This article is written by Lucy Makinson, a Research Advisor at the Behavioural Insights Team.

Whenever there is an election, predictions about who will “win” attract substantial media attention — understandably! But for those of us who are interested in civic participation, we tend to turn our attention to another metric — election turnout.

While civic participation itself is hard to define and measure, election turnout is not - making it an important outcome metric for research into civic participation.

In the US, where individual data on voters is publically available, the ability to match Get Out the Vote (GOTV) campaigns — non-partisan campaigns encouraging voter registration and turnout — to individual actions has given rise to numerous trials evaluating what works for voter mobilisation.

The most crucial insight from these trials is that voter turnout is malleable. While many trials have found no effect on turnout, it is possible to change someone’s decision to participate with relatively small interventions even in cases where the deadline to register to vote has passed. The second insight is that how outreach is done can matter as much as whether it is done at all. Here are three things we can learn from election turnout studies, and what they can tell us about civic participation:

1. Be clear about what you want people to do

This might seem obvious, but one of the hardest and most important aspects of behaviour change is being clear and specific about the behaviour you want to encourage.

Sending multiple messages can muddy the waters and make any action less likely. This poses a challenge for political parties. Whilst single-purpose political campaigns, such as that of Get Out the Vote, have a good record of mobilising voters, partisan campaigns, which are generally concerned with a far greater multitude of issues, yield murkier results. What that means is that if you want to entice people to vote, it is better to make that your only request.

The most effective approaches to behaviour change are timely, and voting is no different

In the context of civic engagement more widely, it is more effective — and more likely to shift behaviour — to focus on promoting specific actions such as inviting citizens to attend a local council meeting or signing a petition, than a looser request to become involved. And those requests are more likely to be successful if the event is politically neutral.

2. Help people to form a plan

The most effective approaches to behaviour change are timely, and voting is no different.

One study which looked at phone canvassing in the four weeks before an election found that only canvassing in the final week had an effect. But speaking to people right before they need to take action isn’t always feasible and, for many forms of civic participation, there isn’t a specific day on which action needs to be taken.

Instead, encouraging people to make their own plans and elaborate on them by specifying the when and where can make it more likely that they will follow-through on a desired behaviour. For example, using a telephone survey to ask people when they would vote, where they would be coming from, and what they would be doing beforehand significantly increased turnout in the 2008 US Presidential election.

This also seems to apply outside the scope of democratic elections. Whether it’s getting someone to write their CV, attend a cancer screening or a training session, encouraging someone to elaborate on their plans increases follow-through. At BIT, we asked people to tick a box to indicate which day they would do the census, which (along with some other changes to communications) doubled response rates.

These "implementation intentions” can be effective, but are by no means a cure-all. It probably won’t, for example, affect people who already have a plan: voters in multi-voter households were more likely to have made a plan to vote and, on average, talking through their plan had no effect. In addition, the approach is designed to increase follow-through when the intention to do something is there, so [initial motivation is important](http://psych.nyu.edu/oettingen/Oettingen,%20G.,%20Hnig,%20G.,%20&%20Gollwitzer,%20P.%20M.%20(2000%29%20Effective%20self.pdf). Whilst voter registration should indicate some motivation for election turnout, this might not be the case for other civic actions.

3. Reduce the information gap

It’s probably not a surprise that people are more likely to vote when they know about the issues at stake. But measuring the size of the effect is difficult, and establishing the causal direction, that more information makes people more likely to vote, is even harder.

A 2004 study aimed to address this by using a pilot information campaign on decentralisation in four Copenhagen districts to evaluate its impact on a subsequent referendum. They found that the information campaign increased voter turnout by just over 3%.

This was driven by people being more likely to hold an opinion on the issue, with those holding an opinion being over 25% more likely to vote and completely explaining differences in turnout due to education levels.

If you want people to engage with information, make it easy

Raising the level of information is, of course, not straightforward. Especially when faced with the catch-22 where those who are not informed are less likely to engage. That means that the people we are most trying to reach are also the least likely to listen to us.

But reducing the information barriers can take place on a much simpler and smaller scale. BIT recently worked with Hackney and Hull councils and the UK Electoral Commission to encourage households to return their Household Enquiry Forms (HEFs). These are sent to households each year to make sure that residents are registered to vote and encourage residents that are not registered to do so.

In total, we tried nine different versions of the letter sent with the HEFs. The two best interventions resulted in a 2% increase in return rates — and led to substantial cost savings. However, the most effective approach in both areas was a simple change to the envelope, to include the message “DO NOT IGNORE THIS ENVELOPE. If you live here, your response is legally required”. It increased response rates by 8.5%, saving time and money in door-to-door follow-up. In summary, if you want people to engage with information, make it easy. — Lucy Makinson

This is just a snapshot of the full literature on experiments in election turnout. For those who are interested in reading more, Gerber and Green have written a fantastic summary which you can find here.

(Picture credit: Unsplash)