www.torfaen4us.co.uk

STV for Torfaen 2012

4th April 2010

STV for Torfaen 2012?

(This document is also available in Word compatable .rtf format )

This article explains why STV (Single Transferable Vote) is better than the First-Past-The-Post (FPTP) voting system currently used in England and Wales.

STV is not new: it has been used throughout Ireland (since 1921) and in local elections in Scotland (since 2007). It’s also the popular choice of many trade unions and “friendly societies”. It’s more complicated than FPTP for the people counting the votes but a lot simpler for the voters - more on this later.

I’ve written this in four sections.
1          Problems with First-Past-The-Post (FPTP)
2          How STV works (with technical details).
3          Why STV solves all the problems of FPTP.
4          Proposals for using STV in Torfaen elections. 

Most of the arguments apply to both national and local government elections - although some are more appropriate to Westminster than Torfaen. Rather than restrict my scope, I have included arguments relevant to all these electoral scenarios.

1     Problems with First-Past-The-Post (FPTP)

(a) In General Elections fewer than 150 “marginal constituencies” change hands and, within these, only the votes for the incumbent and the leading challenger matter. In practice, more than 90% of British voters might just as well stay at home: their votes are irrelevant and will not affect the outcome.

(b) The percentage of “irrelevant” voters is less for local elections but also exceeds 60%. The wonder is that most people still do bother to vote – even when they know it to be a waste of time and effort.

(c)Because at least three candidates contest all Westminster seats (in Wales at least four), the losers normally win more votes (in total) than the winner. This happens also at National Assembly elections and, often, at local government elections. So a majority of voters normally ends up dissatisfied or aggrieved with the results. 

(d) Because FPTP is a winner-takes-all system, a small change in the voting share can make a big change in the seats won. This is universally true (in Torfaen in 2008, one ward switched from 3/0 Labour/others to 0/3 - where a fair result would have been to switch from 2/1 Labour/others to 1/2).

(e) Winner-takes-all also means that it is rare for parties to win seats in proportion to their votes (Proportional Representation). One inappropriate way to achieve PR is to vote for party lists (as in Israel) as this eliminates constituency representation.  

(f) Another bad feature of winner-takes-all is that it encourages parties with similar policies and the same target voters to see each other as the main enemy – instead of as potential allies in coalitions. This makes for unstable government.

(g) Of the several other undesirable aspects of FTPT, I will mention just one – it tends to encourage parties to choose ‘safe’ candidates rather than a representative cross-section of the public as a whole. 

2    How STV works
With STV, you rate candidates in their order of preference – 1, 2, 3, 4, etc. -(until you give up). You don’t have to worry about tactical choices between a not-so-good winner and an excellent no-hoper – the system deals with this by counting your first choice first, eliminating the candidates with the least votes and then counting the second (and third and fourth) preferences, etc
.
This means that no votes are wasted on minority candidates – you can afford to register a protest vote for someone you prefer but don’t expect to win, safe in the knowledge that your second and third preference votes will be counted if your first choice is eliminated. That must be good because it means that the first round count really does reflect what people really would prefer.

The first step is to count the “1” votes – and, if any candidates achieve the so-called “quota”, they are declared elected. I will first explain the “quota” for a single member constituency – although, strictly speaking, that is not what STV is about. 

The “quota”

If only one person is to be elected, the quota is a simple majority - half the votes cast. With the FPTP system, winning candidates can get elected with as low as 30% of the votes. With STV, if no one achieves 50% on the first count, the candidate with fewest votes is eliminated and his/her second preferences votes  (the second preference “2” votes) of the eliminated candidate counted instead.
This process is repeated until – possibly when only two candidates are left – the winning candidate tops the 50% quota. For a single member constituency, STV is better known as the Alternative Vote (AV) system – although this fails to deliver most of the merits of STV that derive from its use in multi-member constituencies.

The common virtue of AV and STV is that a most voters must have voted for a winner (either as their first preference or a subsequent preference) – so less than half go away completely disappointed. So the quota can also be defined as “the number of voters left disappointed”. If a constituency elects 2, 3, 4 or 5 members, the quotas are one third, one quarter, one fifth or one sixth respectively – fewer and fewer being left wholly disappointed by the result. The quota, by definition, is the “total votes cast divided by one more than the candidates to be elected”.

To illustrate the counting process, I’ve created a very simplified illustration of a tally sheet to count STV votes in a hypothetical 4-member constituency contested by 8 candidates.

Counting the votes (see voting tally table below)

There were 1000 votes cast. Because four candidates are to be elected, all four must exceed the 20% quota by one vote (= 201) – and less than 20% of the voters will be left “wholly dissatisfied” (i.e. they didn’t vote for any winning candidate).
In my illustration, no one got 201 on the first count and, as Harry had the least votes, he was eliminated and his second preferences counted. After the second round, Gina had least votes and her next preferences counted. After the third round, Amy reached the quota (precisely – this is unusual, but it simplifies my explanation) and was declared elected whilst, simultaneously, Fred had least votes and was eliminated.

* Nominees of the same political party – see discussion below.

      After the fourth count, Emma had the least votes and was eliminated whilst Bob reached the quota and was declared elected. That left only two candidates still in the running so, in this instance, the fifth count was not needed.

Party candidates

You will note that my simplified illustration makes no mention of political parties. This is because the ballot paper would be essentially identical to that for our current voting system – we vote for individuals by name and their party affiliation is on the ballot paper as a “description”.  With STV, voters are obliged to rank each candidate of their preferred political party - instead of voting for the entire slate.
Suppose, in my illustration, Amy, Emma and Harry were the candidates of the largest political party and could expect 30% of the votes. Against a divided opposition all three would be elected by FPTP but, with STV, only one (Amy) would be. This is a better result because it is more proportional. It also means that the voters – not the party - chose between the party’s candidates. 

Excess votes

Besides the potentially wasted votes for minority candidates, there’s a second and usually more important type of wasted vote in FPTP elections - huge majorities. STV deals with these by redistributing “excess votes” (excess votes above the quota) according to the voters’ next preferences. I ignored this scenario in my illustration by making all the winners hit their quotas exactly. This rarely occurs in practice.
Suppose Amy or Bob, had obtained (say) 251 votes when they passed the 201 votes quota. As this is 50 more than needed, these extra votes would be wasted unless we were to take account of the next preferences of the 251 contributing to this excess.
What happens with STV is that the next preferences of the 251 votes would be counted and redistributed in the ratio 50/201 among the other candidates - so no votes are wasted and every vote has equal value. This is crucial for a political party because, if their votes all went to a leading candidate, these excess votes (over the quota) then contribute again (in proportion) for the counts using the next preference votes.
That’s enough on the technicalities - now to the advantages of STV.

3    Why STV solves all the problems (of FPTP – listed as (a) to (h) previously)

As explained above, STV is designed to achieve a level of proportionality in each multi-member constituency. Although it cannot deliver precise proportionality in every ward and constituency, the overall result using STV will be far better that from the corresponding winner-takes-all FPTP elections (q.v. objection (e) in section 1).

As a corollary, with STV there are no ‘safe’ seats. Even if it is obvious who will top of the poll, there is invariably a contest for the last and usually the last two places. In this sense, every STV constituency is marginal and every vote, in every ward and constituency, is of equal value (q.v. objections (a) and (b) to FPTP).

As to objection (c) to FPTP, the quota concept used for STV means that it is impossible for any but a small minority to be wholly dissatisfied with the outcome.  This should go a long way to reducing the antipathy to elections so prevalent today.
 
With STV, no party can afford to take any constituency for granted and send its flying canvassers away to marginal target seats. The political implications of this are far-reaching – notably that if we had had STV elections for Westminster for the past decades, we would not have witnessed the spectacle of the Labour Party putting all its efforts into wooing middle-class middle-England floating voters whilst ignoring the needs and desires of its core voters in the valleys.

With STV, our votes in Torfaen would have been as important as the 10% of swing voters that the Labour Party has targeted in English marginal constituencies. That has to be good both for us and for democracy – and for the political process.

Finally, a brief comment on two less obvious aspects of STV, deriving from the reality that STV elections are won and lost by the second and third preferences. This fact obliges parties to choose candidates with a wide public appeal and obliges political canvassers be friendly to any possible second preference voters. I’ll explain  the first side benefit of STV with a hypothetical example.

Supposing a constituency has a significant ethnic minority with a traditionally little interest in elections yet thought sympathetic to a party’s policies. It would make sense for that party to choose at least one of its candidates from that ethnic minority. With STV the onus is on parties to choose candidates that maximise their total appeal - not to rely just on its core support (q.v. objection (g) earlier).

Finally, dear to my heart, one of the saddest aspects of politics today is the unreasoning antagonism between activists of parties of a similar persuasion. When a canvasser for Plaid Cymru or Labour meets a ‘lost cause Tory voter’ on a doorstep, they move on quickly. If, alternatively, they meet a potential voter for the other party, they almost invariably hang around to rubbish the perceived competitor party.

In contrast, with STV the more appropriate response is to say something along the lines “yes, we do have similar policies, would you consider giving us your second preference vote?” So, even at grassroots level and before the election results come in, STV has the potential to instil a more positive and constructive approach to politics - in contrast to the “Yah Boo” atmosphere currently prevailing.

4    Proposals for using STV in Torfaen elections

If STV is to be introduced into Torfaen, without a new Westminster law, there will need to be a Torfaen Council decision to instruct the Officers to open discussions with the Electoral Commission to arrange for a pilot STV scheme prior to its potential introduction throughout Wales. Given the success and all-party approval for the use of STV in Scotland, it is quite possible that the Electoral Commission might welcome the opportunity to do so.

The only changes needed would be to make all wards large enough to warrant 3-5 members. Currently our 44 Councillors are elected from seven 3-member wards, six 2-member wards and eleven single-member wards. For STV to work effectively in Torfaen, we need wards of 3-5 members - by merging the single and dual member wards. My suggestion is for nine 3-member wards, three 4-member wards and one 5-member ward.

The mergers could be based just on geography but there are major anomalies - for example, the ratio of electors to councillor range from 1077 (Cwmyniscoy) to 2252 (Llanyrafon South) - more than double. This is because the wards have not kept pace with the the population growth of Cwmbran and relative decline of Pontypool.

To overcome this problem, Cwmbran needs another Councillor and Pontypool one fewer. This is not crucial for STV but seemed worthwhile doing in its own right. It leaves the average number of electors/Councillor and number of Councillors the same but reduces the electors/councillors range to 1213-1990 (still not as good as I would have liked).

My suggested new wards are as follows (other combinations are just as possible):

1) Leave the six of the seven existing 3-member wards exactly as they are.
2) Merge the seventh (Panteg) with Cwmyniscoy to make a 4-member ward (of average size).
3) Merge Greenmeadow & St. Dials to make a 4-member ward.
4) Merge Pontnewynydd, Snatchwood, Wainfelin, Brynwern and Pontypool to give a single 4-member ward with 1578 electors/Councillor.
5) Merge Trevethin & St. Cadocs/Penygarn, Croesyceiliog North & South and Fairwater & Coed Eva to make three new 3-member wards.
6) Merge Llantarnam and the two Llanyrafon wards to make a new 5-member ward (with 1567 electors/Councillor).

Other combinations are possible and probably just as good - but this illustrates that neither the arithmetic nor the geography present insurmountable. The map below shows the proposals graphically.

Submitted by JJR 19th March 2010

Comments invited on chatroom or by email.