Voting
Good God, y’all – what is it good for?
Absolutely Nothing?
Perhaps not, with apologies to Edwin Starr…and whoever wrote that song.* But I do have my doubts. Voting behavior and what influences it was a big part of what I taught and researched during my career. I have been writing down notes about it since before I started this blog, with an eye toward trying to write a loooong piece about modern democracy. Probably too long, so here I will record just a few thoughts, prompted by our looming Canadian federal election. There will be more to come.
It is a much noted and much lamented fact that turnout rates in elections in advanced democracies are ‘too low’. Whether and in what sense that is a bad thing deserves to be discussed, but I will set that aside until later. To start with some facts, below is data from a few countries of which you have heard. I note hastily that such data is hard to compare across countries, for reasons I will indicate.
Canada: 62.6% of those registered to vote in the 2021 Federal election did so. Although it is easy to register in Canada by international standards, not everyone does. I note also that in a labour force survey done just after said election, 75.6% of those surveyed reported voting. More on that discrepancy, which is very common, later.
USA: 62.8% of the Voting Age Population (VAP) is estimated to have voted in the US 2020 election – ‘estimated’ because the VAP is an estimate, although a pretty accurate one, based on the US census. This was a recent high for the US, VAP turnout has been below 60% since 1984 and was below 50% in 1996. Note that it is rather more difficult to register to vote in the US than in Canada.
UK: In the 2024 election, 59.9% of registered voters voted. It is estimated that 8million who are of voting age were not registered to vote, while some 48 million were registered, and 29 million voted. If the 8million number is accurate (and I have no idea if it is), this makes VAP turnout approximately 29/56 = 51%.
Germany: It is reported that in the 2025 federal election turnout was 82.5%, but % of what I have not been able to determine. So far as I can tell it does seem to be easy to register to vote there.
Finally,
France: In 2024, 63% of French voters turned out to vote for the second round of the country’s snap parliamentary elections on Sunday, July 7, but again it is not clear what that is a percentage of. And, that figure is from Le Monde, Wikipedia says 66%.
Clearly there is a fair bit of variation here, although some of it is due to differences in how turnout stats are defined. Still, the ‘best’ of these recent numbers is from Germany, and even there up to 20% of those who could vote did not do so.
Why is this the case? I don’t think I have a good answer, but a place to start is to ask – why do people vote at all? Simply put, if we knew for sure why people vote, it would make it easier to determine why many do not.
Those who study this sort of thing talk about two possible motivations for voting. One is referred to as being ‘instrumental’, and it is the desire by citizens to influence who governs us. Voting is how we do that in democracies, and so people vote in order to influence who will hold important government offices.
The other motivation goes by a variety of names, I will refer to it as ‘internal’ or ‘personal’. These are motivations that come from inside us. Thus, if we vote because we feel we are fulfilling a duty, that is a personal, internal motivation. [There are a few countries in which one can be fined for not voting. For the time being, I am going to leave ‘avoiding a sanction’ out of the discussion of motivations. I’ll come back to it later.]
What separates these two motivations fundamentally is whether they are influenced by the way in which we think others will vote. The instrumental motivation is affected by how other people behave, because their votes – or failure to vote – also influences who wins office.
As an example, people often speak of ‘voting strategically’ in an election in which there are more than two candidates. For example, a Canadian might vote for a Liberal candidate in their federal riding, even if they believe the NDP candidate would make a better MP. They do this because they expect the Conservative and Liberal candidate will be the top two finishers, and they prefer the Liberal to the Conservative. They would not do this if they believed the Conservative and NDP would be the top two vote-getters; in that case they would vote NDP. Strategic voting only makes sense if your reason for voting is instrumental, so your vote is aimed at influencing who takes office.
There are at least two ways in which your motivation to vote might be personal. One, already mentioned, is that you vote because you believe it is your duty as a citizen to do so. If your motivation were one of what I will call ‘pure duty’, then it does not matter for whom you vote, so long as you do your duty and vote. You will feel better for having done so. I doubt anyone feels exactly that way, but if you think it is your duty to vote, and also to vote for the ‘best’ candidate, whatever ‘best’ means to you, then it is still an internal motivation. You will do this – that is, go to the polls and vote for the best candidate – no matter what you think other citizens are going to do.
I claim that no one really votes because of a purely instrumental motivation. That is, no one really thinks they have any influence over who governs us, because everyone realizes that their vote has no impact on the outcome of the election. That is, they realize that whether they vote or not, and who they vote for, is irrelevant to the outcome of the election. Or, to put it in probability terms, they realize that their going to the polling booth and voting for candidate A changes the likelihood of candidate A winning the election not at all. It will change the outcome of the election if and only if candidate A and some other candidate would get the same number of votes, and more than any other candidate, if they did not vote. The probability of the candidate you favour and some other candidate tying for the most votes is, at the point at which you go off to vote, equal to zero out to many decimal places.
Another way to express the same idea: on the morning of voting day, you are thinking to yourself – ‘should I vote or should I spend that time doing something else?’ It is the decision we all have to make on that day.
But you are no fool, and you know that if you vote the probability is .00000xxx that the outcome of the election will be different from what it will be if you don’t vote. You don’t know exactly how many zeros come after that decimal point, and you surely don’t know what those xxx values are, because no one does. But you know that if there are some tens of thousands of registered voters in your riding, there are too many zeros after the decimal point in that probability for the exact xxx values to matter.
Hence, my claim is that probability of making a difference in the outcome cannot be what motivates any rational citizen to go and vote.
On the other hand, the probability that you will feel better about yourself if you vote is 1, if you have a (internal) sense of duty. That is a reason to vote, to feel like you certainly did your duty, to certainly not feel guilty that you did not, to feel for sure that you supported the best candidate with your vote.
So, the consequence of my claim that no one really votes for instrumental reasons implies that they must be voting for internal reasons, reasons which have nothing to do with the actual or predicted outcome of the election.
Now, if you work back through my argument above, it may occur to you that I am arguing that only a personal motivation makes sense as a reason to vote, a reason to take the time to go to the polls. Once that decision is made and you are in the polling booth with the ballot in front of you, then considerations of who might win do become relevant. At that point, it may be perfectly logical to vote strategically, as I described above, for example.
There is a cost to going to the polls, because you could spend that time otherwise, so my point is that the instrumental motivation of influencing the election’s outcome cannot be the reason for paying that cost. No matter how small that cost of voting is, the instrumental payoff is smaller – too many zeros. However, once you are in the booth, that cost has been paid, and it is certainly logical then to think about who you want to see in office. Once you are the booth, it makes no sense to put an x next to the name of the candidate you think would make the worst office-holder.
I have at this point barely scratched the surface of voting behaviour, and I plan to write more. At this juncture, my one point is that if we are to understand why people vote (or don’t) we need to understand what influences their internal motivations to do so. Understanding which candidate they vote for when they do vote is a separate matter. Not completely separate, of course, but I maintain that it is useful to separate these two decisions – should I vote vs for whom should I vote – in order to understand why many do not vote.
This leaves many issues unconsidered, to put it mildly. As just one example, I have referred to the cost of getting out and voting, and that clearly includes the cost of registering to vote, which varies by country. However, there is in addition the cost of becoming informed about the candidates and parties. It is not controversial to say that people vary greatly in how much effort they put in on that front, but it is also costly, because it takes time and effort.
As I say, more to come. No doubt the results of the Canadian vote on Monday will provide more about which to write. Will turnout be higher or lower than in 2012 – and why?
Stay tuned.
* I checked, and it was written by Norman Whitfield and Barrett Strong, who also wrote ‘I Heard it Through the Grapevine’, ‘Papa was a Rollin’ Stone’, ‘Just My Imagination’ and other great Motown hits. I don’t know whether they voted.