With two weeks to go until the independence referendum it is perhaps worth looking at what was happening in the final weeks of the 1995 referendum on sovereignty partnership in Quebec to see if there are any interesting parallels or useful lessons.
At this point in the Quebec campaign the opinion polls were predicting a close race, with the Yes side on 44%, the No side on 46% and 10% undecided (Leger and Leger 6 October). The referendum, unlike the one on 18 September, had a longer question, one which offered sovereignty partnership and, if negotiations over that partnership failed, the prospect of independence. In a sense there was greater uncertainty in 1995. What, exactly, would sovereignty partnership look like? If I cast a ballot for the Yes side did that mean I would get sovereignty partnership or independence? In either case it was not clear what the outcome of partnership negotiations would be. Then, as now, the Prime Minister and his government warned that claims by the Yes campaign were just that, claims rather than guarantees, while Yes campaigners argued that the opposing side was painting an unduly dim view of the future.
Some elements of the Quebec debate would therefore be very familiar to observers of the Scottish referendum. Business leaders warned of the risks of independence, including large firms like Bombardier and Standard Life, who both threatened to leave Quebec in the event of a Yes vote. Also in the final weeks of the campaign the federal finance minister cautioned that independence for Quebec would cost one million jobs.
Other debates have no Scottish parallel. There were repeated challenges from the No side about the territorial integrity of Quebec: Would an independent Quebec keep its current borders? Its smaller borders from 1912? Smaller still from 1867? What of the predominantly No-voting island of Montreal? First Nations Aboriginal people living predominantly in northern Quebec had organised two referendums of their own, to occur before the final province-wide referendum on 30 October, with Inuit and Cree voting overwhelmingly (95%) to stay in Canada
With less than two weeks to go every poll was predicting more than 50% support for the Yes side among decided voters, but with figures of undecideds ranging from eight to fourteen percent. Polling embargos meant that the 72 hours before the referendum, on 27 October, the last published poll predicted a Yes win. That same day, the No campaign held what it called a Unity Rally in Montreal, encouraging Canadians to send a message that they didn’t want Quebec to leave. Other organisations then jumped in. VIA Rail, Canada’s national train company, offered discounted fares for passengers to travel to Montreal. Student organisations in neighbouring Ontario rented school buses to bring university and college students to the rally. These signs of ‘support’ were later deemed to run afoul of the clear spending limits in the referendum.
The result was famously close, 50.6 % for the No side and 49.4 % for the Yes side, with a turnout of 93.5%, more than ten points higher than the previous provincial election that had brought the separatist Parti Québécois to power. This was, of course, not the first time the Yes side had lost a referendum. In 1980, then Premier and Parti Québécois leader Rene Levesque announced that it was a defeat that had the air of victory given how much had been accomplished by Yes supporters. In 1995, when the Yes side lost narrowly, then Premier and PQ leader Jacques Parizeau was less magnanimous, blaming the loss on ‘money and the ethnic vote’ For some the reference to money invoked the federal government who, it was claimed, had violated spending rules in the referendum. For others it was perceived as a reference to the Jewish and immigration populations in Quebec, predominantly living in Montreal, who had prevented ‘ethnic’ Quebecers from achieving their goal, a view which was at odds with the PQ’s formal rhetoric of inclusiveness. Both interpretations compare poorly to Levesque’s gracious reaction, and Parizeau resigned afterwards (which in a pre-recorded interview he had indicated he would do if the Yes side lost)
Given how close it was, what is remarkable is the level of quick acceptance of the results. This was not an equivalent of the Florida recounts and arguments over dangling chads that bedevilled the Bush-Gore election of 2000. That said, there were official complaints about the behaviour of different actors. The No side complained about vote count violations, pointing out that there were significantly higher numbers of spoiled ballots in constituencies expected to produce high No votes, a fact later attributed to the training received by particular returning officers. The Yes side complained both about spending violations related to the Unity Rally and the fact that the federal government had seen a surprising increase in its efficacy in processing citizenship applications (approving the equivalent of one quarter of a year’s applications the month before the referendum). Those in new receipt of Canadian citizenship, it was assumed, were less likely to back independence for Quebec.
Such a close result and questions about violations could have posed considerable problems, including increased distrust and dissatisfaction, disengagement with politics, or low levels of voter turnout in subsequent elections. But none of that happened.
Conspicuous by its absence was any claim of impropriety on the part of the body charged with the conduct of the referendum, the Directeur générale des élections, the provincial equivalent of the Electoral Commission. In a close race, the ability of the losing side to accept the result as fair is key, and central to this acceptance is a perception that the process has been as fair as possible. While there were criticisms of particular actors, there seemed to be a belief that the body charged with investigating those criticisms and bringing any charges would act impartially.
Another reason for the reaction is likely due to the fact that there is something of a status quo bias with close results. If one side has lost by a narrow margin, better for it to be the side that can campaign again for further change than the side forced to live in conditions that are irreversible. Such is usually the thinking behind those arguing for thresholds in referendums.
For its part, the Parti Quebecois government, led by the accomplished politician Lucien Bouchard after the departure of Jacques Parizeau, lowered the temperature when it promised that it would not hold another referendum until it could satisfy certain winning conditions: positive economic conditions - including no fiscal deficit and favourable levels of unemployment - and a winnable campaign - including a popular leader and a clear lead in the polls. Until then, it would get on with the act of governing. For this tactic they were rewarded with another majority government in 1998. In that election they did, however, earn less of the popular vote that their Liberal opponents, which ensured debates about electoral reform surfaced in the following decade.
The Canadian government, however, had a different reaction. In its secession reference to the Supreme Court it asked if it would be possible for any province to secede, unilaterally, from Canada. The Court, with Solomonic wisdom, replied that while there was nothing in international or domestic law that would allow a province to secede unilaterally both sides would be obliged to negotiate if a clear majority of Quebecers expressed their preference for independence in response to a clear question. The federal government responded with a Clarity Act, which set out the terms of any future referendum on independence. The legislation says that any referendum about self determination must refer to independence and only independence and not to any other constitutional arrangement, it must be won by a clear majority (which has subsequently been interpreted in other referendums to mean 60% support) and a Yes vote would be treated as necessitating an amendment to the Constitution, which would mean all the provinces would need to be included in the eventual negotiations. Some of the language in the Clarity Act and the tactics behind it appear similar to that behind the UK’s approach to the Edinburgh Agreement.
The Clarity Act imposes a number of clear limits on constitutional referendums, so on the one hand raises the bar for future referendums and the Quebec government reacted with an act of its own. For all its effort to establish federal rules over provincial referendums it does recognise the right of provinces to hold referendums on any issue. It is perhaps this that has prompted the Parti Québécois to float the idea of holding separate referendums on individual policy areas it wishes to control, rather than on sovereignty partnership as such (thus effecting self determination by stealth).
Are there any notes of caution here for Scotland? I am often asked about the prospect of a neverendum in Scotland, with voters and businesses subjected to the perpetual uncertainty about the political future. This does not appear likely (or rather, not with respect to independence – an EU referendum is another mater). That a second referendum was held in 1995 had more to do with the promises made by federalist politicians in the 1980 referendum than with a desire to routinely subject the electorate to independence referendums. In the first referendum then Prime Minister Pierre Trudeau promised future changes if voters backed the No side: a vote for the status quo did not therefore mean that no change was possible. The resulting changes – both the process employed to effect them as well as what was on offer – were never seen to live up to those vague promises (and here there is a note of caution for Scotland). It was dissatisfaction over those failed promises – patriation of the Constitution, the Meech Lake Accord, and Charlottetown Accord - that led to the second referendum. Holding one referendum on self determination does not, however, mean that more are necessarily on the way.
Certainly none look likely in the near future. The PQ government lost the 2014 election, and failed to retain almost half its seats, including the one contested by the sitting Premier Pauline Marois. Partly this is tied to declining support for independence, particularly among young Quebecers. When a high profile new candidate mentioned in the middle of the campaign that a win for the PQ could mean another referendum, party support dipped. The poor performance has also been linked to its support of a Jacobin-style Charter of Values which, among other things, sought to prevent public employees from wearing religious symbols, including crosses, hijabs and niqabs; a sign of intolerance to multiculturalism from what had, in its formal rhetoric anyway, typically been an inclusive and civic party.
If there is a note of caution, though, it is about the way the preferences of voters map onto referendum options. Polling after the 1995 campaign showed that while Quebecers who voted Yes viewed independence as their preferred constitutional outcome, a) their levels of knowledge about what it actually meant underestimated the degree of separation involved and b) Yes voters were equally happy with additional powers for a Quebec that remained within Canada. Critically, they didn’t really care if other provinces got them as well. Indeed the most popular constitutional option among both Yes and No supporters was greater powers for Quebec in Canada. As close as the Yes side came to independence, its supporters would have been just as happy with more powers to Quebec, at least as far as the 1995 Quebec Referendum Study data are concerned. Referendums offering binary choices are not always the best way to match voters with their most preferred constitutional options.
As a democratic exercise, however, the 1995 Quebec referendum was a clear success and the features that made it a success are present in Scotland. Then, as is happening now, voters were interested, happy with the process, engaged with the issues and respectful of the views of others.