Should a referendum on a fundamental policy or institutional change require more than 50%+1 of the votes cast? Should it require some minimal voter turnout in addition to a majority? These are important questions regarding direct democracy, and different jurisdictions have different rules on this question of referendum approval thresholds (as I alluded to in my discussion of the odd Iraqi rule.)
The question just came up (again) this week in Canada in connection with an upcoming referendum. On November 28, the Atlantic Canadian province of Prince Edward Island will vote on a proposal of an independent commission on electoral reform to adopt a form of mixed-member proportional (MMP) representation for that province. The vote will be the second Canadian referendum on adopting proportional representation, following that of British Columbia in May. (Other provinces will be following suit soon.)
According to the Guardian of Charlottetown, the PEI premier, Pat Binns, may be having cold feet. Despite the fact that this same premier appointed the Commission on P.E.I.’s Electoral Future after a campaign promise to that effect, and despite the fact that the Commission recommended that a majority of votes cast be sufficient to adopt MMP, the premier on October 20 said:
Government is not comfortable with 50 per cent plus one in a low turnout situation. If we were to have a very low turnout for some reason, and only 50 per cent plus one supported change, I would hardly think that that would be enough cause to change the system.
I think Binns has a fair point here (though I would note that one of the purposes of direct democracy is presumably not to make governments comfortable). Major changes to law, especially laws under which future lawmakers will be elected, arguably ought to receive widespread endorsement. (The possibility that major changes in California, such as the law on political spending of union dues or our electoral re-districting process, might be passed in a low-turnout special election disturbs me–more about that in some future posts.) That said, it is troubling that just over a month before the referendum, the premier is changing rules that supposedly had already been set.
What is worse is that he also is taking other actions that appear calculated to keep the turnout low, as noted in the Guardian article:
One of the factors that may affect voter turnout is the number of polls that will be set up across the province on plebiscite day. The Plebiscite Act says the vote should be conducted â€œas nearly as may be possibleâ€ to a provincial general election. But Binns said he has instructed Elections P.E.I. to cut the number of polls across the Island to cut costs. Binns said he expects the number of polls where Islanders can vote in the plebiscite to be cut to two-to-three per district. In a provincial election, there could be as many as 14 polls. [my emphasis]
In a previous referendum (in 1988, on constructing a bridge to the island from New Brunswick), the same number of polling places were provided as for a regular general election.
It seems like the premier is “shifting in the goal posts”, as Fairvote Canada put it (in a press release I received via e-mail).
In the BC referendum in May on adopting single transferable vote for provincial legislative elections, the requirement was 60% and majorities in 60% of the ridings (districts) and everyone knew that well before the vote. (The proposal itself was crafted by an independent Citizens Commission, chosen more or less at random, sort of like a grand jury.)
While the approval threshold for the BC electoral-reform referendum established the principle that such a fundamental change should only be made by consensus, it led to an embarrassing result: The BC-STV proposal won 57.7%, meaning the proposal was narrowly defeated. Yet the governing Liberal party that was re-elected on the same day received a manufactured parliamentary majority on only 45.8% of the vote. (And turnout was high, in part because the referendum was combined with a general election.)
In other words, the BC Liberals received less voter support than did an alternative electoral system that would have deprived the party of a governing majority had it already been in effect. Not surprisingly, the BC government announced in the Throne Speech in September that it will allow a second vote, in November, 2008, on electoral reform. In other words, they recognize that nearly 58% “yes” does not mean “no.”
That is precisely the problem with imposing thresholds on referenda, which supposedly are devices for discerning the “popular will”: When you get more than 50%, but less than the threshold (either in yes votes or in voters showing up), the legitimacy of the thus-expressed “popular will” is ambiguous.