Loud Is Not the Same as Many: What the 2020s Elections Actually Did

In 2024 more people voted than in any year in history. More than seventy countries went to the polls, covering over half the population of the planet, and the total electorate ran past four billion people. It was treated in advance as a referendum on the decade: a chance to find out, all at once, which way the world had turned.

It produced an answer, and the answer was not ideological.

Of the ten major countries tracked by the ParlGov research project that held national elections that year, every single governing party lost vote share. Not most. All ten. Comparable records go back to 1905, and it had never happened before. Outside that set the picture is messier, and a few governing parties held their ground or gained.

That is the finding to hold on to, because it does not say what the decade's most popular story says. Governments of the right lost. Governments of the left lost. What they had in common was not what they believed. It was that they were the ones in charge.

The premise is only half right

Start by checking the claim itself, because it is repeated more often than it is examined.

In the United Kingdom, the 2024 election ended fourteen years of Conservative government and returned Labour with a large majority. Australia elected a Labor government in 2022 and re-elected it in 2025. Mexico elected Claudia Sheinbaum in 2024, from the governing party of the left. Brazil returned Lula to the presidency in 2022. Poland's opposition coalition removed a governing conservative party in 2023 after eight years.

So the strong version of the claim, that conservatives now govern everywhere, is not true, and anyone repeating it has not looked at a list. What is true is more specific and more interesting, and it needs the weaker version of the claim to state it properly.

What actually moved the votes

The single best predictor of how a governing party did in 2024 was not its ideology. It was whether people had watched prices rise on its watch.

The post-pandemic inflation shock ran through most of the world's economies between 2021 and 2023. It is unusual among political events in that it requires no explanation, no briefing and no partisan framing to notice. A person who has never read a manifesto knows exactly what a supermarket receipt used to look like. It is the most legible thing in politics, and it lands on everyone.

Governments that presided over it were punished more or less regardless of what they believed. In the United States the incumbent party of the left lost. In the United Kingdom the incumbent party of the right lost. In France, Japan and South Africa, governing parties lost vote share. The pattern is not a swing to the right. It is a swing at whoever answered the door.

That does not make it meaningless. Anti-incumbency of that scale tells you something real, which is that a very large number of people concluded their governments could not deliver a stable cost of living. But what it judges is competence at holding down a cost of living, which is a different thing from agreeing with whoever replaced them.

The part of the claim that survives

Now the harder half, because dismissing the whole thing would be its own kind of dishonesty.

Something has genuinely shifted, and it shows up over decades rather than in any one night. Pippa Norris and Ronald Inglehart, measuring what they classify as authoritarian populist parties across Europe, found their average share of the vote in parliamentary elections has more than doubled since the 1960s and their share of seats has roughly tripled. Parties of that family have entered government in countries as different as Austria, Italy, the Netherlands, Poland and Switzerland.

The clearer effect is not who holds office but where the argument sits. On immigration in particular, the centre of gravity across most of Europe has moved, and it has moved in governments of the left as well as the right. Parties that lose elections can still win arguments, and a party polling in the teens can drag the whole field towards it by making an issue impossible to avoid.

So the honest version is this. Conservatives are not in power everywhere. But the right has been setting more of the agenda than its seat count suggests, and it has been doing so for longer than this decade.

Why a feed is a broken instrument

Which leaves the question underneath the question. If the left is more visible than it has ever been, why does the visibility keep failing to convert?

Part of the answer is that visibility was never measuring what people assumed it measured, and there is data on precisely this. In 2018 the organisation More in Common surveyed 8,000 Americans for a study called Hidden Tribes. It found that the group furthest to the left, which it labelled Progressive Activists, made up 8 percent of the population and was by a distance the most prolific political group online: 70 percent had shared political content in the previous year. The most conservative group followed at 56 percent.

Those two groups together are about 14 percent of the country. They generate a wildly disproportionate share of what anyone sees.

The consequence is straightforward once stated. A social feed is not a sample of the electorate. It is a sample weighted heavily towards the fraction of people who enjoy arguing about politics in public, which is a small and unusual fraction. Most people never post about politics at all. They are not undecided or disengaged, they simply do not perform it, and they vote in exactly the same numbers as the people who do.

This cuts both ways, which is worth saying plainly. The right's online presence is also unrepresentative of the right's voters. Anyone reading their own timeline as a poll is going to be wrong, and the direction of the error depends only on who they follow.

Intensity is not the same as breadth

There is a second mechanism, and it is about how strongly people hold a view rather than how many hold it.

A position can be held with enormous intensity by a minority and mild indifference by a majority. Intensity is what produces marches, donations, volunteering and posts, and it is genuinely powerful for changing what a country talks about. It is much weaker at producing votes, because a vote counts the same whether it was cast with passion or with a shrug on the way home from work.

Movements built out of high intensity tend to overestimate their size for the same reason a stadium sounds unanimous. Everyone in earshot agrees, and the people who do not are not in the building.

The backlash argument, and its limits

The most cited explanation for the decade is Norris and Inglehart's, set out in their 2019 book Cultural Backlash. In their account, a long values revolution across rich democracies, more secular, more socially liberal, more urban, produced a reaction among people who found their assumptions moved beneath them without being asked, and that reaction has been available for parties to organise.

It is a serious argument with a lot of evidence behind it. It is also contested. Other researchers argue the pathways to the same result differ too much between countries to be explained by culture alone, and that economic insecurity, the decline of industrial employment and the collapse of local institutions do at least as much work.

The honest position is that both are running at once and the mix differs by country, which is unsatisfying and probably right.

The room in the photograph

The picture at the top of this piece is a polling place in Iceland, set up in a shopping centre for early voting: a few ballot boxes, some booths with curtains, chairs nobody is sitting in.

That room is where all of this gets settled, and the striking thing about it is how quiet it is. Nothing that happens there is performed. Nobody sees what anybody else does. The loudest person in the country and someone who has never expressed a political opinion in their life put an identical piece of paper into an identical box, and the box does not weight them differently.

Every explanation above is really a version of the same point. Volume is a measure of how much noise a group makes, and elections do not count noise. They count people, one at a time, in a quiet room, and it turns out those are very different quantities.

Sources

• Inside Story, "The year of governing precariously" (quoting John Burn-Murdoch of the Financial Times on ParlGov data: the incumbents in every one of the ten major countries tracked that held national elections in 2024 were beaten back by voters, and the exceptions outside that set).
• John Burn-Murdoch, on the ParlGov series (that 2024 was an exceptionally bad year for incumbents worldwide, and that the record of every governing party losing vote share runs back to 1905).
• Wikipedia, "Anti-incumbency" (the 2024 pattern of governing parties losing power or vote share across the United Kingdom, the United States, France, Japan, South Africa and India).
• More in Common, "Hidden Tribes: A Study of America's Polarized Landscape" (the 2018 survey of 8,000 Americans; Progressive Activists at 8 percent of the population with 70 percent sharing political content in the previous year, against 56 percent for the most conservative group, and the two poles totalling about 14 percent).
• Norris and Inglehart, "Cultural Backlash: Trump, Brexit, and Authoritarian Populism", Cambridge University Press, 2019 (the values-revolution and backlash argument, the more than doubling of authoritarian populist vote share in Europe since the 1960s, the roughly tripled seat share, and the countries where such parties have entered government).
• Democratization, "Beyond the cultural backlash: exploring diverse pathways to authoritarian populism in Europe" (the scholarly case that cultural backlash alone does not account for the pattern).
• Wikipedia, "2024 United Kingdom general election" (Labour's victory ending fourteen years of Conservative government).
• Wikipedia, "2024 Mexican general election" (Claudia Sheinbaum's election).
• Wikipedia, "2025 Australian federal election" (the re-election of the Labor government first elected in 2022).
• Wikipedia, "2022 Brazilian general election" (Lula's return to the presidency).
• Wikipedia, "2023 Polish parliamentary election" (the opposition coalition's removal of the governing party after eight years).