One year in high school we had a fundraiser. Don't remember what it was for, doesn't matter. There was a table set up by the cafeteria with cans on it, one can per grade level. And there was a contest to earn your grade the most points. To earn points, you dropped pocket change into the can for your grade level. A nickel was five points, a dime was ten points, etc. Except that each penny was one negative point. So pennies went into the can for the grade that seemed to be farthest ahead, or that you just didn't want to win because your brother would never let you hear the end of it. You see where I'm going with this? One person, one vote. But, you can either vote for a candidate, which is worth one point, or you can vote against a candidate, which is worth negative one point. This has several advantages over our current system. One is that it would increase voter turnout. There are a lot of people who are not necessarily uneducated but who are burnt out on politics and just stay home because "my vote doesn't matter" or "I don't want to vote for any of these sleazeballs." But some of them might be willing to come to the polls to vote against the candidate they hate the most. Another is that it would send a clearer message to politicians as to just how little they're liked. Gone will be the days of the victorious politician claiming a "mandate" just because he/she got a whopping 60% of the vote from the 70% of eligible voters who showed up. Imagine how entertaining it will be when a candidate for the top executive position in a nation of over 300 million wins the election with a net score of just a few thousand votes. (Does the electoral college still exist? I don't care. Whatever it takes to make this happen.) Imagine a truly loathsome candidate getting a negative vote total instead of an automatic 30-40% cut of the ballots. Okay, it'll be hilarious, but how will it actually mend the voting system? Consider the possibilities in, say, a primary, from the viewpoint of a voter, Alice. And let's say there are three "serious" contenders for the nomination, A, B, and C. Hypothetically. And let's say that, hypothetically, Alice most prefers C, would find B acceptable, and hates A. The problem is that the news media tells Alice that A and B are the most likely to win. Under the current system, where Alice can only vote for a candidate, she faces the choice of either compromising her beliefs and voting for B, or bravely voting for C and hoping the polls are wrong, and that B didn't need her vote to defeat A. Under the high school system, she has a third option: voting against A. This helps both B and C and is, perhaps, the view Alice holds most strongly anyway. Now, it might turn out that C needed her positive vote in order to defeat B. But the risk here is far less: that an okay candidate wins instead of a terrible candidate winning if Alice votes her beliefs under the current system. Bob is also helped by the new system. He's a low-information voter who has no firm preference between B and C but also hates A. Instead of blindly voting for B or C based on who the media tells him is more likely to win, he can vote against A and hope that better-informed B/C supporters make the correct choice between the two. Also note: this is simpler than ranked choice. You just cast one vote, either for or against someone. The Bobs of the world just have to watch out for butterfly ballots. Campaigning might also improve. Under the high school system, candidates are less incentivized to spend money on an attack ad against a single opponent. In a general election, if the Republican and Democrat spend most of their time fearmongering against each other, swing voters may well simply vote against the candidate they fear the most (or against the candidate whose attack ads irritated them slightly more, without having to vote for the other candidate). Leaving a third party candidate to walk off with the win because nobody bothered voting against him. So you can see why the current duopoly in the U.S. will never let this happen in a general election, but let's continue. Candidates would therefore be wise to use their money to run more positive campaigns about themselves, in order to attract positive votes, and maybe politics would get a smidge less toxic in this country. Of course, nobody ever accused the average politician of being wise, or of having positive characteristics to promote, so there's also the possibility that attack ads would multiply as candidates sought to tear down every single other name on the ballot, just in case. I am willing to allow this possibility, under the heading of "let them show what they're really made of." So there you have it. Better turnout, better accuracy in reflecting the electorate's opinions, better options for voters in the booth, better chances for third parties to win once in a while, better campaigns maybe, and humiliations galore. One can argue other systems are better, but this is certainly better -- and more entertaining to imagine -- than what we've got in the U.S. right now. Happy Memorial weekend.
What happens when too many people who would vote for you instead vote against your opponent and that costs you the election? Your opponent knows that a good chunk of people hate him but guess what? He still won because the people who supported him outnumbered, by just enough votes, the people that voted for you instead of voting against him. In fact we would see a new strategy emerge where a politician enrages people he knows wouldn't vote for him in a bid to get them to vote against him. Especially if he believes he can prevent his supporters from voting against you.
This could make outsiders very powerful. Most people vote for or against the two major candidates, leaving each at a result close to zero. Compare that to the one issue outsider whom most people never even hear about, except for those who support his issue...
A far better way to change the system would be to force third party nominees into debates. The reason the third party candidates get so few votes is because no one knows who they are because coverage is always given to democrats and republicans. Force debates to include third party candidates and you will see things change.
Up to 3 candidates, this is mathematically equivalent to approval voting, but perhaps better psychologically. There’s also the issue of write-in candidates. It would be a weird situation indeed where a write-in candidate wins because no one knows to vote against them. You’d likely see extremists winning all the time whenever there’s a close election between two relative centrists. All the Nazis/commies need is a few hundred people to guaranteed vote for them (say, a hidden Facebook group) while the rest of the city squabbles over vanilla A and vanilla B for Mayor. Making conspiracy an effective part of the voting system doesn’t seem salutary for democracy.
This sounds familiar to what some of my schools did growing up. I believe it was called Penny Wars, but it was slightly different. Pennies counted as +1, whereas any other denomination (nickels, dimes, quarters, dollars, etc.) counted as negative (e.g. nickels were -5...dollars were -100). But, I don't think this would improve voter turnout. The only clearly effective method of improving voter turnout is compulsory voting, even if it's just submitting a blank ballot. I also don't think this method of voting would create effective governance. There are other more proven electoral systems (e.g. ranked voting) that work much better. The fact of the matter is that the US had a good run with our beta democracy, but it has been more than improved upon by other countries. We should really scrap our whole system and rewrite it to make sense for the modern era.
Ranked voting accomplishes what you desire for third party participation (and for what it's worth, I wholeheartedly agree with you that we need more substantive third party participation). Our style of electoral system favors two parties only. There is a ton of real-world data to back this up.
I love the idea of someone doing a write-in vote against someone. "I don't know if anyone has heard of Billy, but he's my neighbour and I just wanted to say that I am really against him ever holding any political office..."
I already see an easy exploit. Let us say I am not popular with the black people. So I take some of my money and help an open klansman run. I really would have just sucked most of the positive black votes away and put them on someone else. This would be pretty easy to do, and very deflective and beneficial for a person with a racist attitude who is not too racist. In this case a person who just treats black people with respect and dignity is not positive enough to attract the votes of the anti-racial crowd who is going to throw negative votes at the klan dude because they fear he might get too many positive votes from his racist buddies. Maybe this sort of problem could be addressed by a positive and negative vote. I would personally like to see some system to take second choice into account. Like in this election where Bernie and Biden have a pretty extreme love/hate thing going on. The second choice might be the more uniting candidate.
I would imagine there would be a number of votes like that against dayton kitchens that was really close to the number of active WF members.
Another good point, but to answer your speculation, no. A positive and a negative vote only means you need to run 2 klansmen instead of 1 to perform your exploit. Not much better.
It is true, trying to get over exploitation of the vote is problematic. It is one of those times where a better solution eludes me so I just sort of complain without having any idea of how to fix the situation entirely. I do think we could make it a lot better and more reflective of societal wants and needs. I also think that negative voting has to be some part of the equation to eliminate the trump factor. There really has to be a point where the overwhelming negativity of a candidate is so obvious that society has to be able to say no to the popular idiot. I think it was really shitty we had two candidates who were huge negatives in the 2016 election. The only problem I could see with a negative vote is it would give a lot more power to racism, sexism, and cultural biases. Obama would have probably never been elected, nor would we ever see a female candidate if we had a negative vote. Those types of people would actually cut off their own nose to spite their face.
Ranked voting definitely gets treated by many as an opportunity to vote against candidates you don't like. It's common here in Australia for people to vote by initially putting their last preference against their most disliked candidate and working up from there. It is pretty satisfying to put a dickhead senate candidate at number 87/87 (or whatever) of your preferences.
That may be a way to go. Especially now that we have better computers that can crunch bigger numbers. This might not have been possible back in the day, but we have some decent algorithms now. The problem would be deciding on an algorithm that would be best, but I think it would be better than the way we do it now.
@Order2Chaos could talk your ear off, but there are plenty of ranked systems that don't need computers (although they definitely make results quicker and easier to count and assuming all the data and formulas are available for scrutiny there's no reasons to avoid computers in counting). Australia has had it for national elections since 1918, and it was first used as early as 1892.
It's the definition of "best" that has to be settled, not which algorithm meets it. From an algorithmic cleanliness point of view, Meek's method is probably most correct for STV (Warren might be better but it's new enough that that's a matter of opinion), but it's computationally expensive. Not like, a hard problem, but there are a ton of multiplications and divisions that make it infeasible to carry out by hand as a guard against malicious software. Even Australia doesn't use it; they use Hare-Clark. Gregory's is an improvement on H-C, and probably the last hand-computable one. ETA: To clarify, Hare-Clark is potentially exploitable because the order the ballots are counted in makes a big difference. Unless they're randomly shuffled (and that's hard to verify), geographical clustering relative to the counting center can radically change whose votes count for 2nd and subsequent seats. Edit 2: And then there's the question of IRV vs IRV-avoid-worst, which removes the candidate with the most last-place votes first, instead of the candidate with the fewest first-place votes first. This tends to eliminate very divisive politicians at the primary stage, since they have to be able to get some fraction of the 2nd-place vote over third-party alternatives in order to win. Edit 2.5: I could go on. A primary pairwise count followed by IRV if nobody wins that is probably best for expressing opinions, but not nearly as easy to do in practice, so is it really "best"?
