In his justly famous essay Two Cheers for Democracy (1951),1 E.M. Forster has withheld from it his full approval. He was not ungenerous. Hindsight since he wrote shows some of the depressing results of confidently putting democratic principle in practice. Its mechanism confers political power by anonymous voting. This forces rivals for power into periodic bidding contests. The bidders offer resources to one part of society in exchange for its votes, and take these resources from other parts of society, a type of transaction that democratic principles find irreproachable and perhaps praise worthy. In this repetitive process, governments are driven to keep expanding the state, using it as a machine for absorbing resources, transforming them, and handing them out again to others in exchange for electoral favours. In the bidding, the sums involved do not stop at taxable capacity. They spill over into public debt. Ever heavier indebtedness renders government difficult and the economy sluggish. Welfare promises made to attract votes cannot be fully met. The electorate grows bitter and bloody-minded and willing to listen to hare-brained extremists.

Classical liberalism is widely understood to offer the remedy to this deeply discouraging prospect. It promises limited government, or more precisely a limit to the size of the state that is strictly fixed by the operation of timeless, permanent principles acting as a firm barrier. If government is so limited the democratic process is held at some agreeable resting point and does not run its ominous course. The cheers classical liberalism deserves depends on how well, if at all, it can reliably serve to complement and restrain the democratic hypertrophy.

The Harmful Harm Principle

Political doctrines tend to be labelled with a brief form of words that suggest their basic drive. The dictatorship of the proletariat or the public ownership of the commanding heights of the economy point to varying types of socialism. Free exchange between consenting adults, or maximum liberty compatible with the same for others, plainly invokes liberal tendencies, though not as clearly as might be. Friedrich Hayek’s “minimum necessary coercion”2 invokes any regime that does not apply more coercion than it finds necessary in pursuit of its objectives, what ever they may be. One often suspects that the road to unpleasant consequences is paved with well-intentioned slogans. One of the most celebrated forms of words intended to serve as a foundation of liberal doctrine is the Harm Principle enunciated by J.S. Mill: “The only purpose for which power can be rightfully exercised over any member of a civilised community, against his will, is to prevent harm to others.”3 Unfortunately, almost any prima facie harmless act that uses resources is harmful to every body else who lose the opportunity to use these resources. If this hair splitting generalisation strikes the reader as too abstract, he might be more impressed by the down-to-earth reason that if no power is exerted on a person’s doings he might do good to persons or groups of his own choice, leaving others by the wayside. If this defined the rightful liberal order, many people would reject it as being grossly unfair and capricious, failing to realise that liberal doctrine may not have the forcible establishment of fairness, an almost undefinable notion, as one of its prime objectives.

For more discussion of John Stuart Mill’s Harm Principle, see The Awful Mill, by Bryan Caplan on EconLog, March 23, 2012.

Be that as it may, the Harm Principle is taken seriously a quarter, a half, or three quarters of the way. Thus it conditions and colours policies that public opinion considers desirable liberal features. Two of the main ones are security and solidarity. Security, imposed on people and their institutions to avoid harm, discourages risk-taking, breeds bureaucracy, and gives free rein to nannying. Solidarity, imposed on people who might not want to practice it of their own freewill, lends the state an immensely powerful mandate to seek equality of resources and its moral rightfulness is more than dubious.

Rules and Rules-Enforcement

Even a lenient judge must condemn the term “government for the people by the people” as a malignant means of fooling the simple-minded. Collective choice is made for a collectivity but not by it. It is made by a part smaller than itself. The choice is made by the whole collectivity only if it is unanimous. Unanimity is a limiting case where collective and individual choice merge and the individual can veto or can consent, but has no other choice. At the other end of the scale the dictator can decide for the collectivity, failing which individuals can make their own choices. In the nature of the case, failing unanimity, collective choice is always stronger than individual choice and can override it. As a convenient simplification the universe of feasible states of affairs can be divided into a collectively and an individually chosen part. The collectively chosen part can encroach on the individual one, but not vice versa. At best, the two halves can stand still on either side of a barrier. This barrier is of central interest to classical liberalism.

A dictator’s choices can be random and half-crazy, but they are in an informal sense rule-bound in that they must all be made by the dictator in his dictatorial capacity. Generally, a collective choice is made by a choice-rule. For instance two houses of a parliament must vote for it in some particular manner. The two houses must come from somewhere. For example, their membership may originate in the system of land ownership. The general form, however, is that of a rule that regulates the rule of choice making. In modern history, this rule of rule-making tends to be embedded in a constitution, of which it forms what, for our present purpose, must be regarded as its core. The scope of this essay is deliberately limited. It does not treat two basic questions whose relevance extends well beyond classical liberalism. One of them asks how (except by supposing a miraculous unanimity) a rule of rule-making that has no rule higher than itself, can be endowed with legitimacy. The other, and even more fundamental question is whether collective choice can be legitimate at all. These questions are side-stepped here and it is taken to be the case that most existing states have a constitution whose function it is to lay down what the state, steered by its government may, must, and must not do by making collective choices.

How The State Would Enforce The Thwarting of Itself

State actions have three aspects: enactment, execution, and enforcement. Means to provide the latter, such as courts, police, and army, are available and controlled by it to some reasonable degree. Enforcement entails punishment, reparation, or both. It is not necessarily perfect, but probabilistic.

A major dilemma in enforcement is the case when the interest of the enforcing agent is that the successful enforcement of a certain rule should have very low probability. To illustrate, consider the case of a constitution which forbids the government to drive the share of collective choice in the universe of potential choice beyond a certain barrier by increasing public expenditure. Increasing public expenditure in an almost permanent fashion is, of course, the essential means used in the bidding process for political power by which a government or its rival can gain and maintain control over the apparatus of the state and its capacity for rule-enforcement.

Protecting individuals and their resources from collective choice by rule-based means would entail a constitutional clause limiting public expenditure. Constitutions do not have such clauses. They have a vacuum where the clause should be. (One exception of sorts is the clause in the German constitution4 which limits the federal deficit, though it does not limit public expenditure).

This is perhaps just as well, for the operation of such a clause is hard to conceive. It would thwart persistent objectives of the government pursued by the state and would be enforced by the state. The state would be disobeying the state and would be punishing itself for doing so. Requiring the enforcer to suppress what it wishes to be the case is a logic-cripple, a deformity, and a generator of deep disorder. It would cover under a tissue of lies the shrinkage of the area of freedom. In fact, it does not take place for constitutions simply do not ask for it.

The potential of perverse enforcement shows constitutional limitation of government to be deeply untrustworthy. The logical misfit that spoils it can be eliminated if we see the constitution, as Thomas Schelling suggested that we should,5 not as a rule but as a vow. The vow can be amended, abandoned, and renewed by the subject, as it only exists while his own will maintains it. No real enforcement problem arises. Logic is satisfied but classical liberal doctrine is revealed to suffer from great weakness in its very base, the constitution that would limit government. Not very many cheers seem to be due to it.

For further information about the ideas Jasay explores in this essay, Anthony de Jasay, The State, Library of Economics and Liberty (Indianapolis: Liberty Fund, 1998) ; Justice and Its Surroundings, Online Library of Liberty (Indianapolis: Liberty Fund, 2002); Social Contract, Free Ride: A Study of the Public Goods Problem, Online Library of Liberty (Indianapolis: Liberty Fund, 2008).

The critic is under no obligation to add alternative proposals for improvement. A quite hint, however, might perhaps be added to the criticism. Game theory tells us that repeated interactions is apt to generate reciprocal accommodations among players which enhances their payoffs. Discipline is enforced by players, in defence of their co-operative payoffs, punishing and excluding free-riders. The resulting system of conventions maturing into rules promises to serve no worse than centrally enforced collective choices, and constitutes ordered anarchy. Skeptics would presumably be all too ready to say that “this would never work in practice”, but how do they know?


E.M. Forster, Two Cheers for Democracy (London: E. Arnold 1951).

For Hayek’s views about coercion see Chap. 9 “Coercion and the State” in Friedrich A. Hayek, The Constitution of Liberty (The University of Chicago Press, 1960).

John Stuart Mill, On Liberty (1859), Chap. 1 “Introductory”, paragraph I.9; also available in The Collected Works of John Stuart Mill, Volume XVIII—Essays on Politics and Society Part I, ed. John M. Robson, Introduction by Alexander Brady (Toronto: University of Toronto Press, London: Routledge and Kegan Paul, 1977).

The German Constitution (Grundgesetz) was amended in 2009 to include a balanced budget provision known as the “Schuldenbremse” (debt brake) which would limit the deficit to no more than 0.35% of GDP. It comes into effect in 2016.

On vows, see Thomas C. Schelling, “Ethics, Law, and the Exercise of Self-Command”, The Tanner Lectures on Human Values, delivered at the University of Michigan, March 19 and 21, 1982. PDF.


*Anthony de Jasay is an Anglo-Hungarian economist living in France. He is the author, a.o., of The State (Oxford, 1985), Social Contract, Free Ride (Oxford 1989), Against Politics (London, 1997), and Justice and Its Surroundings (Indianapolis, 2002). His most recent publications include Political Philosophy, Clearly (Indianapolis, 2010) and Political Economy, Concisely (Indianapolis, 2010). His next volume, Economic Sense and Nonsense: Reflections from Europe, 2007-?2012 (a volume in The Collected Papers of Anthony de Jasay), edited and with an introduction by Hartmut Kliemt, is forthcoming from Liberty Fund.

The State is also available online on this website.

For more articles by Anthony de Jasay, see the Archive.