Voter turnout has been declining across many democracies for decades. This trend has raised a familiar concern: if fewer people vote, do elections still represent what citizens want?
A common assumption is that non-voters think differently from voters. According to this view, low turnout leads to distorted election outcomes because only a specific segment of the population shows up at the polls.
But how true is this assumption?
In recent research, I examined whether voters and non-voters in Europe actually differ in their policy preferences, and what this means for democratic representation.
What people usually assume about non-voters
People often draw a simple conclusion: those who abstain must have very different political views. This belief is intuitive, especially given that we know that non-voters are more likely to be younger, less educated, and economically less secure. It is therefore tempting to assume that elections mainly reflect the preferences of a socially and politically distinct group.
If that were true, low turnout would be a serious and constant threat to democratic representation.
What the data show across Europe
To test this assumption, I used survey data from 29 European democracies, covering a wide range of political issues:
- left–right ideology
- economic policies such as redistribution and public spending
- social issues such as immigration and same-sex marriage
- attitudes toward the European Union
The first and perhaps most important finding is reassuring:
In most countries and on most issues, voters and non-voters tend to hold similar policy preferences.
Large gaps between the two groups are not the norm. Moreover, even when statistically detectable differences appear, they are usually small.
This suggests that low turnout does not automatically mean that election results are completely disconnected from what citizens want.
The problem is not constant differences, but occasional bias
However, stopping the analysis here would miss the main point. While differences between voters and non-voters are usually modest, they are not evenly distributed across contexts. When differences do appear, their consequences depend heavily on turnout levels.
The logic is straightforward. If turnout is high, non-voters make up a small share of the population, and even noticeable differences in preferences will have limited impact on overall representation. But when turnout is low, even small differences can matter.
In such cases, elections start reflecting the preferences of regular voters more than those of citizens as a whole.
Why this matters for real politics
This dynamic helps explain why people can feel underrepresented even in functioning democracies.
Elections do not simply reveal “public opinion.” They send signals, and those signals come disproportionately from people who vote consistently. Policymakers, parties, and governments respond to these signals when they design policies and strategies. Over time, this can create a subtle feedback loop:
- policies reflect regular voters more closely
- non-voters feel less represented
- disengagement persists or deepens
Importantly, this does not require dramatic ideological divides between voters and non-voters. Representation can become skewed even when differences are relatively small.
Not all non-voters are the same
People who never vote differ from those who vote occasionally, and both groups differ from regular voters. Peripheral voters, those who vote intermittently, are particularly important because when they turn out, overall representation improves noticeably. This is one reason why high-turnout elections tend to produce outcomes that better reflect the broader population, even if some citizens still abstain.
For example, let’s look at the British case. The differences between regular voters, peripheral voters and perpetual non-voters are real, but they are not dramatic. In the United Kingdom, non-voters (both peripheral and perpetual non-voters) are, on average, (1) more left-wing, (2) more supportive of redistribution, (3) more pro-Europe, (4) more supportive of environmental protection, but, at the same time, (5) they are slightly less positive toward immigration.
This is not a simple story of “non-voters are more progressive.” The differences can move in different directions depending on the issue. But, the centre of gravity of the electorate shifts depending on who turns out.
In Britain, perpetual non-voters make up a relatively small share of the population (around 7%), which limits how much their abstention alone can distort representation. But the more consequential group is peripheral voters. Peripheral voters are generally closer in their policy preferences to perpetual non-voters than to regular voters. This means that when peripheral voters participate, the electorate’s average position moves closer to that of the broader citizenry. When they abstain, the electorate reflects more strongly the preferences of regular voters. In other words, the representational effect of turnout depends less on the small group that never votes, and more on whether occasional voters show up.
What this means for debates about turnout
Voters and non-voters are usually not far apart in what they want. In most European democracies, their policy preferences are broadly similar. For that reason, moderate declines in turnout do not automatically distort representation in a dramatic way. When turnout remains reasonably high, the electorate still provides a fairly accurate signal of citizen preferences.
The issue becomes more relevant when turnout drops to very low levels. In those contexts, even modest differences between regular voters and the broader population can begin to matter. Elections then reflect more strongly the preferences of those who participate consistently.
Given that governments and parties respond to the signals elections send, policy will gradually align more closely with regular voters’ preferences if those signals disproportionately reflect their preferences.
This is not a claim that democracy fails whenever turnout declines. When turnout is reasonable high, elections still provide a reliable signal of citizen preferences. The representational risk emerges when turnout becomes so low that habitual voters effectively define the electorate.
That said, in highly competitive elections, even incremental differences in who is mobilized to vote can be decisive for who ultimately takes office.
If low turnout begins to skew the electorate toward regular voters, policies such as compulsory voting become relevant as institutional ways of broadening the signal elections send.
Author note: This post is based on cross-national research published in the European Journal of Political Research examining policy preference differences between voters and non-voters across European democracies. To see the full paper, click here.