Forcing people to vote doesn’t make them more engaged citizens, study finds

A recent study published in Public Opinion Quarterly provides evidence that requiring citizens to vote does not automatically make them more interested in politics or more supportive of democracy. The research suggests that removing penalties for not voting drastically reduces electoral participation but leaves a population’s political engagement and democratic attitudes largely unchanged.

Compulsory voting is a system where eligible citizens are legally mandated to participate in elections. In many of the roughly 25 countries that use this system, failing to cast a ballot can result in a fine or other penalties. Proponents of mandatory voting often argue that forcing people to the polls encourages them to learn about political issues and fosters a sense of civic duty.

This issue has become increasingly relevant due to shifting electoral habits worldwide. “This debate is held in part because of declining turnout rates in elections, so in this way we do also address a real-world problem,” said Dieter Stiers, a postdoctoral researcher for the Research Foundation Flanders at KU Leuven. The theory suggests that if everyone has to vote, political parties might expand their campaigns to educate a broader audience.

“The main motivation was the fact that there is a recurrent argument made in the debate on compulsory voting, that has been proven to be very difficult to test empirically: that by forcing people to vote, they will become more engaged and interested in politics,” Stiers explained. He noted that under voluntary systems, citizens can simply choose to ignore politics. “However, by making turnout obligatory, some kind of effort is mandatory. As they will need to vote anyway, it is argued, they will also be more likely to do at least some effort to learn about the parties and candidates.”

Critics argue that the system might breed resentment among those forced to the polls against their will. They also warn that compulsory voting might encourage unengaged citizens to vote at random, potentially lowering the overall political knowledge of the voting public.

Prior research on this topic has produced highly mixed results. Stiers and co-author Shane P. Singh, the Joshua W. Jones Professor of Public and International Affairs at the University of Georgia, noted that past studies often struggled with methodological trade-offs. Cross-national surveys comparing different countries can be influenced by cultural or economic differences that have nothing to do with voting laws.

Laboratory experiments, on the other hand, might not accurately reflect how people behave in the real world. “Testing this is very difficult, and previous research designs face important challenges,” Stiers said.

To overcome these hurdles, the authors took advantage of a rare policy change in Belgium. Belgium has enforced mandatory voting since 1893. In 2021, the regional government of Flanders chose to abolish this requirement for local elections starting in the year 2024. The rest of the country, including the regions of Wallonia and Brussels, kept the mandatory voting rules in place.

This regional split provided a unique real-world laboratory to test how changing election rules impacts the electorate. “We identified a unique opportunity to test this argument in the real world,” Stiers said. “With the study we hence want to contribute to the debate on compulsory voting with a strong empirical test.”

The researchers tracked the attitudes of Belgian citizens across the 2024 election season. They utilized the Belgian Repeated Elections Panel, which is a dataset gathered through a survey company. The survey was distributed in four separate waves to capture changes in public opinion over time.

The initial wave included 6,067 respondents, the second had 4,515, the third retained 3,343, and the final wave collected data from 2,744 individuals. The sample was matched to population quotas for age, gender, education, and region to ensure it represented the broader Belgian public.

The timing of the survey waves was specifically chosen to isolate the effects of the voting law change. The first two waves took place before and after national and European elections in June 2024. During these June elections, voting was still mandatory for all Belgian citizens.

The third and fourth waves took place before and after the local elections in October 2024. For these October elections, voting was voluntary for residents of Flanders but remained mandatory for residents in the rest of the country.

The survey measured several dimensions of political engagement and democratic attitudes. Participants were asked to rate their overall satisfaction with democracy on a scale from zero to ten. They also indicated their preference for democracy as a system of government on a five-point scale. Political interest was measured by asking respondents to rate their interest in politics from zero to ten.

The researchers also measured internal and external political efficacy. Internal efficacy refers to a person’s belief in their own ability to understand and participate in politics. External efficacy refers to the belief that the political system actually listens to and responds to regular citizens. Both were measured using levels of agreement with specific statements.

Finally, the survey tested objective political knowledge with four factual questions about the government and assessed trust in political institutions. To analyze the data, the researchers used a statistical technique known as difference-in-differences modeling. This approach does not just compare the two regions at a single point in time. Instead, it measures how the attitudes in each region changed over time, and then compares those changes against each other.

For this statistical method to be reliable, the two groups must have demonstrated similar trends before the policy change occurred. The researchers verified this by looking at the survey data from the June elections, when voting was still mandatory everywhere. They found that political attitudes in Flanders and the rest of Belgium fluctuated in near-perfect parallel during this period. This provided strong evidence that any sudden divergence in October could be directly attributed to the new voting laws.

The policy change produced a massive divergence in voter turnout. Before the rule change, local election turnout was similar across the country, sitting at around 92 percent in Flanders and 87 percent in the rest of Belgium during the 2018 elections. In the October 2024 elections, turnout in the regions that kept mandatory voting dipped only slightly to about 84 percent. In Flanders, where voting became voluntary, turnout plummeted to roughly 63 percent.

“When compulsory voting was abolished for the local elections in the Flemish Region, voter behavior in the election was very close to what you would expect from any election without compulsory voting,” Stiers said. “Whereas there was some expectation that the level of turnout, for instance, would still be rather high and then might decrease gradually in future elections, there was an immediate and sharp decline. Also other analyses we conducted were in line with voluntary elections. This change was more abrupt than expected.”

Despite this massive drop in electoral participation, the researchers found no significant differences in how political attitudes evolved between the regions. Across the entire country, citizens tended to show slightly stronger democratic attitudes and engagement immediately following the elections. This post-election boost is a known phenomenon in political science.

However, the data revealed that this increase in engagement was virtually identical in Flanders and the rest of Belgium. The removal of compulsory voting in Flanders did not negatively impact citizens’ satisfaction with democracy, nor did it reduce their political interest or knowledge. Similarly, being forced to vote in the rest of the country did not yield any extra educational or civic benefits compared to the voluntary system in Flanders.

These findings suggest that the act of voting itself, whether required or voluntary, does not drive deep changes in a person’s core political attitudes. “Taking the study at face value, we do not find any evidence that people’s democratic attitudes and engagement change differently in the run-up towards elections dependent on whether they are obligated to turn out or not,” Stiers said. Stiers added that because the study found essentially null effects, there is no real magnitude or size of the effect to discuss.

The researchers also investigated whether the rule change had different effects on specific types of people. They ran additional analyses to see if factors like gender, income level, educational background, or age influenced how a person reacted to voluntary voting. They found no evidence of any subgroup-specific effects.

The authors even looked specifically at individuals who admitted they stopped voting once the requirement was lifted. Even among these newly nonparticipating citizens, there was no measurable drop in political interest or democratic support compared to the rest of the population.

An important caveat is that the study relied on an online panel of individuals who opted into a survey company’s pool of participants. This means the sample, though demographically representative, was not drawn entirely at random. This factor could limit how perfectly the findings generalize to the absolute entirety of the Belgian public.

It is also uncertain whether these results apply to countries outside of Belgium. Cultural norms around voting and civic duty vary widely across the globe. A population that has experienced mandatory voting for over a century might respond differently to its removal than a country trying to introduce the policy for the first time. The findings are also limited to local elections, which might carry different emotional or political weight than high-stakes national contests.

“As any scientific study, there are of course important nuances and caveats,” Stiers noted. “The most important caveat is that we examine effects immediately after the very first election without compulsory voting. Democratic attitudes and engagement are likely rather stable and do not change that easily or quickly.”

A possible explanation for the stability in attitudes is that the Flemish public still had to vote in the mandatory national elections just a few months prior. This recent participation might have kept their political engagement artificially high during the local election season.

Future research will need to track citizens over a longer period to see if a permanent shift to voluntary voting eventually erodes political interest. “Even though we see that some changes, such as the participation rate, were rather strong, it is possible that attitudes change more slowly than behavior and that differences will appear in the longer run, something that future research can try to uncover,” Stiers explained.

Scientists might also investigate whether introducing a new voting requirement in a traditionally voluntary system produces a different psychological effect than removing an old one. “The main goal of the larger project this article is a part of, is to uncover the consequences of compulsory voting beyond turnout,” Stiers said. “Examining these spillover effects can inform the debate on compulsory voting by providing a more holistic view on its effects on voters and elections.”

The study, “Does Compulsory Voting Improve Democratic Attitudes and Engagement? Quasi-Experimental Evidence from Belgium,” was authored by Dieter Stiers and Shane P. Singh.