Legitimacy
Why People Obey
Introduction
Do the arithmetic on any state and it stops making sense. A country of sixty million people has perhaps a hundred and fifty thousand police officers. If one adult in twenty decided on the same morning to stop paying tax, ignore court orders, and disregard licensing, there would be no mechanism capable of responding. There are not enough enforcers, there have never been enough enforcers anywhere, and every government in history has known it. Rule does not run on force. Force is what gets used at the margin, against the small minority who do not comply for other reasons.
What produces the compliance of everyone else is legitimacy: the widely-held sense that whoever is giving the orders has some standing to give them. It is the least tangible thing in politics and by a wide margin the most important, because it is what converts a group of armed people into a government. When it is present, a state can govern a continent with a modest bureaucracy. When it goes, the same state discovers that its apparatus was resting on something it did not control and cannot manufacture on demand.
This page is about where that standing comes from, why regimes with completely different foundations can each be stable, and why the loss of legitimacy is so consistently invisible right up until it is total.
Three Answers to "By What Right"
Max Weber's century-old typology is still the starting point because it identifies three genuinely different foundations, and most real regimes turn out to be mixtures. Traditional authority rests on continuity: this is how it has always been, and the arrangement predates everyone currently alive. Charismatic authority rests on a specific person: their perceived exceptional qualities, and the loyalty those qualities generate. Legal-rational authority rests on rules: the office has powers, the person occupying it has those powers only while occupying it, and the procedure for occupying it was followed.
The most useful thing in the typology is not the categories but their differing failure modes. Traditional authority is stable until it is interrupted, and does not recover well from interruption, because its whole claim was continuity. Charismatic authority has a built-in expiry date, since the person dies - Weber's routinization of charisma describes the scramble to convert personal loyalty into an institution before that happens, and how often it fails. Legal-rational authority is the most durable, because it survives the death of any individual, and it is vulnerable in a way the others are not: it collapses if the rules are visibly applied unequally, since the rules were the entire claim.
This is why a personalist regime and a rule-bound one fail so differently. The first has a succession crisis. The second has a legitimacy crisis every time a powerful person is seen to be exempt.
Legitimacy as a Coordination Problem
There is a second account that explains the dynamics better than any belief-based one. It treats legitimacy as something closer to a traffic convention than a moral judgment. On this view, developed by Russell Hardin among others, you comply not because you believe the regime is right but because you expect everyone else to comply. Defying a rule everyone else is following is costly and pointless. The government is a focal point: the answer everyone converges on because everyone expects everyone to converge on it.
This account earns its place by explaining the thing belief-based accounts handle worst - the speed of collapse. If legitimacy were a stock of accumulated approval, it would drain gradually and regimes would decline visibly. What actually happens repeatedly is that a regime looks solid for decades and then goes in weeks. That is the signature of a coordination equilibrium rather than an accumulated belief: nothing changes for a long time because everyone expects everyone else to keep complying, and then the expectation flips and everything changes at once. The 1989 collapses across Eastern Europe, the fall of Ceaușescu, and the sudden failure of the August 1991 coup in Moscow all have this shape - the underlying discontent had existed for years, and what changed in each case was the expectation about what everyone else would do.
It also explains why regimes are so preoccupied with public displays that seem to persuade nobody. A parade, a rigged election won by ninety-eight percent, a mandatory rally - these are not aimed at convincing you the regime is good. They are aimed at showing you that everyone else is complying, which is the belief the equilibrium actually rests on. The purest expression is a mass demonstration of support that everyone privately knows is coerced: it works precisely because everyone can see everyone else participating.
Process Matters More Than Outcome
The most robust empirical finding in this whole area is also the most practically useful. Tom Tyler's work over several decades found that people's willingness to accept a decision, comply with it, and regard the deciding body as legitimate depends far more on whether the process felt fair than on whether the outcome favored them. People who lose a case they consider fairly heard accept it. People who win a case they consider rigged do not gain confidence in the institution.
The components turn out to be specific and unglamorous: being given a voice before the decision, being treated with basic respect, evidence that the decision-maker is neutral rather than aligned, and an explanation of the reasoning. None of these require a favorable outcome and none are expensive. Their absence is what turns an ordinary adverse decision into a grievance against the institution rather than against the decision.
The finding replicates across courts, policing, tax authorities, regulators, and workplaces, which is unusual. It is worth being clear about its limits, since it is often overclaimed. Procedural fairness raises compliance and institutional confidence; the effect sizes in field experiments, particularly in policing, are frequently modest, and attempts to improve legitimacy through training alone have produced mixed results. Fair process is necessary and it is not sufficient, and it cannot substitute indefinitely for outcomes that a population regards as intolerable.
Rule by Delivery, and the Trap Inside It
A regime with no procedural claim can still generate real compliance by delivering results: rising incomes, order, functioning infrastructure, national standing. Political scientists call this performance legitimacy, and it is not a lesser or fake version. Rapid and broadly shared improvement in material life produces genuine acceptance, and several of the most durable non-democratic governments of the past half-century have rested substantially on it.
The trap is structural rather than moral. A regime whose claim is delivery has, in effect, signed a contract that is renegotiated continuously, and it holds no reserve for the period when delivery falters. Democracies survive bad decades because the claim is procedural: the government failed, and the mechanism for replacing it is the thing that is legitimate, so a recession removes a government rather than a system. A performance-based regime has no such separation - the downturn is an indictment of the arrangement itself, which is why such regimes are so sensitive to slowdowns that a procedural system would absorb as ordinary politics.
This produces a predictable set of moves as growth slows, visible in more than one country: a shift toward nationalism or historical grievance, which supplies a claim that does not depend on this year's figures; tighter control of information about performance; and a heavier reliance on the coordination mechanisms above. None of that is irrational. It is what you do when the foundation you built on turns out to be a variable.
Preference Falsification, and Why Rulers Are Blind Too
Where expressing dissent is costly, people conceal what they think. The resulting public picture is not a slightly noisy version of private opinion. It is systematically wrong in one direction. Timur Kuran's analysis of preference falsification draws out the consequence that makes this more than an observation about surveys. Because everyone conceals, everyone also infers from the surrounding silence that they are unusually dissatisfied, which raises the cost of speaking and deepens the concealment. The equilibrium is self-reinforcing, and it can hold for a very long time on almost no genuine support.
The consequence people usually take from this is that outside observers cannot assess the stability of closed regimes. The sharper and more interesting consequence is that the regime cannot either. Its own reporting chain is staffed by people whose careers depend on reporting satisfaction, its intelligence services return what the leadership has signaled it wants, and its elections are designed to produce a number rather than to measure one. A government that has successfully suppressed the expression of discontent has, by the same act, blinded itself to the level of it. This is a large part of why authoritarian collapses surprise the authorities involved as much as anyone.
It also explains the disproportionate response to small public defections. A single person standing alone with a sign is materially irrelevant and is nonetheless treated as a serious threat, because what they demonstrate is that defection is possible and survivable. In an equilibrium held together by everyone's estimate of everyone else, one visible counterexample is an attack on the load-bearing element.
What Elections Do
The usual account of elections is that they select good leaders and translate preferences into policy. They do both imperfectly and neither is their main structural contribution. The load-bearing function of an election is that it produces a result which the losing side accepts, and it does this by being a procedure agreed in advance whose outcome nobody controls. That is what converts a contest for power into a transfer of power without violence, which is historically the hardest problem in politics.
Seen this way, the critical moment in a democracy is not the vote. It is the concession - the point at which the losing side, which usually retains supporters, money, and often armed sympathizers, declines to contest the result by other means. Everything else in the system depends on that becoming routine, and the routine is what is actually fragile. A single refusal that goes unpunished changes what future losers believe is available to them, which is exactly the coordination logic from earlier running in reverse.
This is also why procedural legitimacy is worth more than it looks, including to winners. A government installed by a process the losers accept can govern with a fraction of the coercive capacity any other route requires. It can also hand over power later without its leaders facing the fate that awaits a deposed ruler in a system with no agreed exit. The value of the procedure is highest to the people who expect to be on the losing side eventually, which is everyone.
How Legitimacy Is Lost
The popular model is a scandal so severe that people withdraw their consent. Real cases rarely look like that. Legitimacy is usually lost through a slow decoupling of the rules as stated from the rules as applied, and the decisive perception is not that the powerful behave badly but that the rules are applied to some people and not to others.
This is why visible impunity is more corrosive than visible corruption. A system understood to be corrupt but predictable retains a kind of working legitimacy, because people can still model it and act accordingly. A system that proclaims equality before the law while demonstrably exempting a class of people has attacked its own foundation, since legal-rational authority claims nothing except that the rules bind. Each visible exemption is not merely one injustice; it is evidence about what kind of thing the system is.
The second common route is the failure to deliver a function so basic that no other claim survives its absence: physical safety, a currency that holds value, courts that resolve disputes, borders that mean something. Populations tolerate a great deal of ineffectiveness. But there is a floor beneath which the question shifts from whether the government is good to whether it is a government. Once that is being asked openly, alternatives that were previously unthinkable become discussable. People begin looking for whoever can actually provide the missing function, which is how militias, cartels, and religious authorities acquire quasi-governmental standing.
The Measurement Problem
The concealment described earlier is the deepest reason legitimacy resists measurement, but it is not the only one, and the practical problem is worth setting out separately. Survey questions about trust in institutions are the standard proxy and they are weak ones. They are highly sensitive to wording, they conflate approval of the current government with acceptance of the system, and they move with the news cycle in ways that genuine legitimacy does not. Declining trust figures across most democracies over several decades are a real signal of something; what they are a signal of is not established.
Behavioral measures are better and harder to obtain. Voluntary tax compliance where audit probability is low, willingness to use official courts rather than informal alternatives, whether people call the police when a crime occurs, and whether losing parties concede all reveal something a survey cannot, because they cost the person something. The most informative single indicator may be the simplest: what fraction of enforcement capacity a state needs to deploy to achieve ordinary compliance. A government that needs a great deal is, by definition, running on something other than legitimacy.
The uncomfortable implication is that legitimacy is most reliably observed at the moment it is tested, and that a system can appear entirely healthy on every available measure right up to the test. That is not a defect in the research. It is a property of anything that rests on mutual expectation, and it is the reason confident forecasts of regime stability - in either direction - deserve more skepticism than they usually receive.
Every government is a very small number of people issuing instructions to a very large number who could, in principle, decline. What closes that gap is not force, which is far too scarce. It is a shared expectation that the instructions will be followed - held in place by tradition, by procedure, by delivery, or simply by everyone's estimate of everyone else. That expectation is the most valuable asset any state has. It appears on no balance sheet, and its owners routinely discover its true size only on the day it is tested.
