Of all the objections levelled against direct democracy, the most philosophically serious is also the most emotionally resonant: the fear of the majority. James Madison articulated it in the eighteenth century. John Stuart Mill refined it in the nineteenth. And across the twentieth century, as civil rights movements confronted popular prejudice encoded into law and upheld by ballot, the fear seemed to be vindicated over and over again. The argument runs as follows: if you give the majority unchecked power to decide policy, minorities — ethnic, religious, sexual, political — will be stripped of their rights by popular vote. Democracy, on this view, is not protection but threat.
It is an argument worth taking seriously. And yet, if we examine it carefully, we find that it proves too much. It is an argument not merely against direct democracy but against democracy itself — and against the political equality of citizens. Understanding why requires us to disentangle several threads that are often woven together: the historical record, the structural problem, and the question of what safeguards actually work.
The Historical Record Is Mixed
The most frequently cited example is the United States. Between the late nineteenth and mid-twentieth centuries, a number of ballot initiatives in American states were used to restrict the rights of racial minorities, immigrants, and LGBTQ+ individuals. In 1964, California voters passed Proposition 14, which effectively repealed a fair housing law protecting Black residents from discrimination. In 2008, Proposition 8 eliminated marriage equality for same-sex couples — a right that had only recently been won in court.
These are real cases. They deserve to be taken seriously. But they also need to be placed in a broader context.
The same era that produced Proposition 14 also produced a Congress that enforced Jim Crow, a Senate that had blocked civil rights legislation for decades through the filibuster, and a series of presidents who had either supported segregation or tolerated it. The American representative system was not a bulwark against the oppression of minorities — it was, in many cases, the instrument of that oppression. If the historical indictment is of direct democracy specifically, it has a great deal of explaining to do about why representative institutions performed so much worse.
Switzerland — the country most associated with direct democracy — offers a more complicated picture. The Swiss did delay women’s suffrage at the federal level until 1971, a decision enabled partly by a male electorate voting in referendums. This is a genuine mark against the Swiss model. But Switzerland has also used referendums to pass some of the most expansive social protections in Europe, and its system includes constitutional protections for minorities that cannot be overridden by popular vote. The lesson may be not that direct democracy is inherently hostile to minority rights, but that it reflects the values of the society practising it — for better and for worse.
The Structural Problem and Its Solutions
Behind the historical examples lies a structural concern. In any society where a majority holds prejudice against a minority, giving the majority the power to make binding decisions on matters affecting that minority creates a structural risk. This is not a problem unique to direct democracy. It is a problem of democracy per se, which is why every serious democratic theorist from Madison onward has argued that democracy requires constitutional constraints.
The key insight is that the question is not whether to have constraints on majority power, but what form those constraints should take, and how they should be enforced.
In representative systems, the constraint often takes the form of judicial review — courts that can strike down laws passed by elected legislatures. The United States Supreme Court, for example, struck down laws criminalising sodomy in 2003 and mandated marriage equality in 2015. This model has real advantages, but it also has a significant weakness: it transfers immense power to an unelected judiciary, whose composition is determined by political appointment and who are themselves drawn from a narrow social stratum. Court-based rights protection is reliable only as long as the courts are willing to protect rights — and as recent American history has shown, that willingness cannot be taken for granted.
Direct democracy systems can and do incorporate equivalent constraints. The Swiss constitution contains provisions that cannot be altered by referendum, including fundamental rights protections aligned with the European Convention on Human Rights. Several proposals in Switzerland have been rejected or modified precisely because they conflicted with these constitutional safeguards. A 2010 initiative that would have required automatic deportation of foreign nationals convicted of certain crimes was challenged and its implementation shaped by the courts. This is not a perfect system — no system is — but it demonstrates that constitutional protection of minority rights is compatible with direct democratic practice.
Deliberation as a Safeguard
There is a further consideration that goes beyond institutional design. The objection to direct democracy assumes that majority opinion is a fixed fact — that if a majority currently holds a prejudiced view, that view will simply be expressed in a vote. But political opinions are not fixed. They are formed, shaped, and sometimes transformed through deliberation.
This is where the design of direct democratic institutions matters enormously. A referendum held immediately after a heated campaign, with little structured public deliberation, may well reflect the worst impulses of the moment. But a citizens’ assembly — where randomly selected citizens hear from experts, affected communities, and each other over weeks or months — tends to produce more considered, less reactive, and more rights-respecting conclusions.
The evidence from citizens’ assemblies on precisely this kind of sensitive question is striking. The Irish Citizens’ Assembly on same-sex marriage, held in 2016-17, was given a question that polls suggested was nearly evenly divided in the population. After deliberation, the assembly recommended legalising same-sex marriage by a margin of 79 to 21 percent. The subsequent referendum passed with 62 percent of the vote. The assembly had not simply aggregated existing opinion — it had changed minds, by creating the conditions for genuine deliberation.
This does not mean deliberation always produces rights-expansive outcomes. It is not magic. But it suggests that the relationship between direct democratic mechanisms and minority rights is more contingent than the simple majority-oppresses-minority model implies.
Reframing the Question
The real question is not whether to protect minority rights, but how. All serious democrats agree that minority rights require protection. The disagreement is about which institutions are most reliable guardians of those rights.
Representatives have a poor track record. Courts are reliable only conditionally. Supranational institutions can provide a check, but they raise their own legitimacy questions. Constitutional protections are essential, but constitutions must be interpreted and enforced by humans who operate within political systems.
What direct democracy adds to this picture is not a solution to the problem of minority rights — no institution solves that problem entirely — but a different configuration of risks and opportunities. The risks include majority prejudice expressed through popular votes. The opportunities include the transformation of public opinion through genuine deliberation, the expression of minority voices that are systematically excluded from representative politics, and the development of civic cultures in which citizens take responsibility for collective decisions.
The tyranny of the majority is a real danger. But so is the tyranny of the minority — the domination of political systems by elites who are insulated from popular accountability precisely because they are not chosen by popular vote. Both dangers need to be taken seriously. Any political system that takes only one of them seriously while ignoring the other is not a system committed to democracy, but to the particular distribution of power that already exists.
The task is not to choose between majority rule and minority rights. It is to build institutions that make both possible simultaneously — and to remain honest about the fact that no institution, representative or direct, has yet fully achieved that goal.
