Election Exception Proves the Rule


On November 4, while votes were being counted all across the country, a strange thing happened in Mississippi. A woman named Janie Sisco won election as Lincoln County elections commissioner, unseating a 20-year incumbent — by just one vote.

The final tally had Sisco beating Charles Monroe Smith 1,580 to 1,579. Assuming Sisco voted for herself, she may have cast the deciding ballot in her own favor.

Sisco’s odds of having affected the outcome of the national election, of course, were vanishingly small, making the great Sisco/Smith 2008 race to the commissionership an exception that proves the rule: individual votes almost never decide elections. So why do we vote?

(Hat tip: Jim Cropcho)


In these issues of a whether a vote "matters", I think the assumption is being confused with it "affecting the outcome"

If a parent tells a stressed out child before a test they're not prepared for "I love you honey, no matter how you do things will be ok" that matters, even if it doesn't affect the outcome the child gets on the test (we'll assume a quantitative test here for those who say "reducing the stress will improve scores", its hard to fake your way through a math exam). If you have to look at things from an end result only aspect, talking to anyone during an office break who cannot directly help you to achieve a raise or promotion isn't something you should do either.

More importantly, at which point does it make statistical sense by the definitions of "why do we vote?" to do something? Only when we're the marginal vote? When we're at a 1% outcome? I mean technically 1/35,000,000 still somewhat counts (unless you're arguing against the electoral college system in general), so if you're saying that doesn't count is there a level between that and being the marginal vote that is "logical" to vote at again?


jeff b.

This is interesting because I was just thinking about this today but from a somewhat different perspective (since it was before reading this). I was thinking about voting and people being so upset with the other side winning. Most people (at least at a national level) know who they're going to vote for many months ago for a number of reasons (party loyalty, single issue, prejudices, etc) yet I believe it's followed (at least by myself) as more of a sporting event at some point (which then becomes partisanship to some extent). Voting allows you to help your team (candidate) win, which is much harder in other sports. But what I was really thinking about is how upset people (myself included) get with people who don't agree with you (politically) and I came to the conclusion that we know our vote isn't going to be the one to make the difference so we want to control other people and how they vote, which is impossible, but that's democracy.



You can say, "Don't blame me. I voted for the other guy."


I'm inclined to think along the lines of 20 and 22.

In the 1580 to 1579 outcome you mentioned, every single one of those 1580 votes was the deciding vote...

And then, in the land of polling we live in, margin of victory counts. The perceived 'mandate' effect seems to grant the inbound candidates political momentum or political 'capital.' And the aggregate effect of this can (may?) have a large effect on what policies actually get enacted.

Consider the gay marriage proposition 8 in California. Had they lost 50.5 to 49.5, they might come back with the same strategy next cycle or so, but losing by the margin they did may send them back to the drawing board to re-design a 20-year strategy.

So, I'm not sure that the prima facie assumption that most votes don't matter is really standing on anything substantive. Can we hear a cogent explanation of why votes don't matter again?

Chris M

Voting is a sham. It's a tool used by the state to keep its citizens docile and "in line." Read some Cicero, he had it right. Might DOES NOT make right.


Voting should be considered a responsibility and not a right. I still make my bed because I'm supposed to and not because I gain anytihng from it.


Here's a good question for veterans day. One soldier will never decide a war, but thousands still enlist to "protect their family" despite the large potential cost. Why?


Suppose 66 voters voted for candiate "x" and 58 voters voted for candidate "z." One would naturally calculate that candidate x won by 8 votes. However, if 4 would have changed their vote and voted for z the vote would've been tied.

As it takes only 1/2 of the vote differential to tie a vote a more meaningful method to me is use 1/2 the differential and divide this by the total vote count. In this case 3% of the population determined the outcome of this example election, as it did in our general election.

3% of the electorate decided the outcome of our general presidential election. It would've only been necessary for McCain to convince an additional 3% of US voters that he would've been the better candidate. This is what Obama did.


#26 above, you're correct. I left out the electoral college issue, which is of paramount importance. My bad.

#28 above, you're also correct. However, the indy candidates sucked off a small percentage and probably the greens sucked off the left votes and the "libertarians" sucked off the right, probably close to equal net. Nadar 686k, Barr 505K, Green 140K


To #24 -- that assumes that the alternative would be the other candidate, which is not always the case. In this case, with so many newly registered voters, the choices for some were 1) vote for Obama or 2) don't vote at all. So, you can't always use half the differential as a meaningful number.

Raj Pandravada

I agree - why DO we vote?

There's a fantastic science fiction story by Isaac Asimov called 'Franchise', where, in the not-too-distant future, the great supercomputer Multivac decides EVERY election, municipal to presidential elections, by analyzing the mindset, outlook and vote of a single representative American.

I LOVE this idea....it would save everybody a ton of time....

Asimov rules.


To #24 - the Electoral College messes up that nice analysis. In theory, McCain had to convince far less than 3% of the electorate to vote for him in order to win, but of course he did not.

Because of the Electoral College and the robust science-based polling and reporting available this time around, the outcome wasn't in doubt since September - it was in the media's interest to keep the "horse-race" meme alive.

Witty Nickname

I think it was in 2003 I cast a ballot in a city council race in Marshall Texas that was won by one vote.


I have another example of the "Exception Proves the Rule", but where the implied rule isn't correct.

Near my town there is an intersection with traffic lights where the right hand turn only lane has a sign reading "Right Turn on Red Only After Stop". When I learned to drive, I took this to mean that you can turn right on red without stopping first when such a sign is not present.

Years later I found out that this wasn't correct. Which raises the question of why that sign was there in the first place?


But candidates who win elections by only one vote likely must work harder to be perceived as legitimately elected and are not perceived to have the mandate to make big changes. Part of the reason people dislike Bush was because he just barely won (if that). But Obama's decisive win already has already reduced animosity. It is not important to me to be the deciding vote in an election because the margin of win also carries information. And for that, every vote counts.


How do you know which vote was the deciding vote?


After thinking about the economist's arguments of vanishingly small effect that people's votes have on the national elections, I've come to this conclusion on why one should vote:

When I was in middle school, I was taught that local elections and local politics have a far greater direct effect on one's life than national politics, to the degree of whether or not you can go to the pub on Sunday, or how confident in your kid's educators you'll be. So, always vote local, because that's where the race is smallest, you'll have the most impact, and you'll see the greatest direct effect on your life.

Then, if you've voted locally, why not just go the extra step and vote nationally, since there's little extra cost once you're already at the booth.


I'm sorry, but this is a classic example of missing the forest for the trees: "individual votes almost never decide elections."

Individual votes absolutely decide elections. I'll give you that rarely are elections decided by a one vote margin, but every election is made up of individuals. Say what you want about Obama, or his campaign, but the get out the vote effort this election intimately recognized that individuals absolutely decide elections. The war was won by the actions of individuals.

It's like wars. The actions of any particular individual are usually inconsequential taken in isolation, but the decision of every individual to give a little less, or a little more, is often the margin of victory.

Don't fall into the common economist's trap of assuming that because you are dealing with the actions of a large population, you therefore can ignore the individual. That's a shorthand. It is only a lack of tools that prevents the analysis of markets by the aggregation of individual analyses.


Al Marsh

I'm not so sure that by voting one is "doing his/her bit". By abstaning, surely one is just expressing satisfaction with the status quo and perhaps genuinely doesn't care who gets in. I don't think there is anything wrong with this.


Regarding denis's (#8) objection, an excellent discussion of this usually-stupidly-used phrase "the exception proves the rule" is found here:

Straight Dope

Scroll down to the subsection called "The Last Word on Exceptions" for the ultimate explanation of why, and when, this expression makes sense (and why it's stupid the rest of the time).