How a close race is bringing Scots to the polls

One thing is certain about the outcome of the Scottish independence referendum on September 18: turnout will be among the highest in UK history. Evidence from Switzerland suggests the recent increase in turnout is connected to predictions of a very close race, says political scientist Oliver Strijbis. 

Based on voter registration, the Electoral Commission estimates that about 80% of the electorate will cast their ballot. But opinion polls are seeing even higher figures: about 85% of voters say they will definitely turn out to vote. And in the latest surveys, this has risen to about 90%. 

While the generally high level of mobilization can be explained by the issue at stake, the recent increase in the number of citizens willing to cast their vote is due to other factors. 

As evidence from Switzerland suggests, this increase in turnout is most likely connected to the fact that polls are predicting a very close race. 

Past research into parliamentary and presidential elections has shown that turnout is particularly high when voters expect a close outcome. 

While this finding might seem unsurprising at first glance, it isn’t if you consider the fact that the potential impact of an individual’s vote is minimal. 

But why should a citizen chose to rather vote in a tight race? Because the probability their ballot is the decisive one is no more likely than winning the jackpot in a national lottery. 

Political scientists have come up with two explanations why turnout might increase with the expectation of a close race. First, voters may heavily overestimate the probability that the vote they cast is the crucial one. 

As cognitive psychologists have demonstrated, this overestimation of low probabilities does not only take place during elections, but is a human trait observed in many other situations. 

The other explanation is linked to the behaviour of the political elite. If a vote is expected to be tight, politicians may try to make the difference via last-minute campaigning to try to mobilize as many voters as possible. 

Also, lobby groups may be more willing to spend on political campaigns if they expect their money will make a difference. And the media is much likelier to cover a campaign if they can present it as an exciting horse race. A more intense campaign and neck-and-neck race have positive effects on turnout reminding citizens to vote and giving them the feeling there is something important at stake. 

While these explanations are well known for elections, it is unclear whether they also play a role in direct democracy votes such as the Scottish independence referendum. 

To find out more about this phenomenon Laurent Bernhard (University of Zurich), Sveinung Arnesen (University of Bergen) and myself have looked closely at the relationship between expected outcomes and turnout among voters in Switzerland’s largest canton – Zurich. 

Our major challenge has been to find a good way of measuring expectations about an outcome. You may assume that these can be measured using survey data. But polls are generally interested in how people are most likely going to vote and not what they expect the outcome to be, so they are not useful in this case.

For our study we instead measured the expected outcome with data from a betting market. This is a virtual market where people bet on the outcomes of events such as football matches or the results of elections.

Since the winning outcome depends on where participants put their money, a betting market gives valuable information on what outcome they expect – in our case the percentage of ‘Yes-votes’ in a referendum or initiative.

In order to establish whether the expectation of a close race – as measured with data from the betting market – brings people to the polls, we therefore compared the data from the betting market with official turnout figures.

From our analysis we can deduce two results. First, the expectation of a tight race also triggers turnout in direct democracy decisions.

Second, we have found evidence of the two mechanisms described above: people tend to participate at higher rates when they expect the outcome to be tight because they think their vote might be decisive and because the campaign is especially intense in such an instance.

Although the citizens of Zurich vote on more mundane issues than independence, we are confident that our observations are also valid for the Scottish electorate.

Oliver Strijbis has a PhD in Political Science from the University of St.Gallen. He is a senior researcher at the Social Science Research Centre in Berlin (WZB) and, together with Laurent Bernhard, author of the direct democracy blog www.50plus1.ch.

Opinion series

swissinfo.ch publishes op-ed articles by contributors writing on a wide range of topics –Swiss issues or those that impact Switzerland. Over time, the selection of articles will present a diversity of opinions designed to enrich the debate on the issues discussed.