Why People Vote: C (Voting #2) Transcript

Transcript (PDF)

Economists are kept up all night by the puzzle of voting. If voting is irrational, then why do millions regularly show up to vote?

In this and the next few videos, we’ll try to resolve the puzzle. In other words, explain why people vote. We’ll do so by poking holes in the economist’s argument. An argument which consists of three ingredients.

First, the cost of voting, C, may be small, but it ain’t zero.

Second, your vote matters — swings the election result — only in two wildly-improbable scenarios. And so the probability, p, that your vote matters is infinitesimal.

Third, the value, V, you place on your vote mattering — on your vote swinging the election result — may be large, but it ain’t infinite.

The economist argues that for just about any plausible values of C, p, and V, C exceeds pV. The cost of voting exceeds its expected benefit. And so voting is irrational. [“Very nice!”]

In this video, we’ll focus only on C. In the last video, we argued that voting costs at least 10 minutes. And so, even if your time is worth only the US minimum wage, C exceeds $1. In this video, we’ll present three arguments that C is, au contraire, much smaller than $1.

Argument 1. In many countries, election day is a non-working day.[1] And if you’re not working, 10 minutes of your time has no value, right?

Wrong. Even if you’re not working, your time isn’t worthless. There’s an opportunity cost. Those 10 minutes could’ve been put to some alternative use. You could’ve had 10 more minutes for brunch, YouTube, or sleep.

So this argument is wrong. Your time is always valuable, even if you’re not working. That’s because your time can always be put to some alternative use.

Next. Argument 2. In countries where election day is a workday,[2] employers often give workers time off to vote. Some jurisdictions even make this the law. For example, Missouri lets you take three hours off work to vote! So you could spend 10 minutes voting and watch a movie with the remaining time.

And even if there isn’t extra time for a movie, the cost of voting is zero or even negative. That’s because the time spent voting would otherwise have been spent working. And for most people, work sucks.

Now. This argument is correct but has limited applicability. It applies only if election day is a workday and workers are given time off. Moreover, it doesn’t apply to weirdos who love their jobs or high-powered executives. Because such workers actually value their working time.

Next. Argument 3. Economists assume that people are like Homo economicus. A hyper-rational creature who carefully weighs costs against benefits and always makes the best possible decisions.

But in reality, people are more like Homer economicus. They do not carefully weigh costs and benefits, especially when said costs and benefits are tiny.

Here’s an analogy. Imagine this is your restaurant bill. The total is $98.58. If a 15% tip is customary, most people would simply tip $15. Not so the economist.

“First of all, tipping is irrational. But if you must tip, be sure to tip only on the pre-tax amount. 15% of $93 is $13.95. So the correct tip is $13.95. Not $15.”

At this point, most people will simply have left a $15 tip and punched the economist in the face.

The argument here is this: When the costs and benefits involved are tiny, most people don’t bother weighing them.[3]

To the economist, this is irrational. But if brainpower is itself a scarce resource, then maybe this isn’t so irrational after all. If voting costs only 10 minutes, then you know what? “Just do it!”

Unless you’re Homo economicus, it would take less time and brainpower to simply go out and vote, than to listen to the entire “voting is irrational” argument, evaluate its merits, and then carefully weigh up the costs and benefits of voting.

Let’s end this video by giving the economist the last word. We’ve been saying that including traveling time, voting takes only 10 minutes.

But realistically, it’s gonna take more than 10 minutes. Indeed, the wait time alone is probably more than 10 minutes.

For example, in 2008, the average US voter waited 16.7 minutes. In 2012, she waited 13.3. Note that these were the averages. Across regions and *ahem* demographics, there were dramatic variations. Floridians waited 26 times as long as Vermonters, while blacks waited twice as long as whites.

In Lee County, Florida, the average voter waited more than two hours![4] [“What!?”]

Note that unlike Missouri, Florida law doesn’t give workers any time off to vote.[5] Yet in Lee County, more than a quarter million showed up to vote! It’s probably safe to assume that at least some of these voters had to burn their own free time. And very few of them would dismiss a two-hour wait as a tiny cost.

To explain the behavior of Lee County voters, we’ll probably have to look somewhere else. And that’s what we’ll be doing in the next video.


[1] In South Korea, election days are usually declared “temporary national holidays”. For example, the most recent presidential election on Wednesday, May 9th, 2017 (source).

In New Zealand, the Act of 1950 requires that elections be held on Saturdays (source).

In France, Article L55 of Code électoral requires that elections be held on Sundays (source).

[2] In Canada, “holding election day on a Monday … has been the tradition for many years” (source). Workers “must have three consecutive hours to cast their vote on election day” (source).

In the US, Statute II of the 28th Congress (1845) requires that “the electors of President and Vice President shall be appointed in each State on the Tuesday next after the first Monday in the month of November” (source). the biannual elections are held on Tuesdays. Time-off-to-vote laws vary by state (source).

In the UK, “Thursday has become the traditional day for general elections”, although this need not necessarily be the case (source). “The last general election not to be held on a Thursday was on Tuesday 27 October 1931.”According to The Guardian, “no legislation exists to allow employees to take time off work to vote”.

[3] The political scientist André Blais is a strong advocate of this argument. See To Vote or Not to Vote? (2000, pp. 89-91 and 139-140) and “Rational choice and the calculus of voting” in Handbook of Social Choice and Voting (2016, pp. 62-63).

[4] The wait time statistics cited here are all from Stewart (2015), “Managing Polling Place Resources” (PDF). See pp. 3, 5, 7, 37.

[5] According to this, this, this, and this webpage.