The Evolution of Modern Liberalism

From Locke's property rights to algorithmic accountability: how the liberal tradition rebuilt itself four times in three centuries, and what the fifth rebuild will have to solve.

Escrito por Ottavio Malatacca · Publicado 2026-02-11 · Actualizado 2026-08-30

Liberalism is the most successful political tradition of the modern era and the one that is hardest to pin down. It has been the creed of revolutionaries and of the establishment, of free traders and of welfare states, of secularists and of religious minorities seeking protection. That elasticity is not incoherence — it is the consequence of a tradition organised around a question rather than an answer: how much power may be exercised over an individual, by whom, and on what justification?

A tradition defined by a question, not a programme

Most political families are defined by a destination. Socialism points at collective ownership; conservatism points at the preservation of an inherited order; nationalism points at the sovereignty of a people. Liberalism points at a constraint. It says that whatever destination a society chooses, there are things that may not be done to individuals along the way, and that power must always explain itself.

This is why liberalism keeps reappearing on opposite sides of political arguments. A liberal in 1830 was a radical demanding the vote for the middle classes; a liberal in 1930 was arguing about unemployment insurance; a liberal in 2026 is arguing about who controls a recommendation algorithm. The constraint stayed constant. The concentrations of power it was aimed at kept moving.

Understanding liberalism therefore means tracking the target. Each time a new form of concentrated power appeared, the tradition split: one wing insisted the original enemy was still the state, the other insisted the enemy had changed. Those splits produced the four rebuilds described below.

The first rebuild: the state as the enemy (1689–1830)

Classical liberalism was written against absolute monarchy. John Locke's Two Treatises of Government (1689) made three moves that everything afterwards depends on. He argued that political authority is not natural but constructed; that it is constructed for a limited purpose, the protection of life, liberty and estate; and that when it exceeds that purpose it forfeits its claim to obedience.

The practical programme that followed was narrow and specific: habeas corpus, jury trial, freedom of conscience, an end to arbitrary taxation, and the security of property against seizure. Property mattered less as an economic doctrine than as a political one. A person with independent means could not be starved into compliance. Property was, in the seventeenth-century argument, the material basis of the capacity to say no.

The limits of this first liberalism are as instructive as its achievements. Its universal language coexisted comfortably with property qualifications for voting, with the exclusion of women from civic life, and with colonial rule and slavery. Locke himself invested in the Royal African Company. The tradition would spend the next three centuries being held to its own stated premises by people it had excluded — which is, in practice, how it grew.

The second rebuild: the market as a new concentration (1830–1945)

By the middle of the nineteenth century, industrialisation had produced a form of power Locke had not anticipated. A worker in a Manchester mill was formally free — no monarch commanded him — and materially unable to refuse almost anything. The state was not the only institution capable of coercion.

John Stuart Mill is the hinge figure. On Liberty (1859) contains the most quoted formulation of the classical position, the harm principle: the only purpose for which power can rightfully be exercised over any member of a civilised community against his will is to prevent harm to others. But Mill also insisted that liberty is worthless to someone without the education, health and security to use it, and in later editions of his Principles of Political Economy he moved steadily towards what he called a qualified socialism.

This produced the split that still organises the tradition. If freedom means only the absence of interference — what Isaiah Berlin later labelled negative liberty — then taxation to fund schooling is itself an infringement. If freedom means the actual capacity to direct one's own life — positive liberty — then a state that provides schooling is enlarging freedom, not shrinking it. Both readings are recognisably liberal. They generate opposite budgets.

The only freedom which deserves the name is that of pursuing our own good in our own way, so long as we do not attempt to deprive others of theirs.

John Stuart Mill, On Liberty, 1859

The third rebuild: the embedded compromise (1945–1980)

The post-war settlement was an attempt to hold both readings at once. Its architecture was distinctive: open trade between countries, capital controls and full-employment policy within them, and a welfare state to absorb the shocks. The economic historian John Ruggie called this embedded liberalism — markets deliberately embedded in social obligations rather than left free-standing.

John Rawls gave the arrangement its most rigorous philosophical defence in A Theory of Justice (1971). His device of the original position asks what principles you would choose for a society if you did not know which position in it you would occupy. Rawls argued you would choose extensive basic liberties for all, and would permit inequalities only where they improve the position of the worst off. It is a liberal argument — it starts from individuals and their rights — that arrives at a substantially redistributive conclusion.

The counter-argument was made with equal force. Friedrich Hayek's objection in The Constitution of Liberty (1960) was epistemic rather than moral: no central authority can assemble the dispersed, tacit, local knowledge that a price system aggregates automatically, so planning fails not because planners are wicked but because the information does not exist in a collectable form. Robert Nozick's objection in Anarchy, State, and Utopia (1974) was about consent: redistribution overrides transfers that people freely chose to make, and no pattern of holdings can be maintained without continuously interfering with those choices.

The fourth rebuild: neoliberalism and its exhaustion (1980–2016)

From the late 1970s the Hayekian argument won the policy contest across most of the democratic world. Deregulation, privatisation, independent central banks, and the removal of capital controls became the shared programme of parties that called themselves both left and right. Its defenders point to falling global absolute poverty and the integration of hundreds of millions of people into world markets. Its critics point to stagnant median wages in rich democracies, the 2008 financial crisis, and a sharp rise in the share of income accruing to capital.

The politically decisive fact was not inequality as a statistic but the loss of a promise. The embedded compromise had offered security in exchange for accepting market outcomes. When the security thinned but the market outcomes remained, the bargain looked one-sided. The populist wave of the 2010s is unintelligible without that broken exchange, and we examine its mechanics in a separate essay on populism in the digital age.

The fifth problem: power that is neither state nor market

The concentration of power now in question fits neither of the tradition's two historical targets. A platform that mediates the speech of three billion people is not a government: it holds no territory, commands no army, and answers to no electorate. It is also not a market participant in any ordinary sense: it sets the rules of the space in which other participants compete, and it can change those rules unilaterally and without notice.

Liberalism has no settled vocabulary for this. The tradition knows how to constrain a sovereign, through constitutions and courts, and how to constrain a firm, through competition law. It has never had to constrain an institution that behaves like a sovereign in some respects and a firm in others. The live proposals divide along familiar lines:

• Structural remedies, which treat the problem as one of market concentration and respond with antitrust enforcement, mandated interoperability and data portability.
• Procedural remedies, which treat platforms as quasi-governmental and demand due process: notice before removal, reasoned decisions, and an independent route of appeal.
• Fiduciary remedies, which borrow from the law of doctors and lawyers and impose a duty of loyalty on anyone holding sensitive personal data.
• Public-option remedies, which propose publicly funded infrastructure — search, identity, cloud — as a competitive floor rather than as a replacement for private provision.

Each has a coherent liberal pedigree. They are also mutually constraining: aggressive structural separation makes fiduciary supervision harder, and strong procedural rights entrench incumbents who can afford compliance departments. The tradition is at the beginning of this argument, not the end of it.

What the compass captures, and what it misses

On a two-axis compass, liberalism does not occupy a point. It occupies the libertarian half of the vertical axis and spreads across the whole economic axis, because the disagreement between the negative and positive readings of liberty is precisely a disagreement about the economic axis. A classical liberal sits in the libertarian-right quadrant; a Rawlsian social liberal sits in the libertarian-left quadrant. They share a method and divide on a conclusion.

This is a genuine limitation of any two-dimensional model, and it is worth being honest about it. The compass measures positions, not reasoning. Two people can arrive at the same coordinates from incompatible premises — one because they think redistribution violates rights, another because they think it is inefficient — and the model will not distinguish them. What it does well is show that the economic and social questions are separable, which the older single left-right line concealed. If you want to see where your own answers place you, our fifteen-question test maps both axes; the methodology page sets out how each question is weighted.

Conclusion

The recurring obituary for liberalism mistakes a rebuild for a collapse. The tradition has been declared finished after 1848, after 1917, after 1929, and after 1968, and each time it re-formed around whatever power had grown large enough to need constraining. Its current difficulty is real and specific: the institutions that now shape the conditions of individual life are private in form and governmental in effect, and liberalism's two inherited toolkits were each built for one of those and not the other. Whether it produces a fifth rebuild or finally exhausts itself is genuinely open — and it is one of the most consequential open questions in contemporary politics.