Here’s a question that sounds like it should have an easy answer. What is a democracy?
The casual answer most people give is: a country where citizens vote for their leaders. The voting seems to be the defining feature. Countries that hold elections are democracies; countries that don’t aren’t. This is the folk theory of democracy, and it’s tidy, and it’s deeply incomplete.
Political scientists who have studied what actually makes democracies function — as opposed to what makes them superficially resemble democracies — have concluded, with unusual consistency, that voting is necessary but far from sufficient. A country can hold regular elections and still be, in most meaningful senses, not democratic. A country can be genuinely democratic for periods where its electoral procedures look, from outside, much like those of countries that aren’t. Voting is a small part of a larger institutional pattern.
Understanding the larger pattern matters because it makes it possible to notice when democracy is being eroded in countries that still hold elections, and — more practically for most readers — to understand what kind of citizenship democracy actually asks for, beyond the annual or quadrennial trip to the ballot box.
Dahl’s framework
The most influential modern analysis of what democracy requires came from the American political scientist Robert Dahl, whose book On Democracy, published in 1998, distilled decades of his earlier work on democratic theory. Dahl, who died in 2014, argued that a political system was democratic to the extent that it met a specific set of conditions — of which free elections were only one.
Dahl’s core conditions:
Elected officials. Public decisions are made by, or under the accountability of, elected representatives. This is the voting condition most people start with, but it’s only the first.
Free and fair elections. Elections are conducted in ways that allow the public will to be meaningfully expressed, without coercion, fraud, or systematic exclusion.
Inclusive citizenship. All adult residents (or nearly all) have the right to participate. Democracies that exclude large categories of their population from voting — women, minorities, the poor, the illiterate — are, in Dahl’s framework, only partially democratic. Historical regimes that had elections but excluded most of their populations don’t count as full democracies even when contemporaries called them that.
Freedom of expression. Citizens can speak about political matters — including criticism of those in power — without fear of official retaliation. Systems that hold elections but imprison critics don’t meet this condition.
Alternative sources of information. Citizens have access to information from multiple independent sources rather than only from state- or regime-controlled media. A population fed only one version of events cannot meaningfully choose between versions, even if it’s allowed to vote.
Associational autonomy. Citizens can form organisations — political parties, trade unions, interest groups, civil-society bodies — independent of the state. Democracy requires the existence of institutions between the individual and the government, and systems that suppress such institutions can hold elections but are not really democratic.
Dahl’s point was that these conditions have to operate together. A country with free elections but without free press becomes a system where elections produce the appearance of democratic choice without the reality. A country with free press but without inclusive citizenship excludes part of its population from the democratic conversation. Each condition is necessary; no single one is sufficient. The word “democracy” names the whole configuration, not any single element.
This framework has been influential partly because it gives a language for what can go wrong in countries that look democratic from outside but aren’t functioning that way. The hybrid regimes or electoral autocracies of recent political science are systems where some Dahl conditions are met while others are systematically weakened. The elections happen. The media are state-controlled. The opposition is harassed. Civil society is constrained. Citizens can vote, but the information and institutional conditions that would make the vote meaningful are absent.
Consolidation and its absence
A related body of work has examined what makes young democracies stable or unstable — the process called democratic consolidation. The central researchers here include Juan Linz, a Spanish-American political scientist, and Alfred Stepan, an American, whose collaborative work on transitions from authoritarian rule influenced a generation of scholarship.
Linz and Stepan argued that democracy consolidates when it becomes, in their phrase, “the only game in town” — when all major political actors, including those who lose elections, accept democratic procedures as the legitimate way to compete for power. Democracy is genuinely consolidated when the losers of an election respect the outcome and wait for the next election, rather than seeking to overturn it through force, fraud, or institutional subversion.
This framing is important because it suggests that the most dangerous moments in a democratic system are the ones where the losers of an election decline to accept the result. The machinery of elections can run; the machinery of democratic consolidation is about whether the results are accepted by those who don’t like them. A democracy where losers always accept results is consolidated. A democracy where losers sometimes don’t is in trouble, regardless of how well-run its specific elections were.
In the decades since Linz and Stepan wrote, the global picture has shifted in ways their framework helps illuminate. The American political scientist Larry Diamond has documented what he calls the democratic recession — the pattern, since roughly 2006, of democratic quality declining in many countries around the world. Formal elections still happen in most of these places. But the Dahl conditions that make elections meaningful, and the Linz-Stepan conditions that make democratic procedures durable, have been eroding. The erosion is usually gradual, often legal, and rarely involves a dramatic coup. It’s the quieter process of systematically weakening the conditions that surround voting while keeping the voting itself.
Diamond’s diagnosis, supported by indices produced by organisations like Freedom House and V-Dem (Varieties of Democracy), suggests that roughly a third of the world’s countries have experienced democratic backsliding in the last two decades. This isn’t an argument about any specific regime — it’s a broader statistical pattern, visible across many countries on every continent. The overall direction of democratic quality globally has been downward, after decades of progress following the 1970s and 1980s.
The minimalist counter-view
Before accepting the rich-definition view wholesale, it’s worth taking seriously a counter-tradition in political science that argues democracy is simpler than Dahl’s framework makes it sound.
The Austrian-American economist Joseph Schumpeter, writing in 1942, proposed what has come to be called the minimalist conception of democracy. Democracy, in Schumpeter’s definition, is simply a political system in which ruling parties can be voted out in periodic elections. Nothing more. The free press, the inclusive citizenship, the associational autonomy — these might be desirable, but Schumpeter argued they weren’t part of the definition of democracy. Democracy was specifically the procedure of electoral turnover.
This view has had significant recent support from the Polish-American political scientist Adam Przeworski, whose book Democracy and the Limits of Self-Government makes a sophisticated case for the minimalist position. Przeworski’s argument, roughly: attaching too many requirements to the definition of democracy makes the concept unwieldy and risks disqualifying political systems that most people would recognise as functioning democracies. A narrow definition — can the ruling party lose an election and actually leave office? — has the advantage of being clear, testable, and historically meaningful.
The debate between maximalist and minimalist conceptions of democracy matters because it affects what we think of as democratic erosion. Under Dahl’s framework, restrictions on press freedom or civil society constitute democratic erosion even if elections remain competitive. Under Schumpeter’s, they don’t; democracy is fine as long as incumbents can lose elections. Both views have thoughtful defenders, and neither is obviously correct.
For most practical purposes, though, the richer Dahl framework is probably more useful. Elections without the surrounding conditions tend, historically, to converge toward incumbent dominance over time — the press becomes less free, the opposition becomes less viable, the information environment becomes less pluralistic — until eventually even the minimalist condition (can incumbents lose?) becomes harder to meet. The additional conditions aren’t decorative. They protect the core electoral condition from gradual capture.
What this asks of citizens
The most useful implication of the richer framework, for ordinary citizens, is that democratic citizenship involves more than voting. Voting is a small slice of what the Dahl conditions require. The rest of the conditions are maintained, or eroded, by the ordinary behaviour of many citizens acting across many institutions.
A free press depends on citizens who are willing to pay for journalism, read it, distinguish better outlets from worse ones, and support independent investigation when it happens. A culture without these habits produces a press system that survives on clickbait and propaganda.
Alternative sources of information depend on the persistence of non-state institutions — universities, research organisations, libraries, archives, independent think-tanks. These require funding and defence, often against political pressures to suppress inconvenient findings.
Associational autonomy depends on citizens actually joining and participating in civil-society organisations — religious communities, professional associations, unions, sporting clubs, volunteer organisations, political groups. A society whose citizens belong to nothing between themselves and the state has lost something important about how democracies work.
Inclusive citizenship depends on continued vigilance against forces — economic, cultural, administrative — that exclude categories of people from meaningful participation. This isn’t primarily about extending the franchise; most democracies have long done that formally. It’s about ensuring that formal inclusion translates into practical capacity to vote, speak, and organise.
Freedom of expression depends not only on legal protections but on cultural habits — the willingness to defend speech you disagree with, the tolerance for disagreement within ordinary life, the refusal to personally destroy people for opinions.
None of these are things that can be done by voting. They’re done — or not done — in the ordinary texture of political and social life, by millions of small decisions made by citizens who mostly aren’t thinking about democracy at all. The democracy is the aggregate of these decisions, not the elections that happen on top of them.
The question that remains
The deepest thing Dahl’s framework teaches, and what the quieter political-science tradition behind it confirms, is that democracy is not primarily an electoral procedure. It’s a whole configuration of institutions, habits, and cultural commitments — of which elections are one visible part. Countries that keep the elections but lose the rest become what political scientists now call illiberal democracies or electoral autocracies. The name is an acknowledgement that the label “democracy”, used casually, no longer means what it used to.
This has implications for how you think about the country you live in and the ones you hear about. A country’s democratic quality isn’t captured by whether it holds elections. It’s captured by whether it holds the broader configuration Dahl described — the press, the pluralism, the inclusion, the associations, the expression, the acceptance of electoral outcomes by losers. Any of these can erode without the elections changing. All of them together are what being a democracy actually means.
The question to carry, as you think about your own political life:
Beyond voting, what are the specific institutions and habits that keep a democracy working — and which of them are you currently supporting, and which are you not thinking about at all?
Key research referenced: Robert Dahl, On Democracy (1998) and Polyarchy (1971); Juan Linz and Alfred Stepan, Problems of Democratic Transition and Consolidation (1996); Larry Diamond’s democratic-recession research; Joseph Schumpeter, Capitalism, Socialism, and Democracy (1942); Adam Przeworski, Democracy and the Limits of Self-Government (2010).