Beyond our Island Republic
Democracy, in the narrow and honest sense of the term, is a procedure. It is a method by which a community selects the persons who will occupy public office, together with a rule, in practice the rule of the numerical majority, by which disagreements about the use of that office are settled. Everything else the word has come to carry, liberty, equality, prosperity, human dignity, the will of the people, is cargo loaded onto the procedure by two centuries of rhetoric, and none of it follows from the mechanism itself. The claim I make in this essay is that democracy is a decision procedure that has been mistaken for a moral destination, that its apparent successes in the advanced Western states are in truth the successes of institutions that are not democracy operating beneath it, that in the Maldives the procedure was imported in 2008 without those institutions and has produced, in under two decades, precisely what the theory of the procedure predicts, and that the remedy for a small, dispersed, clientelist type nation is not a purer democracy but a smaller one. I would try to introduce an argument in stages, first tracing how the procedure evolved and how it came to dominate the modern world, then assembling the philosophical and economic case against it, then applying that case to the Maldivian condition specifically, and finally proposing a constitutional design, a decentralised state of self-governing atolls beneath a small central government, the whole standing under a constitutional monarch who holds oversight and the integrity institutions but no executive power, together with the pillars on which the design rests and the route by which it could be reached.
Hereinafter when I use the term democracy, it refers to mass electoral democracy under universal adult suffrage, the form adopted by nearly every state in the world after 1918 and by the Maldives in 2008, unless specified otherwise. The direct assembly of Athens, the deliberative council of an island, and the shareholder meeting of a company are all majoritarian devices, but they are not the object of this essay except where they illuminate it.
It is essential for my argument to fix, at the outset, the distinction on which everything that follows will turn. One must separate democracy considered as a method from democracy considered as an end. As a method, it is one candidate among several for answering a practical question, namely how a society should select and dismiss its rulers without bloodshed, and like any method it can be evaluated by its results and compared against alternatives. As an end, it is beyond evaluation, since an end is the standard by which other things are evaluated, and this is exactly the status democracy has acquired in the modern mind. The word has migrated from the vocabulary of mechanisms into the vocabulary of morals. To question the method is now heard as an assault on the end, which is roughly as sensible as hearing a critique of a particular bridge design as an attack on the idea of crossing rivers. The whole of this essay is an exercise in keeping the two apart, in asking of the method the questions one asks of any method (what does it optimise, what does it cost, compared to what), while leaving the genuine ends, peace, liberty, prosperity, the restraint of power, exactly where they stand.
The procedure is old, and so is its critique. Fifth-century Athens governed itself through a direct assembly of citizens, selection of most offices by lot, and majority vote on matters up to and including capital punishment and war. It should be noted that the two most careful minds that ever observed that system at close range both rejected it. Plato watched the assembly of his city vote to execute Socrates, and in Book VIII of the Republic he set out the sequence he believed democracy must follow, an escalating auction of freedoms and gratifications, the erosion of every authority that might discipline appetite, and finally the arrival of the demagogue who rides the auction to tyranny (Republic, 562a ff.). Aristotle, the more empirical of the two, classified constitutions in Book III of the Politics by asking two questions, who rules, and in whose interest. Rule by the many in the common interest he called politeia and counted among the correct constitutions. Rule by the many in the interest of the many, that is, of the poor majority as a faction, he called demokratia and counted among the deviant ones (Politics, 1279a–b). The distinction is worth a thought or two. Aristotle did not object to the many having power. He objected to the interest of a majority being mistaken for the interest of the whole, and he saw, 23 centuries before the term rent-seeking was coined, that a majority is perfectly capable of behaving as a faction, and that when it does, its numbers make it the most dangerous faction of all.
Rome answered Athens with the mixed constitution, consuls, senate, and popular assemblies arranged so that no single principle, the one, the few, or the many, held the whole. Polybius, writing in the second century BCE, credited this mixture with Rome's stability and described the cycle (anacyclosis) through which unmixed constitutions degenerate, kingship into tyranny, aristocracy into oligarchy, democracy into mob rule and thence back to a strongman. The Roman lesson, absorbed by every serious constitutional thinker down to the eighteenth century, was that the popular element is a necessary ingredient of good government and a fatal recipe for it when unmixed. Then followed the long monarchical interlude, thirteen centuries in which the European state was, in effect, a piece of property, inherited, husbanded, occasionally ruined, and it is a fact of some importance for what comes later in this essay that the men who built the modern doctrines of liberty, the rule of law, natural rights, the sanctity of contract and property, built them under monarchs and against monarchs, not under democracies. The rights of Englishmen were asserted against kings and were centuries old before any Englishman of ordinary means could vote.
The moderns who reintroduced the popular principle did so with visible anxiety. Locke grounded government in consent but hedged it with natural rights that no majority could touch. Rousseau supplied the doctrine that would do the most damage, the general will, the notion that beneath the noisy disagreement of actual individuals there exists a single, discoverable will of the people, of which the state can be the instrument. I shall return to this notion, because it is the philosophical load-bearing wall of modern democracy and it happens to be false in a sense that can be demonstrated mathematically. The American founders, for their part, used the word democracy almost exclusively as a term of abuse. Madison, in Federalist No. 10, distinguished a democracy, in which the majority governs directly, from a republic, in which representation, federal structure, and enumerated powers filter and fracture the majority precisely because "measures are too often decided, not according to the rules of justice and the rights of the minor party, but by the superior force of an interested and overbearing majority." The machinery of the American constitution, the separation of powers, the Senate, judicial review, the Bill of Rights, is a machine built by men who expected majorities to behave badly and designed accordingly. The nineteenth century then ran the experiment the founders feared. The franchise widened decade by decade, and the two most perceptive observers of the widening, Tocqueville and Mill, converged on the same warning, that the new despotism would not be a man but a number, that the tyranny of the majority would operate through statutes, through opinion, through conformity, and the slow disciplining of every dissenting mind (Democracy in America, vol. I; On Liberty, ch. I). They were not enemies of the democratic age. They were its most honest friends, and they described its failure mode before it had fully arrived.
The decisive break came in 1918. Hans-Hermann Hoppe, whose Democracy: The God That Failed (2001) I will engage at length below, dates the modern era precisely there. Before the First World War, most of the world's states were monarchies of one form or another, and the democratic republics were exceptions. After 1918, with the Habsburg, Hohenzollern, Romanov, and Ottoman crowns swept away at a stroke, the democratic republic became the default form of the Western state, and after 1945, and again after 1991, it became the default form of the state as such, exported by decolonisation, imposed by victory, and certified by international institutions. Whatever else one says of this transformation, one should be clear that it was not the result of a controlled comparison. Democracy did not win a laboratory trial against the alternatives. It won two world wars and a cold one, and the winners wrote not only the history but the political science.
An honest critique must begin by explaining the dominance it proposes to criticise, and the explanation cannot be that the entire world was simply mistaken. Democracy became the prime system of the advanced West for reasons, and some of the reasons are good. I shall state the three strongest, because the argument that follows is only as valuable as the case it answers.
The first and best reason is the one Ludwig von Mises gave in Liberalism (1927), and it is characteristic of Mises that his defence of democracy is the most durable one ever written while claiming almost nothing for it. Democracy, Mises argued, has one function, and one only, that matters. It makes possible the peaceful adjustment of government to the wishes of the governed. Every other system settles the question of succession, in the end, by violence or the threat of it. Elections are a substitute for civil war, a mechanism by which the faction that would probably win a fight is identified by counting instead of fighting, so that the fight itself becomes unnecessary. This is a real achievement and I do not intend, anywhere in this essay, to belittle it. It should be understood, however, what a narrow achievement it is. Mises's argument establishes democracy as a technique for changing rulers without bloodshed. It establishes nothing about the quality of what those rulers will do, and Mises himself insisted that the value of the technique depends entirely on the framework, a liberal order of private property and limited government, within which it operates. The peaceful-succession argument is an argument for elections. It is not an argument for majoritarian control of economic and social life, and the whole pathology of the modern democratic state lies in the distance between those two things.
The second reason is a conflation, and it is the most consequential conflation of the twentieth century. The democracies of the West were also, and first, market economies under the rule of law, and they grew rich. The socialist states were also, and first, command economies, and they stagnated and collapsed exactly as Mises's calculation argument (in Socialism, 1922) said they must, because without market prices for capital goods there is no way to compute what anything costs, and an economy that cannot compute costs destroys wealth with every decision it makes. The West's victory in that contest was a victory of property and prices over planning. It was recorded, in the popular mind and in a good deal of the academic mind, as a victory of democracy over dictatorship. The two claims are not the same, and the difference between them is the difference between Singapore and Venezuela. Milton Friedman spent a career on precisely this point. In Capitalism and Freedom (1962) he argued that economic freedom is not a by-product of political freedom but its precondition, that the market is itself a system of representation more refined than any ballot, since each person "votes" with every unit of currency and receives, in what he called the proportional representation of the market, exactly what he voted for rather than what 51% of his neighbours voted for. A ballot compresses a million preferences into one winner, while a price answers each preference individually. The customer in a market chooses among alternatives and bears the cost of his choice, and the voter in a mass election does neither, a point to which I will return with some force, because it is the hinge of the economic critique.
The third reason is the one Francis Fukuyama gave its canonical form in "The End of History?" (The National Interest, Summer 1989). Fukuyama's essay is remembered by people who have not read it as a piece of triumphalism, and it is nothing of the kind. His claim, drawn through Hegel and Kojève, was that the great ideological competitors to liberal democracy had exhausted themselves, that fascism had been destroyed in war and communism had been destroyed by its own economics, and that no remaining ideology could answer, as completely as liberal democracy answers, the human demand for recognition, the demand of each person to be acknowledged as a being of equal dignity. Democracy dominates, on this account because it legitimates best, not because it is the best form of governance. It is the only system left whose story about itself, every citizen an equal author of the law, is one that modern people, formed by equality, will accept for long. I regard this as the strongest philosophical case for democracy ever assembled, and my verdict on it, which I will defend later, is the charitable one it deserves. Fukuyama's theory is sound and valid as an account of legitimacy, and it is too impractical to serve as what it became in the hands of his enthusiasts, a prescription, because a story that legitimates power is not the same thing as a structure that restrains it, and exporting the story without the structure has been the characteristic policy failure of the last 35 years. The Maldives is one of its case studies.
So the dominance is explained. Democracy is the reigning system of the advanced West because it solved succession, because it stood next to capitalism when capitalism won, and because it is the only remaining story about power that modern people accept. None of these three reasons, and I ask the reader to verify this against the preceding parts of this essay, is a claim that majority voting produces good government. That claim requires separate demonstration, and it is that claim which now goes under examination.
The case against democracy as a method of government divides naturally into four parts, an epistemic part concerning what elections can know, an incentive part concerning what elections reward, a moral part concerning what majorities may do, and an economic part concerning what the whole arrangement costs. I will take them in that order, and I ask the reader to notice that none of the four depends on any claim about the character of politicians or voters. Every argument below goes through on the assumption that every participant is honest, intelligent, and well-meaning. That is what makes them structural arguments rather than complaints.
Rousseau's general will presupposes that there exists, beneath the disagreements of individuals, a single coherent social preference which voting can extract. In 1951 Kenneth Arrow proved, in Social Choice and Individual Values, that no such extraction is possible. Arrow's Impossibility Theorem states that once three or more alternatives are on the table, no method of aggregating individual rankings into a social ranking can simultaneously satisfy a short list of conditions so mild that each reads as a restatement of the word "fair", among them that if everyone prefers A to B the society must prefer A to B, that the social choice between A and B must not be flipped by the arrival of an irrelevant third option C, and that no single voter's preferences may simply become the social answer regardless of everyone else. Violate none of them and the system contradicts itself. The theorem has an ancestor, the Marquis de Condorcet's paradox of cycling majorities, in which three entirely rational voters produce a collective preference that runs A over B, B over C, and C over A, a circle with no resting point, so that the "will of the people" in such a case is not hard to discover but does not exist, and the outcome is determined by whoever controls the order in which the questions are put rather than the electorate itself. Agenda control, in other words, outranks preference. Anyone who has watched a coalition government assembled in a place (let's say a place like Sri Lanka), has seen the theorem in business attire.
It should be understood what this result does and does not establish. It does not establish that elections are useless, and it does not overturn Mises's peaceful-succession defence, which needs elections only to count the factions, not to reveal a general will. What it establishes is that the central moral claim of modern democracy, that the winning policy carries the authority of the people as a collective author, is not simply doubtful but incoherent whenever the options number more than two. The people, as a unitary chooser, is a mathematical fiction. There are only individuals with preferences, and rules that convert those preferences into outcomes, and different rules convert the same preferences into different outcomes, from which it follows that every electoral result is an artefact of the rule as much as of the electorate. I have written elsewhere, in an essay on the Maldivian elections specifically, about how our own two-round presidential system and first-past-the-post parliament manufacture supermajorities out of pluralities and liquidate the political centre on schedule, and I will not repeat the demonstration here. The point that carries forward is philosophical. A procedure whose output is an artefact of its own design cannot confer the kind of authority that the phrase will of the people claims for it, and every constitutional structure that hands unlimited power to that output is handing unlimited power to an artefact.
To the impossibility of aggregating the voters' knowledge one must add the poverty of the knowledge itself, and here the argument owes its rigour to Anthony Downs. In An Economic Theory of Democracy (1957) Downs observed that information about public affairs is costly to acquire, that the probability of any single vote deciding a national election is for practical purposes zero, and that the rational voter therefore invests almost nothing in political knowledge, not because he is stupid but because he is mathematically competent. The expected return on a hundred hours spent mastering the national budget is the probability of decisiveness multiplied by the benefit of the better outcome, and since the first factor is on the order of one in the hundreds of thousands, the product is indistinguishable from zero. This is rational ignorance, and Bryan Caplan extended it in The Myth of the Rational Voter (2007) to rational irrationality, the further observation that since holding a false political belief costs the individual voter nothing, voters indulge beliefs that feel good, tribal loyalty, scapegoats, painless remedies, in a way they never would where their own money is at stake. The same man who researches a motorcycle purchase from Sheesha for three weeks votes his party's colour on the strength of a rally and a rumour, and he is not being inconsistent. He is responding to prices. The motorcycle error lands on him. The ballot error lands on everyone, which is to say, on no one he can identify. Democracy is thus the only major institution of modern life in which the decision-maker bears none of the cost of his decision, and the theory of it has assumed, for two centuries, that this would not matter. Every market on earth is organised on the opposite assumption.
I come now to the argument I regard as the deepest in the modern literature, the one at the centre of Hoppe's Democracy: The God That Failed, and I want to state it with the care it deserves, because it is routinely dismissed by people who have encountered only its conclusion, which is scandalous, and never its mechanism, which is a plain application of property theory. Hoppe's central distinction is between privately owned government and publicly owned government. A traditional monarch, whatever else may be said against him, owns the realm in the sense that matters economically. He holds its capital value, the future stream of revenue the country can produce, and he can pass that value to his heirs. A democratic officeholder owns nothing of the kind. He is, in Hoppe's exact and damning term, a temporary caretaker, entitled to the current use of the state's resources for a term of years, entitled to none of their capital value, and certain to be replaced by a rival.
Then it becomes important to ask what each arrangement rewards, and the analysis is the standard economics of ownership applied without flinching. An owner, precisely because he holds the capital value, has an interest in maximising it, which is to say an interest in the long run, in moderate and predictable taxation, in the prosperity of the productive population from whom all future revenue must come, for the same reason the owner of a house repaints it and the tenant of a house does not. His time preference, the rate at which he discounts the future against the present, is structurally low. The caretaker's position is the tenant's. Whatever value he does not extract during his term accrues not to his children but to his opponents, and every year of restraint is a gift to the rival who replaces him. His time preference is structurally high, not because democratic politicians are worse men than kings, some of them can be considerably better, but because the institution assigns them the incentives of a tenant with a 5 year lease and no deposit to lose. From this then it follows that democratic government will systematically consume the state's capital, running debts whose repayment falls on citizens not yet born, expanding present expenditure against future solvency, and preferring in every choice the visible benefit today over the invisible cost tomorrow, and it will do this regardless of who wins, because the incentive belongs to the office and not to the man. When the public debt figures of virtually every mature democracy are set beside this prediction, the prediction does not look daring. Hoppe adds a second observation of particular relevance to what follows in this essay, namely that open entry into government, celebrated as democracy's virtue, functions in practice as a selection mechanism for precisely the talents of the auction, the making of promises, the assembly of coalitions of beneficiaries, the disciplined shortening of every horizon to the next election. Competition is a discovery procedure, as the Austrians never tire of saying, and what democratic competition discovers is the most efficient redistributor.
I should be honest about where I part company with Hoppe. His conclusion is not constitutional monarchy but the abolition of the state in favour of a natural order of private law and insurance-based defence, and between his mechanism and his terminus there is a gap he never, to my satisfaction, bridges. His comparative analysis of monarchy and democracy is sound and valid, and his prescription is too impractical to be applicable to the condition of any actually existing society, a verdict I will make use of later, because one can accept everything true in Hoppe, the ownership analysis, the time-preference mechanism, the praise of small jurisdictions and of the right of exit, without following him off the edge of the state altogether. It is the mechanism I am taking from him, not the eschatology.
The same mechanism, run at the retail level and pressed to its conclusion, is the substance of Frank Karsten and Karel Beckman's Beyond Democracy (2012), and since the book bears on this essay at more points than any other single source, I will give it more than a passing citation. Its full title states its whole case, that democracy does not lead to solidarity, prosperity, and liberty but to social conflict, runaway spending, and a tyrannical government, and the body of the book is organised as the demolition of thirteen myths by which, the authors argue, the word democracy holds on to goods it never earned. Beneath the thirteen demolitions sits a single diagnosis. Democracy, Karsten and Beckman argue, is a species of collectivism, the last one still in good odour. Socialism collectivised the means of production and was found out, while democracy collectivises the means of decision, the whole society constituted as one compulsory association in which everything brought before the ballot belongs to everyone and therefore to no one, and it has not been found out, because it is defended the way faiths are defended rather than the way arrangements are defended. They observe, and the observation survives contact with any dinner table, that democracy functions in the modern West as a secular religion, that its tenets are recited rather than examined, and that the man who questions the arrangement is heard not as proposing an alternative but as committing a profanity. The present essay is, in their terms, a profanity committed deliberately and at length.
Their thirteen myths sort naturally into three families, and each family lands on the Maldivian case with uncomfortable precision. The first family is epistemic, the myths that every vote counts, that the people rule, and that people get what they want. Against the first stands the Downsian arithmetic already given. Against the second stands the observation, as old as Michels' iron law of oligarchy and as fresh as any coalition negotiation in one of our resorts, that the people cannot rule because the people is not an agent, and that what rules in every democracy is an apparatus, party machines, organised lobbies, and the permanent brokers who assemble majorities out of purchased blocs, with the voter's role confined to ratifying, at intervals of five years, one of two menus he did not compose. Against the third stands Arrow, and nothing further is needed. The second family is moral, the myths that the majority is right, that democracy is freedom, and that democracy is tolerance. Karsten and Beckman's treatment of the freedom myth is the sharpest passage in the book and deserves its paraphrase here. Freedom, they say, is the condition of not being coerced, of disposing of one's person and property according to one's own judgment, and democracy is not the absence of coercion but a procedure for organising it, a rule specifying whose preferences the coercive apparatus will enforce this term. A man wholly governed by his neighbours' votes is not free in any sense of the word that a slave would recognise as a change of condition, and the equation of the two ideas, democracy and freedom, is the most successful substitution of a mechanism for a value in the history of political language. Tolerance receives the same knife. A tolerant society is one in which what my neighbour does is none of my business, and majoritarianism is the institutionalisation of the opposite premise, that everything my neighbour does, his school, his diet, his wage bargain, is exactly my business, since I have a vote on it. Once a question is made collective, everyone must fight over the single answer, and the fight is permanent, whereas questions left private are answered by each in his own way without anyone needing to defeat anyone. Every extension of the ballot's jurisdiction manufactures a new front, and the observed rancour of democratic societies, ours conspicuously among them, is not a failure of civility but the designed output of a machine that keeps putting to a vote questions that need never have been voted on at all.
The third family is consequential, the myths that democracy produces prosperity, that it protects the poor, and that it restrains corruption and war. Prosperity has been dealt with above, it belongs to property and prices, and the democracies that possess it acquired its foundations before they were democracies. On the poor, Karsten and Beckman make the observation that transfer systems built by majority coalition do not flow to the weakest, who are few and disorganised, but to the best organised claimants of the broad middle, and the observation is the welfare-state ratchet seen from below. Each new benefit creates a constituency organised to defend it, while its cost is scattered across taxpayers too diffuse to organise, so programmes accumulate in one direction like sediment, the state grows under every party, and governments of the nominal right and nominal left differ in the beneficiaries they favour while agreeing perfectly on the direction of the total. To this they add the political economy of the promise itself, that the democratic politician structurally cannot say no, since the refusal is punished at the next auction while the cost of the assent matures after it, and a system in which no one empowered to spend is capable of refusal has runaway spending not as a risk but as a property, the property our own ledger exhibits at 135% of GDP. And on corruption their conclusion anticipates the Maldivian record exactly, that where the state is the largest prize in the society, democratic competition is not a check on its capture but the tournament by which the captor is selected.
The thirteenth myth, that there is no alternative, is the one the book exists to break, and their answer is the one this essay will build with in later parts of this essay. The alternative to rule by the national majority is not rule by a national minority, which is the false dilemma the myth trades on, but the shrinking of the domain over which anyone rules at all, decentralisation of decision to the smallest workable scale, competition among many small jurisdictions, and the right of exit as the citizen's standing vote, with Switzerland and Liechtenstein entered as exhibits that the arrangement is not a thought experiment but a going concern with the best governance record in Europe. I give their book the charitable verdict its seriousness has earned. Where the authors follow the logic to a wholly privatised society of contract, they arrive, like Hoppe, at a terminus sound in its internal logic and too impractical to be applicable to the condition of any people that actually exists, and an island nation of about 350,000 is not the place where mankind's first stateless order will be piloted. But nine-tenths of the book requires no such terminus. Its diagnosis, that collectivised decision is the disease and the shrinking of the collective domain the cure, and its prescription, small jurisdictions in competition under a right of exit, survive scrutiny intact, and they happen to fit the geography of our country as if drawn for it.
The moral argument is older and simpler, and it can be stated in a paragraph because the whole liberal tradition stands behind it. If it is wrong for one man to seize another's property or silence his speech, it does not become right when fifty-one men out of a hundred vote to do it, because the number of those who commit an act is not among the properties that determine its rightness. Whoever says otherwise must explain at what headcount injustice transmutes into justice, and no one has ever produced the number. The ballot, this is to say that, confers power and cannot confer authority over things no one has the authority to grant. Locke knew this, which is why his majorities operate inside a fence of natural rights. Madison knew it, which is why his republic breaks majorities into pieces before they reach the levers. Tocqueville and Mill knew it, and said that a majority acting as a faction is more dangerous than a king acting as a tyrant, because the king can be resisted in the name of the people, while the majority claims to be the people, and resistance to it is made to look like treason against everyone. The twentieth century then supplied the experimental confirmation, elected majorities that voted minorities out of their property, their citizenship, and in the limit their existence, and it supplied the institutional lesson too, that every constraint we now regard as civilised, entrenched rights, constitutional courts, supermajority requirements, exists precisely to stop the demos from doing what it voted to do. Modern liberal democracy, examined honestly, is a system nine-tenths of whose architecture is devoted to frustrating democracy, and its admirers call this arrangement democratic the way a man calls his wild pet friendly because it is muzzled.
The economic critique gathers the preceding three and prices them. Milton Friedman's taxonomy of spending, set out in Free to Choose (1980), does most of the work in four lines. A man can spend his own money on himself, in which case he economises and seeks value. He can spend his own money on others, in which case he still economises, or other people's money on himself, in which case he seeks value but does not economise. Or he can spend other people's money on other people, in which case he neither economises nor seeks value. Democratic government is the fourth category constituted as a permanent institution and staffed, per Downs, by men whose tenure depends on directing the flow toward organised claimants. To this add the asymmetry that public choice theory, from Buchanan and Tullock's The Calculus of Consent (1962) onward, placed at the centre of the discipline, concentrated benefits and diffuse costs. A subsidy worth millions to a small group and costing each taxpayer a few MVR will be fought for by professionals and opposed by no one, and a thousand such measures, each individually trivial, aggregate into the fiscal condition of the modern democratic state, which is a machine that borrows from the future to purchase consent in the present, precisely as the caretaker analysis predicts. The dispersed majority that pays is rationally ignorant of each item, the organised minority that receives is rationally obsessed with it, and the politician who brokers the transfer is rationally indifferent to a cost that will mature after his term. Every actor behaves correctly, and the system fails. That sentence, it should be noted, is the signature of a design flaw rather than a moral one, and it is the sentence this essay exists to apply to my own country.
The critique now assembled faces an obvious rejoinder, and it deserves to be stated at full strength. If democracy is epistemically incoherent, incentivally perverse, morally unbounded, and fiscally ruinous, why are the most democratic societies on earth also the freest, richest, and best governed? Denmark votes, and Denmark works. Switzerland votes on everything, practically at the level of the village, and Switzerland is the best-run polity in the recorded history of the species. Either the critique is wrong, or something else is doing the work.
I believe something else is doing the work. The advanced Western states are not well governed because they are democracies. They are democracies that can afford their democracy because they are well governed, and the assets that govern them well were accumulated, in every single case, before the mass franchise arrived. England had the common law, secure property, independent judges, parliamentary control of taxation, and a professional administration long before it had anything resembling universal suffrage, and the suffrage, when it came, arrived by instalments across a full century, each cohort of new voters socialised into an operating constitutional culture before the next was admitted. The frame preceded the crowd. Fukuyama himself, in his later and greater work (The Origins of Political Order, 2011), made exactly this point in the form of a sequence, first the state, an administration competent to execute, then the rule of law, a fence the administration cannot cross, and only then accountability, the mechanism by which the governed replace the governors. Where the sequence runs state, law, accountability, one gets Denmark. Where accountability is installed first, onto a society without an impersonal state or an operative rule of law, elections do not discipline the patronage networks. The patronage networks metabolise the elections, which become the mechanism by which clans and patrons compete for the state's resources, and the vocabulary of democracy provides the ceremony while the distribution proceeds underneath. He gave the problem its name, getting to Denmark, and his honest conclusion was that nobody knows the route, because the countries that arrived did so over centuries and left no map.
It is then informative to restate this in the language of my distinction. What the West possesses is the substance, a low average time preference in the population itself, formed by generations of security and stability, a civic culture that discounts the demagogue, courts that will rule against the government and be obeyed, bureaucracies with professional pride, newspapers and universities and professional bodies, the thousand institutions that stand between the individual and the state and answer to neither. What the West exhibits on election day is the form. The form is the visible 10%, and the export model of the last 30 years, the model applied to my country in 2008, consisted of shipping the visible 10% and assuming the rest would condense around it. The assumption has failed everywhere it has been tried without the substance already in place, and the failure has a literature too large to survey and a pattern too consistent to deny. Elections without the substance select for the auction. The auction consumes the state. And the consuming is then blamed, by each faction, on the personnel of the other, so that the diagnosis never rises from the man to the mechanism.
One more Western datum belongs in the record, because it cuts the other way and honesty requires it. Even with the substance, the mature democracies exhibit exactly the drift the critique predicts, debt ratios that climb across decades under governments of every colour, entitlement programmes no politician can touch and survive, regulatory sediment that accumulates and is never dredged. The substance slows the mechanism, and it does not repeal it. The West is not the refutation of the caretaker analysis. It is the caretaker analysis running at low speed on a very large inheritance, and the inheritance is being spent. This matters for my argument because it forecloses a lazy conclusion. The remedy for the Maldives cannot be "become Denmark first", both because no one knows how and because Denmark itself is running down the asset. The remedy must be structural, and it must work with the electorate and the political class a society actually has, a requirement I will hold myself to in the latter part of this essay.
I have argued the mechanism in the abstract. My country has spent seventeen years performing it as a play, and I have had the professional misfortune of a seat near the stage. The Maldives was a kingdom for the better part of its history, an arrangement that survived the Portuguese, the Dutch, and the British, and was abolished in 1968 in favour of a republic that spent its first forty years as an autocracy in republican dress. In 2008 we adopted, with genuine hope and to genuine international applause, a modern liberal constitution, separation of powers, multi-party elections, independent commissions, devolved councils, a professional civil service. I have written three essays on what followed, one tracing the system dynamics, one the electoral mathematics, one the institutional anatomy, and I will compress their findings here into the shape the present argument requires, because each of them is a local instance of a theorem stated above.
Begin with the fiscal record, Hoppe's prediction in a single national ledger. Public debt is projected to reach 135 per cent of GDP by 2026–27. The Sovereign Development Fund, nominally our buffer against external repayment, holds tens of millions of dollars against obligations running to multiples of that, which is to say that it is a fund in the accounting sense and not in any operational one. The state-owned enterprises function as the true welfare system, disguised unemployment distributed as patronage, and the audit of one of them, Fenaka, disclosed 1,501 staff hired before the September 2023 presidential election and a further 1,468 before the April 2024 parliamentary one, hiring driven by the electoral calendar with the shamelessness of a seasonal industry, which is what it is. A Member of Parliament earns roughly MVR 80,000 a month, some MVR 4.8 million across a five-year term, while a competitive campaign costs between MVR 3 million and MVR 10 million, from which any rational-actor analysis concludes what mine concluded, that the seat is purchased as an investment and the return is collected from sources other than the salary. The caretaker does not even need to be theorised here. He can be interviewed.
Continue with the electoral mathematics, Arrow and Condorcet staged with local costumes. The 2013 presidential election produced a first round, an annulment by a Supreme Court whose intervention rearranged the field, and a runoff whose result owed as much to the sequencing of the contest as to any distribution of preferences, agenda control outranking preference exactly as the paradox says it must. The two-round system then spent a decade manufacturing a kingmaker class, third-place candidates auctioning their endorsements between rounds, until polarisation liquidated the centre altogether and Duverger's law collected its debt, a parliament of ninety-three first-past-the-post seats converting modest pluralities into supermajorities, 65 seats for one party in 2019, 66 and in time more for its opposite in 2024, each supermajority read as a national consensus and each in fact an artefact of district arithmetic silencing the fraction of the country that voted the other way. The will of the people, in the Maldivian record, has been whatever the rule of the moment computed, and the rule has been chosen, contested, and litigated by the very actors it would go on to crown.
Complete the triptych with the institutional anatomy. The 2008 constitution drew the full Westminster-style architecture, and the architecture exists on paper at every point and in operation at almost none. The Permanent Secretary who is meant to hold the administrative line against the Minister is appointed, in substance, by the political authority she is meant to constrain, and may be removed at its pleasure, so she signs. The Anti-Corruption Commission, the Auditor General, the Elections Commission, and the courts are appointed by the political class whose conduct they exist to police, and the April 2024 suspension of three Supreme Court justices, in the middle of a constitutional review they were conducting, demonstrated the arrangement's operating truth to anyone still requiring demonstration. The MMPRC affair of 2014–15, in which upward of 50 islands and lagoons, the finite natural capital of the nation, were leased at discounts through a state vehicle whose proceeds flowed to private accounts and thence to parliamentarians, judges, and enforcers, was not corruption as friction but corruption as capture, the watchdogs purchased with the proceeds of the theft they existed to prevent. And in May 2026 the elected atoll councils, the one genuinely decentralising element of the 2008 design, were abolished under the 17th Amendment, the centre reabsorbing the periphery precisely when the argument for dispersal had never been stronger. Form without substance, at every joint of the machine.
Now I must do more than recite the record, because the argument of this essay requires a stronger claim, not that democracy has failed in the Maldives but that it was always going to fail, and will continue to fail, in a place configured as the Maldives is configured. The claim rests on five structural features, and I present them as a set because their force is conjunctive.
The first is scale. Every theorem I have written concerning the anonymity of the voter runs backward in a society of our size. Rational ignorance assumes the individual vote is invisible, but in an island constituency of a few hundred families, votes are inferable and promises are enforceable, so the ballot, secret in law, is transparent in practice, and the electoral relationship collapses from representation into transaction. Vote-buying, which in a large polity is a retail inefficiency, is in a small one the wholesale structure of the market, because the price of a decisive bloc is within the reach of a single patron. A concentrated interest does not need to capture a distant legislature through lobbying, since it can capture the constituency directly, name by name. Smallness, which I will argue below is our greatest constitutional asset, is under centralised majoritarianism our greatest liability, because it prices every vote and identifies every seller.
The second is geography. A polity of scattered islands has the highest natural transaction costs of any state form on earth, and a centralised government amplifies them, every decision travelling to Malé and back, every island's particularity averaged into a national programme designed for none of them. Distance, as I have written before, is an administrative problem in its own right, and centralisation does not solve it but postpones its consequences at compounding cost. The unitary state and the island nation are a category error joined by a constitution.
The third is the economy. Tourism rents plus sovereign borrowing flow through a single treasury, which makes the state itself the largest prize in the society and control of the state the only business plan with unbounded upside. Where the economy's commanding heights are a treasury, Aristotle's deviant form is not a risk but an equilibrium, the majority coalition governing in the interest of its members with the treasury as the instrument, and the opposition differing not on the principle but on the roster of beneficiaries.
The fourth is formation. I have made this argument at length elsewhere and will state it here again. The Westminster architecture presupposes an electorate formed, over generations, in the habits that discount the auction, the tracing of consequences past the first step, the suspicion of the painless promise, the treatment of institutions as held in trust rather than owned by the living majority. Schooling we have, and that formation we do not. It is not that Maldivians are deficient, but it is because the formation is the work of generations, and no people on earth has acquired it in 17 years. An electorate at our stage rewards the auction, rationally, and the political class conducts the auction, rationally, and the arithmetic of the two rationalities is the national debt.
The fifth is the absence of exit. Hirschman's distinction in Exit, Voice, and Loyalty (1970) is the relevant instrument. When an organisation decays, its members can exit to an alternative or exercise voice to reform it, and voice without the possibility of exit is bargaining without a walk-away, which is to say, hardly bargaining at all. In the unitary Maldivian state there is one jurisdiction, one police, one treasury, one auction. A citizen of Addu or Haa Alif who despairs of government has no Maldivian alternative to move to, no neighbouring jurisdiction whose better rules can shame his own, and so the only competition our system permits is the quinquennial contest for the single prize, which is precisely the competition the whole preceding analysis convicts. Every mechanism of discipline that operates in functioning federations, comparison, migration, yardstick competition among jurisdictions, is structurally absent, by constitutional choice, in the one country whose geography offers them more naturally than any other on the planet.
The design I propose takes its materials from the thinkers already engaged, Hoppe's ownership analysis and his praise of small jurisdictions, Karsten and Beckman's decentralism, Friedman's insistence that choice disciplines what voting cannot, Fukuyama's sequencing, and one further source I now introduce, Patri Friedman's theory of competitive governance, stated in "Beyond Folk Activism" (Cato Unbound, April 2009). The younger Friedman's argument is that political systems improve the way industries improve, through competition among providers and low barriers to customer exit rather than through the persuasion of incumbents, and that the scarcity of new entrants in the governance market is the root explanation of its stagnation. His own application of the idea, floating autonomous communities at sea, invites the smile it receives, and my verdict on it is the one his grandfather's colleagues would have given, that the theory is sound and valid and the vehicle too impractical to be applicable to the human condition as presently housed. But it should be noted, and this observation is the seed of everything that follows, that the Maldives does not need to build the seastead. We are the seastead. Nature has already arranged us as some 26 natural laboratories separated by water, each with a coherent community, a harbour, and centuries of practice at governing its own affairs at the island scale. The competitive-governance thesis, which is speculative engineering everywhere else on earth, is in the Maldives a description of our geography waiting for a constitution to notice it.
I will now state the design, and I ask that it be read as a single mechanism rather than a list of preferences, because each element exists to carry a load the others shed. Let us call it, for convenience, the Decentralisation of Atolls.
The first element is the devolution of general competence to atoll governorates. The inhabited atolls, singly or grouped by contiguity and population into perhaps 7 to 10 governorates, become the primary governments of the Maldives, each with an elected council and an elected governor, each with plenary competence over everything not expressly reserved to the centre, land use, economic regulation, licensing, education, health administration, local infrastructure, and, deliberately and in full knowledge of the weight of the choice, its own police. Policing is the state function most intimately entangled with daily liberty and most dangerous when it answers to a distant patron, and a governorate police answerable to a local council, watched by neighbours who know every officer by name, is corrigible in a way no national force commanded from elsewhere has ever been. Each governorate levies and retains its own taxation on the activity within it, including a defined share of the tourism revenue its own lagoons and islands generate, a provision that converts every governorate at a stroke from a supplicant at the treasury into a proprietor of its tax base, with the proprietor's incentives toward the health of that base. Here Hoppe's mechanism is put to work rather than lamented, for the governorate, unlike the national caretaker, lives beside its tax base, is watched by it at the market, and inherits its own mismanagement in a way no central ministry ever has.
The second element is the deliberate shrinking of the centre. The national government retains an enumerated and short list, foreign affairs, defence, currency and the monetary system, nationality, the supreme court, inter-governorate standards where physical networks genuinely require them (aviation, telecommunications, maritime safety), and nothing else, with the constitutional presumption running against the centre in every contested case, the inversion of the presumption that has governed us since 2008. The national parliament shrinks with the jurisdiction. It should be understood what this element is for. It is not administrative tidiness. It is the removal of the prize. The whole pathology traced in previous sections, the auction, the patronage hiring, the capture of watchdogs, the MMPRC, feeds on a single treasury worth capturing, and no reform that leaves the prize intact will change the competition for it. Make the centre small and poor and the capture of it unrewarding, and the energy now spent capturing it must go elsewhere, ideally into the governorates, where it is disciplined by the third element.
The third element is exit, the disciplining principle of the whole compact. Charles Tiebout demonstrated in "A Pure Theory of Local Expenditures" (1956) that where citizens can move among many small jurisdictions offering different bundles of taxes and services, migration performs the function that voting performs badly, it reveals preferences honestly, because moving is costly and nobody signals falsely with his own removal van, and it disciplines providers continuously rather than quinquennially. A Maldivian dissatisfied with his governorate's schools, taxes, or police would possess for the first time a domestic alternative, and, which matters more, every governor would know it. The governorates compete for residents, for enterprises, for the mobile professional class, and the competition is precisely the discovery procedure that the single national auction suppresses. One governorate will try school vouchers and another will not, one will court guesthouse tourism and another exclusive resorts, one will fail conspicuously, and the failure will be quarantined by water and made instructive to the rest, an arrangement engineers will recognise as modularity and fault isolation, the design principle by which sound systems prevent a local fault from propagating across the whole. Under the present constitution, error is national, mandatory, and synchronised. Under the compact, error is local, optional, and instructive. I regard this single property, before any other virtue, as sufficient reason for the design, because a small country cannot afford national experiments, and the present system permits no other kind. And the mathematics of it improves in the same motion, for Arrow's conditions bind aggregation across heterogeneous multitudes, and the smaller and more homogeneous the electorate, the fewer the dimensions in contention and the nearer the median voter to a real person rather than a statistical ghost. Democracy, this is to say that, is not abolished by the compact. It is repatriated to the scale at which it functions, the scale at which the voter knows the candidate, sees the budget's consequences, and lives among those his vote affects. What is abolished is the pretence that the same procedure aggregates honestly across 350,000 people and nine hundred kilometres of sea.
The fourth element is the crown, and I present it with the care owed to the proposal most certain to be misread. Above the political structure, national and governorate alike, stands a constitutional monarch, restored rather than invented, since the Maldives was a kingdom for the better part of its history and has been a republic for fewer than sixty, holding no executive power, no legislative power, and no role in policy, and charged with exactly four functions. The first is the appointment, upon the binding recommendation of technical constitutional commissions whose composition is fixed in law, of the integrity institutions, the judges of the supreme court, the Auditor General, the Elections Commission, the Anti-Corruption Commission, the Human Rights Commission, and the Civil Service Commission, together with the protection of their tenure against every authority, national or governorate, whose conduct they exist to police. The second is a reserve power, exercisable only through formal mechanisms specified in law and reviewable by the court so appointed, to require any government within the state to remain inside its constitutional limits. The third is the command of the military, held by the crown precisely so that no elected faction, national or local, ever holds it, with deployment requiring the assent of the national government and the force itself professionalised under a service commission beyond political appointment. The fourth is continuity, the embodiment of the state across electoral transitions, which sounds ceremonial and is not, in a country where each incoming administration has treated the state as conquered territory and its files as spoils.
The reasoning is the one I have defended elsewhere and will compress here, because it follows directly from Hoppe's analysis rather than from any sentiment about thrones. Every failure catalogued reduces to a single structural fact, that the bodies meant to constrain the political class are appointed and removable by the political class, which holds a standing interest in their non-functioning. The constraint must therefore be anchored outside the electoral cycle, in an institution whose time horizon is generational and whose interests run with the capital value of the state rather than with any faction's term. It is essential for my argument to pause here and distinguish the monarch as a person from the monarchy as an institution, because the case rests entirely on the second. The institution does not stand for re-election, cannot profit from the auction, and is delegitimised by exactly the decay, corrupt courts, captured commissions, a politicised army, that a caretaker government can profitably tolerate for a term. Its incentives, when the architecture confines it to guardianship, are the owner's incentives of Hoppe's analysis harnessed to the one task where an owner's horizon is wanted and an owner's discretion is not, the keeping of the referees. Bhutan, whose constitution of 2008 is the working precedent, assigns its king precisely this appointing and protecting function above an elected parliament and government, and Bhutan is, by the Corruption Perceptions Index, the cleanest state in South Asia, an outcome I take as evidence for the architecture while conceding, as one must, that a single comparator proves possibility rather than necessity. The crown, in short, is the compact's answer to the question every constitution must answer and ours has answered wrongly for 17 years, who appoints the umpires, and its answer is that it must not be the players.
The fifth element is the fiscal constitution, without which the rest is furniture. Hard statutory ceilings on debt and deficit at both tiers, certified quarterly by an independent fiscal council reporting to the Auditor General. The senior civil servant of each treasury made personally the accounting officer, liable for breaches, and protected in refusal of unlawful instruction by the Civil Service Commission the crown appoints, so that the refusal can be made and held against any minister's displeasure. No bailout of any governorate by the centre, stated in the constitution in words any person can parse, because the first rescue converts every governorate budget into a claim on the national treasury and reassembles, five rescues later, the very centralised auction the compact exists to dismantle. A governorate that ruins itself must restructure under the supervision of the fiscal council, its citizens free, this is the discipline of exit doing its work, to walk to the ferry. Hard budget constraints with mobile citizens, as against soft constraints with captive ones, are the difference between Switzerland's cantons and the departments of every failed unitary state that has decentralised expenditure while centralising the guarantee, and the Swiss comparator carries particular weight here, seven and a half centuries of confederated cantons, most of them smaller than our larger atolls, conducting the longest-running and most successful experiment in dispersed government the species has produced.
Assembled, the compact answers each failure mode in its own terms, and the correspondence is worth stating compactly. The auction is starved because the treasury worth auctioning no longer exists at the scale the auction requires, and the governorate treasuries are watched by neighbours and disciplined by departure. The tyranny of the majority is dissolved because there is no single national majority with jurisdiction over everything, only many small majorities with jurisdiction over their own affairs and none over their neighbours', which was Madison's remedy for faction restated in salt water. The welfare-state ratchet loses its pawl because redistribution funded by a governorate's own visible taxation, among people who know one another, is self-limiting in a way that redistribution funded by national borrowing is not, Karsten and Beckman's point made operational. The caretaker's time preference is bounded above by the crown's tenure of the referees and below by the governorate's proprietorship of its own tax base. And the formation problem, the generation-long absence of a critically reasoning electorate, is not assumed away, which was the 2008 constitution's founding error, but designed for, the institutions carrying the constraint the electorate cannot yet carry, while the governorate scale, where cause and consequence share the same road, performs the civic education that no national campaign ever has. The compact does not require better Maldivians. It requires only the Maldivians we have, responding to prices they can finally see.
I will now restate the compact's foundations explicitly, as pillars, philosophical, economic, and governance, so that each design choice above can be traced to the principle that requires it, and so that a person reading this can at least know which principle he is spending.
The philosophical pillar is built of four commitments. The first is subsidiarity, the principle, older than its name, articulated by Althusius in the Politica (1603) out of the experience of federated Calvinist towns and given its modern formulation in Catholic social teaching, that every decision belongs to the smallest community competent to take it, and that a higher order may assume a function only upon demonstration, not assertion, that the lower cannot perform it. Subsidiarity is not administrative decentralisation, which is the centre choosing to delegate what it may recall at pleasure, and the 17th Amendment has just demonstrated what delegation at pleasure is worth. It is a claim about where authority originates, from the family and the island upward, so that the governorate does not receive its competence from a central location but retains what it never surrendered, and the burden of justification runs downhill forever. The second commitment is consent made real through exit. The social-contract tradition from Locke forward has rested government on a consent nobody ever actually gives, inferred from residence, imputed to the newborn, and the compact replaces as much of this fiction as geography allows with the genuine article, the standing capacity to leave one jurisdiction for another without leaving one's country, language, or faith. Where exit is real, remaining is consent in the only sense that binds. The third commitment is the separation of guardianship from rulership, the recognition, as old as the quis custodiet problem and as fresh as our suspended justices, that the power that governs and the power that keeps the governors within their commission must not be held by the same hands, nor appointed by them, nor dismissible by them, and that the second power must live on a longer clock than the first. The fourth commitment is trusteeship across generations, Burke's partnership of the dead, the living, and the unborn, entering the design not as sentiment but as structure, since every mechanism above, the crown's horizon, the debt ceilings, the entrenchment rules, exists to give the unborn, who vote in no election, a party at the table. It should be noted what is absent from this pillar. There is no appeal to the general will, no claim that any procedure reveals what the people want, and no promise of equality beyond equality before the law. The philosophy of the compact is negative where the philosophy of 2008 was affirmative, it is the lesson and a deficiency.
The economic pillar is the discipline of the whole arrangement, and it rests on three mechanisms. The first is proprietorship, Hoppe's insight installed at every level where it can operate benignly. The governorate owns its tax base and lives beside it, the civil servant owns a protected career whose value compounds with the institution's health, the crown owns, in the institutional sense, the long-run standing of the state, and at each level the owner's low time preference is set to work against the caretaker's high one, so that the structure contains, for the first time in our constitutional history, actors whose rational interest is the year 2075. The second mechanism is competition with hard budget constraints. Tax and regulatory competition among governorates performs for government what market competition performs for firms, the discovery of what works and the punishment of what does not, but the discovery procedure functions only if failure is permitted to hurt, which is why the no-bailout rule is written as an absolute, why governorate borrowing is capped by statute and certified by the fiscal council, and why the equalisation transfers that every federation is tempted into are confined, in the compact, to a single transparent per-capita grant fixed in the constitution itself, beyond annual political adjustment, since discretionary equalisation is the solvent in which every federal discipline has ever dissolved. The third mechanism is the separation of the state from enterprise. The state enterprises are transferred to a single statutory holding company on the Temasek model, the state as sole shareholder, boards appointed for competence on staggered terms by the holding company and not by the executive, subsidiaries run commercially or wound up, and payroll expansion statutorily frozen in the four months before any election with personal liability on the accounting officer, so that the shadow welfare state, the patronage machine I have described elsewhere as an entropy machine, is dismantled and such welfare as the society chooses is delivered as welfare, visibly, locally, and on budget, where its cost can be seen by the people paying it. Money remains national, MVR and its peg administered by a central monetary authority whose governor the crown appoints on commission recommendation, because a currency is a network good that does not decentralise, and because monetary discretion in political hands is the auction conducted by other means.
The governance pillar assembles what the preceding sections have specified into a single framework. Two elected tiers, governorate and national, each supreme in its own enumerated domain, neither able to invade the other's, a national tier confined to genuinely national functions, a crown above both holding guardianship and only guardianship, integrity institutions appointed through the crown on the binding recommendation of technical commissions and removable only by supermajority impeachment before the supreme court, a national civil service posted across the governorates and protected by a commission the political class does not appoint, a judiciary with governorate courts of first instance and a single national appellate pyramid, and an amendment rule requiring a double majority, a national referendum plus ratification by a majority of governorate councils, with the compact's load-bearing walls, the enumeration of central power, the no-bailout rule, the guardianship functions of the crown, and the right of internal movement, entrenched behind a higher threshold still, three-quarters of the governorates, so that no transient national coalition can reassemble the unitary prize by ordinary politics. The reader will notice that every element of this pillar is intentionally made to a fence. The 2008 constitution distributed powers, and the compact distributes vetoes, which are what a society at our stage of formation can operate, because a veto, unlike a discretion, does not require virtue in its holder, and it fails safe.
The compact resolves into eleven chapters:
The State and the Crown: The Maldives as a decentralised unitary state of governorates under a constitutional monarch, the crown's four functions enumerated exhaustively, succession and regency fixed in law, the monarch's civil list fixed and audited, and an explicit clause voiding any act of the crown outside its enumerated functions.
Fundamental rights and the limits of the collective: The negative rights, person, property, contract, expression, worship within the constitutional place of Islam, due process, and the guarantee of internal free movement and establishment across governorates, the compact's charter of exit.
The Governorates: Their formation and boundaries, plenary residual competence, elected councils and governors, governorate police under council authority with national inspection standards, taxation powers and the fixed per-capita grant, and the procedure by which islands may petition to change governorate, exit fractally applied.
The National Government: The enumerated list, foreign affairs, defence, currency, nationality, justice, inter-governorate networks, a parliament sized to the shrunken jurisdiction, and an executive, presidential or parliamentary, on which the compact is deliberately agnostic, since either can operate within fences.
The Judiciary: Governorate courts, the national justice pyramid, appointment through the Judicial Service Commission under the crown, tenure to retirement, removal only by supermajority process.
The Integrity Institutions: The Auditor General, Elections Commission, Anti-Corruption Commission, Human Rights Commission, Civil Service Commission, and Fiscal Council, each with fixed composition, crown appointment on binding commission recommendation, protected tenure, and budgets charged directly on the consolidated fund beyond annual appropriation.
The Civil Service: The officer class defined, selection by competitive examination, senior posts capped by statute, the support class distinguished, postings across governorates, and the protected right of refusal of unlawful instruction.
The Fiscal Constitution: Debt and deficit ceilings at both tiers, the no-bailout clause, the accounting-officer liability, quarterly certification by the Fiscal Council, and the pre-election payroll freeze.
The Military and Security: The defence force under the crown as commander-in-chief, deployment on the advice and assent of the national government, a service commission for appointments, and an absolute bar on domestic political employment of the force.
Amendment and entrenchment: The double majority for ordinary amendment, the three-quarters governorate threshold for the entrenched walls, and the prohibition of amendment by emergency decree.
Transition: The schedule of devolution, the disposition of existing institutions, debts, and personnel, and the sunset provisions under which the transition itself expires.
Three objections deserve their full strength, and a fourth deserves its honest concession.
The first is that the compact is undemocratic, and in the narrow sense the objection is correct, three functions, guardianship of the integrity institutions, command of the military, and constitutional enforcement, are removed from electoral determination altogether. I answer that these are exactly the functions whose subjection to electoral determination has been tested in this country for 17 years, with results catalogued above, and that a democracy which cannot protect its own referees from its own players is not made more democratic by the repetition of the failure. The compact leaves the legislative and executive functions of both tiers fully elected, and it multiplies rather than reduces the occasions of meaningful voting, since the governorate ballot decides matters the voter can see, verify, and escape. One might rightly say that the people are sovereign and may not be fenced from any function they choose to claim. But this proves too much, since it indicts every entrenched right and every constitutional court in every democracy admired by those who make the argument, and nine-tenths of what they admire, as I observed in the earlier part of this essay, is fencing.
The second objection is that monarchy is a romantic anachronism, that a monarch is a man, that men are corruptible, and that I have simply relocated the single point of failure rather than removed it. This is the serious objection, and I decline to wave it away. A corrupt or ambitious crown could, over time, colonise the commissions that nominally bind it, and a contested succession could stage a constitutional crisis of the first order. The selection of the first incumbent, in a country whose crown was set aside in 1968, is a genuine problem, though a more answerable one than it first appears, since the setting aside is recent, the genealogical record of the royal houses survives, and the procedure I defend later, which is the election of the first monarch by the constituent assembly from among the documented descendants of those houses, with customary succession entrenched in law thereafter, resolves it without inventing anything. The question is never whether an arrangement can fail but what it costs when it does, compared to what, and at what frequency. The republican single point of failure, the capture of the integrity institutions by the elected executive, is not a risk in the Maldives but the operating record, exercised at will, at zero cost, within every single administration since 2008. The monarchical failure mode requires the slow suborning of an institution whose entire standing depends on not being suborned, against commissions it does not compose, a court it cannot instruct, and governorates it does not fund, and history's constitutional monarchies, the Scandinavian, the Japanese, the Bhutanese, exhibit that failure rarely where the enumeration is tight. I am offered a certainty of capture under the present design and a possibility of capture under the proposed one, and I take the possibility, while writing the enumeration as tightly as drafting can make it. He who demands a design with no failure modes demands a design that does not exist, and usually, in my experience, defends the incumbent one by that demand.
The third objection is capacity, that atolls of some thousands of people cannot staff serious government, and that devolution will deliver them to local magnates, the "Rashu Dhaadhaa" replacing the "Sarukaaruge Verin". The concern is real. The professional civil service I have argued for elsewhere, small, selective, well paid, appointed and protected by the Civil Service Commission under the crown, is a national service posted to the governorates, so that the officer in Fuvahmulah carries the same selection, protection, and portable career as his counterpart in Malé, and local politics directs local administration without owning its personnel, the German and Nordic pattern, in which small municipalities govern through professionals they could never individually cultivate. As for the magnates or the so called "Mahujanun", they exist now, and under the present constitution they buy constituencies wholesale and sit in a national parliament where their vote prices legislation for the entire country. Under the compact their theatre of operations shrinks to a governorate their neighbours can watch and their victims can leave, and a magnate whose subjects can emigrate at the cost of a RTL ferry ticket holds a wasting asset. Capture cannot be abolished, but it can be shrunk, exposed, and made escapable, and that is the whole of what constitutional design has ever honestly promised.
The fourth objection is feasibility, and here candour costs me nothing because I have already spent it. No incumbent national political class votes to dissolve the prize it holds, and I labour under no illusion that this essay's proposal is politically proximate. But the trajectory of the public finances has a terminus that is now visible without instruments, and constitutional moments arrive in this country's history the way its storms do, suddenly and with the furniture rearranged, 1953, 1968, 2008. The purpose of working out a design in advance of its moment is that moments do not wait while designs are drafted, and the last such moment, 2008, was met with a design borrowed from societies whose assets we did not hold, while we did no justice to reflecting on our state's thousand year old administrative history. When the next one arrives, there should exist, written down and argued through, at least one alternative that was drawn from our own geography, our own history, and an honest theory of what voting can and cannot do.
The 2008 transition failed in a particular way, and the route to the compact must be designed against that failure specifically. What happened in 2008 was a constitutional moment without a proper preparatory decade, the form adopted without proper dues and the substance left to condense afterward, which it did not. The compact inverts the order, in obedience to Fukuyama's sequence, and the inversion is the route's governing principle. The substance is built first, under the constitution we have, by ordinary statute, and the constitutional moment, when it arrives, ratifies and entrenches an architecture already partly running, rather than commissioning one from nothing.
The first phase is statutory, requires no constitutional amendment, and could begin in any parliamentary session in which a working majority chose to begin it. Its contents are the reforms I have argued in this essay and elsewhere, each valuable on its own terms even if nothing further ever happened, which is what makes the phase politically imaginable. A Fiscal Responsibility Act with debt and deficit ceilings, and an independent Fiscal Council. A Civil Service Act constituting the officer class, competitive entry, capped senior posts, the protected refusal, and the inversion, over time, of the compensation hierarchy between the political class and the professional one. A single statutory holding company for the state enterprises on the Temasek model, with the pre-election payroll freeze. The restoration and strengthening of elected local government, reversing the direction of the 17th Amendment and returning to the decentralisation clauses of the 2008 constitution their intended weight, with local councils given genuine fiscal proprietorship over defined local revenues. And electoral repair in the meantime, since elections will continue while the larger argument proceeds, the replacement of our ordinal ballots with approval voting for national office, which I have defended elsewhere as the reform that extracts the most consensus from the least constitutional disturbance. None of this is the compact. All of it is the compact's load-bearing masonry, and every month it operates it trains officers, generates records, and accumulates the institutional class on which everything later depends. I estimate this phase honestly at five to eight years, because the operating takes so long, and the point of the phase is operation, not enactment.
The second phase is experimental, and it is the phase on which the whole route wagers its plausibility. Under an organic law passed within the existing constitution, two or three willing atolls, and there should be a genuine application process, since willingness is itself the first datum, receive devolution in substance, plenary local competence, fiscal proprietorship including a defined share of locally generated tourism revenue, a locally directed police under national inspection standards, and the posted professional civil service, the whole under the hard budget constraint and the Fiscal Council's certification. These pilot governorates are the seastead built on land we already own. They will be imperfect, one may fail outright, and the design must let it fail without rescue, because the demonstration that failure is quarantined is worth more to the eventual constitution than the demonstration that success is possible. What the pilots produce, across 5 to 10 years, is the one commodity no argument can produce, comparison. A Maldivian family watching a pilot governorate pave its roads on its own revenue, discipline its own police, and publish its own audited accounts is receiving the civic formation that no curriculum delivers, and the demand for admission to the arrangement, if the theory of this essay is sound, will become the political constituency for the third phase, a constituency the present constitution cannot generate because it permits nothing worth joining. I would sooner stake the compact on this mechanism than on any campaign of persuasion, mine included, and it should be recorded that this is Patri Friedman's deepest point, generalised, that arguments do not move incumbents, working alternatives move populations, and populations move incumbents.
The third phase is the constitutional moment itself, and about its timing I will not pretend to knowledge no one has. It may be summoned by the debt reckoning whose arithmetic I have twice stated, or by an electoral crisis of the kind our voting mathematics periodically manufactures, or by the slow accumulation of the pilot evidence, and the drafting must be ready before whichever summons arrives. The mechanics I would specify are these. A constituent assembly seated for the single purpose and dissolved upon its completion, composed in equal parts of delegates elected from the governorates-designate and members nominated by the professional and judicial commissions built in phase one, so that the room contains both consent and competence. Ratification by double majority, a national referendum and a majority of atoll electorates voting severally, so that the periphery whose charter this is cannot be overridden by the capital's weight of numbers, and I would regard a narrow national yes with a broad atoll yes as a stronger mandate than the reverse. On the crown, the assembly must settle the question this essay has so far left procedurally open, and I state my position with its reasoning. The throne is to be restored from its roots, from the historical royal lines of the Maldives, and not invented by constitutional fiat. It should be understood that the monarchy here was not extinguished in some unrecoverable antiquity, it was set aside in 1968, within living memory, and the genealogical records, the chronicles, and the registries of the late kings are sufficient to establish who the remnants of the royal houses are and which of their descendants stand eligible today. Let the constituent assembly, assisted by a genealogical and historical commission constituted for the purpose, identify the eligible candidates from those documented lines and elect the first monarch from among them, so that the foundation joins the two legitimacies the office requires, the legitimacy of descent, which connects the institution to its thousand year old history that give it its meaning, and the legitimacy of the constitutional act, which binds it to the compact it exists to guard. Thereafter the crown passes not by election but in the manner in which it has always passed in this country, under the customary succession the kingdom practised across its many centuries, now written into an entrenched succession law so that custom and statute state the same rule and no political organ holds any part in the devolution of the throne. This is deliberate, and it follows from the institution’s own logic. The whole case for the crown, is the long horizon, and an office whose authority is its continuity should carry its history with it rather than begin at year zero, a throne standing in demonstrable continuity, in the eyes of the people whose trust it must hold, a title no assembly’s invention could confer. The royal upkeep, it must be said plainly, is to be narrowed to the royal family proper and to an extent established in law, a fixed and audited civil list covering the sovereign, the designated heir, and a defined inner circle of the house and no one beyond it, with no titles, pensions, or precedence multiplying outward through the wider kin, because the authority of the institution is moral before it is legal, and a court that grows numerous and rich grows illegitimate at exactly the rate it grows. And the transition chapter must sunset itself, a fixed decade of scheduled devolution, institution by institution and governorate by governorate, at the end of which the transitional powers expire of their own terms, because open-ended transitions are how centres keep what they have promised to give up.
Summed honestly, the route is 20 to 25 years, which is to say it is the same generation-long horizon I have attached elsewhere to the institutional reconstruction we never carried out, now given a destination. I do not regard the length as a defect. The length is the argument, since every shorter programme on offer in our politics consists of changing the personnel of a machine whose design guarantees that the personnel do not matter. The men and women who would complete this route are in school today, and the ones who must begin it are not, and the beginning requires no constitutional moment at all, only a parliament willing to bind its own hands slightly and a government willing to let one administrative atoll show what the rest could be.
In conclusion, I have argued that democracy is a method and not an end, that as a method its one secure achievement is the bloodless replacement of rulers, that the grander claims made on its behalf dissolve under the aggregation theorems, the incentive analysis, and the fiscal record, that its successes in the West are the successes of substance accumulated before the franchise and now being drawn down, that in the Maldives the form arrived without the substance and has produced auction, capture, and insolvency on schedule, and that a country configured as ours is, small, dispersed, transactional, and unformed, should stop asking the procedure to do what it cannot and redesign the state around what it can. Decentralisation keeps the ballot where the ballot works, in communities small enough to know what they are choosing, restores the guardianship functions to an institution built to hold them across generations, and lets the sea, which has always been our distance and our defence, become our discipline, the quiet competition of many small governments watched by citizens who can leave. The trade-offs are real and I have priced them where they stand, unequal provision across governorates, a crown that must be fenced with care, a transition of two decades that will be resisted at every step by everyone the present arrangement feeds. Against them stands the attainable alternative, which is not Denmark, and was never on offer, but the present republic continuing as designed, consuming the future to purchase the present until there is nothing left. Between a difficult architecture and an impossible one, the choice does not seem to me close. We have tried western democracy as it comes and it has failed. It should now try a method built and informed by its culture, its people, its history, and its geography.