Posts Tagged ‘alternative vote’

My Dilemma With The Alternative Vote System

Sunday, February 20th, 2011

With the impending referendum on the Alternative Vote System, it seemed appropriate to rehash my thoughts on this area.  I had my first foray into considering the election system during the General Election in May 2010, and having favoured the first-past-the-post system was roundly criticised for my lack of real thought, which was harsh but perhaps fair.  I followed this up with a couple of blogs in August on different voting systems in the first and a discussion of the key issues relating to electoral systems in the second.  Since then I have continued to mull over the issues in my mind and it leaves me with a dilemma.

While first-past-the-post (FPTP) is not a fair system, however (in my opinion) the alternative vote system (AV) is not much advance on it in terms of fairness.  In fact, although AV is arguably fairer than FPTP, it is perhaps not worth the effort in changing from the devil you know for a new system unless it is palpably better and AV really is not good enough.

Conversely, we do need a fairer voting system in England and proportional representation is the only sensible system, with the single transferable vote method (STV) being the best.  I like the proportionality in STV and the idea of multiple members for enlarged constituencies.

So what to do?  Can AV be regarded as a good stepping stone that moves the electoral debate a smidgeon forward? Or is it a sop that will legitimise a new status quo centred on either AV or even worse FPTP, without the argument moving into the territory of proportionality and a fairer voting system?

In the end it is a gamble, or at least a game of tactics, in which I remain undecided and perhaps worse relatively unexcited. 

I feel there are bigger issues than the voting system: is the “old” Western world losing out to the new and rising stars in the East? Has Britain wasted its capital by investing its wealth in non-productive assets and putting itself in hock with its creditors who all sit in the Middle East, India and China?  Are we wasting our intellectual capital by letting them get involved with consultancy and investment banking rather than building the UK into something great?  Does the taxation system force Britons to focus on short term returns rather than investing for the longer term to build/rebuild equity and productive assets?  And so on…or should we fiddle like Nero as Rome burns.

Electoral Systems For The UK (Part 2)

Sunday, August 15th, 2010

…This follows on from last week’s blog

Are these systems used in the UK?

First-past-the-post (“FPTP”) is the system that has been central to British voting in elections for many years and remains the status quo method for General Elections in the UK and local elections in England.  Alternative Vote (“AV”) is the main system used in Australia and for by-elections in the Northern Ireland Assembly, and is the proposed alternative to FPTP being proposed for UK General Elections.  Party Lists (“PL”) is used in Britain for European Elections, i.e. to elect MEPs.  Single Transferrable Vote (“STV”) is used for most elections in Northern Ireland and has been used for local elections in Scotland since 2007, as well as being the main system used in the Republic of Ireland.  STV is also the preferred system of the Electoral Reform Society.

Is it easy to understand?

Complexity is one of the arguments used to argue for the status quo.  However, I feel that it is an intellectually arrogant position to hold and which effectively says most of the electorate is too dumb to understand some of these alternative systems, so we must not use them; football league tables, betting odds, the rules of cricket, Strictly Come Dancing and how to use a mobile phone are all mysteries to me, but no-one has ever said that they should be stopped. 

From my point of view, the FPTP is undoubtedly the simplest system, but it is also a result the crudest – a truly blunt instrument.  However, I understand the rationale for AV, PL and STV and what these electoral processes are aiming for, and I can work out simple scenarios for how individual constituencies could play out, even if I would not want to work out the detailed algorithms.  Therefore, while the detail can get a bit complex, I do not think that the concepts are that difficult, and isn’t a bit of sophistication in order for such an important influence on our day-to-day lives?

Linkage from representative back to constituency

For me, this is an important one, and even though many MPs have no genuine affinity back to their constituents, it remains for me one of the crucial strengths of FPTP, and so would be of AV.  However, while I originally felt this was a make-or-break point, I now feel that it is really a matter of balance, so I would not want to see vast multi-member constituencies as in Israel where there is one constituency for the whole country nor would I want to see a closed party list where voters cannot make a choice for a particular individual, albeit we don’t really ever know that much about them in the UK anyway.  Whereas I am no longer averse to having two or three member constituencies that better reflect the political views of most of the people within a geographic area.

Are votes valued?

This question covers a number of issues, but at its heart is a very important problem – while we are all told it is important to vote, most members of the electorate believe that their vote has no real influence on which party wins the election.  This is because under FPTP the winner takes all, even if the winner gets less than 50% of the votes (in fact almost all MPs are elected on less than 40% of the vote); therefore, the views relating to the “losing votes” are discarded and their votes “wasted”.  Therefore, even if a party consistently wins 20% of the electoral votes across the country, it may get no MPs into parliament if it comes second or third in every seat.  This results in general dissatisfaction with the whole political system, but also creates many of the undesirable side effects of FPTP:

  • Small amounts of votes can have big impacts on election results, so many policies are targeted towards capturing these marginal “floating” voters rather than the greater body of public opinion, while boundary changes take on special political poignancy and can encourage “gerrymandering”;
  • Tactical voting can become the order of the day, whereby voters vote against who they dislike rather than for whom they actually want, i.e. voting becomes negative rather than being a positive choice.

Therefore, if voting is so important, it then must follow that everyone’s vote should be valued.  As a result, FPTP cannot be advocated as the most desirable system, because it ignores the majority of votes in almost all constituencies.  The questions then move on to which system best balances the need to give value to each vote, and so most voters political viewpoints, while keeping some physical link back to a geographic place.

There are many detailed points for and against individual systems, but the above are the key criteria for me, and I feel that the debate boils down to the following key questions:

  • Do the various proposed systems work? Yes
  • Can the logic of the systems be explained in relatively simple terms? Yes
  • Are most votes valued? No for FPTP, but yes for the other three
  • Is there a link from a geographic location to representative? Yes for FPTP and AV, and can be for PL and STV
  • Should we have a single member constituency or multi-member? Now this is the real question and this is where the political debate should really be, rather than on which system is best.  My own view is that we should have multi-member constituencies of three MPs which would give all parties the potential to get a seat in each constituency, so each part of the country would be worth fighting for.  More than this and you start to lose the linkage back to constituency.  But in the end it becomes a matter of individual judgment.

The big negative against PL and STV seems to be the argument about unstable governments and that you do not get a definitive result for one party.  However, my counter-arguments would be that surely it is more important to have votes that have value and are not wasted than governments that are voted in on low percentage votes of the electorate, and that the coalition in Britain at present happened under the FPTP system and it seems an eminently mature and sensible bunch of politicians.  My biggest issue with the STV and AV system is that I personally do not think that your second or third preference votes should have as big a weighting as your 1st preference, but then there follows a hair-splitting debate about by how much?

So let’s look at a practical example.  For my own benefit, I have assumed that you merge my three local constituencies and I have used the 2010 results:

2010 results for Thirsk, Skipton/Ripon and Harrogate/Knaresborough

  Thirsk/Malton Skipton/Ripon Harrogate Total
Conservative     20,167 53%     27,685 51%     24,305 46%     72,157 49%
Liberal Democrat       8,886 23%     17,735 32%     23,266 44%     49,887 34%
Labour       5,169 14%       5,498 10%       3,413 6%     14,080 10%
UKIP       2,502 7%       1,909 3%       1,056 2%       5,467 4%
Liberal Democrat       1,418 4% 0% 0%       1,418 1%
BNP 0%       1,403 3%       1,094 2%       2,497 2%
Independent 0%          315 1% 0%          315 0%
Youth 0%            95 0% 0%            95 0%
Currency 0%            84 0% 0%            84 0%
      38,142 100%     54,724 100%     53,134 100%   146,000 100%

The first thing you notice are the variations in number of voters, however Thirsk & Malton was a quirk in that this constituency was more like a by-election in that voting was one month later, and so after the result of the General Election in 2010.

The second point is that while it is strongly Tory in this rural area, the Liberal Democrats do get a very good section of the electorate and are especially strong in Harrogate & Knaresborough.  So if you were to divide the enlarged constituency up to give 3 MPs, you would definitely give 1 to a Conservative and another to a Liberal Democrat, giving each one-third an MP to voice their political views, whereas currently you have 3 Conservative MPs.

The final point is what do you do about the third MP.  Now that’s where you need to get a mechanic that is fair in the distribution of the final chunk of votes.  Under STV, the balance of Conservative votes over the threshold (36,500) would be transferred to other candidates, which would go where?  There’s the rub, as they might actually all go to UKIP rather than Labour.  Under the PL system, I would have thought that you would get 2 Conservative MPs and 1 Liberal Democrat MP, but please correct me if I am wrong there.

Overall, I am pleased that I have looked in more detail at these different electoral systems as my point of view has changed.  Whereas I was an advocate of FPTP, I now feel that it is a broken system that must be changed.  However, I also think that this referendum is a waste of time, because while the sop is that this is potentially the start of changes to the electoral system, I feel that the questions being asked are wrong and do not really address the core issues.  Furthermore, I do not think that the detailed mechanics of the electoral system is actually something that should go to a referendum, rather it should be hammered out, debated and equations worked out by a committee of experts.

I think a referendum is needed, but that the question should be different, but absolutely fundamental to how Britain is governed.

All the major electoral systems have been devised and work, plus many of them are practised in the UK and other parts of the world.  Similarly, all systems have their issues, but none of them insurmountable, and while they are interesting for politicos, they are pretty boring for most people and (I believe) not crucial to the debate.  Therefore, whichever system is chosen can probably do a good job, so long as fair and sensible criteria are set for determining which system to chose.  So the electorate should not debate the intricacies of each system, but they should be asked to set the agenda for the bureaucrats.

So the question comes down to what should be the brief.  I feel that some of this has already been debated by the 1998 Jenkins Committee, which was set the following eminently sensible criteria:

  • The maintenance of a geographic link between MP and constituency
  • The need for stable government
  • The desire for broad proportionality
  • An extension of voter choice

I am not convinced by the last point as I feel that voter choice is pretty wide already, rather the issue is that, because of lack of proportionality and wasted votes, minority views do not get representation.  So I would change extension of voter choice to “minimisation of wasted votes”.

So you might ask what is there left to debate by the electorate.  Well there is one fundamental question and I feel this is the key question: 

  • “Does the electorate want multi-member constituencies, or not?”

We all want fair elections.  We all want our votes to mean something.  But the key systemic debate is should we have single or multi-member constituencies.  And while I believe multi-member constituencies would help fairness and proportionality, it would be a big change, from which would flow how best to run an election. 

A vote for single member constituencies would mean a debate between FPTP or AV, while for a multi-member system, the debate would be PL or STV.  Once you have decided on this key point, therefore, it becomes simply a matter of mechanics, so while the Electoral Reform Society prefers STV over the PL system, both work, are fair and provide proportionality, so would be better than the status quo.

My own view is for three member larger constituencies, but thereafter I am not especially concerned about whether we vote via the PL or STV system, so long as these work, which they do.  I find PL easier to understand, but am really fairly ambivalent between PL and STV.

Electoral Systems For The UK (Part 1)

Sunday, August 8th, 2010

I was roundly castigated for being a political ignoramus with my first look at electoral reform, which was probably sound.  However, far from being deterred, I still want to continue to try and understand the debate in spite of the heckling, and see if I can get to grips with the issues, arguments, and general blah! blah! blah! about this crucial issue. 

I am not sure that I have progressed much further, but here is how I see it; I ask for some patience as you read it as it will be posted over a few blogs rather than just one, with the first being about the main types of voting, then the next a general discussion and my conclusion.

Nuts And Bolts Of The Voting Systems


First-Past-The-Post (“FPTP”) is the current system in the United Kingdom.  With FPTP, you divide the country up into as many constituencies as you want representatives (i.e. one representative per constituency), then get voters to make their choices and the elected representative in each constituency is the one that gets the most votes, however small the margin between first place and second place.

Alternative Vote

The Alternative Vote (“AV”) system is used in Australia for its House of Representatives and most of the Legislative Assemblies of it States and Territories.  AV is an extension of FPTP in that you still divide the country up into as many constituencies as you want representatives, so AV still results in one representative per voting region, but it enables voters to express their preferences for alternative representatives if their own initial choice cannot win.  So when voting, you rank the candidates in order of preference until you can no longer express an opinion, ranking them 1, 2 etc.  When the votes get counted, if one of the candidates gets a clear majority, i.e. someone has been ranked 1st by more than all the other 1st preferences combined, then they are elected.  However, if no-one has a clear majority, the count starts analysing the preferences of the weaker candidates: in reverse order, you take the candidate that came last and then determine the 2nd preference, i.e. alternative vote, for those who voted for the bottom candidate and allocate those to the remaining candidates and recount, continuing this process until one candidate has over 50% of the votes, and so becoming the elected representative.

Party Lists

Proportional Representation is what I always called the Party List system per the Electoral Reform Society.  PL is, also, the system already used in England, Scotland and Wales for electing MEPs, taking an open party list approach.  The basics of PL are simple: starting with a multi-member constituency, the voters vote, then you work down the list of votes cast to elect representatives in proportion to how many votes each gets until all the representatives’ positions have been filled. 

The complexities in this relatively natural system relate to how you actually construct the system:

(i)                  Voting lists – these can be open or closed, i.e. you vote for candidates who are named on the ballot paper (“open party list”), or you vote for a political party without knowing who the candidates are (“closed party list”), then after the election the political parties work out which candidates they want to represent you in their order of preference;

(ii)                Shape of the constituency – in PL, all constituencies are multi-member, and the larger the constituency and so the greater the number of potential representatives, the more proportional the end result, i.e. the more representative the MPs are of the voters’ actual voting preferences;

(iii)               Minimum voting percentage – most countries (except for example the Republic of South Africa) set a minimum threshold that the minority parties need to exceed before they can get any representation, which seems to be predominantly in the range of 1.5% – 5%. 

(iv)              The final wrinkle is how you actually calculate the number of seats to give each party, which (while fundamental to the actual system) is largely irrelevant to the discussion of the best voting system as the detail can be decided afterwards through an analysis of the various mathematical calculations together with a bit of political horse trading.

Single Transferrable Vote

Single Transferable Vote (“STV”) is used in Northern Ireland for Assembly, European and local elections, most elections in the Republic of Ireland and local elections in Scotland.  In 1917, STV was, also, chosen by the House of Commons for roughly half of constituencies with the remainder to use AV, but this never passed through the House of Lords and was dropped.  It is, also, the preferred choice of the Electoral Reform Society.

STV works on the basis of multi-member constituencies with representatives found via a quota system; representatives are determined by calculating a quota that successful candidates must reach to be elected for each constituency and then working out those candidates that get over that threshold in the constituency. 

Under STV, voters put a number “1” against their first choice, a “2” against their second choice and so on until they no longer have any views.  They can stop at any point, so do not need even to make a second choice.  All the valid ballot papers are then counted up and the threshold calculated as the number of valid ballot papers divided by the number of people to be elected plus one.  So per Electoral Reform Society, “with 100 ballot papers and 3 places to be filled, the quota would be 25”, i.e. 100 ÷ 4 (3+1).

Next, the votes have to be allocated to candidates and to available places to be filled.  This is done by sorting the ballot papers firstly into first preferences.  If any candidate has more first preferences than the quota they are immediately elected.  The next stage is to transfer any surplus votes for those elected candidates, being the difference between a candidate’s actual vote and the threshold, i.e. if I got 33 votes, then 8 of my votes could be transferred to other candidates.  But to prevent the argument as to which votes to transfer, all my votes are transferred but with a reduced value per vote and then allocated to second preference candidates.

After allocating all second preferences, the votes are counted up again and you see who has passed the threshold and then allocate them to places.  If they have not, then the candidate with the fewest votes is excluded and their votes transferred to voters’ second preferences.  This process of reallocating surpluses and excluded candidates’ votes, plus re-counting continues until all the places have been filled in descending order.

Interestingly, the Electoral Reform Society suggests reallocating the whole of any surplus from the first round, but only the last batch of surpluses from later rounds.  It views this as a matter of expediency, which seems bizarre as it inconsistently claims that taking a bit more time and using a computer programme is a small price to pay for the increased fairness and complexity of STV in the first place.  Also, I am not sure why this is called Single Transferrable Vote as while an elector does have a single vote, it is not really that single vote that actually counts as it can be reallocated after being counted once (albeit at a reduced value), while it may, also, be your second or third preference that is chosen.

Hybrid Systems

There are various hybrid systems ranging from AV+ to Total Representation, however these seem to be overly complex and do not really improve on the systems as above.  I have not considered them further, but you can find out more about all the systems at the Electoral Reform Society.


It is interesting that the Coalition is including a question about constituency sizes in the referendum questions for 2011, and that the Labour Party is getting itself into a tizz about this second question.  Also, being in an area that has been boundary-changed twice in the last 15 years, I feel particularly sensitive about this issue.

Having thought a bit about electoral reform now, the nature of your constituency is vital.  To make it fair, each elected representative must relate to almost equal numbers of constituents, so (taking into account movement of people) the shape and size of constituencies should be checked every 2 elections or 10 years.  Secondly, the size of the constituency must not get so large that constituents become so diverse that their very specific local issues get lost in the bigger picture.  So it is a balance between number of constituents, equality and overall size.

That is easy in principle, however you then need to make sure that there is some geographic logic to it as people (or at least me) feel a regional kinship to certain places and geographic regions and you must pay heed to these.  For example, you could create a huge constituency of Yorkshire, but I feel no linkage to South Yorkshire and the issues for Sheffield are not the same as for little old Ripon.  On a more micro scale, we are now part of the Harrogate constituency having been part of the Vale of York, yet my issues are rural, small town rather than those of Harrogate which are more suburban and looking towards Leeds, so for me Vale of York was better.

Also, and I will come to this towards the Summary & Conclusion stage, the referendum question should perhaps be simplified to one of single member or multi-member constituencies, rather than what the voting system is itself.

Voting Itself

This is another key issue.  In the end, not enough people are engaged in the political system often enough, which then causes questions of legitimacy of elections.  Which is the most popular political party when most people do not vote, even though politicians impact all our lives hugely?  Is a government’s mandate legitimate if turn out is low?

It would be possible to make it a legal obligation for people to vote, however this has implications on personal freedom.  The only way I can see that us, the people, will become more engaged in politics and care about voting is if politicians engage with the electorate more, respect them more and make the political system smaller, less bureaucratic and recreate it on a more human scale rather than being a huge, amorphous beast that has no master and no heart.

Discussion follows in next blog…