Ailsa Henderson, Liam Delaney and Robert Liñeira discuss the findings of their report on Risk and Attitudes to Constitutional Change. The report is available for download here.
Any referendum on independence must confront the issue of risk and loss. For some voters, the prospect of pursuing a radical constitution change carries with it risks so great that the very idea is unthinkable. Possible losses could include security or shared cultural ties. For others, the risk might be framed as a lost opportunity, whether for better policy or achieving statehood. The two campaigns have employed the language or risk and loss at times, highlighting the consequences of fiscal uncertainty for an independent Scotland or limited autonomy for Scotland within the United Kingdom. Our study, Risk and Attitudes to Constitutional Change, shows that voters are also thinking of the risks associated with a Yes or a No vote and that risk plays a far larger role in determining vote choice than other things we typically look to such as gender, age or even national identity.
Our research, funded by the ESRC Future of Scotland and the UK programme, is part of a larger project coordinated by the Scottish Centre on Constitutional Change. In June 2014 we conducted a survey of 2063 Scottish residents eligible to vote in the 18 September referendum. The results give us a useful update on the latest voting intention figures but they also give us new data on engagement with the referendum campaign as well as attitudes to risk and uncertainty.
In our sample 51% said they would vote No and 38% said they would vote Yes in the 18 September referendum. Our findings suggest that the number of undecided voters is no more than 12%, with really only five percent of voters truly no more likely to vote Yes than to vote No. Potential voters are more engaged with the referendum campaign than with other areas of politics (including Scottish and UK politics), are satisfied with the referendum process and feel they have a high level of knowledge. Actual levels of knowledge, however, are fairly low. In response to ten true/false statements about the Yes proposals none was answered correctly by a majority, although knowledge about the desired currency (46% correct) and Head of State (44% correct) is highest. There is also uncertainty about the level of information provided by the campaigns. 70% believe neither campaign can accurately estimate the true consequences of independence. Potential voters are more likely to be seeking information from newspapers and television than the two campaign sides.
Our research also shows that perceived risks play a far greater role in determining vote choice than previously assumed. General tolerance for risk relates to constitutional preference – so that those more willing to take risks in general are also more supportive of constitutional change – but this general relationship disappears when we look at the specific risks that people anticipate in the event of a Yes or No vote. A clear majority of No voters believe that a UK government won’t allow an independent Scotland to keep the pound (70%) and that Scotland won’t retain membership of the EU on the same terms as the UK (63%). An equally large majority of Yes voters, by contrast, believe that the UK government will cut the Scottish Parliament’s budget if Scotland remains within the UK (73%) and a smaller majority believe that the UK will vote itself out of the EU in a referendum (51%). No voters who anticipate a Yes win have also engaged in a number of ‘risk mitigating’ behaviours, including preparing to move from Scotland, seek employment outside Scotland or move bank deposits out of Scottish banks.
When it comes to predicting voting intention in the referendum, risk matters: if you think an independent Scotland will be able to use the pound sterling as currency, if you think it will retain membership in the EU, if you expect that a No vote will result in the UK government cutting spending for the Scottish Parliament and, critically, if you feel certain about the consequences of independence, you’re more likely to vote Yes. National identity matters – those who feel Scottish not British are more likely Yes voters - but it matters less than attitudes to risk. Other demographic variables, such as age, gender or employment do not influence voting intention at all once we control for attitudes to risk.