Turnout information and voting
Project description
This project evaluates the impact of information provided by a non-partisan third party on voter turnout/non voting. The aim is to see whether information about the voting/non voting history of the reference group affects voting in elections. The idea is to compare these voters to those in a control group and to those in a group receiving just a persuasive message.
We focus on first-time voters, who have had very low and declining turnout in recent years. Successive government reports and expert studies have highlighted their decrease in political participation. There are many local initiatives seeking to promote the turnout of young people and it is hoped this project will provide an understanding of what messages can appeal to these potential voters.
The first part of this project is a field experiment examining the impact of a mailshot to the home addresses of a random sample of young people. The sample is drawn from those who have registered to vote for the first time and are identifiable by local authorities. We will randomize these people into three groups. One receives a letter with positive turnout information among young people and encouraging them to vote, another group will receive a neutral letter just encouraging them to vote and a final group gets no correspondence (control). The appendix contains a summary of the messages we could use.
Our first partner is Shropshire County Council, who is supplying the project with about 3,000 first-time voter addresses. The council will mail out the voters with our letters at cost (paid for by the research project). For Shropshire residents, we use Shropshire turnout information in recent elections and it is anticipated we will use the same information from other councils who participate in the study.
With a similar sample size from two other councils we will have enough participants to compare voter turnout from electoral registers after the 2009 elections. We would seek to use regression models using covariates at the super-output level to produce results of the effect of the treatment.
To develop the intervention we are running a focus group study of young voters in Shropshire and Manchester aged between 17 and 18. During the focus groups we present information on turnout among young people using different phrasing and local or national turnout figures and ask the participants to explain their reactions to the information and likely effects. We would need to use these discussions to ensure the messages are easily understood by all participants, phrased as meaningfully as possible, and not missing any stated information that participants’ views to be pivotal.
We are also running a small internet survey among a non-random sample of people (students, friends and family) using the website www.surveymonkey.com. This is a widely-used, professional and inexpensive way of collecting data among non-random samples. Here we could pilot the turnout information and ask participants to provide feedback, in an open ended format, on their reactions to this information.
The final aspect of the study is the survey experiment in the British Election Study Continuous Monitoring Survey, to be run in March 2009. This is an opportunity to test the basic expectations about the effects of turnout information and also an expression of support for the importance of the idea and the research project. The survey experiment provides the ideal opportunity to test this central premise – we can evaluate the effects of providing information about abstention in a context of hypothetical voting, or likelihood to vote and this has been approved by the BES team for inclusion in their Continuing Monitoring Survey. With a sample size of 1,200 respondents, if there is a large effect size, it is possible to separate by treatment group, and then again by treatment group and by age (or gender, locality, etc) though the opportunity to do this will be limited.
The project team is: Jane Green, Sarah Cotterill and Peter John from the University of Manchester (the contact is jane.green@manchester.ac.uk).
The outputs of the project will be a report and a policy briefing for the Ventures project and at least two academic papers to be given at conferences in 2010.
