1. Agreements and Conventions

Sugden reminds us that for Hume (1978), a convention is an agreement of sorts and since agreement and contract are cognate, ‘it should not be surprising that someone who finds Hume's analysis attractive is also attracted to social contract theory, or vice versa.’Footnote ² The two are not in conceptual conflict.

There are, of course, several sorts of agreements. If I meet a friendly stranger and he tells me that it is a lovely day, I will agree that it is indeed lovely. The agreement commits to nothing and stirs to no action. Let it be called Agreement 0. If I see all drivers driving on the right, I conjecture that they find such coordination a good thing and I agree with them, so much so that I, too, will do as they do. My agreement stirs me to emulation. But it is not an agreement with them. Call this Agreement 1. The corresponding convention yields mutual advantage, but that is no part of the reason why I adhere to it. The reason is my own advantage. If I did not recognise it and deviated from the convention, most others would still adhere to it. Under Agreement 2, I conjecture that all my neighbours would find it a good thing if their neighbours stole less, or not at all, from them, and they would rather steal less frequently, or indeed not at all, if the neighbours responded to such restraint by like restraint. I agree, and will join them in the corresponding convention by stealing less often or (if the most efficient of the alternative equilibria gets selected) not at all. Note that Agreement 2 is confined to a recognition of identical interests and an identity of responses to be expected from one another. The agreement does not entail a binding undertaking to perform an action or forbear from one. Mutual advantage results from the convention, but that is not the reason why I adhere to it. It is not an exchange of performances, nor of undertakings to exchange performances in the future, though it may look as if it were one. Finally, in Agreement 3 which serves as the keystone of the Humean edifice, we all agree to keep our promises. This is an undertaking that reacts upon itself to make it binding, and to render binding all other explicit undertakings. Agreement 3 has two variants, the unilateral or unrequited promise and the reciprocal promise or contract. Hume pays little attention to the former. The latter, rendering exchanges of non-simultaneous performances feasible, expands the scope for generating mutual advantage, beyond the constraint of the cash-on-the-barrelhead type and is a gift of incalculable value to civilisation.

By establishing the binding undertaking, Agreement 3 lays what looks like a footbridge leading from non-cooperative to co-operative game theory, and also to social contract theory. This terminates the crescendo of agreements.

2. Social Contract, or the Agreement to Submit

For all that conventions, contracts and the social contract are all agreements and linguistically cognate, there are gaps of discontinuity between each and a chasm of unfathomable depth between contract and social contract.

Linguistic cognateness is no excuse for ignoring the incompatibility in what two closely related terms signify. In convention, including its most stringent form, the contract, individual choice prevails throughout. The social contract, on the contrary, is an hypothetical agreement by all members of a society (the ‘all’ invoking an idealised unanimity, to be treated with Wicksellian tolerance) to adopt a meta-rule, a rule of rule-making or constitution, by which non-unanimous decisions can be reached that are binding for all. The social contract legitimises collective choices to which all individuals have committed themselves to submit.

The social contract differs in its outward form according to who, in the long line of its authors from Aquinas to Rawls, is providing the text. In Hobbes, the contracting parties create Leviathan who enforces their covenant with the sword. In Locke, they deal with a sovereign in place who agrees to do certain things and not to do certain others, his subjects agree to obey him as long as he keeps his side of the bargain and may revolt if he does not. In Rousseau, all agree to conform to a General Will that is not necessarily or even probably a will of their own. In Rawls, a duty of fairness persuades all to purge their minds of all that might otherwise induce them to disagree about how the ‘background institutions’ should distribute the basic good things of life. Since they contrive to ignore all relevant differences between them, their social contract is an agreement of one standardised person with himself. In Nozick, all entrust their quarrels and powers to protective agencies that are destined to merge and form the state.

Real-world social contracts in our time, embedded in constitutions, all provide for rule making by non-unanimous vote, the result passing for collective choice entitled to pre-empt or override individual choice. Individuals accept this principle of submissions. They do so in the putative interest they have in the common good (maximisation of ‘the’ social welfare function or other words to a similar effect) that collective choice pursues.

3. Alternative Redundancies

For disciples of Hume, collective choice should be anathema for two major and some minor reasons. One major reason is that the idea of individuals all agree to surrender their autonomy and to endow government with power to impose collective choice – instead of the Humean idea of acquiescing in the government that, by ‘long possession, present possession or conquest’Footnote ³ is already in place, – goes right against the Humean grain. The other major reason takes a little longer to state. Hume's great trinity of spontaneous conventions, ‘the stability of possession, its translation by consent and the performance of promises’Footnote ⁴ satisfy the enabling condition for society to exist. Each convention is brought forth by an equilibrium selection mechanism, a ‘game’ whose solution is payoff-enhancing, advantageous to the ‘players’. In some of these game-like interactions, a free rider option is a temptation to deviate from the equilibrium. Since ascertainable historical and empirical facts testify to the survival of such conventions, there must be enough ‘players’ using contingent retaliatory strategies to spoil the free rider payoff and keep deviation in check. The mechanism of generating advantageous coordination equilibria that are self-enforcing thanks to their incentive structure or to retaliation that deters free-riders, calls forth not only the great Humean trinity of ownership, transfer and promise, but a much larger set of greater or smaller conventions that can generate mutual advantage. The set protects against torts, discourages nuisances and scolds incivilities. Such a set of conventions is the foundational institution of ordered anarchy. In a closed society, this set of convention suffices to keep the peace and to uphold the social order, and makes the Hobbesian government redundant, just as the creation of such government by social contract makes the set of conventions redundant. Humeans would have no hesitation in deciding which of the two is redundant.

4. The Contractarian Evasion

Objectors to this condemnation of government usually refer to the small group-large group problem. (Hayek, for one, does so, perhaps making too much of it. Large groups, except crowds, are usually agglomeration of small groups). Hume was perfectly aware of the small group-large group problem. He nevertheless judged that a society ruled only by its conventions could function as an ordered anarchy if left alone. He declared unambiguously that ‘…I assert the first rudiments of government to arise from quarrels, not among men of the same society, but among those of different societies.’Footnote ⁵

Interestingly, this squares with the experience of modern-day philosophical anarchists whose main difficulty seems to be in showing how a society without government could provide for its own defence against foreign aggression.

Can Humean social theory make the social contract redundant yet be compatible with it? Sugden evades the obvious fault of trying to have it both ways separating social contract theory proper (incompatible) from contractarianism (compatible). The separation, if genuine, saves his thesis, but inspection easily reveals that the separation proves impossible where it would really matter.

Briefly, for Sugden contractarianism is the social theory of James Buchanan. It is defined mainly by what it is not: it is not a social contract with the object if pursuing the common good, maximising some conception of welfare and ensuring the supply of public goods. On these professed negatives, it does look separate from the standard understanding of the social contract. In contractarianism, we are told, ‘the role of government is to implement voluntary exchanges.’Footnote ⁶ Presumably, exchanges are ‘implemented’ by the parties to the exchange. Does Sugden mean that the role of government is to exchange, or is it to facilitate exchanges by individuals? Be that as it may, no protestations of self-restraint and modesty of ambition alters the fact that contractarianism, like the other social contracts, entails agreement on a constitution permiting to make rule-bound social choices to which the contracting parties accept to submit. By applying the collective choice rule, one part of society can pre-empt individual choices in order, a.o. to extract resources from the other part. No constitution known to man fails to do so. It would deny its own raison d'etre if it did fail. The redistributive function is consistent with the principle of submission at the heart of any social contract, and it has enough merit to attract some bona fide defenders. Buchanan himself asserts that even those whose resources are habitually extracted approve of this being done, for tier own future is shrouded in the ‘veil of uncertainty’ and they might well expect redistribution to work in their favour one day. (One wonders what an insurance underwriter would make of this argument). In any event, the contractarian constitution not only ‘implements voluntary exchanges’, but keeps engendering collective choices that progressively reduce the scope of these voluntary exchanges relative to the mandatory ones the choice mechanism is geared to multiply.

Imputing to individuals a rationally calculated consent to a constitution, – any constitution – that must produce this result and has in modern history invariably produced it, is inseparable from the contractarian hypothesis. Maybe people do rationally and near-unanimously seek such an outcome; one cannot conclusively deny it. But we should not be asked to believe, as Sugden asks us to do, that contractarianism is something apart from social contract theory is immune from the latter's vices, so much so that one can without intellectual embarrassment be both a Humean and a contractarian.