Tag Archives: Alternative Vote

The voting system that I am advocating for parliamentary elections

I previously discussed a few possible ways of how you could have a fair proportional national election, and in my last post I promised to commit to a single system to advocate. I had five criteria for such a system:

1. Voters should be able to vote for any candidate from anywhere in the country. It would be a national election.
2. The system should be proportional at a national level.
3. There should be no systematic bias against independent candidates.
4. The ballot paper and voting system should be relatively simple for the electorate, so people can still just turn up on the day and vote as they do now.
5. The system should, as much as possible, favour a reasonable geographical spread of MPs.

For it to be possible to vote for any candidate standing in the country – of which there would be thousands – there would have to be some increase in complexity. We just have to make sure that it’s not too great an increase, and that any increase in complexity is offset by an even greater improvement to our democracy.

The system I am advocating is essentially a simplified version of Jameson Quinn’s PAL system, which you can find described here and here. I will describe my simplified version in this post.

The ballot paper would be largely as it is now. All candidates would have a nominated constituency and their name would be printed on the ballot papers for that constituency (with their party name if applicable). People would be able to vote for one of these candidates exactly as they do now, and the process would be just as simple for them. Alternatively they could write in underneath the name of any other candidate standing nationally to cast a vote for them. Each candidate could also have their own code to distinguish between those with the same or similar names. The other option open to voters would be to rank the local candidates listed on the ballot paper.

Writing in candidates’ names on the ballot is not a new phenomenon, and it has been done frequently in American elections. For example, in 1928, Herbert Hoover won the Republican Massachusetts presidential primary election as a write-in candidate on his way to becoming president of America. Similarly in 1940, Franklin D. Roosevelt won the Democratic New Jersey presidential primary election as a write-in candidate on his own way to retaining the presidency.

The election would work as a Single Transferable Vote (STV) proportional election. In STV, voters would normally rank the candidates, and their vote can be transferred from their top preference to their lower preferences if their top preference is eliminated or elected with more votes than needed. But in this system, it would be unrealistic for someone to rank a large number of candidates whose names are not even printed on the ballot paper.

Instead, candidates would all have their own pre-declared public ranking of the other candidates. Then, instead of your vote being transferred to your own second or third choice and so on, your vote would be transferred according to your favourite candidate’s ranking. Voters would have the option of ticking a box to say that they do not want their vote transferred in this way. These voters can still take part in the ranking process, but they would be limited to ranking the local candidates printed on the ballot paper, by writing the rank number next to each name. The entire country would be treated as one super-constituency, and the 650 seats would be allocated according to the STV method.

Instead of candidates individually ranking all other candidates, candidates from the same party would be ranked equally in a given candidate’s list. However, independent candidates can still be ranked anywhere. This simplifies the process somewhat. This also stops parties from effectively having a priority list that all members of the party would be expected to use, with the party leader top and so on. A party candidate’s ranking would always have the other candidates from their own party ranked immediately behind themselves. For example, a party candidate’s ranking might look like this:

1. The ranking candidate (every candidate is automatically top of their own list)
2. The rest of the candidates from the ranking candidate’s party
3. An independent candidate
4. All the candidates from party A
5. All the candidates from party B
6. Another independent candidate

And so on. Candidates would not be expected to rank every other candidate standing nationally. They can stop at any point, and any ranking below an entire party of candidates would be largely irrelevant after the vote is diluted into an entire party. The above list would effectively be:

1. The ranking candidate
2. The rest of the candidates from the ranking candidate’s party

An independent candidate’s ranking could look slightly different as they might rank several independent candidates before the candidates for a particular party. For example:

1. The ranking candidate
2. An independent candidate
3. Another independent candidate
4. Yet another independent candidate
5. All the candidates from party A

And as before, any ranking below the candidates of an entire party would be unlikely to affect the result.

Is it reasonable to allow candidates to rank the other candidates on your behalf? I think it is. If you vote for a candidate, then it would generally be because you like their policies and would trust them to vote for what you would agree with in Parliament. Allowing your vote to be transferred according to their ranking list is simply the first act of representation that you are conferring onto them. If you agree with a candidate’s policies, it is likely that you’ll also agree with the policies of candidates that they also agree with and that they therefore also rank highly. Also, because candidates’ rankings would be published in advance of the election, voters can use this information as part of their decision process in determining who to vote for. And as said, voters can also choose not to allow their vote to be transferred, and rank just the local candidates themselves, or simply vote for one of them.

Who would be my MP? People wouldn’t have one official MP; they would be able to contact any one of them. Supporters of a party may want to write to the geographically closest MP to them from that party, and parties might have their own boundaries defining which of their MPs covers which area. But there would be nothing official within this system that determines who your MP is. This gives people the freedom to write to the MP that they think is best suited to their particular concern. Currently, there is a high chance that your MP is someone that you didn’t vote for and don’t think represents you very well. This is a big downside to having one local MP who officially represents you. The system I am advocating is much more flexible.

If this system is implemented, then in the future, further features could be added. Ideally it would be possible for voters to rank any of the candidates standing nationally, and how to implement this simply could be discussed once the main system is in operation. Perhaps this would be possible if computer voting ever came in. Another alternative would be to give candidates scores rather than ranks. But that is for another day. To summarise the system:

The ballot paper for a constituency lists all candidates who have nominated that constituency. Voters have the following three options:

1. Cast a vote for one of the listed candidates
2. Cast a vote for any other candidate in the country by writing their name at the bottom
3. Rank the listed candidates

Before the election, every candidate will make a publicly available ranking list of the other candidates. Candidates from a particular party would always be ranked equally in a given candidate’s list. Candidates from the ranking candidate’s own party would always be top of the candidate’s ranking.

If a voter chooses option 1 or 2, then they also have the option of ticking a box to say that they do not want to use their preferred candidate’s ranking list. In such cases, their vote will not be transferred if their preferred candidate is eliminated or elected with a surplus of votes. If they do not tick this box, then their preferred candidate’s ranking is used as the voter’s own ranking list. The country is treated as one super-constituency with 650 seats, and the election proceeds as an STV election.

Let’s have a look again at the five criteria that I wanted an electoral system to fit.

1. Voters should be able to vote for any candidate from anywhere in the country. It would be a national election.

Allowing voters to vote for candidates by writing their names on the ballot paper, as I am advocating, is the simplest way to obtain a single national election where every voter is free to vote for whichever candidate they like.

2. The system should be proportional at a national level.

Having the whole country as one super-constituency and using STV with votes transferred according to candidates’ ranking lists or voters’ rankings of the local candidates would allow this to happen.

3. There should be no systematic bias against independent candidates.

Unlike party-list proportional elections, this system allows independent candidates to stand equally alongside party candidates. Because votes can be transferred, your vote is not wasted by voting for someone who is unlikely to be elected. This means that unlike First Past the Post, voters are not forced into voting for one of the main parties if they want their vote to count.

4. The ballot paper and voting system should be relatively simple for the electorate, so people can still just turn up on the day and vote as they do now.

The system would still be very simple for voters. They can vote exactly as they do now. They would also have the added option of writing in any candidate standing nationally, or ranking the local candidates, as they would have done had Alternative Vote been adopted.

5. The system should, as much as possible, favour a reasonable geographical spread of MPs.

Local candidates would still appear on the ballot paper for each constituency, and voters would be able to rank these candidates, meaning that it would still be very easy to vote purely for local candidates if people wanted to vote on local issues.

This system retains enough simplicity and dramatically enhances democratic power, and that is why I am advocating it for use in our Parliamentary elections.

Advertisements

Electoral reform 2 – Why we need proportional representation

Welcome to my second post in my section on electoral reform. In my last post, I discussed why First Past The Post (FPTP) is such a bad voting system, and that we lost out by choosing to retain it over Alternative Vote (AV) in the referendum in 2011.

However, I would argue that the main problem with our electoral system is that we have 650 individual geographical elections with one candidate elected from each. This is a problem under either FPTP or AV, or indeed any system that would elect a single candidate from each constituency.

The problem is that general elections are trying to achieve two different things: elect local representatives and also elect a national Parliament. But just adding up the local representatives is a very crude way of forming a national Parliament and, unsurprisingly, it does a very poor job, and results in a fairly arbitrary national distribution of MPs, which doesn’t represent the national vote in any meaningful manner. How close a national election result is isn’t just down to how close the parties are in total votes, but it’s also in large part down to the geographical spread of these votes.

If people’s voting preferences were more evenly spread around the country than they are now, then most constituencies would return similar results to each other at each election. This would mean that the most popular party would always be likely to win the national election with a landslide. This is regardless of whether they had 50% of the national vote, 40%, 30% or whatever. Any diversity in Parliament relies on geographical voting diversity.

The current system favours parties whose support is polarised into strong support in some areas and weak support in others, rather than those parties that have a consistent level of support across the country. A good example of this is the 1983 UK general election. Here is John Cleese discussing it. Labour received 27.6% of the national vote and the SDP-Liberal Alliance received 25.4%, whereas they won 209 and 23 seats respectively. It would even be possible for the most popular party to lose out in this way, especially if the top three were all fairly close in terms of national support, and the most popular party had consistent support across the country rather than pockets of high and low support.

Because of the importance of the geography of the votes, single-winner constituencies are susceptible to gerrymandering, where constituency boundaries can be altered to favour one party despite making no difference to the total number of votes received by any party. Here is a video to explain it.

Even if we could all agree on the best voting system for electing a single candidate in a single election, whether FPTP, AV or something else altogether, you cannot simply put together the results of 650 individual elections like this and expect it to add up to a reasonable democratic system at a national level. It simply does not work. And this is why we need a better system: specifically proportional representation (PR). If a certain proportion of the population have the same political views as each other, then they have the power to elect that proportion of the MPs together. That’s how democracy should work. PR is simply more democratic than what we have now. There are various forms of PR, each with their own pros and cons, and I will discuss some of these. But first I want to discuss some of the arguments people use against PR.

PR is unlikely to lead to strong majority government. Unless a party has over 50% of national support, which is unlikely, parties have to form a coalition if we are to have a majority government. This is seen by some people as a bad thing. However, many other successful countries, including Germany, have PR, and it hasn’t caused the world to collapse in on itself.

Watch this video by John Cleese on PR. It’s the full-length version of the video I linked to above. Some of it is out of date, but he makes some good points. In particular he points out that majority single-party governments that change from election to election result in less stability than coalitions because every time a new government gets in they are likely to make big changes, which can then swing back and forth. PR also encourages honest voting rather than just voting for who think can win in your constituency.

To have a majority government that doesn’t have a majority of the support is less democratic than PR. So, you are trading in a certain amount of democracy for your precious majority. And as I discussed in my last post, there is no guarantee of a single-party majority government anyway with 650 single-winner elections. It has become less likely nowadays because fewer people vote for the main two parties than in the past. Indeed, the 2010 UK general election failed to return a majority for any party. So if it’s not proportional and it’s not a majority for one party, what is it? It’s an undemocratic mess that no-one wanted.

It’s often argued that PR would allow extremist parties in. Well, only in the proportion that they are elected. That’s democracy. If they only have a small amount of support, they will only have a small amount of influence in Parliament. There are also many minority parties whose inclusion in Parliament could benefit the country.

The bottom line is that proportional representation is more democratic than any other way of creating a national Parliament and the arguments given against it aren’t strong enough to counteract this.

There are various systems of PR. For example, there are party-list systems, where everyone votes for a party, and each party then gets the proportion of the seats equal to their proportion of the vote. This can work at a national level to ensure a high level of proportionality by dividing up all 650 seats in the most proportional way, rather than splitting them regionally first. However, this would do away with the concept of a local constituency MP. It also discriminates against independent candidates who would not be able to stand in a party-list system. I have previously argued the case for independents being able to stand on an equal footing with parties, so I see this as an unacceptable failing.

Multiple-seat constituencies are an improvement on party-list PR in some ways. Candidates can stand either for a party or as an independent, as they can in our current system. However, there would be larger constituencies and more than one MP elected per constituency. For example, it could be eight MPs per constituency. This would mean having constituencies eight times the size that they currently are, to keep the total number of MPs the same. The system normally used for this is Single Transferable Vote (STV), explained in this video. It is the multiple-winner equivalent of AV, which is why many saw AV as a stepping stone to proportional representation. Under STV, parties generally field more than one candidate in a constituency. For example, if a party has a quarter of the vote, then they would expect to get two MPs in an eight-seat constituency so it would make sense for them to field at least two candidates. Constituents would have eight local MPs that they could potentially write to, so there would be more chance that they would have at least one MP that is sympathetic to their views.

However, because there are still regional constituencies, STV only offers limited proportionality. If a party has 5% of the vote nationally, then out of 650 MPs they should get 32 or 33, but under this system they might still get none, because they need to reach a higher threshold to get a single seat out of eight, so it’s not proportional at a very fine level. It would not be feasible to make STV work with a single country-sized constituency because every candidate in the whole country would have to be listed on every ballot paper. So while it is better than party-list PR for independents, it gives less proportional results.

There are also combined “mixed-member” systems, explained in this video, where you get two votes – one for a party and one for a candidate in your constituency. Half of the seats go to the constituency candidates who are elected in the usual way, so it can allow for independent candidates. The remaining half of the seats are allocated to parties in a proportional manner, so that all parties end up with the number of seats that relates to their proportion of the votes. The proportion calculated includes the constituency seats, so it is possible that a party already has too many seats before the proportional part of the system kicks in. This system is a bit of a mess, forcing together two different systems in an ugly manner, and it is still party-centric with a bias against independent candidates in half of the available seats.

Ideally, we’d have a system that isn’t biased against independent candidates and also gives proportionality at a national level. Because of this, I think it is important that candidates should be able to stand nationally. An independent may have supporters across the country, not just in their geographical constituency. They may have enough national support to get elected in a fully proportional system, but not enough in a smaller geographical area. This is also a problem for minor parties, but at least they can field candidates in several constituencies, whereas an independent candidate only has one of themselves.

From the voters’ point of view, I also think that they should not be limited to the candidates standing in their constituency. The democratic choice available to voters varies wildly from constituency to constituency. Someone might be in a constituency with only mainstream party candidates, or they might be in the Speaker’s constituency and have very limited options, since the main parties have an agreement not to stand against the Speaker – don’t worry about voters’ democratic rights or anything. There might be a few independent or minor-party candidates dotted around the country that someone supports, and it’s just luck whether they happen to live in the right constituency to be able to vote for one of them. A fair system should not favour voters who are lucky enough to live in the same area as a candidate that they support.

We actually have 650 local elections, rather than one national election, and your democratic power is highly dependent on where you happen to live. Having a single national election would be a more democratic system. But this brings us back to the problem of having a local MP to contact.

MPs vote nationally so don’t just deal with local affairs. If I haven’t voted for my local MP (currently in my case the Conservative Brooks Newmark), then I don’t feel represented by him, and he’s not really the person I’d want to write to. I think we should be able to write to an MP that we consider to represent us, not just the one who just happens to be the most local. I will discuss the idea of local MPs further, however, in the context of what I am going to propose in the next post.

Our current system tries to elect local MPs and also a national Parliament while using a voting system that’s far too crude to manage it properly. Are you supposed to vote for the candidate who is your favourite at a local level, or the one who represents the party machine that you prefer at a national level? There’s no single correct answer to this, and people are left to their own devices to make their way through a confusing and bizarre system. We need to start again with a new system.

In summary, here are my criteria for a good electoral system:

1. Voters should be able to vote for any candidate from anywhere in the country. It would be a national election.
2. The system should be proportional at a national level.
3. There should be no systematic bias against independent candidates.
4. The ballot paper and voting system should be relatively simple for the electorate, so people can still just turn up on the day and vote as they do now.
5. The system should, as much as possible, favour a reasonable geographical spread of MPs.

And how to implement that is what I’m going to discuss in the next post.

Electoral reform – why First Past the Post is not fit for purpose

In this post I am going to address the shortcomings of our current voting system and the awful attempts to defend it in the run-up to the referendum in 2011.

I have already argued that our current voting system is not fit for purpose, but I want to look at it in more detail, and explain why our system is so bad, as well as present some reasonable alternatives. This will take two or three posts, so settle in.

As a nation, we had the chance to change the voting system through the referendum in 2011, unfortunately without success, but this is still an important issue and not one I intend to drop. The voting system we have in the UK to elect our Parliament is commonly known as “First Past the Post” (FPTP). Interestingly, among people who have an academic interest in voting systems, FPTP has virtually no support whatsoever. For example, it received no support in this survey.

Under FPTP, if two similar candidates stand, then it decreases both of their chances of winning against a candidate with distinct policies, through vote-splitting. A voter can only cast a single vote for one or the other, so has to choose between them. Because the votes are split between the two candidates, they are each harmed by the presence of the other.

Often a minor candidate, with no chance of winning themselves, can change the result by taking away votes from one of the favourites. A candidate who affects the outcome in this way is known as a spoiler candidate. Here is a nice video explaining some of the main problems with FPTP.

FPTP keeps minor parties as minor parties by encouraging people to vote for their favourite of the top two in the polls or risk wasting their vote. So it entrenches the two-party domination, according to Duverger’s Law. It encourages strategic rather than honest voting by putting people off from voting for their favourite candidate, and can even put lesser-known candidates off from standing at all. The two main parties become like two brands that dominate the market, regardless of quality. So I see it as a very undemocratic system.

Even when the two-party stranglehold fails and other parties do get a look in, FPTP can produce pretty bad results anyway due to the vote-splitting that I’ve just discussed. FPTP isn’t equipped to deal with more than two candidates in a reasonable manner.

Alternative Vote (AV), is less vulnerable to the problems that plague FPTP. Under AV, you rank the candidates in order, and candidates are successively eliminated, so if your favourite is eliminated, your vote still carries on into the next round. Ranking a minor candidate top does not render your vote irrelevant.

Even though David Cameron and the Conservative Party were against the change to AV, it is a very similar system to how Cameron was elected as their leader. David Cameron, David Davis, Kenneth Clarke and Liam Fox all stood in the first round, in which Kenneth Clarke was eliminated as he had the fewest votes. Liam Fox was then eliminated in round two, and finally David Cameron beat David Davis in a head-to-head.

The AV referendum failed arguably, at least in part, due to the circumstances at the time. It was seen as a Liberal Democrat policy, and they, along with leader Nick Clegg, were very unpopular at the time, especially after university tuition fees debacle.

Also, many of those who wanted electoral reform didn’t want AV. Nick Clegg himself had previously referred to it as a “miserable little compromise”, which was then used against him.

On top of this, we had to put up with a lot of rhetoric, nonsense and outright lies from the “No” campaign, which I will now look at in a bit more detail.

This video says that implementing AV would cost £250 million. However, this claim has been discredited – see here and here. In any case, £250 million is still less than £5 per person in the UK. I’d be willing to pay more than that for a better system of democracy.

Here is the “AV Sports Day” video posted by the Conservative Party. There’s a lot of build-up in this video, but essentially what happens is that a boy wins a running race but the trophy is awarded to someone else. A voiceover then says “Under AV, the person who finishes third could end up winning.” This isn’t just a bit of political distorting of the truth. It is, plain and simple, an outright lie. It is true that the candidate who would have finished third under FPTP could win under AV, but this just means that the systems aren’t the same and don’t always elect the same winner. If they did always elect the same winner, there would be no point in changing. Equally, if a country was voting on whether to change from AV to FPTP, someone could make the same argument against that change. “Under FPTP, the person who finishes third [under AV] could end up winning.” To be clear, under AV, there would be a winner. That person would not have finished third in any sense because AV would be the system being used and they would be the winner under that system. The argument boils down to “Different systems can elect different winners.”

The racing analogy is poor, and the name “First Past The Post” is misleading. The system is nothing like a race with a finishing post. There is no post. Whoever gets the most votes wins. A finishing post implies a set distance, or set percentage of the votes. A more analogous situation to FPTP would be a race that ends as soon as the total distance reached by all the competitors added together reaches 100 metres (or any particular distance). That would be a bit of a weird race. In fact, it is AV that is more like a system with a finishing post. If no-one reaches 50% of the votes (the post), there is another round, and it continues until a candidate has more than 50% of the votes in a round, among the voters who have participated in that round (expressed a preference among the remaining candidates).

This is the most watched video posted by the official “no to AV” campaign, and it has most of the main points that were used against AV.

It starts off with Alan B’stard (the main character in 1980s and 1990s sitcom “The New Statesman”, played by the late, great Rik Mayall) announcing some ridiculous policies and then saying “And of course, with AV, even if they don’t vote for me, I’ll probably still get in.” This is mindless rhetoric.

“You know, the really great thing about a fudged coalition is that neither of us needs to carry out a single promise of our election manifesto.” Well, it turns out that we have a coalition now under FPTP. But I will come back to this at the end because I have a bit more to say about hung parliaments and coalitions.

“Of course the best thing about AV is never having to say we’re sorry or to pretend that we care.” More mindless rhetoric.

The video then has a horse race, which makes the same argument as the sports day video. There is then a confused teacher trying to explain AV to some confused students in an overly complicated manner, when clearly it isn’t complicated in the least.

It then has typed on the screen “Democracy is One Person One Vote.” However, I would take “One Person One Vote” to mean that everyone can cast a vote of equal power: i.e. some people don’t get to vote twice with others just once in the same election. If you take it to mean “Name one candidate then shut up” (a phrase I got from this site) then I fail to see why that’s a principle you’d want to hold onto. It results in just about the least expressive system possible. There may be several candidates standing, but you only get to say anything at all about one of them.

There are no serious points in this video. It’s bad for FPTP as a system that they had no sensible arguments and had to rely on rhetoric, nonsense and lies to defend it. The video is nothing short of moronic.

In this video, David Cameron argues that supporters of fringe parties get their votes counted more times than voters of mainstream parties, because once their first preference is eliminated, their later preferences are also counted. On the other hand, a mainstream party may not get eliminated at all so, according to Cameron, their supporters only get the one vote. However, this is also nonsense.

If there isn’t a majority winner in the first round, the candidate who has the fewest first preferences is eliminated and it goes to a second round. Everyone’s votes are counted a second time, or at least everyone who has indicated a preference among the remaining candidates. Those whose favourite candidate has been eliminated cast a vote for a different candidate in the second round, but everyone else is still casting a vote in this second round. For those whose favourite candidate hasn’t been eliminated, the two votes both happen to be for the same candidate. Everyone has one vote per round. And as said, this is very similar to how David Cameron was elected as leader of the Conservative Party. Those who voted for Kenneth Clarke in the first round and then had to change to one of the remaining candidates did not get more votes than those who were able to stick with their original favourite in the second round.

Having made you watch those horrific videos, here and here are couple in favour of AV to redress the balance.

AV is not a proportional system, so if it’s hung parliaments and coalition governments you’re worried about, then it wouldn’t make much difference. Majority governments are largely the result of a single-winner system, which AV and FPTP both are. Not that I’m a fan of majority governments anyway – I favour proportional representation – but I’ll address that in the next post.

Fewer people vote for the main two parties now than used to, so we’re more likely to get hung Parliaments than in the past even under FPTP, which could help explain why the 2010 election ended with a hung parliament. The graph here shows how the proportion of the vote going to the top two parties has declined over the last few decades. Here is an interesting video by Dan Snow on hung parliaments, which also makes the point that hung parliaments are more likely nowadays even with FPTP. FPTP has been shown to be an awful system for electing an individual to a single position, but its saving grace was supposed to be that it’s good at ensuring a party majority when many of these elections (650 or so) are put together. However, it has been shown to fail here too, such as in the 2010 general election.

If having a majority government is such an important thing, then we could indeed do a lot better than 650 individual elections each with a winner decided by FPTP, such using as a system that actually guarantees it. We could have a system that automatically gives the most popular party a majority of the seats. For example, we could have a “mixed-member” system. In a mixed-member system, you vote for a local candidate and also a party. The local candidates are elected as usual, but then there are “top-up” seats that go to parties ensuring that each party gets the right number of seats overall based on the national votes for that party. Normally, mixed-member systems are used to ensure proportional representation, but it could just as easily be used to ensure an outright majority for one party. By the way, I wouldn’t support this system, partly because I’m not so keen on majority governments, but it just goes to show that even if majority governments were an important criterion for you, you could do a lot better than FPTP.

This post has largely been to show what a poor system FPTP is, and also how poor the “No” campaign was in the 2011 referendum, relying as it did on rhetoric, nonsense and lies. I haven’t devoted too much time to showing why AV is better (although if you watch the videos that I have already linked to (once more here and here), you will clearly be able to see it) because while I want electoral reform, I would not advocate using AV given a free choice. So in the next post, I’m going to talk more about where I think we should be going. I’m going to make a case for proportional representation.

AV lost against FPTP in the 2011 referendum, but electoral reform per se didn’t lose. Regardless of the result of that referendum, FPTP is still as awful a system as ever, and we need to continue the fight against it.