Friday, 18 June 2010

List PR

List based proportional representation is the system used by nearly all developed nations in the world, except the Anglo-Saxon nations.

There are variations and hybridised versions that meld with first-past-the-post while keeping the proportionality, but basically modern democracies have list PR.

The most equal, prosperous, democratic, environmentally friendly, politically engaged and least corrupt nations use List PR. The disadvantage is that there is less geographic link between elected representatives and where some toff has drawn some lines on a map. This also means that boundary changes have little effect on the result under list PR systems. The result cannot be gerrymandered like it is with FPTP or STV.

List PR elected parliaments contain more women, minorities and lower socio-economic groups, much more than FPTP. Even after the advent of all women shortlists, only 20% of Westminster MPs are women. In comparison, both the Scottish Parliament and the Welsh Assembly elected by List PR hybrid systems have around 50% female representation.

The so called disadvantages of List PR pale into insignificance when you look at the superior governance countries with List PR enjoy. The higher quality, better funded and better managed public services, the responsiveness and diverse representation within parliament.

Like all PR systems, List PR is likely to result in coalition government not single party rule. Contrary to Anglo-Saxon myth, this actually allows for more decisiveness demonstrated by the speed with which infrastructure projects are agreed and difficult long-term issues like keeping on top of government debt are tackled.

The UK's experience of List PR is the closed list system used for elections of our MEPS to the European Paliament. There are two criticisms that are unfair here when comparing the election of MEPs to the election of MPs to Westminster. For a start there are only 78 MEPs covering the UK compared to 650 MPs so obviously MEPs will have to cover more voters than MPs even before multi-member constituencies. Tory MPs sneer about people not being able to name two of their MEPs, but ask people to name their MP and one of their 3 local councillors and most voters will draw a blank. Also a surprising amount of MEPs are well known. Think of Nigel Farage of UKIP and Caroline Lucas of the Greens in my region - the South East (pre-May when she became an MP).

The other criticism of list PR is that only the party has a say over their elected candidate because they order the list. This is of course true for the closed list system used for the Euro parliament, but very much not true for the open list systems that are in use throughout a lot of the developed world where the voter very much chooses the candidate and not the party. There are also two further rather subtle points that need to be made.

One, because list PR unlike FPTP, gives people an effective choice of much more than just 2 parties, parties have to more democratic, open and transparent about how they choose their canddidates, so even under closed list PR, candidates tend to be more responsive to the electorate and democratically chosen. If parties are not democratic under list PR, it is easy for voters to go elsewhere and still find their vote effects the result.

Supporters of FPTP should not be smug about closed lists and remember that 80% of MPs under FPTP are chosen in safe seats by around 100 party members in a room with little chance of defeat in the general election because their party is so entrenched in that area. There is nothing more closed and chosen by 'the party' than that.

But as already said List PR does not have to be closed list. One of the easiest open list systems is the one used in Sweden, where voters first pick up the ballot paper from the party they support and then mark an X next to the party candidates they want. This way they choose both the party and candidate with just one vote. This is as easy and quick as FPTP voting in terms of both the count and for the voter, but gives completely fair and proportional results. There is none of the complexity of the vote that STV invokes.

I think you might be able to guess which electoral system I think is best after reading this.

Wednesday, 16 June 2010

Single Transferable Vote

In the second of my series on voting systems (after questioning whether first-past-the-post even qualifies as a democratic electoral system), I now turn my unique eye to the single-transferable-vote.

STV is the favourite voting system of a lot of reformers, most influentially the Electoral Reform Society and the Liberal Democrats.

STV is the voting system used in Ireland, Malta, the Australian Senate (upper house) and...err that's it!* (*Oh, to be fair, it is now also used in Scottish and some New Zealand local government elections as well).

STV is listed amongst the family of voting systems described as proportional representation, although strictly it isn't a PR system, although it can deliver reasonably proportional results (if quota is low enough).

STV is a preferential voting system (like the Alternative Vote - AV is in fact STV in a single constituency), i.e. instead of 'simply' putting an X against one candidate as you do with most List PR voting systems or for electing an MP to Westminster with first-past-the-post, you 'simply' list candidates 1,2,3 in order of preference.

Now, it sounds silly, but some voters will have a problem in writing 1,2,3 on the ballot paper. The spoilt rate will go up, but we are talking about very small numbers here. For those who oppose STV, you would be scraping the barrel using this argument.

I suppose the biggest difficulty with preferential voting is that it requires a higher amount of knowledge amongst the electorate - always a dangerous assumption. I do however believe that people can rank their Green or UKIP preferences with a bit of practise).

One problem is the rules on how many candidates to rank. If you say ALL the candidates have to be ranked (like in Australia) you increase the spoilt ballot rate and 'donkey vote'(those going eeny meeny miny moe), whereas if you say people can rank as many or few as they like, you give a more powerful vote to those who know 'politics'. Unfortunately this tends to be the more wealthy. We could get into arguments about how voting per se favours the wealthy, but lets stick to STV.

Like FPTP, STV has a strong geographic link, but unlike FPTP that is not the ONLY thing that matters. Where you draw the boundaries can have a huge influence on the result under FPTP turning a narrow defeat into a landslide. This is less of a problem with STV, but it is still a problem. The degree of the problem depends on how proportional you decide STV should be. This depends on the number of MPs per constituency which determines the quota. This is where a brief description of how STV works is needed. You see, what the voters have to do is easy, the counting process however is another matter.

Firstly, it takes much longer to count preference vote systems (this also applies to the Alternative Vote but lets be honest, having to wait a few extra hours for a result is hardly a big problem). It takes longer because you might have to count the same ballot paper seven or eight times as 'surplus' votes of winners and losing candidates votes are re-allocated.

Basically pretty much every voter is guaranteed a vote that counts but this might not be their first preference, it might be their second, third or Nth preference depending on how many candidates there are. In each constituency, a pre-determined number of candidates are elected. You might decide five MPs is a good number for an area. This determines the 'quota' which would basically be the total number of votes divided by six (one more than the number to be elected), plus one vote - meaning a candidate would need around 17% of the votes to be elected. In practical terms, the quota does the same job as the 'threshold' does in a list PR system. It creates a barrier for smaller parties to get over for them to win any seats at all, so parliament is not overly fragmented. In list PR systems a reasonable threshold is usually judged to be around 4-5% as this limits parliament to no more than 8 or so parties represented.

In STV systems the quota threshold can be higher because we are not just counting first preferences, so for example, Greens are elected in Ireland despite their first preferences only reaching 5% because they pick up a lot of second and other preferences to take them over 17% in some areas.

Gerrymandering possibilities are more difficult under STV, but Ireland shows it can be done. One outrageous way available is to follow the Irish example of having 3 winners in some areas and 5 in others, this means the quota is different in different areas - an obvious way to alter the value of the vote. If you are to have an unequal voting system this is the way to go about it. But the boundaries like in FPTP are also a way of cheating. The fact that where you draw the boundaries is such an easy way to cheat in FPTP was one of the reasons I gave for it being semi-democratic and why it is the favourite system of dictators from Zimbabwe to Taiwan and this also remains a problem with STV, but less severe.

STV is however, fairer than FPTP to smaller parties who are not geographically concentrated and is more proportional. However it tends to be the third placed centre parties than benefit the most, which might explain why the Lib Dems like it so much.

One of the big disadvantages of STV that it shares with FPTP is how it under-represents women, minorities and lower socio-economic groups. This makes me very suspicious of STV. List PR tends to do wonders for women and lower socio-economic group representation and this is reflected in the fact that developed countries with list PR tend to have much lower inequality than similarly developed countries with FPTP or STV.

To conclude STV is better than FPTP, much better, but still has some serious deficiencies. It is not my ideal system, but I would knock on many doors if we the people were ever given an opportunity to choose it over FPTP.

Tuesday, 15 June 2010


This is the first of a series of posts with my unique take on different voting systems. I start with the favourite system of dictators the world over - first-past-the-post.

First-past-the-post is used to elect councillors to local government in England and Wales and of course to elect our MPs to the UK Parliament in Westminster. It is also used in Canada, India and the US, and a host of other semi-democracies and dictatorships around the world.

For those who don't know how it works, in general elections the country is divided into a number of areas called constituencies with each electing one MP. The voter is provided with a ballot paper with an alphabetical list of candidates standing in their constituency next to the name of the party they are standing for (if any) and the voter 'simply' places an X next to the candidate they want to vote for. The voter is allowed one vote. The candidate with the most votes in that constituency is elected its MP.

In local government a smaller area called a ward is used and usually (for no reason that I can ascertain) 2 or 3 councillors are elected for each ward. The voters are allowed the same number of votes as the number of councillors to be elected in their ward - so put Xs by the 2 or 3 candidates they are voting for. Depending on how many are to be elected, the councillors who come in the top 2 or 3 places in number of votes become councillors.

This all sounds beautifully simple and democratic and we can see why voters on first inspection might think there is no need to change any of this. Only when they realise that most votes have no impact on the actual results and that most voters can spend their entire lives electing no-one, do people start to think something is not quite right.

People listened with incredulity at this election when they were told that the Lib Dems could top the popular vote but get only half the seats of the Tories in second place and less than half the seats of Labour in third.

Such a result is obviously absurd and people immediately think that this must be down to differences in electorate size between constituencies (because this is the only unfairness the Tory press highlight - David Cameron did it again this week at PMQs comparing the Western Isles with 30,000 constituents with the Isle Of Wight with 110,000 - you will hear a lot about the Isle of Wight because it is so unique - indeed it was the Tories who opposed it being split into two constituencies and made smaller at the recent boundary review before the 2010 election. A cynic might wonder whether this example is so useful for them that that is precisely the reason they have kept it so large. There are only a handful of seats that have more than 80,000 constituents, the target is 68,000, it makes little difference to the bias in the system - which is explained below)). But constituency sizes are actually already relatively equal (despite what is implied by the Tories) and they could actually be completely identical in size and produce the same ridiculous results.

It is the system itself that discriminates mainly against those parties unable to get over 30% of the vote and this accidental bias can also happen between the two big parties this system favours, no matter how hard you try to draw the boundaries fairly. Of course, if the 'gerrymandering' is deliberate the possibilities for unfairness are even more enormous. Enlarging the boundaries as the Tories are proposing to do, is one way of making gerrymandering easier.

Where you draw the boundaries can have a huge influence on the result under FPTP - the gerrymander wheel demonstrates how by just moving the boundaries around a 'winning' party can win 80% of the seats or 40% of seats without changing a vote or moving a single voter and crucially without altering the size of constituencies either, which always remain completely equal. For the smaller parties even 20% of the vote can vary between 0% and 20% of the seats by just changing boundaries. This is the power of the boundary commission under FPTP.

Which brings us to the crux of the unfairness about first-past-the-post - boundaries. There are 'independent' quangos instructed by parliament to do this 'fairly'. As already noted this is an impossible task even with the best will in the world, but inevitably there is tremendous pressure from political parties as to where these boundaries are drawn and the big two parties dominate proceedings. The current rules dictate quite sensibly that county boundaries, geographical, administrative and social ties should also be taken into account when drawing up boundaries - so for instance wards should be wholely contained within constituencies and not divvied up. This makes sense because people vote on a variety of issues including how well their local council is performing. This all helps keep the 'constituency link' real, and this is supposedly one of the main strong points in FPTP's favour. There are many considerations that have to balanced when designing an electoral system and one of these is a link between a geographical area and an MP - the 'constituency link'. Peculiarly FPTP puts this consideration far above ALL other considerations including the fairness between representation and votes nationally. Geography is the only factor that really counts under FPTP, even who wins the popular vote nationally is unimportant. Yet even this geographical link is illusory as where you draw the boundaries is so important and yet so arbitrary - most people do not realise how important the boundary commission is for our democracy - politicians know how important it is though. FPTP is a strange beast indeed as we shall come to see.

The Tories in particular (and also Labour MPs in safe seats) regularly trundle out this mythical link as a strong point of the system. Yet it is the Tories that want to enlarge boundaries weakening accountability and the Tories also want to disregard the current boundary rules that give the link any semblence of credibility. New constituencies will not be tied to administrative boundaries or within city limits. You may want to throw out your MP, (or support them), only to find they have been 'moved', even after just one parliamentary term. This makes a mockery of being able to 'throw them out'.

Your vote will no longer be able to show what you think of the party running the local area, because your MP's constituency could easily be moved elsewhere by the next election. The Tory proposals mean the regularity of change and fluidity of where the boundaries will be drawn in the future will destroy any pretence of a constituency link. All of this is being done at a large cost to accountability. It will be much harder for MPs to build local support (it is already pretty clear that national issues and party leaders have a bigger impact on voting behaviour anyway but it is still just possible to build support locally to some extent under the current boundary sizes).

We need to remember that since 1945, the adult population has grown by a third, but the number of MPs has remained the same. So constituency size has risen from a manageable 50,000 per MP to around 75,000 and will go over 85,000 if the Tories have their way. Imagine wanting to stand as a candidate and being faced with the huge task of 40,000 doors to knock on - it would take many years for an MP to meet all their constituents. Face to face contact is perhaps the only way the smaller parties and independents can make an impact.

Under the new rules, boundaries will change more regularly and smaller parties will be particularly hard hit as canvassing an even larger area will take even more resources. The Tories are doing all this for one reason only - 36% of the vote in this election 'only' gave them 47% of the seats.

This might sound pretty generous towards the Tories, but as any Tory will tell you, 36% of the vote gave Labour 55% of the seats in 2005. The only way the Tories can achieve this same distorted result in their favour is to to drastically change and enlarge the boundaries. I hope people following this can see what is being done here and it hasn't got anything to do with fairness and democracy. To say a system is 'unfair' to a party that gets 47% of the seats with 36% of the vote is beyond parody. The real unfairness is obviously that 35% of people turnout to vote for parties other than Tory or Labour but only get around 10% of the seats in parliament to represent them (and even that representation is skewed - for instance, there is only one MP (Caroline Lucas - Green) in England who is not Tory, Labour or Lib Dem, despite around 10% of votes going to other parties and independents).

Why does Labour do even better out of this system than the Tories? I don't want to bore you with too much detail on this but basically there are 3 factors that have increased bias over the long term - vote concentration, differential turnout and surburban drift, in that order of importance.

Firstly, Labour's vote is more concentrated - this is crucial if you want to win seats under FPTP.  Labour waste less of their votes in seats they cannot win - mainly in rural areas. The Tories problem is they have significant amounts of votes in southern urban areas, the North of England and Scotland, but not enough to win many seats in these areas - this effectively wastes millions of their votes.

Secondly, urban turnout is much lower. Labour win most of their seats in urban areas and this lower turnout means they get less votes in doing so. Tories can pile up masses of votes in rural seats that they will win easily anyway.

Thirdly, and this is the only argument you will hear from the Tories about 'unfairness'. Urban constituencies tend to 'shrink' over time and rural seats 'grow' as people move out to the suburbs and rural areas. This means that without more frequent boundary reviews (there are major reviews every 10-12 years), urban constituencies will tend to be smaller (by about 4,000 potential voters on the electoral register (or around 5-10% less than more rural seats). This actually is by far the smallest factor in why Labour do better, making maybe 5 to 10 seats difference nationally. The other two factors are far more important making around 50 seats of the difference (Labour can get the same vote share as the Tories but around 60 seats more. If you think that is unfair compare the Tories to the Lib Dems where equal shares of the vote for these two would deliver the Tories around 200 seats more and no jiggling of the boundaries will change that). Note also that because registration is lower in urban areas - the actual number of over 18s is actually probably fairly similar between Tory and Labour seats. In fact Labour seats tend to have more people overall because they contain more under 18s. This is probably why the Tories oppose votes for 16 year olds and want to introduce more stringent registration procedures which will deter over 18s from registering - the lower the registration the bigger the constituency size as boundaries are drawn on registered voter numbers not by how many over 18s entitled to vote it contains). The MPs with the biggest constituency workload are in urban areas because they have more poverty, social problems AND more people. Tories don't win these seats and so would not appreciate this.

The actual main advantage of first-past-the-post probably has nothing to do with the 'constituency link' at all, it might actually stem from one of FPTP's perceived weaknesses. The fact that FPTP reduces the number of voters that can affect the overall result to a few thousand voters in key marginals and that modern targeting techniques can find them, might mean it is easier for the main parties to argue their case in more detail. Perhaps having fewer voters that count will allow for a more informed 'electorate that matter'. This is a doubtful argument, but probably the only one that is legitimate when talking about the advantages of our present voting system.

Under a more proportional system these swing voters would be in every area of the country,  and run into millions, making it harder for the parties to win their vote. Under first-past-the-post, a few thousand voters in a certain type of area can be 'targeted' with numerous mail, telephone and face to face canvassing. It could be argued that maybe a higher level of debate could ensue, although the mushiness and risk averse policies of the major parties under FPTP undermines that argument. Maybe more proportional systems do encourages more radical debate even though more voters need to be persuaded.

If we were to keep FPTP, one way of improving it would be to have more MPs and smaller constituencies not less and bigger. Maybe go back to 1945 when each MP had no more than 50,000 constituents compared to the current 75,000. MPs would have more time to contact constituents and be closer and potentially more accountable. Smaller parties and independents would stand more chance of bucking the national media trends. It would also mean that the executive would be proportionally smaller in parliament and backbenchers better able to hold them to account. Instead in the name of democracy, the Cameron Tories are proposing the opposite. I wonder why that is then?

Thursday, 10 June 2010

It is First Past The Post That Is Weak And Indecisive.

This Blogger thinks he has found the perfect electoral system - basically tweaking STV to avoid coalition government, but is coalition government as bad as the right-wing mythmakers make out it is?

Did you know that the countries with the smallest levels of borrowing have coalition governments? And the highest levels of democracy. I would also argue coalition government is more decisive not less. Think of the wartime coalition in this country but also look at the long-term infrastructure in European countries that have had long-term coalition government - look at the quality of their public services - their healthcare, their transport infrastructure etc. Look at the speed with which decisions are made compared to our parliament (half unelected) where important decisions can take years to pass. But most important of all, look at their much higher levels of public engagement in politics and the representation of all groups proportionally in parliament. You must also remember that all parties are coalitions - think of a Labour supporter voting for Jeremy Corbyn in the same party as Alan Milburn, or the 1922 back bench Tories in the 'Monday Club'. I think it better the public decides the proportions each coalition gets - the negotiations we saw after this election is more open than party manifestos decided by a few party members behind closed doors. People rarely vote on manifestos anyway. I have a feeling people know more about their government's policies after the election than they did of the party they voted for policies before it.

Saturday, 5 June 2010

FPTP, AV, PR, Hung Parliaments Now Likely Whatever Voting System Is Used

Via Mark Thompson's blog I have been alerted to this excellent examination of FPTP (first-past-the-post) by Prof John Curtice in Parliamentary Briefing.

John is the psephologist from Strathclyde University who in 2005 said the most likely result of the next election would be a hung parliament.

Basically, it is getting increasingly difficult to achieve one party rule under first-past-the-post for 3 main reasons - 1. the declining share of vote of the two main parties. 2. the dwindling number of marginals, and 3.  the ineqitable treatment of the 2 main parties.

John's analysis sadly doesn't go on to analyse the boundary changes the Tories are planning. The enlargement of constituencies will make it less rewarding to vote other than the 'big two' but as John points out the main change is regional difference between North and South and this is virtually impossible for the Tories to gerrymander away.

A lot is made of the 'unfairness' to the Tories of 'only' winning 47% of the seats on 36% of the vote. Wrongly identified as the cause of this is that their constituencies have on average 4,000 more registered voters than Labour seats, but the clue there is in the word 'registered'.

When the lower registration of voters is taken into effect, Labour constituencies actually contain more potential voters and also have much more people because they have larger under 18 populations, not to mention higher workloads for MPs having to deal with the higher poverty and associated social problems.

The Tories with Lib Dem assistence are planning to make registration harder thus exarcerbating these problems and to ignore the fact that unregistered voters make Labour constituencies bigger. Lower turnout in poorer Labour seats is also the other reason why Labour can get less votes than the Tories and still not lose as many seats as Tories would on such a low national vote. 

The real big problem for the Tories is the fact they cannot get more than 40% of the vote and most of that is concentrated in rural and suburban areas of England. In reality the Tories need to win some Billy Bragg types over to get large majorities, this seems unlikely to ever happen. We still remember the 80s.

The Tories want more marginals but seem unwilling to sacrifice some of their votes in their ultra-safe seats which is the only way of achieving this. It will also be interesting to see how they cull the number of MPs - Tories will want only non-Tory MPs culled - whether that is achievable I doubt.

I am going to be looking at each voting system in more detail over the next few months in the run up to the possible referendum. Starting soon with the glorious first-past-the-post system.