Understanding Populism in the Digital Age

Populism is a structure of argument, not a set of policies. Here is how that structure works, why platform architecture rewards it, and how to tell it apart from ordinary democratic opposition.

Verfasst von Ottavio Malatacca · Veröffentlicht 2026-02-26 · Aktualisiert 2026-08-30

Populism is among the most used and least defined words in contemporary political commentary. It is applied to governments of the left and the right, to movements in power and in opposition, and often simply to any politician the speaker dislikes. Used that way it explains nothing. Used precisely it explains a great deal — because populism is not a set of policies at all. It is a structure of argument, and that structure can be filled with almost any policy content.

A definition narrow enough to be useful

The most workable definition in political science comes from Cas Mudde, who describes populism as a thin-centred ideology. It makes only two claims. First, that society is ultimately divided into two homogeneous and antagonistic groups: the pure people and the corrupt elite. Second, that politics should be an expression of the general will of the people.

Thin-centred means it cannot govern alone. It has nothing to say about interest rates, school curricula or defence procurement, so it always attaches to a fuller host ideology. Attached to socialism it produces left populism, which defines the elite economically as bankers and multinationals. Attached to nationalism it produces right populism, which defines the elite culturally as cosmopolitan professionals, and usually adds a second enemy below in the form of immigrants. The shared structure is why left and right populists sound alike in their rhetoric while proposing opposite budgets.

The three moves

The populist argument has a reliable internal grammar, and recognising it is more useful than any checklist of policies.

1. Construct the people. The people is never the actual electorate, which is plainly divided. It is a moral category: the authentic, ordinary, hardworking majority whose views are assumed to be unified and self-evident.
2. Construct the elite. The elite is defined by betrayal rather than by income or status. This is why billionaires can credibly lead anti-elite movements: the accusation is disloyalty to the nation or to ordinary people, not wealth as such.
3. Claim exclusive representation. This is the decisive move. The populist leader does not claim to represent a constituency better than rivals do. The claim is to represent the people as a whole, uniquely and completely.

Jan-Werner Müller's argument in What Is Populism? (2016) is that the third move is the one that matters for democracy. Anti-elitism is ordinary democratic politics and always has been. Anti-pluralism is not. Once a movement claims to be the sole legitimate voice of the people, opponents are not merely wrong — they are outside the people, and their exclusion becomes a democratic act rather than a violation of democracy.

Populists do not claim we are the 99 per cent. What they imply is: we are the 100 per cent.

Jan-Werner Müller, What Is Populism?, 2016

Why the argument is winning now: the demand side

Populism is not new — the American People's Party of the 1890s and the Latin American movements of the mid-twentieth century are earlier instances. What requires explanation is the scale and simultaneity of the current wave. Two families of explanation compete, and the evidence supports a combination.

The economic account points to the hollowing out of middle-income employment through automation and trade exposure, to two decades of near-flat median real wages across much of the OECD, and to the post-2008 asymmetry in which financial institutions were rescued while household balance sheets were not. Regional analyses consistently find populist voting concentrated in places that lost industrial employment and never replaced it.

The cultural account, argued most fully by Pippa Norris and Ronald Inglehart in Cultural Backlash (2019), points to a value shift among younger and university-educated cohorts on questions of gender, sexuality, race and national identity, and to the reaction of older and less-educated cohorts who find that positions they held uncontroversially for decades have become socially costly. On survey data, cultural variables generally predict populist support more strongly than income does.

The accounts are less opposed than they appear. Economic decline in a region does not automatically produce cultural grievance, but it dissolves the sources of status that were not economic — stable employment, a trade, a settled place in a community — and status loss is what cultural grievance is usually about. The essay on identity and political alignment develops this mechanism in more detail.

Why the argument is winning now: the supply side

Grievance is a constant in democratic politics. What changed is the cost of assembling people around it. Three features of platform architecture systematically favour populist communication, and none of them was designed with politics in mind.

Disintermediation

For most of the twentieth century a national political message had to pass through editors, party structures and broadcast regulators. Those filters were slow, often self-serving, and they screened out a great deal of legitimate opinion along with the illegitimate. They also imposed a shared factual baseline. Their removal is genuinely double-edged: it lowered the barrier for outsider movements of every kind, and it removed the friction that had made unfalsifiable claims expensive to circulate.

Engagement optimisation

Ranking systems trained on engagement learn that moral outrage and in-group solidarity are unusually reliable predictors of sharing. Research on large samples of social media posts has repeatedly found that language expressing moral-emotional content about an out-group circulates substantially further than neutral language about the same subject. No editorial intent is required. A system optimising for attention will discover the people-versus-elite frame because that frame reliably produces attention.

Sorting

The echo chamber thesis is often overstated — most people encounter more cross-cutting content online than the popular account suggests. The better-supported finding is about intensity rather than exposure: online environments sort people into affectively coherent communities where disagreement carries a social cost. The result is not that people never see the other side, but that they see it in its least sympathetic form, quoted by allies for the purpose of ridicule.

The hard part: telling populism from legitimate opposition

This is where the term is most often abused. Calling a movement populist is frequently a way of refusing to engage with its complaint, and that refusal has itself been a reliable engine of populist growth. Two distinctions do most of the work.

The first is between the diagnosis and the structure. A movement can be entirely right that a region was abandoned, that a bailout was unjust, or that an immigration policy was adopted without meaningful consultation, and the correctness of that diagnosis says nothing about whether its argument is populist in form. Conversely, a movement can be structurally populist and factually wrong about everything.

The second is the anti-pluralism test, and it is largely behavioural. Does the movement accept that its opponents are legitimate participants in the same political community? Does it accept electoral defeat as final? Does it treat courts, statistical agencies and the press as institutions with independent standing, or as obstacles staffed by enemies? A movement that is fiercely anti-elite but answers yes to these questions is doing democratic opposition. A movement that answers no has crossed a line that has nothing to do with how radical its policy platform is.

Where populism sits on the compass

It does not sit anywhere, and this is the most common error made with two-axis models. Because populism is thin-centred, it has no coordinates of its own — it takes the coordinates of whatever host ideology it attaches to. A left-populist movement lands in the left half of the economic axis; a right-populist movement lands in the right half. Both tend upward on the authoritarian axis, but that tendency comes from the anti-pluralist move rather than from populism as such.

If a compass test tells you that you are populist, the test is measuring something else — usually anti-establishment sentiment, which is a different variable and one that a large majority of citizens in most democracies would register on. Our own test deliberately measures positions on economic and social questions rather than attitudes toward elites, for exactly this reason.

Conclusion

Treating populism as a disease to be eradicated misunderstands it. It is a signal that a substantial group of citizens believes the ordinary channels of representation have stopped working for them, and that belief is sometimes accurate. The response that has historically worked is not better messaging against populists but the repair of the channels: representation that visibly includes places and occupations that had dropped out of political calculation. The response that has historically failed is the one most often reached for, which is to treat the complaint as illegitimate because of the form in which it arrived.