I got a call from a friend in the Canadian media on the morning after the UK’s referendum on EU membership. “Were you surprised by the outcome?” she asked. It wasn’t an easy question to answer. Like many of my colleagues, I had expected a narrow win for Remain.
I was also cognisant of the warnings issued by pollsters in the final few days of the campaign that the result was “too close to call”. And, as a student of referendums, I was also aware of the volatility and uncertainty that have been the hallmarks of referendum campaigns in other places and times.
But what about “LeDuc’s Law”? The tendency of undecided voters to gravitate towards the status quo or the less risky option near the end of the campaign is a pattern that has often been observed in other referendums. That’s particularly the case for votes on complex constitutional issues, EU treaties, or proposed institutional changes. Some scholars attribute this to a natural tendency of voters towards risk aversion, and this has certainly been a factor in some of the referendums studied.
However, it’s hardly a law, and more than risk aversion is typically involved. The more complex the issue being voted on, the more there is a tendency on the part of some voters to just vote “no” to whatever change is being proposed.
Often this reflects misinformation, a lack of understanding of the issue, or just a desire to chastise the government or the individuals or groups behind the proposal. These “second order effects”, as they are often called, can be as important in determining the outcome as risk aversion.
In the case of the Brexit referendum, voters were never really able to assess the risks involved. Both sides were emphasising (and exaggerating) the dangers of what the other side was proposing. In other words, there was no risky option to shy away from – both the status quo and the change were being portrayed as a huge gamble. The Remain side focused on the economic risks of withdrawing from the EU and the Leave side essentially argued that staying in posed unacceptable security risks. The virulence of the campaign did little to encourage rational deliberation of the issues.
This is not an argument against referendums as a means of making decisions. Citizen engagement in some form is an important component of modern democracy. But it should be possible for such institutions and processes to promote the deliberation of important public issues rather than impede it.
The New Zealand votes on electoral reform in the 1990s played out over a longer period of time and involved multiple sources of public information. And Switzerland, with its extensive experience in direct democracy, has often been able to successfully manage campaigns on contentious issues.
There are better ways to do direct democracy than the high emotion, high risk affair just witnessed.