Tue Feb 13

Which day of the week gets the most people to vote? We analysed thousands of international elections to find out

Written by Ferran Martinez i Coma, Senior Lecturer in Political Science, Griffith University

In the aftermath of elections, one of the issues usually discussed in the media is the amount of people who turned out to vote. This is known as “participation” or “voter turnout”.

Several factors, such as the weather, can affect turnout. For example, the Republican primaries in Iowa on January 15 were held in very cold temperatures (subzero wind chills and a blizzard). Commentators have identified the cold as a factor that negatively influenced ^[1] turnout, as many Republican voters decided to stay at home, even though Iowa is (almost) always cold in January.

The Republican primaries were held not only on a cold day, but on a working Monday. Yes, a Monday. This may not sound all that strange to the US public, who are used to voting on Tuesdays in their general elections, but it could for Australians who are used to voting on Saturdays. Australia is one of only a few countries that vote on Saturdays, along with Cyprus, Malta, Iceland, Latvia, Slovakia, Taiwan and New Zealand.

But, does it matter when we vote? Does it affect voter turnout? Do we know if more people vote during the weekend than, say, on a Tuesday? We analysed data from thousands of elections across the globe to find out.

What’s the most popular day to hold an election?

We looked around the world to see when people vote. We collected turnout data for 3,217 national elections between 1945 and 2020 in 190 countries. We then collated the data and created an original dataset ^[3] on turnout.

The first thing we can assess is which day of the week most global elections are held.

The graph shows, in general, voting takes place on weekends (more than 60% of elections), with Sunday being the preferred day. The day on which the fewest elections are held is Friday.

We could also examine how many countries choose a given day of the week to hold their elections. The graph below shows that 94 countries chose a Sunday for polling day, while just eight went with a Friday.

Interestingly, this preference for Sunday elections is not evident in countries with a significant Protestant Anglo cultural influence, in which public activities other than going to church tended to be restricted on Sundays. For example, in Australia, everything used to be closed on Sundays: bars, cinemas, shops, and there were no sporting events (the restrictions were gradually lifted from the 1980s).

How does that affect voter turnout?

So is there any relationship between the day on which you vote and participation?

The studies currently available show varying results. For example, a 2004 study ^[4] that considered 29 countries found that when the election was held on a Sunday, participation was higher. However, when the analysis was expanded to 63 countries, the day of the election did not seem to affect participation.

As the graph above shows, the median voter turnout is around 70% for every day of the week.

For example, the average participation on Sundays was 71.6% while on Fridays it was 70%.

Therefore, it does not appear that the day on which the election is held is related to the level of participation.

This answer is simplified, of course. We are mixing democracies and authoritarian countries, places where there is mandatory voting and places where there is not, presidential and parliamentary systems, and countries that hold elections with either one or two rounds, among many other factors.

Why does this matter?

When to vote (and whether to vote or not) is an issue that matters. Participation is unequal and is used strategically, especially in countries where voting is not compulsory. In some countries, wealthier voters tend to show higher participation rates ^[6] than poorer voters. This is a pattern that has been identified ^[7] in the United States and Europe but not necessarily in other countries such as India or Indonesia.

Participation is strategically used by political parties promoting (or disincentivising) voting in different ways and to differing extents. There are blatant examples of parties strategically managing voting around the world. In Kenya, polling booths in some areas have more staff than others ^[8], skewing how many people are able to cast a vote before closing time. In the US, strict voter ID laws have acted to suppress the votes ^[9] of some racial and ethnic groups.

Some instances are more insidious. In 2008, Spanish campaign director Elorriaga Pisarik, in referring to undecided socialist voters, declared ^[10] “if we can generate enough doubts about the economy, immigration and nationalist issues, maybe they – the socialist voters – will stay at home”.

Participation also has an intrinsic value. Imagine two scenarios: one in which the candidate wins the election with 51% support, in an election that had a 90% turnout. Then imagine another election where the candidate wins by the same margin but in an election with a 30% turnout. Although both victories are valid, we tend to attribute greater legitimacy to the one that has brought more people to the polls.

In a year when more than half the world’s population will vote ^[12] in a national election, it’s worth including data in the global discussion.