Compulsory voting

I have a few things to say about compulsory voting, since scrapping it has been put on the table by Queensland state government. (Oh Queensland, what would we do without you?)

I happen to be a fan of compulsory voting, not because it’s the status quo, nor simply because it’s “our duty” (although that is a convenient way of looking at it). First, I’ll tear down some of the arguments against compulsory voting.

“Voting is a right, not a duty”. Voting-as-a-right — an individual right — makes no sense, because individuals do not benefit from their own individual votes. The benefit only occurs collectively. We benefit — in the sense of having a functioning democracy — from each other’s vote, not our own. This, at the very least, makes voting a service that we perform for the benefit of others.

“It’s not democratic unless I have a right to not vote”. There are two parts to this. First, you do in practice have the right not to vote — simply insert your blank or defaced ballot paper in the box. Opponents to compulsory voting will then re-phrase their argument as follows: “It’s not democratic unless I have the right to not attend a polling station”. But this is a non-sequitur, and confuses democracy with libertarianism. Democracy does not imply a “right not to attend a polling station” any more than it implies a “right to not do jury duty” or a “right to not pay tax” or a “right to not go to school” or a “right to not stop at red lights”. We all have legal obligations of various kinds, many of which are much more onerous than the requirement to vote, or at least attend a polling station. And what do you gain by not voting (bearing in mind that emergencies and other reasonable exceptions are already allowed)?

“Idiots are forced to vote”. No question. Under compulsory voting, every apathetic idiot trudges down to the polling booth and unleashes their special brand of wisdom on the rest of us. But we’re falling for a highly simplistic psychological theory here. Can you demonstrate empirically that stupid people would actually be less motivated to vote? What if that’s not true? There are a lot of very silly minor political parties representing the tip of the iceberg of politically-engaged insanity. There is a party — and associated supporters — who think the British royal family is behind the world’s “fascist” environmentalist movement. They vote (and not just because it’s compulsory). And conversely, although apathy may be stupid in some sense, it does not necessarily imply ignorance or an inability to conduct sound reasoning. Indeed, the requirement to vote is probably enough in many cases to overcome apathy and engage people’s brains. Thus, voluntary voting — self-selection — does not necessarily make for sensible decision making. It could just as easily be the other way around.

“Apathy leads to informal and donkey votes”. This is certainly true, though informal (invalid) votes are not particularly a problem. The tally of such votes — particularly those with messages written on — is a useful measure of disaffection with the current political situation.

Donkey votes are somewhat trickier. If a voter numbers candidates in the order in which they appear on the ballot paper, this may or may not be a donkey vote. It’s a donkey vote if the voter doesn’t care, but of course there’s no way to know that after the fact, and so donkey votes are counted like any other formal (valid) votes. For the same reason, it’s difficult to estimate the overall number of donkey votes, though a recent paper (King and Leigh, 2009) puts the advantage gained by being the first candidate at 1%. This is actually a real problem, but one that is fixable without abandoning compulsory voting. If there were many different versions of the ballot paper, each listing the candidates in a different order, donkey votes would largely cancel each other out. (This is already done in the ACT and Tasmania, for instance.)

Having failed to be convinced by the arguments against compulsory voting, here are some arguments for it.

Compulsory voting more accurately determines the sentiment of the people. Obviously, voter turnout will be higher, which lends the result increased credibility. But it’s not just the magnitude of the turnout that matters; it’s the relative certainty in the breakdown of that turnout. With compulsory voting, the breakdown of voter turnout (by demographic) should neatly reflect the breakdown of the overall population. If 34% of the whole population (excluding children, non-citizens, etc.) belongs to a particular group, then 34% of the voters will too. Random chance plays very little role.

Under voluntary voting, external (non-political) factors can influence the turnout of different demographics. In the United States (and presumably elsewhere), the weather is known to influence the ratio of Democrats to Republicans who show up to vote. This paper (Gomez et al., 2007), for instance, makes the following finding:

For every one-inch increase in rain above its election day normal, the Republican presidential candidate received approximately an extra 2.5% of the vote. For every one-inch increase in snow above normal, the Republican candidate’s vote share increases by approximately .6%.

You may notice that the estimated 2.5% advantage per extra inch of rain, for voluntary voting, is substantially larger than the estimated (and fixable) 1% advantage for being listed first on the ballot paper under compulsory voting.

Alternatively, suppose that there’s an outbreak of influenza or another illness; there are specific demographics more likely to be affected, and so more likely to stay home. All of this is just random chance, and while there are worse forms of government than flipping a coin, that’s not democracy and not what we’re aiming for.

Compulsory voting is logistically simpler and more robust. This is related to the previous point. Since turnout can be predicted with a small margin of error, electoral officials will have a good idea in advance of how many workers, polling booths and ballot papers are needed. Accordingly, we are less likely to suffer long queues, ballot paper shortages, vote counting delays, etc. Thus, there is less scope for frustration, confusion and controversy.

Further, since virtually every adult votes, there ought to be greater overall awareness and understanding of the voting process. It is more likely, for instance, that a first-time voter will be able to consult their friends and family if in any doubt as to how things work.

Compulsory voting encourages political discourse. Compulsory voting necessitates at least some level of political awareness on the part of those who would otherwise not take an interest. Political debate benefits greatly from the involvement of people who are not locked into a particular way of thinking. They themselves benefit from greater understanding of issues that may affect them. This segues into the next point.

Compulsory voting mitigates against political polarisation. With voluntary voting, a political party must expend a great deal of effort motivating its “base” to turn up, not just winning over undecided voters. This occurs by means of “no-compromise” policies and rhetoric that appeal to ideologically-committed hard-liners. This has an inherently polarising effect on politics. Each party shuts itself off from the other, except to hurl abuse calculated to make its own supporters cheer more loudly.

With compulsory voting, the only way to increase your share of the vote is to win over those who previously voted for someone else. Elections can only really be won or lost in the centre (and preferential voting helps here too, by ensuring that “vote splitting” cannot happen). The centre isn’t always sensible, because it’s merely arbitrarily half-way between ideological extremes, but it keeps the ideologues at bay and the major political parties in touch with each other. It gives us a better chance at good government.

Compulsory voting is more inclusive. There are a number of people for whom presenting identification is problematic, and who would therefore be dissuaded or prevented from voting if ID checks were needed. Under compulsory voting, they are not. You must merely state your name and address, and be on the electoral roll. At first glance, this seems wildly open to abuse. Strictly speaking, there is nothing stopping someone else from falsely claiming to be you at the polling station, and even voting multiple times under your name. But if you think about it carefully, nothing can be achieved by this. Electoral officials can — and do — deduce the overall number of duplicate votes and compare this to the winning margin. In the extreme case that the former exceeds the latter, where there is the theoretical possibility of voter fraud changing the outcome, the result is annulled and thus another election held — expensive, but unlikely to be required because nobody would benefit. Thus, voter fraud can be neutralised without needing to verify each voter’s ID.

This isn’t so easy in a voluntary system, where potential vote fraudsters may be able to predict who is less likely to vote, based on demographic information gleaned from the electoral roll. In the absence of ID verification, you could vote under someone else’s name — someone who you suspect will not vote. For electoral officials, there’s no way to count the number of false votes cast like this where the actual registered voter hasn’t shown up.1

Reforming electoral laws is something that must be taken especially seriously, because they underpin our democracy. Electoral laws have a great deal to say, indirectly, about who gets to run things, and so on the face of it all political parties have a conflict of interest. We must tread carefully, and keep those conflicts of interest in mind when we hear our politicians speak on the subject.

  1. However, even in a voluntary system, the risk of voter fraud tends to be overblown. In an established democracy, it would be a very complex, risky and labour-intensive exercise, and is probably impractical in all but the narrowest elections. []