Resolving a tied election

[AY1975]

There have been a few occasions when two (or maybe even three or four?) potential winners in an election have been found to have received the same number of votes, even after a few recounts.

I'm not aware of it ever having happened in a general election or parliamentary byelection (though it wouldn't surprise me if it happens at next week's Gorton & Denton byelection!), but it has very occasionally happened in council elections and byelections.

If this happens, the returning officer has to determine the winner by some form of random selection. This is known as "drawing lots", and the returning officer has to decide how to resolve the situation when this happens. There are various methods of doing this that have been used (or could potentially be used), including:

(a) tossing a coin.

(b) taking one already counted ballot paper marked with each of the potential winners, or writing the names of each of the potential winners on separate blank slips of paper, putting them in a ballot box or some kind of container, mixing them up and drawing one of them at random.

(c) taking two blank slips of paper, marking a cross on one of them and leaving the other one blank, mixing them up, then asking each of the potential winners to take one of the slips and the one who takes the one with a cross marked on it is the winner.

(d) asking the two potential winners to each throw a die (or maybe two dice each), and whoever gets the highest score is the winner.

(e) taking two drinking straws or matchsticks, cutting or breaking one of them in half and keeping the other at full length, mixing them up, concealing them in the hand with only the top bit showing, and asking each candidate to pick one, and then either the one who takes the shortest or the longest one is the winner. This is commonly known as "drawing straws" or "drawing matches".

(f) asking the two potential winners to do rock, paper, scissors.

Does anyone know of any other method that has been used (or can anyone think of any other way it could be done), or even have any first-hand experience of this, either in an actual parliamentary or council election or any other election such as a students' union election?

 

[Gloster]

I have read of somewhere using the method of drawing a card from a shuffled pack. I have a feeling that the Electoral Commission has either rules or strong recommendations as to which methods to be used.

 

[Jan Mayen]

Coin toss is the only one I've heard about in this country, with the winner getting to keep the coin.

 

[gg1]

I've always thought that in the result of a tie and where one of them is the incumbent, that person should be re-elected, or the incumbent party be re-elected if the tied candidates are both new.

Doesn't help of course where the incumbent candidate/party doesn't finish in the top two.

 

[The exile]

I've always thought that in the result of a tie and where one of them is the incumbent, that person should be re-elected, or the incumbent party be re-elected if the tied candidates are both new.

Doesn't help of course where the incumbent candidate/party doesn't finish in the top two.

Couple of observations (three actually). 1.Officially we vote for candidates, not parties (however much the press likes to say different). 2. Likewise officially, by the time an election is held there is no incumbent, merely a set of candidates for a vacant seat 3)A tie is just as likely to be brought about by a shift in votes for other candidates as it is to be directly connected with any incumbent.

 

[pdeaves]

I personally would favour a (necessarily subjective) decision to award to the candidate who contested the election in the most 'gentlemanly'/honourable way. I accept that in reality that could be fraught with its own problems.

 

[Gloster]

What about mortal combat using broadswords (claymores in Scotland). That should make a few chancers think a bit more about standing.

 

[PGAT]

Assuming there are more parties running, would it not make sense to do another vote with just the tied candidates? Random selection doesn’t feel fair and doesn’t sit right with me.

 

[35B]

Assuming there are more parties running, would it not make sense to do another vote with just the tied candidates? Random selection doesn’t feel fair and doesn’t sit right with me.

I'd personally favour the French system of two round elections. But in the absence of that, breaking a tie with something absolutely random seems entirely fair - and, as importantly, proportionate to the cost of running an election.

An isolated rerun is also potentially very unfair, as it risks lower turnout and a statistically distorted electorate. The value of highly motivated supporters will be a lot greater.

 

[Trackman]

In recent years its happened in Bury and Bolton and Blyth. Bury was a drawing lots affair, Bolton it was cutting a pack of cards, Blyth was the shortest straw.

 

[Gloster]

On the Electoral Commission’s website there is a page titled Equality of Votes. It does seem a bit fluid: one would have thought that single, rigid practice would have evolved and would be used unless there were sound reasons to vary it.

 

[AlterEgo]

Greco-Roman wrestling contest.

 

[nap666]

As per the examples already given there is no set method for resolving a tied vote, I had a Returning Officer suggest a game of Ludo or some other board game, managed to persuade them to stick to "2 marked ballot papers in a box" and another person to draw winner.

 

[zero]

For politics, I would prefer a preferential voting system where the top two candidates getting exactly the same result is extremely unlikely.

I did experience this situation in an election for President of a university club. With the top two tied, it was decided to hold another round like in the French system mentioned above. This resulted in another tie, so the candidates were given a chance to canvass votes again, but the third round was also a tie (membership was close to 100, so it was felt that some people might change their vote). As both winners had broadly similar visions for the direction of the club, I suggested they have a joint presidency and a majority of members agreed.

 

[JohnofWessex]

Many years ago I worked for Bristol City Council.

Elections were by thirds, ie a third of the seats came up for election each year with one in 4 years 'off' for what had been the Avon elections.

One year we ended up with an equal split between the number of Conservative, Labour & Lib Dem candidates.

The Councillors respected the decision of the electorate and the result was a 3 party coalition for a year.

 

[DarloRich]

they two tied candidates should fight a dual with the incumbent choosing the weapon

 

[Cloud Strife]

Does anyone know of any other method that has been used (or can anyone think of any other way it could be done), or even have any first-hand experience of this, either in an actual parliamentary or council election or any other election such as a students' union election?

I'm personally of the opinion that if there's a tie, then the only fair and sensible way to do it is to hold the election again with those two candidates. I don't like drawing cards, lots or otherwise: the winner hasn't won a plurality or majority, nor did the loser actually lose. If there's a tie in the second, then the election should simply keep being held until there's a winner.

Having said that, in council/parliamentary elections, I would support some sort of tie-breaker being used to reduce costs of a second election. Probably the fairest way to do it would be to look at the results from individual polling stations: the candidate who won the most votes from the polling station with the highest turnout could be declared the winner.

Whether this is possible in the UK is another question, given that ballot boxes are taken to a central location for counting as opposed to being counted in the individual polling stations.

 

[35B]

I'm personally of the opinion that if there's a tie, then the only fair and sensible way to do it is to hold the election again with those two candidates. I don't like drawing cards, lots or otherwise: the winner hasn't won a plurality or majority, nor did the loser actually lose. If there's a tie in the second, then the election should simply keep being held until there's a winner.

Having said that, in council/parliamentary elections, I would support some sort of tie-breaker being used to reduce costs of a second election. Probably the fairest way to do it would be to look at the results from individual polling stations: the candidate who won the most votes from the polling station with the highest turnout could be declared the winner.

Whether this is possible in the UK is another question, given that ballot boxes are taken to a central location for counting as opposed to being counted in the individual polling stations.

At the level of individual candidates, this is rare. I think the acknowledgement that there has been a tie, and a simple act of chance to determine how two will go into one, is short, sharp, and equitable. It doesn't, unlike reviews of individual polling stations, favour any particular voters, and has the virtue of simplicity.

The example cited by @JohnofWessex from Bristol, where there was a tie between parties, is slightly different, and I respect the ability of the three parties to agree that coalition rather than combine on a 2 vs 1 basis.

 

[Cloud Strife]

At the level of individual candidates, this is rare. I think the acknowledgement that there has been a tie, and a simple act of chance to determine how two will go into one, is short, sharp, and equitable. It doesn't, unlike reviews of individual polling stations, favour any particular voters, and has the virtue of simplicity.

I genuinely don't like acts of chance being used in elections, because there isn't a clear and decisive winner. By using more objective criteria such as polling station turnout, it's much more objective - yes, voters in the polling station with the highest turnout are favoured, but this is a way of rewarding them for their civic duty. But of course, this is exactly why STV or similar is needed.

One interesting issue I've seen before was an argument over exactly which form of STV was to be used. The organisation holding the election simply stated in their rules that STV was to be used, and the returning officer in question had used some program to calculate the results. The problem was that the exact method of counting wasn't specified, and it became clear that an alternative method resulted in a different winner. The winner of the election as originally declared had no interest in a re-run as he had been declared the winner originally, while the loser quite rightfully pointed out that the rules simply stipulated that STV was to be used.

 

[35B]

I genuinely don't like acts of chance being used in elections, because there isn't a clear and decisive winner. By using more objective criteria such as polling station turnout, it's much more objective - yes, voters in the polling station with the highest turnout are favoured, but this is a way of rewarding them for their civic duty. But of course, this is exactly why STV or similar is needed.

One interesting issue I've seen before was an argument over exactly which form of STV was to be used. The organisation holding the election simply stated in their rules that STV was to be used, and the returning officer in question had used some program to calculate the results. The problem was that the exact method of counting wasn't specified, and it became clear that an alternative method resulted in a different winner. The winner of the election as originally declared had no interest in a re-run as he had been declared the winner originally, while the loser quite rightfully pointed out that the rules simply stipulated that STV was to be used.

And this is why “fairness” needs to be based on clarity. I’ve had involvement in elections subject to a proportional system, and the results were difficult to interpret and understand without knowledge of spreadsheets.

Any priority matrix for resolving a tie is necessarily subjective. It uses a value judgement of which factors weigh higher, and codifies those into rules - assuming that they are codified.

The use of chance as a tie-breaker addresses the fundamental reality that there is no difference between two candidates attracting equal votes. It avoids getting drawn into a subjective assessment of who should win, and instead provides a definitive answer. As Alexander the Grwat found, sometimes it’s better to cut a knot than try to untie it.