What Liberalism Is

§1: Liberalism as the solution to a second-order Hobbesian problem

“What may now be meant by the word “liberal” is anyone’s guess.”

— Michael Oakeshott, Rationalism in Politics and Other Essays, pp.439–40

I. Introduction

I consider myself an intellectual heir to a liberal tradition stretching back centuries. Yet as Oakeshott suggests, the term “liberalism” has extended to cover such a wide and often conflicting range of positions that it risks losing any determinate meaning at all.

My claim is that liberalism nevertheless names an identifiable family of ideas: core principles that explain much of this variation, even if they do not encompass every view that travels under the liberal banner. This piece aims to clarify what I take to be liberalism’s first principles, to examine their historical and philosophical sources, and to trace their implications for economic and social life. In doing so, I hope to show why liberalism, as I understand it, is a genuinely attractive doctrine, one that addresses real and persistent problems arising from the human condition as it presently exists.

This is not an attempt to survey liberal theory in its entirety. Instead, I seek to reconstruct liberalism from the ground up: to identify the problem it aims to solve, to sketch the institutional form that responds to that problem, and to draw out the conceptual commitments implicit in those institutions. Along the way, I draw on key figures from the liberal tradition where their ideas illuminate the argument, while also rejecting (implicitly or explicitly) certain tempting but mistaken paths. My aim is not to adjudicate every intra-liberal dispute, nor to systematically engage counterarguments from within or outside the tradition, but to motivate and advance a particular account of liberalism. For that reason, some approaches commonly associated with liberalism — certain contractarian frameworks, for instance — receive relatively little attention here, on account of their limited influence on the version of liberalism I aim to defend.

II. Violence, Scarcity & The Problem of Plural Preferences

Human beings inevitably find themselves in a world with problems: states of affairs that fall short of how they would prefer things to be. Each individual therefore confronts aspects of the world whose modification would, from their own perspective, constitute an improvement. From this mismatch between the world as it is and the world as it is wanted arises a standing impulse to act upon one’s surroundings, reshaping outcomes so that they more closely track one’s preferences.

Yet individuals acting alone are remarkably impotent when it comes to reshaping the world. A solitary human being is not particularly strong, nor fast, nor knowledgeable¹. Our strength as a species comes from our capacity for coordination, for pooling our knowledge and effort to achieve what none of us could accomplish alone.

Unfortunately, coordination is rarely straightforward, because people value different things. Even when our aims aren’t directly opposed, they still collide in practice since resources are scarce. Under these conditions, the potential for violence is built into human co-existence: when someone else’s actions frustrate your plans, or occupy resources you want, this creates incentives to coerce them to get your way. Violence “resolves” disagreement by removing the other party’s capacity to choose and act².

When violence is available as a general instrument for settling conflict, social life tends toward a negative-sum equilibrium. If anyone can seize what others have produced, there is little reason to invest time and effort producing it. Worse, one’s limited time and effort must be spent protecting oneself from predation, rather than generating the social surplus that could advance everyone’s ends through productive cooperation.

III. The Hobbesian Patch

The modern state can be understood as an institutional response to this Hobbesian dilemma. By centralising and monopolising the use of force in a single, publicly recognised³ authority, it deprives individuals of violence as an instrument of bargaining. This Hobbesian “Leviathan” does not eliminate conflict or disagreement, but it changes the strategic environment in which they occur, by raising the cost of predation and lowering the need for constant self-defence. As a result, people can redirect their energy — for example, planting because they can now expect to harvest.

Once this institution exists, however, it commands extraordinary power: it is the lone entity permitted to deploy violence. To fulfil the purpose that motivates its existence, it must use that power to preserve its monopoly. When individuals use force against each other, the state may use force against them; and by definition, there is no further recourse beyond it, closing the loop of violence rather than allowing it to escalate.

Yet violence is not always negative-sum, so the utility of coercion is not exhausted by this narrow, self-maintaining rule. It can also serve as a useful, and arguably necessary, tool for other purposes: defence against external adversaries; enforcing contracts, so that promises can be relied upon, because breaches carry agreed-upon penalties neither party can legitimately inflict privately; and the protection of private property rights, whose stability rests on the credible threat that violations will be met with enforcement. Where these functions require violence, they must fall to the state.

One might infer from this the principle that whenever violence would be useful for solving a problem, the state may — or must — intervene to solve it. Hobbes leans in this direction: so long as the sovereign preserves peace, its authority is legally absolute, constrained by no higher court and no enforceable limit beyond its own judgement of the common good. But this principle is far too strong. If you are the only party permitted to wield violence, coercion becomes an unusually general-purpose instrument, capable of resolving not only disputes, but a wide range of practical obstacles, by compelling others to act in accordance with your aims.

If political authority is allowed to extend this far, and we accept that whoever controls the state can impose their will — whether justified by bare power or by contested claims about what is “good” — then control of the state becomes existentially valuable. The returns to sovereign power become unbounded, as those who possess it can privilege their own views and entrench their values over all others. The original problem of negative-sum conflict is not eliminated, but merely displaced, resurfacing at the higher level of struggle over who gets to direct the state’s coercive apparatus.

IV. From Rule by Command to Rule by Rules

Thus having succeeded in establishing a monopoly on violence, it immediately becomes necessary to place limits on what it can actually do. Otherwise, its absolute authority simply transposes us from the first-order problem of fighting over resources, to the second-order problem of fighting over the authority that decides resource use.

Indeed, one reason that centralising coercive power in a single entity is desirable is that it simplifies this task: if there were many sovereigns, each would need to be constrained separately. As it stands, there is only this one Leviathan to constrain. Rights, on this account, are one instrument of constraint. They mark out domains of individual action into which the state may not intrude, regardless of who holds power or what ends they wish to pursue. Norms, traditions, constitutions, and the separation of powers similarly function to limit how coercive authority may be exercised.

None of these constraints are discoveries of the natural order, deliverances of pure reason, or gifts from God. They are contingent social technologies, adopted because experience suggests they reduce the risk that concentrated coercive power becomes a vehicle for arbitrary domination. Their content is therefore shaped by circumstance: a right to healthcare, for instance, is unintelligible in a society that possesses no medical knowledge. What rights, norms, and institutional limits we recognise at any given time reflects the problems we confront and the solutions practically available to us.

The core proposition of liberalism, therefore, is this: that the state should be indifferent between different conceptions of the good life⁴. This is not the positive claim that all ways of living are equally admirable. Rather, it is a structural restraint on how collective force may be used. Given persistent disagreement about values and priorities, and given that the state’s distinctive instrument is violence, the relevant question is not which conception of the good is “true”, but which principles of political authority can be justified to people who reject one another’s ultimate ends.

For this reason, liberalism is naturally allied with the rule of law. The state exists not to command its citizens towards the good, but to govern by general rules under which each can pursue their own ends. Coercion, where necessary, must operate through stable and knowable procedures rather than ad hoc moral judgement, lowering the stakes of political victory and the incentive to treat government as a prize. Where the state instead acts as a moral executor — using force to realise a contested ideal — it recreates the Hobbesian dynamic it was meant to suppress, turning plural preferences into a fight for command. Liberalism, then, is the practical solution to the problem of living together when no one gets to be sovereign over everyone else’s values⁵.

V. Liberalism After Leviathan

This first-principles, bottom-up framing illustrates how liberalism as a set of institutional practices emerges as a solution to a second-order Hobbesian problem — that is, a solution to a problem generated by the solution to the Hobbesian problem. Historically, this settlement emerged from the European wars of religion, which exposed the need to prevent the state from enforcing contested faiths.

But from this practical solution, and from the reasoning behind it, we can extract a more general conceptual orientation, constitutive of liberal principles. These principles are not prior axioms from which liberal institutions are deduced; rather, they are abstractions distilled from institutional practice. Once articulated, however, they acquire a degree of normative independence, functioning as standards by which existing institutions can be evaluated, revised, or rejected. More precisely, reflecting on why liberal institutions succeed yields a criterion of evaluation grounded not merely in convention, but in the fact that their success appears to track something fundamental about the structure of human coexistence and cooperation⁶.

At the core of this orientation is the recognition of individuals as agents capable of forming, revising, and pursuing their own ends, rather than as mere instruments for the purposes of others. Because coercion operates by overriding an agent’s capacity to choose, its use demands justification. Rights, as previously described, are one institutional expression of this requirement, limiting what may be done to individuals in the name of collective aims. More broadly, they instantiate a prior idea of respect for agency, with the state supplying the coordinating framework within which those limits can be made mutually compatible among many agents with competing plans.

That said, nothing here requires the stronger Kantian thesis that political authority is constitutive of freedom itself, or that rights come into existence only through their legal enactment. Nor does it require, as Rawls does, that the state act solely on reasons all can affirm as correct. The constraint is instead structural: coercion is politically destabilising when it is justified by appeal to a contested ordering of ends.

Once this is acknowledged, the scope of restraint must be both universal and revisable. Universal, because it must apply to persons as such, rather than to members of particular cultures or traditions, since any selective standing would itself encode a disputed hierarchy of value. And it must treat institutions as corrigible, because fixing a single authoritative resolution to questions about the good would entrench one perspective against others, reintroducing the very conflict restraint is meant to contain.

This orientation accordingly gives rise to a recognisable cluster of commitments, articulated by John Gray in his 1998 book Liberalism:

It is individualist, in that it asserts the moral primacy of the person against the claims of any social collectivity; egalitarian, inasmuch as it confers on all men the same moral status and denies the relevance to legal or political order of differences in moral worth among human beings; universalist, affirming the moral unity of the human species and according a secondary importance to specific historic associations and cultural forms; and meliorist, in its affirmation of the corrigibility and improvability of all social institutions and political arrangements.

As the foregoing illustrates, these features are not independent moral postulates, but the conceptual residues of the attempt to govern a world of plural preferences.

VI. Applications To Economic & Social Life

These liberal principles have implications that extend well beyond the formal structure of the state. Applied to economic and social life, the liberal orientation enables us to extend the same anti-violence, pro-coordination logic into non-political domains.

In economic life, market exchange allows individuals to improve their situations by mutual consent, without requiring agreement on ultimate ends. For a buyer and seller to transact, they need not share a conception of the good life; they need only find terms acceptable to both. This mirrors liberalism’s core political insight: that coordination among agents with divergent ends is possible without resolving the disagreements⁷. Prices function as signals that coordinate the plans of millions of dispersed actors, none of whom needs to know more than their own circumstances.

This is Hayek’s central insight: the price system is a mechanism of decentralised communication, transmitting information about the relative scarcity of goods and the intensity of competing demands in a form that no individual or planning authority could aggregate on its own. Markets do not merely allocate known resources among known uses; they discover possibilities that were not previously apparent, as entrepreneurs experiment with new combinations and consumers reveal preferences through their choices. The liberal case for markets is thus an epistemological case: central planning fails not because planners lack good intentions, but because the knowledge required to coordinate an economy is dispersed, tacit, and constantly changing, and is therefore in principle unavailable to any single mind or institution.

In social life, too, liberalism provides a framework for peaceful coexistence under pluralism: a way of living together despite fundamental disagreement about how life should be lived, without resorting to force. If the state refrains from imposing a substantive vision of the good, individuals are free to pursue “experiments in living” — to try out different forms of life, to learn from each other’s successes and failures, and to discover⁸ for themselves the arrangements most conducive to their own flourishing, so long as they don’t harm others doing the same. The conditions under which people find and revise their own conceptions of the good — freedom of thought, freedom of association, freedom of expression, absence of enforced conformity — follow directly from liberalism’s commitment to non-coercive pluralism.

The economic and social applications of liberal thought are thus not separate doctrines bolted onto a political theory, but expressions of the same underlying logic. In each domain, the liberal insight is the same: that the dispersed, experimental, self-correcting processes made possible by individual freedom will, over time, generate outcomes superior to those achievable by centralised direction — not because individuals are infallible, but because decentralised systems are better equipped to detect and correct the errors that inevitably arise. For this reason, the liberal maintains that by binding coercion to general rules and leaving room for many ends to be pursued at once, we can turn the diversity of human aims from a standing cause of conflict into a renewable source of peaceful discovery and shared flourishing.

---

1. If you insist, fine — I concede that I am an exception. But it holds for most of the rest of you!

2. Of course, power (roughly, the ability to privilege your ends over others’) takes many forms besides overt violence or direct coercion. I don’t claim that violence is the only problem, but that it is structurally prior. Where the use of force is unconstrained, disputes over all forms of power are ultimately settled by violence, rendering separate efforts to regulate economic, social, or cultural domination unstable or incoherent. Hence the problem of violence is fundamental, and must be addressed first, else no durable resolution of the rest is possible.

3. As Hobbes argues, the Leviathan’s authority is acceptable to all is because no individual is so much stronger than others that they can reliably dominate by force: even the strong can be brought down by a small coalition. Since anyone can be overpowered by several others acting together, it is in everyone’s interest to remove private violence as a bargaining tool.

4. Note that I am careful here to say, “indifferent between different conceptions of the good life”. That is not the same as being indifferent between all axiological systems! Liberal institutions presuppose structural values, most notably, individual agency/autonomy and the capacity to form and revise ends (which despite my arguments, would certainly be rejected, for instance, by Aristotle). The claim is only that the state should remain neutral among rival substantive accounts of how that autonomy ought ultimately to be exercised.

5. I will address democracy in a separate piece. For present purposes: while compatible with this framework, democracy is not sufficient on its own. Even if free and fair elections decide who wields the state’s power, unconstrained majoritarianism fails to solve the second-order Hobbesian problem: it simply relocates it from “who controls the state”? to “which coalition controls the government?”. This raises the stakes of politics and invites the domination of minorities or dissidents. If democratic mandates licensed the imposition of a comprehensive conception of the good, I doubt democracy is a stable equilibrium, as incumbents would face incentive to entrench their power against successors wielding equivalent authority. Unsurprisingly, no existing polity relies on simple majority rule; stable democracies restrict what electoral majorities may permissibly do, through entrenched limits on authority that are insulated (to varying degrees) from ordinary political contestation.

6. As Samuel Hammond has put it (emphasis mine):

As evinced by the proto-liberalism of Thomas Hobbes or the natural law theory of Hugo Grotius, liberalism was in some sense discovered through the logic of positive-sum games, not invented by Enlightenment philosophers. They merely gave that logic a vocabulary.

7. As written, this liberal case for markets allows some rule-like intervention into markets, and does not entail “laissez-faire”. Here, I have only sketched why market institutions form the default economic expression of liberal commitments, with intervention understood as a secondary modification to, rather than a replacement for, a market framework. In a later piece, I will consider classical liberal political economy in a far more detailed manner.

8. The repetition in these two paragraphs of discovery and experimentation is deliberate (if not subtle). Both applications stem from the same epistemological structure. If we take seriously the claim that people are agents capable of forming, revising, and pursuing their own ends, then we must also recognise that the knowledge of what constitutes a flourishing life for any given person is not something that can be specified in advance, even by that person. It emerges through the lived process of pursuing one plan rather than another, encountering others who live differently, and adjusting in response to what one learns. The knowledge is not separable from the practice: you cannot know what kind of life suits you without actually attempting to live it. Individual freedom, on this account, is not merely a political accommodation but an epistemological requirement: the freedom to experiment is the condition under which the relevant knowledge can come into existence at all.

It is the same point (structurally and substantively) in the case of markets. What should be produced to satisfy people’s shifting, heterogeneous preferences is not fully knowable in advance. The point is not merely that economic knowledge is dispersed and tacit — that a planning authority cannot collect what entrepreneurs already know — but that the competitive process is itself the primary source of such knowledge. That is, the possibilities that market actors discover through rivalrous experimentation are not pre-existing facts awaiting collection, but generated by the process of competition itself — of testing offers against actual uptake, prices, and alternatives — and do not exist in its absence.

In both domains, freedom-enabled experimentation cannot be replaced by directive authority, since the experiment is itself the process through which the ends become known.