given to injustice.

Representative democracy took two main designs. In a parliamentary democracy, which grew up in Britain between the ministry of Robert Walpole (1721–1742) and the 1830s, the government is drawn from the legislature and stays in office only while a majority there supports it. In a presidential democracy, devised at the Constitutional Convention of 1787, a separately elected president serves a fixed term and each branch checks the others. In 1990 the political scientist Juan Linz argued that the presidential design endangers democracy, because president and legislature both claim the people's mandate and fixed terms make government rigid, pointing to the coups in Brazil in 1964 and Chile in 1973. José Antonio Cheibub answered that presidential democracies had died more often mainly because they had been adopted in countries already prone to military intervention. Parliamentary systems have had failures of their own, such as the Weimar Republic, whose Reichstag passed the Enabling Act in 1933. Arend Lijphart's comparison of thirty-six democracies turned the debate to the contrast between majoritarian and consensus systems.

Beyond One, Few and Many

Aristotle's questions assumed a city of citizens who could be counted. Many forms of government answer a different question, how power is held together across kinship, distance and difference, and the political scientist S. E. Finer classified states instead by where power lay: in the palace, the nobility, the church or the forum, or in mixtures of these.

A chiefdom, in the anthropologist Robert Carneiro's definition, unites “a number of villages or communities under the permanent control of a paramount chief”. It is rule by one without salaried officials, police or written law: the chief ruled through kinship, ritual and tribute, and held his followers by generosity in feasts and gifts. Chiefdoms arose independently from Mesopotamia, where many archaeologists place the first in the Ubaid period, to the Americas, where Cahokia grew after about 1050 into the largest chiefly centre north of Mexico, and to Polynesia, where Kamehameha I ruled all the Hawaiian Islands as king by 1810.

The nomadic confederations of the Eurasian steppe joined herding tribes under a ruling clan and its khan. The anthropologist Thomas Barfield has called them “imperial confederacies”, autocratic towards the outside world but consultative and federal within, held together by the wealth their rulers drew from China by raiding, tribute and trade, a pattern that the Xiongnu set after 209 BCE. Assemblies of princes and chiefs, such as the kurultai that proclaimed Genghis Khan in 1206, raised their rulers and could also split an empire: in 1260 rival kurultais elected Kublai Khan and Ariq Böke, and the Mongol Empire broke into four realms. Ibn Khaldun argued in the Muqaddimah of 1377 that such peoples, bound by strong group feeling, conquer settled realms and then lose that cohesion within a few generations.

The city-state is defined by scale, not by the number of its rulers: in Mogens Herman Hansen's definition, “a self-governing, but not necessarily independent, political unit”. It could take almost any constitution. Lords and kings ruled the cities of Sumer, holy lords the cities of the Classic Maya, narrow oligarchies or the whole citizen body the Greek poleis, and elected consuls and councils the Italian communes. Aristotle thought smallness a condition of citizenship, since citizens had to know one another's characters; Alexander Hamilton, arguing in The Federalist for a strong union, recoiled from the “distractions” that had continually agitated “the petty republics of Greece and Italy”.

Feudalism was monarchy in form and divided rule in fact. In the historian François-Louis Ganshof's narrow definition it bound lord and vassal by homage and fealty, the lord granting a fief, most often land, in return for military service and counsel; in practice rights of justice, taxation and command passed from kings to counts, castellans, bishops and abbots. Adam Smith judged that under feudal law the authority of government remained “too weak in the head and too strong in the inferior members”, yet Joseph Strayer saw in the feudal monarchies of England and France the cradles of the modern state, and the vassal's duty of counsel grew into the assemblies from which parliaments developed. Since the critiques of Elizabeth A. R. Brown in 1974 and Susan Reynolds in 1994, historians have debated whether the word fits medieval Europe at all.

Colonial rule lay outside Aristotle's frame altogether: government by a distant state, in pursuit of interests “often defined in a distant metropolis”, in Jürgen Osterhammel's words, over people who were subjects rather than citizens, ruled directly by colonial officials or indirectly through local rulers, as Frederick Lugard prescribed in 1922. Aristotle himself was cited in its defence: at the Valladolid debate of 1550–1551 Juan Ginés de Sepúlveda justified the Spanish conquest with Aristotle's theory of natural slavery, against Bartolomé de las Casas, who had denounced the destruction of the Caribbean peoples. In each of these forms the decisive question was less how many ruled than who belonged.

What the Register Shows

The Institute's register records 5,468 states, each with its dates and a recorded kind: kingdom, empire, republic, city-state, chiefdom, colony and others. It is a register of states, not of constitutions. Its kinds say what a state was called, not how it was governed: a city-state may have had a king and a kingdom may have been governed by its parliament, while a kingdom that became a republic usually appears as two states. Its coverage is uneven, fullest where historians have mapped many small polities, as in ancient Greece, the Holy Roman Empire and the princely states of India, and thinnest for the modern world, where it records fewer states than exist: 188 in 2020, fewer than the 193 members of the United Nations.

In the twenty-fifth century BCE the register's city-states, from Uruk and Lagash in Sumer to Mari and Byblos further west, outnumber its monarchies by 16 to 5, although most of them were ruled by lords or kings: the count measures the scale of states, not their constitutions. From the seventh century BCE to the second, republics and city-states outnumber monarchies in every century, by 176 to 70 in the fifth, most of them Greek poleis whose constitutions ran from the two kings of Sparta to the democracy of Athens. In the third century BCE the largest group is neither: 169 chiefdoms and confederations, most of them the Gauls, Thracians, Illyrians and other peoples of Iron Age Europe. Rome's conquests show next. Between the second century BCE and the first, the republics and city-states fall from 111 to 27, and the chiefdoms and confederations from 151 to 77.

From the third century CE to the twentieth, monarchies are the largest group in every century, and they peak at 837 of the 1,277 states alive in the middle of the eighteenth century, a peak that owes more to South and Southeast Asia, with its princely states and sultanates, and to Africa than to Europe. Between the eighteenth century and the nineteenth the church states fall from 69 to 16 and the republics and city-states from 88 to 40, as Venice, Genoa and the Dutch Republic vanished and the Imperial Recess of 1803 secularised almost all the church states of the Holy Roman Empire and abolished most of its free cities. Colonies, 14 in the fifteenth century, reach 115 in the eighteenth and 111 in the twentieth. Between the middle of the nineteenth century and the middle of the twentieth, colonial conquest, revolution and the end of most of the princely states cut the monarchies from 760 to 155, and in 2020 the register counts 96 republics and city-states against 47 monarchies: the first of its century counts since the second century BCE in which republics and city-states outnumber monarchies.

Lifespans give no simple verdict. Among states with a recorded start and end, the median monarchy lasted 205 years and the median republic or city-state 200. Church states, most of them prince-bishoprics whose rulers were elected rather than born to rule, lasted longest, with a median of 449 years; chiefdoms and confederations lasted 249, and colonies and protectorates only 64. The register also records for each state a mode of rule, which cuts across its kinds: feudal for 2,059 states, tribal for 941, bureaucratic for 922, mercantile for 560, colonial for 449, theocratic for 295 and nomadic for 125.

What Survives

Some ideas from the long argument proved durable. Polybius's mixed constitution passed through Cicero and Machiavelli to Montesquieu, who held up England's separation of legislative, executive and judicial power as the model of political liberty, and from him into the checks and balances of the Constitution of the United States and many constitutions since. The Roman idea of lawful emergency power survives in emergency clauses such as Article 48 of the Weimar Constitution and Article 16 of the French constitution of 1958. Even the Athenian lot has returned, in citizens' assemblies drawn at random in British Columbia in 2004 and in Ireland from 2016.

Political scientists now judge governments less by the number of their rulers than by how rulers are chosen, checked and replaced. Karl Popper defended democracy in 1945 as the system that lets citizens remove their rulers without bloodshed, and studies of dictatorship since Geddes find that whether power lies with a person, an army or a party predicts how a regime lasts and how it ends. The evidence rarely runs one way: Amartya Sen observed that no substantial famine has ever occurred in a democracy with a relatively free press, while studies of economic growth find that dictatorships on average grow about as fast as democracies, though with larger swings. Historians, for their part, have grown wary of the labels: some have called absolutism a myth, feudalism a construct and the chiefdom an archaeological delusion. The forms in the record were seldom pure: Venice combined a doge, a senate and a great council, and Britain kept its crown and became what H. G. Wells called a “crowned republic”.

Aristotle's first question, how many should rule, has been answered in every possible way, and the record gives no clear verdict among the answers: the median monarchy in the register lasted 205 years, the median republic or city-state 200. His second question, for whose good, has proved the more lasting, because no form of government settles it once and for all. Each constitution has had to answer it again in its own terms: in who may vote and stand for office, how long an office lasts, who may overrule the ruler, and how a government that has lost its support gives way to the next.