Can you increase voter turnout?

Cartoon in danish
Extract from comic strip about elections, by Thomas Thorhauge (in Danish).

KASPER MØLLER HANSEN, PROFESSOR AT DEPARTMENT OF POLITICAL SCIENCE

Kasper Møller HansenYoung people are less engaged in local politics than senior citizens. Only 58 per cent of 19 to 21-year-olds vote in local council elections, while the turnout among voters aged 70-79 years is 83 per cent. So how do we get young voters mobilized?

This question was the starting point for the research project 'Can you increase voter turnout?’, launched in 2013 by Professor Kasper M. Hansen.

The project was a joint initiative involving the Danish Youth Council, the Ministry of Economic Affairs and the Interior, the Ministry of Social Affairs, the Danish Parliament, the Danish Research Council and all 98 of Denmark’s local authorities. It was based on data from the local elections held on 19 November 2013, when a number of campaigns were launched aimed at getting more young people to vote.

Method

Researchers conducted a number of field experiments during the local elections in 2013 and followed them up with campaigns at the European parliamentary elections in 2014 and the local elections in 2017 in order to compare the effects of the different types of campaign. They also studied the impact of campaigns aimed at immigrants and people on welfare, two groups who also statistically have way-below-average voting rates.

It is essential for a well-functioning democracy that as many people as possible participate in elections, and it is particularly important that all groups of voters are represented. In the United States and other countries, campaigns to increase voter turnout have been a permanent fixture for many years, but in Denmark it is very new, and therefore no systematic surveys have ever been carried out to test the effect of such campaigns on Danish voter participation. We now know much more about what it actually takes to get young people to vote, and we can see that local authorities and organisations are applying this knowledge on a large scale.
Professor Kasper Møller Hansen, Institut for Statskundskab

The mobilization campaigns took place in the weeks leading up to the election and consisted of various initiatives, such as text messages for young voters encouraging them to vote, letters and postcards, and face-to-face chats (knocking on doors).

Before the election campaign, the researchers divided the target voters into random groups. One group served as a control group and were not subjected to any of the specific campaign initiatives, while the other groups - the so-called stimuli-groups - were exposed to the specific initiatives. In this way, researchers could show the effects that different types of voting campaigns had on young voters.

Campaign 1: The Danish Parliament tries to reach young people using the Constitution and humour

The Danish Parliament's campaign consisted of sending a copy of the Danish Constitution to voters who had turned 18 years old during the six months prior to the local elections in 2013, i.e. first-time voters. The Constitution was accompanied by either a traditional letter or a humorous cartoon encouraging the recipient to go and vote. The letters were received four days before the election. Analysis of the results showed that the voter turnout for those who received the letter was 1.1 percentage points higher than for the control group. Delving deeper, the figures showed that especially the cartoon had a positive effect on voter turnout. Overall, the experiment shows that encouraging voting can actually boost voter turnout, even though there may already be a lot of people in the target group who do use their right to vote. The impact on the young people who had a potentially low turnout was an approx. 5 percentage point increase, a group consisting of a lot of new Danes with immigrant backgrounds, whose families also don’t tend to vote so often.

Campaign 2: The Ministry of Economic Affairs and the Interior used the force of argument

The Ministry of Economic Affairs and the Interior’s campaign targeted first-time voters in the age group 18-21. In this campaign, eight different kinds of letter were tested, each using a different argument as to why it is important to vote. In all, 100,000 letters were sent out.

The total impact was positive with a modest increase of 0.4 percentage points, which is in fact only just statistically significant. There were indications that certain arguments worked better than others and that a letter outlining several different arguments for voting was particularly effective. The impact of this campaign should be seen in the light of the fact that it is relatively expensive to send letters out.

Campaign 3: The Danish Youth Council counted on the effectiveness of text messages

The Danish Youth Council sent a text message with the following wording: "Democracy needs you" to 140,000 young people aged between 22 and 29. Recipients were also reminded about the election date. The text messages increased average voter turnout by 1.8 percentage points. This is quite a considerable effect. Especially compared to how short text messages are and how cheap they are to send. Here, too, the effect on groups with a potentially low turnout was approx. 5 percentage points.

Results

The researchers’ field experiments show that all three voting campaigns increased voter turnout, but that the size of the impact varies, with text messages clearly showing the lowest cost per additional vote. The researchers also measured the indirect effects of the campaign, and it actually turns out that 30 per cent of the effect "rubs off on" other people in the same household and spreads in the voter's social network, so that, for example, family and others in the household also choose to vote.

Table over results

Overall, it appears that campaigns of this type particularly mobilise voters who generally belong to low turnout groups, thereby contributing to reducing the gap in voter participation across different groups of voters.

As a result of the research showing that text message campaigns worked, a large number of local and regional councils across Denmark decided to send text messages out to their voters during the local elections in 2017. And the Danish Parliament has altered its practice so that the Constitution is now sent out to all first-time voters just before an election rather than just before their 18th birthday.

The researchers hope that the results from the project will yield additional knowledge on how to get the most value for money from public sector campaigns in general.

Further research

The researchers are now investigating how to fine tune the impact of text message campaigns by experimenting with when the text messages are sent, who the sender is, and how personal the tone is. The project will also be extended to include not only young people, but also other low turnout groups of voters and will shift the focus to digital post in order to test new forms of communication. Finally, the researchers are also working together with other researchers on ways of using social media for controlled experiments in other contexts involving mobilising groups of people.

The experience gained from the 'Can you increase voter turnout?' project was applied by many Danish local authorities during the local elections in 2013 (and repeated in following elections)

  • Many local councils sent text messages to young voters encouraging them to vote
  • Other organization targeted immigrants with text messages, other people on welfare and yet other again the elderly.
  • In total, more than 300,000 text messages were sent out in an attempt to get more people to vote. Researchers estimate that the proportion of people voting rose by 2 percentage points among the groups that received text messages.

Can you increase voter turnout?

An analysis of a mobilisation campaign at the local elections in 2013
 
Kasper M. Hansen, professor, Department of Political Science, UCPH
Yosef Bhatti, professor MsO, VIVE – The Danish Center for Social Science Research
Jens Olav Dalgaard, assistant professor, Department of Business and Politics, Copenhagen Business School
Jonas H. Hansen, research assistant, Department of Political Science, UCPH