Study of 3 million people finds non-voters tend to die earlier

A new study of more than three million people in Finland has revealed a strong association between not voting and an increased risk of death over a 21-year period. The analysis showed that this mortality gap between voters and non-voters was even larger than the well-documented gap between individuals with the highest and lowest levels of education. The findings were published in the Journal of Epidemiology & Community Health.

In recent years, public health experts and medical associations have begun to frame voting as a social determinant of health. This concept suggests that a person’s health is shaped not just by genetics and lifestyle choices, but also by a wide range of social, economic, and environmental factors. The act of voting can be seen as a measure of an individual’s engagement with their community and their ability to influence policies that affect their well-being, from healthcare access to environmental quality.

Previous research has hinted at a connection between civic participation and health, but these studies often relied on surveys, which can be affected by how people remember their actions or their desire to give socially acceptable answers. A team of researchers from the Helsinki Institute for Demography and Population Health at the University of Helsinki sought a more objective way to investigate this link. They wanted to know if the simple, recorded act of voting was connected to the most definitive health outcome of all: mortality.

To conduct their analysis, the researchers obtained a remarkable dataset from Statistics Finland. They started with the official voting lists from Finland’s 1999 parliamentary election, which covered the entire electorate of the country. This information was then confidentially linked to national administrative records, including demographic data and death certificates, through the end of 2020. This method provided a precise, individual-level record of who voted and who did not, and their survival status over the subsequent two decades.

The study included more than 3.1 million Finnish residents aged 30 and older at the time of the 1999 election. The researchers used a statistical approach known as a Cox proportional hazards model to compare the mortality rates of voters and non-voters.

Free daily newsletter

This type of analysis allows researchers to calculate the relative risk of an event, in this case death, while accounting for other factors, such as age, to ensure a fair comparison. The team first calculated the risk of death for non-voters compared to voters, adjusting only for age. They then added education level into the model to see if it could explain the observed differences.

The results showed a clear and substantial difference in mortality risk. After adjusting for age, men who did not vote in the 1999 election had a 73 percent higher mortality rate over the following 21 years compared to men who did vote. For women, the difference was also significant, with non-voters having a 63 percent higher mortality rate than their voting counterparts.

When the researchers introduced education into their statistical models, the association between voting and mortality remained very strong. The elevated mortality risk for non-voters was only slightly reduced, suggesting that education did not account for the majority of the difference.

The mortality disparity between voters and non-voters proved to be wider than the mortality disparity between individuals with a basic education and those who had completed tertiary education, a finding that highlights the powerful connection between this form of civic engagement and health outcomes.

The association was even more pronounced when the researchers looked at specific causes of death. The risk of dying from “external” causes, a category that includes accidents, violence, and alcohol-attributable conditions, was more than double for non-voters compared to voters. This finding points toward potential links between non-voting and certain behaviors or environmental risks.

The analysis also uncovered nuances related to age and gender. The mortality gap between voters and non-voters was widest among younger adults, those under 50 years old. In another notable finding, women aged 75 and older who did not vote had a higher risk of death than men of the same age who did vote. This is an unusual pattern, as women typically have a longer life expectancy than men at all ages.

The study’s authors point out a primary limitation: their research establishes a strong correlation but cannot prove a direct causal link. It is not clear whether the act of not voting contributes to poorer health, or if underlying health problems and social disadvantages make it more difficult for a person to vote in the first place. For instance, poor health could reduce a person’s mobility, resources, or motivation to participate in an election. The study examined participation in a single election, which might not reflect a person’s long-term voting habits.

Future research could explore these dynamics more deeply. Longitudinal studies that track individuals’ voting patterns across multiple elections alongside changes in their health status could help clarify the direction of this relationship. Such work could better distinguish between individuals who consistently abstain from voting and those who miss an occasional election, potentially revealing more about the mechanisms connecting civic life and physical well-being.

Regardless of the precise causal pathways, the powerful association between voting and mortality has important implications. The researchers suggest that information on voting patterns could serve as a useful indicator for monitoring population health and identifying disparities. The findings also raise concerns about democratic equality. If populations with higher mortality rates are less likely to be represented at the ballot box, their needs and health concerns may receive less political attention, potentially creating a cycle of disadvantage.

The study, “Voting is a stronger determinant of mortality than education: a full-electorate survival analysis with 21-year follow-up,” was authored by Hannu Lahtinen, Isa Yang, Lasse Tarkiainen, and Pekka Martikainen.