Mandatory voting is the next step
Over at Statistics Canada, they’re calling it the “best census since 1666” ( the year of the first census of New France, as I’m sure you knew). I’m talking, of course, of the 2016 census, which boasted a 97.8- per- cent response rate, up from 68 per cent five years ago.
What accounts for this remarkable improvement? Need I even say? In 2011, under the Conservatives, the mandatory long- form census, filled out by one-fifth of Canadian households, was replaced with the voluntary National Household Survey. ( The short-form census, which every household receives, remained mandatory.) Under the Liberals, the long- form census was again made mandatory.
And why was that cause for celebration? Because a voluntary census can never be as representative a sample as a mandatory one: the sorts of people who would agree to fill out a 40- page questionnaire are l i kely to possess certain characteristics in greater proportion than the population at large. Rather than a random sample, you get a self-selected one. Skewed is another word.
As it happens, 68 per cent is roughly the turnout in the past federal election, while 98 per cent is not far off the turnout in Australian elections. You may already have guessed the reason: in Canada voting is voluntary, while in Australia it’s mandatory. That mandatory voting is associated with higher turnouts is not much in dispute. In 1922, the last Australian election before voting was made mandatory, turnout was just 58 per cent. The election following it shot up to more than 90 per cent, where it has remained ever since. Across the 20- odd other countries in which voting is mandatory, it is reckoned to have boosted turnout by 10 to 15 percentage points.
Mandatory voting is the other big issue the parliamentary electoral reform committee has been asked to examine. It would do well, then, to keep the success of the long- form census in mind. Because the case for mandatory voting is essentially the case for the mandatory census. The 68 per cent who showed up to vote in the past election — it has been as low as 59 per cent — are not a representative sample of the population. Who votes, and who does not, are very different.
How do they differ? Demographic factors certainly enter into it: voters tend to be older, richer and whiter than non- voters. But the biggest single factor is the efforts of the various par- ties to motivate sympathetic voters to get to the polls. Elections used to be decided by the “swing voters,” without strong ties to any party. Increasingly in recent elections turnout has been the decisive factor: Elections often turn on which party can better “get out the vote” amongst its base of committed supporters.
Turnout dominates party thinking about elections. It accounts for the messages parties choose, designed to energize their supporters and, as important, depress their rivals’. And it accounts for much of the cost and scale of the modern campaign: the elaborate machinery of voter identification and robocalling and the rest, all in the name of GOTV.
By contrast, when everybody votes, in every election, turnout- based strategies cease to be relevant. Rather than simply riling up the base, parties are more driven to expand it, by converting other voters to their cause. As important, the kinds of voters who are now typically underweighted in the “sample” — young people, aboriginal Canadians and so on — assume their due weight.
Mandatory voting is similar in this regard to proportional representation. Rather than represent only those who voted for the winning candidate in each riding, PR systems seek to represent all the voters. Likewise, rather than just represent the voters the parties see fit to recruit, mandatory voting seeks to represent the whole of the voting-age population. If government is about seeking the greatest good for the greatest number, this is surely elementary.
What’s the downside? Ah yes: the terrible toll on human freedom, that sacred right for which our ances- tors fought — the right not to vote. But the imposition on liberty is trivial: trot round to your local school and mark an X on a piece of paper once every four years or so. Even the short- form census takes longer.
You’re not, to be clear, obliged to vote for a particular party, or for any of them. You’re allowed to spoil your ballot, or decline it, or mark “none of the above.” The fines are minor — in Australia, it’s about $ 20. Plus there are common- sense exceptions for those who are sick, or travelling, or have some other legitimate excuse. But it has succeeded in creating a climate in which voting is the norm — not just a civic obligation but a universal custom.
As for that other great objection, that mandatory voting lets the ignorant and the apathetic in on the game, I wouldn’t be so sure. Voting is objectively an irrational act: the impact of any single vote on the outcome is not worth even the relatively minor cost in time and effort. But if everyone calculated the odds as rationally as the non- voters, elections would collapse. By compelling the free- riders to vote, we’re arguably improving the quality of the electorate.
That, of course, is not really the point. Democracy isn’t just for the better educated or the well-informed. It’s for everybody. Mandatory voting isn’t the only reform we need, nor will it cure all the ills of our democracy. But I can’t think of a single thing that would do more to improve it, at less cost.