Democracy has a hidden assumption. Not the obvious ones about elections, representation, or the rule of law. A deeper one, buried in the foundation: that voters can be sufficiently informed to make reasonable choices about who governs them and how.
This assumption has always been a stretch. It is now impossible.
The scale of information needed to vote competently in a modern state is enormous. A single election requires judgments on foreign policy, domestic spending, regulation, taxation, social policy, environmental management, and a dozen other domains, each of which is the subject of entire academic disciplines. Even a full-time specialist cannot master all of them. A voter with a job, a family, and forty-five minutes to research before election day has no chance.
Political parties have also found that not having policies can make them more electable, not less. If there are no policies with which to disagree, the thinking goes, fewer people will be put off.
Democracy never required perfect information. It required enough information. The idea was that voters could rely on proxies: party affiliation, trusted commentators, interest group endorsements. You do not need to understand monetary policy to know which party shares your values on it. The system routes your general values through representative institutions that handle the technical details.
This was never a great system, but it worked well enough. People had heuristics. Parties had brands. Newspapers had editorial positions. A voter could make a reasonably informed choice by triangulating a few trusted signals.
The internet destroyed these heuristics without replacing them.
Consider what has changed. A voter in 1990 had three television channels, a local newspaper, and whatever books and magazines they chose to buy. The information environment was narrow, stable, and filtered through professional editors who at least pretended to standards of accuracy. The voter could not know everything, but they could know what was considered credible by people whose job was to determine credibility.
A voter in 2026 has access to more information in one hour than their 1990 counterpart saw in a year. Every fact has a counter-fact. Every expert with a PhD has a counter-expert on Twitter with a PhD from a different university. Every news story is immediately met with accusations of bias, spin, or outright fabrication. The editorial gatekeepers are gone. In their place are algorithms that optimize for engagement, not accuracy.
The result is not a better-informed voter. The result is a voter who has seen too much to trust anything.
This is the paradox the internet created for democracy: universal access to information has produced universal distrust of information. When everything is available, nothing is credible. When every source has an agenda, the rational response is to believe none of them - and then to believe whatever feels right, because at least that is honest.
The cognitive burden is large. To be a fully informed voter today, you would need to evaluate not just policies and candidates, but also the credibility of every source you encounter, the algorithmic incentives shaping what you see, the financial interests behind every think tank, the methodological validity of every study, and the editing history of every Wikipedia article. This is not a reasonable expectation. It is a full-time job for a team of specialists. Who will themselves, of course, have biases.
The voter falls back on identity. If you cannot judge the policies, you judge the people. Which tribe are they in? Do they talk like us? Do they dislike the same people we dislike? This is not irrational. It is a sensible response to an impossible situation. It is also the death of deliberation.
The implications for democracy are not theoretical. When voters cannot judge policy, elections become referendums on identity. Who you vote for becomes who you are. Disagreement becomes betrayal. Compromise becomes heresy. The system depends on the possibility of persuasion, on the idea that a voter can change their mind based on new evidence. But if you cannot judge the evidence, you cannot change your mind. Your views become part of your identity, and identities do not update.
This is why polarization is not an accident of modern politics but a structural consequence of the information environment. It is not that people have become more tribal. It is that tribalism is the only usable heuristic left.
Governments understand this. They have always tried to shape public opinion. What has changed is the effectiveness of the tools. A 1990 government could plant a story in a friendly newspaper and hope it spread. A 2026 government can algorithmically amplify selected narratives, suppress inconvenient stories through platform pressure, and monitor the response in real time across every demographic. The capacity for democratic control of public information has never been lower.
The free speech framework assumes a marketplace where truth eventually wins. That assumption is valid only if buyers can evaluate the goods. When they cannot, the marketplace becomes a flea market where the loudest seller wins regardless of what they are selling.
None of this means democracy is doomed. It means democracy cannot function as currently conceived. The system needs a theory of how voters make decisions under radical uncertainty, not a fantasy that they will somehow become well-informed if we just provide enough fact-checks and civics lessons. The standard response to misinformation is more information. That is treating the symptom. The disease is that the voter cannot process the information they already have.
Any serious defense of democracy begins by acknowledging that its foundational assumption is broken. Then it asks: what can we build that works with actual human beings, not the idealized citizens of political theory?
The history of democracy is a history of imperfect solutions to impossible problems. We need another one.
This is part three of a four-part series on ideas, speech, and democracy. Previously: The Battle of Ideas - why bad ideas keep winning. Next: Locke’s Ghost - what the philosophers of liberty would make of this mess.