The general election is officially underway!

No, not that one.  I'm referring to Manitoba, where Premier Brian Pallister recently kicked off an election in the Keystone Province, escaping the notice of most Canadians.

But wait wasn't Manitoba scheduled to hold its next election on October 2020?  Isn't that province's premier constrained by fixed-election-date legislation, meant to prevent opportunistic early elections?

In theory, yes.  In reality, however, there is little to prevent Canada's first ministers from calling early elections despite fixed-election-date legislation.  The problem is that such legislation, implemented across most of the country, lacks teeth.  There are no penalties when such laws are violated or, as some might argue, the laws are written in such a circuitous way as to become a meaningless ornament.  In essence, such legislation has been reduced to aspirational scheduling, still subject to the expedient whims of politicians who seek to remain in power.

Although first ministers require the Crown's approval to trigger an election, vice-regal representatives don't typically thwart such requests.  Again, in theory, the Crown's representative could refuse a first minister's advice to dissolve parliament, but such a scenario would likely only transpire during a brief window: roughly six months following an election.  (And note that such a refusal would likely be based upon constitutional precedent, not the presence of fixed-election-date legislation.)  After that initial period, Crown representatives tend to obediently adhere to a first minister's advice, even if that advice is to draw up the writs for an early election despite the existence of fixed-election-date legislation.

While it is tempting to mourn the hollowness of laws ostensibly meant to prevent snap elections, the current collection of such legislation across Canada is admittedly a considerable improvement to what existed prior.  Most general elections in Canada are now held on dates scheduled years in advance, which removes the unfair advantage of governments quietly planning surprise elections that opposition parties (as well as the news media and general public) can be woefully underprepared for.

Yet despite the noticeable decrease in early elections, they occasionally still occur.  Perhaps the most infamous example in recent history was the snap federal election of 2008, initiated by then-prime minister Stephen Harper despite his government having implemented fixed-election-date legislation the previous year.  The subsequent election in 2011 was also held early after a non-confidence vote.

Unfortunately, Canada's election system can encourage first ministers to ignore scheduled election dates.  In particular, the instability of minority governments is exacerbated by the first-past-the-post voting system, which entices governments and opposition parties alike to trigger snap elections in the hope that a small increase in support can be distorted into a disproportionately large gain in seats.  The 2014 Ontario election is the best such illustration, in which a single-percentage-point gain in popularity for the provincial Liberals resulted in a 20 percent increase in their seats, ballooning the caucus into a majority government.

Knowing that Canada's antiquated voting system turns early elections into high-risk, high-reward games of roulette, it's little wonder governments are constantly tempted to take a spin of the wheel, despite legislation purportedly meant to prevent them from doing so.

Although it would probably be wise to continue to allow legislatures to be dissolved early in the event of insurmountable partisan gridlock, majority governments unilaterally triggering snap elections merely out of opportunism is perhaps the most unpalatable of unnecessary votes and most in need of prevention.  It's worth noting that there have only been two early elections ushered by majority governments where fixed-election-date legislation exists in Canada.  But ominously, both have occurred in the past four years: Alberta in 2015 and the current election in Manitoba.

Is this the dawn of a new trend?  Should we be concerned?  Perhaps.  In the Alberta example, the governing party that called a snap election was punished by the electorate and succumbed to heavy losses.  But if Manitoba's Brian Pallister next month becomes the first premier to dissolve a majority government in violation of fixed-election-date legislation and still earn re-election, other Canadian first ministers will undoubtedly take notice.

Pallister had hinted since December that an early vote might be coming, and unofficially announced the election almost eight weeks before the writs were drawn up.  Compared to typical behaviour of yesteryear, in which early elections weren't announced until they had already officially begun, Pallister's softer approach may ameliorate much of the sting.  But how the Manitoban electorate reacts to Pallister's gambit could prove pivotal for the future of fixed-election-date legislation across Canada.  If flouting the law is met with a collective shrug from the public, other premiers may wonder if their province's voters would respond equally indifferently.  But if Manitobans recoil and hand Pallister a palpable loss, he could go down in history as the final first minister boasting a majority government to attempt to contravene a predetermined election schedule.

Knowing that current legislation across Canada is flawed and doesn't effectively prevent early elections, what is the solution?  One fix would be to strengthen existing laws.  Although locking up first ministers could prove amusing, a more realistic amendment would be to emulate legislation already implemented in the United Kingdom: to require that a first minister must have the support of at least two-thirds of elected members before advising that a legislature be dissolved.

While imperfect, the new UK law requires a first minister to have unequivocal parliamentary support before attempting to trigger an early election.  This not only necessitates securing the agreement of at least one other political party (in most scenarios) before dissolving a legislature, but it also reinvigorates the concept of "responsible government" by strengthening parliamentary supremacy surely a welcomed transformation in Canada, where party leaders and their unelected strategists hoard undue influence.

Needing two-thirds approval to call an election would also radically improve the stability and duration of minority governments in Canada, which typically only survive a couple of years.

Such a change to fixed-election-date legislation still wouldn't conclusively end early elections not even from majority governments.  In fact, the UK's Conservative majority government called an early election in 2017 after garnering nearly-unanimous parliamentary support.  But if early elections must be held, surely a super-majority parliamentary threshold would be preferable to leaving the decision exclusively to a fireside chat between just two people first minister and Crown representative as is what currently transpires.  Such an amendment would also end the particularly problematic practice of minority-government first ministers unilaterally dissolving legislatures, rather than allowing other parties an opportunity to make a parliament work before dragging voters back to the ballot box.

The only potential downside of a British-style strengthening of Canada's fixed-election-date legislation would be if partisan gridlock overwhelmed a parliament and two-thirds support for dissolution could not be mustered.  But what might the long-term consequence of that be: pressure for elected members to work more cooperatively in the future?  That's a risk Canadian politics should be willing to take.