Direct Democracy Will Not Slide Back Into Autocracy, Because Nobody Wants It To

The most common objection to direct democracy is not that it would decide badly. It is that it would not last: the mob would vote itself a strongman, and we would be back where we started, minus the institutions that were supposed to prevent it.

The objection deserves a proper answer, because the history it gestures at is real. Republics have died. Ancient assemblies did hand power to tyrants. Weimar did vote, more or less, for its own end. I want to explain why I think the analogy fails for a direct democracy established now, in a population that already lives inside a democratic order.

Ask Who Would Want It

Start with the question the objection never asks: regression to whom, at whose request?

A regression to autocracy needs a constituency that gains from it. In every historical case, that constituency existed and was identifiable. A general with an army and a grievance. A party that could not win otherwise. An aristocracy that wanted its position back. A population so frightened by disorder that it would trade anything for quiet.

Now look at the parties in a modern direct democracy and ask what each would gain.

The public has, for the first time, the decisions in its own hands. It has nothing to gain by handing them to one man and everything to lose. There is no version of the trade in which the many end up better off, and the many know it, because they have just spent a generation learning what the decisions are worth.

The politicians are the group people expect to want the regression, and they are the group with the least means to bring it about. Under a direct democracy they have already been reduced to administrators of decisions taken elsewhere. An administrator who wants to become an autocrat needs a lever, and the lever, the power to decide, is exactly what has been distributed. You cannot seize what is no longer concentrated anywhere.

Business is the interesting case, because commercial interests have historically been quite willing to live with autocrats. But that willingness was a bet on predictability and access. An autocrat in a country whose public has become used to deciding is neither predictable nor a reliable point of access; he is a man surrounded by a population that has been trained to say no. Business prefers a system it can work with, and it will have spent years learning to work with this one.

So the regression has no author. That is the first half of the argument, and I think it is the stronger half.

Values Do Not Regress, They Are Internalised

The second half concerns what the population has become.

The regressions of the past happened in populations for whom democratic values were new, thin or borrowed. The Athenians had a few generations of assembly behind them and no theory of rights. The Germans of 1933 had fourteen years of a republic imposed by defeat. In each case the democratic layer was a coat of paint on a population that still expected, at bottom, to be ruled.

That is not the population a direct democracy would be built from today. The people who would live under it have been raised inside representative regimes for their whole lives and their parents' lives. They hold, without thinking about it, that a person cannot be imprisoned without trial, that speech is not a crime, that the government is answerable, that an election lost is an election lost. These are not positions they argue for; they are the floor they stand on.

The internalisation of democratic values has already done the work the objection assumes still needs doing. Watch how modern populations react when a government tries to cancel an election, silence a court or postpone leaving office: not with relief, but with the streets. That reaction is the immune system. It exists now, in representative regimes, and a direct democracy inherits it rather than having to grow it.

A system that adds decision-making power to a population with that immune system is adding power to the body that already refuses autocracy. The scenario in which it then votes for a dictator requires the population to abandon, all at once, everything it has believed since childhood. Populations do not do that. They change slowly, and slow change in a direct democracy is visible and reversible at the next vote.

The Real Risk Is the One We Have

I will concede one thing to the objection, and turn it around.

Regression does happen. It is happening now, in several representative regimes, and it happens the same way each time: an elected government uses its concentrated powers to weaken the courts, capture the press, rewrite the electoral rules and stay. The public objects and discovers that its only lever, the next election, has been moved out of reach.

That mechanism needs a concentration of power to work. It needs a small body that can act on all fronts at once and that the public can only remove every few years. The concentration is the vulnerability, and it is a feature of representation, not of democracy. Disperse the decisions and there is no office from which the rules can be rewritten, no single body whose capture delivers the state.

So the honest comparison is not between a stable representative regime and a fragile direct one. It is between a system that has regressed repeatedly through its own central mechanism and a system that has no central mechanism to regress through. On that comparison I know which one I would call fragile.