Featured Post

Why Don’t People Vote?

Hoteling’s Law:

Suppose that there are two competing shops located along the length of a street running north and south, with customers spread equally along the street. Each shop owner wants to locate his shop such that he maximises his own market share by drawing the largest number of customers. In this example, the shop itself is the ‘product’ considered and both products are equal in quality and price. There is no difference in product to the customers. Therefore, each customer will always choose the nearest shop because there is no difference in product or price.
One shop
For a single shop, the optimal location is anywhere along the length of the street. The shop owner is completely indifferent about the location of the shop since it will draw all customers to it, by default. However, from the point of view of a social welfare function that tries to minimize the sum of squares of distances that people need to walk, the optimal point is halfway along the length of the street.
Two shops: halfway
Hotelling’s law predicts that a street with two shops will also find both shops right next to each other at the same halfway point. Each shop will serve half the market; one will draw customers from the north, the other all customers from the south.

This idea has been extended to political races by Anthony Downs. People are thought to make their voting decisions based on which viable candidate comes closest to representing their own preferences.

Parallel to the above logic, both major candidates’ positions will converge on the median voter’s position. Or at least they will campaign that way, fighting to grab that median voter content that everyone to their right or left will behave predictably.

All theories are wrong, but this theory in particular seems importantly wrong. Namely, people don’t vote.

A year ago, Tess in these pages made A Desperate Plea for a Renewal of Citizenship. That and a lot of other much-more-expensive efforts collectively garnered these results:

If staying home were on the ticket, it’d be president.1

The Hoteling model is candidate-focused. The theory focuses on the behavior of candidates, and its most important prediction is about the behavior of those candidates.

But in real life, voters aren’t getting off the couch to buy. These people cannot be treated as an error term. They exist, and their behavior is important. If you have a mathematical model, that predicts who people vote for, it also needs to predict who doesn’t vote. If you have a model that predicts what percentage of the population votes for which candidate, it also needs to predict what percentage does not vote.

Not voting is often chalked up to laziness, indifference, kooky conscientious objections, and other undesirable attributes a person may have.

I submit one other. Namely, large swaths of the voting-eligible population, do not feel they can be adequately represented by any candidate running. They live to far away from any store, so they just don’t buy. Even if they decidedly prefer one candidate to another, they may not be able to vote in good conscience for the better candidate because the distance between their own views and that of the candidate’s is too vast.

Though his objections seem quaint now, this is roughly the logic articulated by Conor Friedersdorf when he refused to vote for Barack Obama in 2012:

I am not a purist. There is no such thing as a perfect political party, or a president who governs in accordance with one’s every ethical judgment. But some actions are so ruinous to human rights, so destructive of the Constitution, and so contrary to basic morals that they are disqualifying. Most of you will go that far with me. If two candidates favored a return to slavery, or wanted to stone adulterers, you wouldn’t cast your ballot for the one with the better position on health care. I am not equating President Obama with a slavery apologist or an Islamic fundamentalist. On one issue, torture, he issued an executive order against an immoral policy undertaken by his predecessor, and while torture opponents hoped for more, that is no small thing.

What I am saying is that Obama has done things that, while not comparable to a historic evil like chattel slavery, go far beyond my moral comfort zone. Everyone must define their own deal-breakers. Doing so is no easy task in this broken world. But this year isn’t a close call for me.

He wrote this despite preferring Obama’s policies to Mitt Romney’s.

Robert Wright would not let this stand and refused to refuse to vote for Obama:

this isn’t a movie, so I have a hard time ignoring the consequences of (implicitly) encouraging would-be Obama supporters to nullify their votes and thereby increase the chances that Mitt Romney will be our next president.

Wright would have Friedersdorf vote strategically. But for something as wrapped up in questions of morality as politics, strategic voting comes at a heavy price. However quaint they seem now, Friedersdorf’s qualms about Obama’s stances on civil liberities were real. Even knowing Romney would be worse, voting for Obama would feel like it’d cost a part of his soul.

Willie Davis in Salon:

In art criticism, the aesthetic quality of the work matters less than what our opinion of the art says about us. In politics, the policy doesn’t matter; it’s what our vote says about us.

slackers photo

Image by PunkToad Why Don't People Vote?

Voters don’t just choose which option is better. They choose an identity. Most people now seem to believe the Democrats ran a bad presidential campaign in 2016, but one thing they got right was the tag line “I’m with her.” Voting is an affirmative act. I read a surprising number of friends express “I’m with her,” and I think this did something to encourage the reluctant to vote for Clinton. But there were still more who weren’t with anyone.

  1. I was among those who stayed home. []

Senior Editor
Home Page Twitter 

Vikram Bath is the pseudonym of a former business school professor living in the United States with his wife, daughter, and dog. (Dog pictured.) His current interests include amateur philosophy of science, business, and economics. Tweet at him at @vikrambath1. ...more →

Please do be so kind as to share this post.

37 thoughts on “Why Don’t People Vote?

  1. I’m with her is a perfect example of identity politics. Hell, the substance doesn’t matter, nor the policies.

    Anyway, you said “do not feel they can be adequately represented by any candidate running”. How about this: “do not think they are represented by any candidate running”.

    BTW, I was all over that post from Tess. I continue to stand by everything I said there. Guess that puts me in the “kooky conscientious objections” category from many’s perspective.


  2. Campaigning without focusing on ISSUES lost Clinton the election. (Romney got ratfucked into losing while still believing exactly what he wanted to hear)
    Please don’t mistake my vote for my identity, I assure you it has little to do with who I am.

    You want a model? Well, we’ve got a model, and it’s house to house, person to person. Knows who you’ll vote for, or if you won’t vote at all.


  3. Do multi-party systems see higher turnout that two-party systems? If you and Conor are part of a statistically significant group, one would expect that to be the case, right?


  4. I agree with your instinct that “laziness, indifference, kooky conscientious objections, and other undesirable attributes a person may have,” is an unhelpful and probably incorrect analysis. A better model would identify rational, intelligent reasons that 29.9% of the electorate behaves the way it does.

    I would submit that voting fails a simple cost/benefit analysis for many people: in the astronomically unlikely event that one person’s vote decides an election, the candidate they elect will change very little, and what they do change, they will change in unpredictable ways whose consequences will not be apparent for many years to come.

    I will set present events aside: although I am a Democrat, in 2012 I had faith that Mitt Romney would not start any major wars, would not end any current ones, and would in general preserve America’s alliances and standing in the world in a roughly similar way to what Barack Obama would produce. On the domestic side, he might have nudged government spending one or two percentage points in a way I didn’t prefer; on the personal financial level I submit that any consequences of his presidency for me would be entirely impossible to predict.

    Setting all of the pieties aside, why would I waste a couple hours – or even thirty minutes – making what is essentially an arbitrary decision that will not affect anything and will either remain secret or make me the object of the scathing hatred of 50% of my family? I work hard and I value my leisure.

    A different model is needed.


    • I think this captures a lot of it. Combine the lack of clear personal benefit/ineffectiveness of the act for many people with apathy, lack of information, and inconvenience and it’s almost a wonder that millions of people still do it. I also agree with your implication that this is not a good thing.


    • Yeah this captures most of the non-voting sentiment. For a lot of the non-voters they don’t vote because the stakes are just too low.
      What gooses voter turn out? Misery. Rage. Unhappiness. Outrage.


    • I would submit that voting fails a simple cost/benefit analysis for many people: in the astronomically unlikely event that one person’s vote decides an election, the candidate they elect will change very little, and what they do change, they will change in unpredictable ways whose consequences will not be apparent for many years to come.

      This is why I sorta feel that Donald Trump is a disaster for Republicans in ways they do not fully comprehend yet. A lot talk has gone talking about how *voters* decide how to vote. But the reality…it’s not just voters that vote. (By which I mean, it’s not just *previous* voters that vote.)

      Voter turnout is based on
      1) the perceived *difference* between the two sides
      2) how much the voter calculates their vote will change the outcome

      Multiple those two together, and you will get a number. Voters will only vote if that number crosses a personal threshold for them. Also, Republicans either have a lower threshold, or, what seems more likely to me, that they have been carefully trained to think #1 is larger than it is. (Because, historically, the people that Republican voters listen to claim that Democrats are horrible left-wing communists, whereas the people that Democratic voters listen to claim that Democrats are horrible centerist quislings.) Or both.

      People should go look at this: https://en.wikipedia.org/wiki/Voter_turnout_in_the_United_States_presidential_elections

      Scroll about halfway down, it has the voter turnout for all elections. Note the big changes in it, changes of about 5 points at once. Starting with, and ignoring, the big jump from 1948 to 1952 (I am not sure what caused that.), the next big change was the drop in 72, which was basically caused by Democrats all staying home because it was clear their candidate couldn’t win. Then it slowly slips down to 50% in 88. And, then, the Republicans had had four years of the unimpressive Bush I, and the Democrats had a good candidate, so another 5% of Americans who were Democrats showed back up at the polls and voted in Clinton. So the Republicans spend 4 years attacking him, and got people who actually *cared* about the results of elections back down to 50% in 96. And then we had an election decided by a few hundred votes and the Supreme Court , which got us Bush…and got a lot of people realizing, hey, wait a minute, maybe voting is important? Which bumped the numbers up another 5%, where they have stayed. (Everyone talks about how 2008 was a record turnout…and it technically was, we hadn’t seen that level of voter turnout since 1968…but it was only 1.4 points higher than 2004.)

      In short, as far as I can tell, large changes in voter turnout are entirely based on Democrats *either* thinking their candidate can’t win (1972) and not showing up , or thinking there is a notable difference between the candidates (1992) or thinking their single vote might make a difference (2000 onward.) so showing up. That is it. That is the entirety of large changes in voter turnout.

      The two *actually important* changes, the ones that alter elections, are the last two. (Democrats not showing up in 1972 didn’t hurt anything because, uh, they were right that their vote wouldn’t matter. I know we’ve all forgotten this due to the reelection scandal stuff, but Nixon got the largest popular vote margin ever in 72.) Those last two changes cause large jumps in voting population, and those jumps, as far as I can tell, consisted of mostly Democrats.

      People following that? Well, two things, including one that everyone seem to have not noticed.

      First of all, we had TRUMP modifying people’s calculation for #1. As are the Republican hyper-right-wing Tea Party idiots having a chance to put their stuff into action for the first time. People have noticed this.

      But we just had sorta had another 2000, where a presidency was decided by not that many votes. Yes, it was more than 2000, but it really wasn’t that much, and *that* sort of thought was helped by a) the polls all being wrong, and b) the Democrats winning the popular vote. (Remember, this isn’t whether their vote would have *actually* mattered, but if they *think* their vote could have mattered.) *Even if* we had ended up with a Republican president and Congress that *didn’t do anything*, that sort of closeness would cause people to #2 recalculate the next election.

      Put them together…well, like I said, I suspect those two things are multiplicative.


  5. I like to fantasize that a large portion of the electorate is responsible and self-aware enough to abstain from voting on the grounds that they don’t understand the issues well enough, but honestly I think it’s mostly the CBA factor.

    I don’t think Conor Friedersdorf can be taken as representative of any nonnegligible segment of the population.


  6. What I’d be curious to see is what proportion of the non-voting population live in gerrymandered safe districts and/or states that aren’t in play in a presidential election. I live in such a place. I vote out of a stubborn and possibly misplaced sense of civic duty combined with the fact that my vote may matter for local initiatives.

    My wife on the other hand tends to sit them out for the simple reason that, other than situations where the democratic party insults it’s own constituency by not even pretending to contest the election (see Anthony Brown, Kathleen Kennedy Townsend), most of the time her vote does not matter. On the other hand she was much more active when she lived across the river where the outcome isn’t always a given.


  7. I think coming up with one unified theory about why tens of millions of people did not vote is a kind of fool’s errand.

    Some people might want to vote but find that they can not take the time off of work or have enough time before or after work. This includes states with laws protecting your right to vote and with generous early voting laws. However, these laws are not uniform across the United States or even in states themselves. There were a lot of stories in 2016 on how states with voter ID laws also did things like close DMVs in minority-heavy areas and/or give minority-heavy areas fewer resources for voting locations and early voting. Sometimes this got reversed after a public outcry. Sometimes not.

    Other people might not feel represented by their political choices and want candidates further to the left or to the right.

    Others might have forgotten or been lazy or just don’t care who is in charge or not. Others might have come down with a really nasty case of the cold.

    The problem with coming up with grand unifying theories is that I often find that such grand unifying theories end up proving said speaker’s priors. The same is true for hypotheticals and this all so wonderfully convenient for the advocate of whatever theory. This might be a very human thing to do but it doesn’t make the grand unifying theory correct.


    • Interesting twist in the “black voter suppression” story: Black voter turnout has nearly converged with white voter turnout, and in fact exceeded it in 2008 and 2012. Given that voter turnout is correlated with age and education, this is surprising. I bet if you regressed voter turnout on race, income, education, age, and other relevant factors, the coefficient for being black would be positive.

      This doesn’t necessarily mean that suppression isn’t real, but it really calls into question how common, effective, and/or electorally significant it is.


  8. It’s interesting how this ties into the last post in that they’re both critiques of people’s attempts to conceptualize systems in terms of markets that don’t really function as such in anything but the crudest sense. Your critique here is valid and really there are plenty of reasons that the original theory doesn’t work either.


  9. Has anyone compared participation rates in Or to other states. In particular since Or has gone to all mail voting, there is no longer a convenience or lack of time argument relating to voting in Or. If I could I would go to all mail voting in all states.


    • IIRC, there’s modest evidence that vote-by-mail increases turnout somewhat. It’s confounded by the fact that all of the vote-by-mail states are also citizen initiative states, and there’s some modest evidence that a ballot initiative that people have strong opinions on increases turnout.


  10. Heck, the question to ask is why anybody reschedules their day so they can drive across town to stand in a long, slow moving line to fill out a bunch of paperwork. Are they the same people who go to the DMV to update their driver’s license photo every time they change hairstyles?


    • I stop on the way home, go a few block out of my way, and spend about 10 minutes at it. Of course, my state legislature isn’t trying to prevent me from voting by making it what you describe.


      • That pretty much describes me (except the polling place isn’t even out of my way, it’s on my way to work). That said, I go very early, when no one is there. I suspect that if one goes during certain hours, it takes a lot longer.

        (((This of course isn’t to criticize your other point, about the fact that some people are harried by the state when they try to vote. Fortunately, I don’t have that problem, but I know others do.)))


    • I live in a state where (a) you’re required to show that you’re actually eligible once (and there are organizations that will help you) and (b) after that they mail you a ballot if you take the trivial steps of returning the ballots (blank is okay) and sending in a postcard or going online when you move. (Aside: are there people who move who don’t give the Post Office a change of address? Lord, am I old. Or naive. Or sheltered. Or privileged. Or something.) And in our best years we still only get 75% or so of registered voters to return the ballot.


  11. I clicked on the Salon link, because it struck me as a stupid quote, and I wanted to bash it but wanted to see it in context first. Huh. That’s a pretty good article. It may not break new ground, but it’s a well thought-out version of the “Both Sides Are In Bubbles” genre.


  12. Well, judging by the pictures, I need to get some new hats.

    That said, I think there is a certain level, or type if you will, of people who are going to vote no matter what, just as there is an opposite of that. I voted, even though my state at the time went uber blue. I did this because I always vote, no matter the known outcome or not. But, there are those who will forget to go, forget to mail it in, etc. It just isn’t that heavy on the conscious for them. And unless they have something weighing on them, something that is affected by the election ie a ballot measure, a cousin running for dog catcher, whatever, they won’t change on this.


  13. In that piece Friedersdorf presents his choice as equivalent to that faced by left-liberals. Where the background preferences/priorities that are balanced against possible deal-beakers are basically the same for him as they are for left-liberals, so his argument that they should not outweigh the deal-breakers should work for let-liberals. But to simply look at the body of Friedersdorf’s record of commentary on public issues, that’s simply not the case. His priorities aren’t those of the left-liberals to whom he addressed that piece, and so this is just a simple case of a similar balancing exercise coming out differently for different people – not a scorching-hot take on what kind of balancing exercise it is, or just how disqualifying the ostensible deal-breakers really should be for “liberals.”

    If the Democratic vision of health care policy, or tax policy, or abortion rights were as important to Friedersdorf as they are to the typical left-liberal, he would have made the same decision in 2012 that 90% of them did. No matter what he was claiming at the time.


  14. I love the intellectual wankery over why people didn’t vote in an election between Hillary Clinton and Donald Trump.

    Like, you really need to do math to come up with the answer to that one?


  15. One other though for each position add an option to vote for none of the above. I wonder if none of the above if offered might have won the presidency last year since both candidates were disliked.


Comments are closed.