?

Log in

No account? Create an account
Polls, both geeky and political - Many a mickle maks a muckle

> Recent Entries
> Archive
> Friends
> Profile

November 1st, 2003


Previous Entry Share Next Entry
11:58 pm - Polls, both geeky and political
First, let's wrap up last week's "geekiest hobbies" poll. 21 votes, mean 5.10, median 5, sigma 2.56, mode 4, no significant difference in voting patterns from men vs. women, from Potter fans vs. non-fans or from people who have met me vs. those who haven't. I think this is more insightful "how I am perceived" data than the old "sum up your impression of me in one word" routine. 15 votes from men and 6 from women, though - a clearly unrepresentative skew.

Because technicalities must alas count, congratulations to you modal folks - Dan, Mark, Rob, Mike and Carl - who are absolutely correct with your estimate of four. I note that it should be three; after all, I only ever played Pokémon once, and it was with someone else's deck. When it comes to counting your geeky hobbies, gotta catch 'em all.

4/10 is not a lot; it's less than I'd've hoped. I would criticise the original list of ten for being very mainstream and American in its definition of geekiness. A British counterpart list would surely include pastimes such as DIY, home brewing and - yes - transport enthusiasm. It seems that all my geeky games-related overenthusiasms are too small to make an impression. Surely lots of little geekinesses should add up to one great big geekiness? Better luck next time, I guess.


As much as I like the three ladies on my Friends list who are proud supporters of the Conservative Party, as much as I feel that some of their policies might be preferable to those of other parties, as much as I feel tradition and history have a place in politics, I still want to see the Tories implode messily and permanently. I suspect that the probable selection of Michael Howard as leader will represent very little change from the current situation, except removing the Betsy Duncan Smith vulnerability. I am taking an interest in the current party political machinations, though my commentary is mostly here in malachan's journal and behind a friends-locked entry in karen2205's journal. I shall adapt and excerpt the latter.

Economically, I'm definitely fairly strongly statist, mostly through a low opinion of people's decision-making ability. Socially, I am somewhat less authoritarian and indeed strongly libertarian on a few issues. I would be likely to prefer a new party which was overtly more statist than those which exist to today's mainstream three. (I would also strongly consider voting for a party which decided to officially take a secular standpoint, though I'm not sure I would favour a violent and messy disestablishment of the Church of England.) I think both the major parties are undesirably broad churches. I would prefer to see smaller parties which were free to take more radical stances, with coalition government being ruled by consensus. Not enough to make me want to move to Italy, but enough to make me admire the model to some extent.

I believe I've always voted Liberal Democrat in the general elections, but I know I've voted for councillors from all 2.51 major English parties over the years, as well as the local independent who has a reputation for being out for himself only. (Our ward had a Conservative councillor who died quite recently who was very kind to Mum in her enquiries about roadside trees - that's my excuse for the blue vote.) Not really convinced that my reasons for LD support are necessarily that sound; the first time I can remember taking an interest in the concept of politics was being attracted to the Alliance party on the grounds that (a) I thought that the concept of alliances in general and an alliance party was cool and (b) I generally approved of the rough concept of centrism - (b) may well have been encouraged by such hard-hitting and incisive political treatises as Goldilocks and the Three Bears. Happily the LDs shift left-of-centre (relative to the other parties...) has roughly coincided with mine.

The snappily-named qwghlmBlog points to a very entertaining Guardian column from Wednesday, proposing one B. Johnson Esq. for the Conservative leadership. While much of the article is left-wing haw-hawing at an opposition in trouble, the situation does warrant it just a little and it's written entertainingly. Nevertheless, in the anteanteantepenultimate paragraph, a fascinating and disturbingly plausible theory of "narrative politics" is discussed: voters, their minds shaped by television and cinema, are drawn to candidates whose triumph represents the biggest plot-twist or most extraordinary final scene. Could there be something to it? A worrying trend.

While we're on politics, Tom Watson points to London Elects, detailing the city's European, Mayoral and Assembly elections taking place simultaneously on June 10th. Entertainingly, all three have different voting systems. Tom and many others also point to the BBC's iCAN, a site aiming to gather activists for local political campaigns. Putting the two together, I think I know what campaign I want to start: Fair Representation for Small Parties in the London Assembly Election.

The European election is run by a slightly complicated system. It's not very complicated, but I wouldn't call it easy. In a country where 15,000,000 adults lack basic maths skills, it's at least a little bit tricky.
British MEPs are elected by a system of proportional representation called the Regional List system, using the d’Hondt formula.

All European states are divided into regions. Each region elects a set number of MEPs to the European Parliament. The United Kingdom is split into 12 regions. For the 2004 election, London is likely to elect nine MEPs.

Each political party prepares a list of candidates ranked in order to match the number of seats to be filled in that region. Once votes are counted, the first seat is allocated to the party or independent candidate with the highest number of votes. If an independent is the highest then the seat is allocated to that individual. If the seat has been allocated to a party it will go to the first candidate on that party’s list: that party’s total is then divided by two and the second seat is allocated to the next party or independent candidate with the highest number of votes. The process continues until all seats have been allocated.

With the list system, the number of MEPs elected for each party roughly reflects the share of the votes for the party or independent candidate received in that region.
Now the London assembly election is slightly more complicated still.
There are 25 members of the London Assembly: 14 Constituency Members, and 11 London-wide Members. The London Assembly is elected using an electoral system known as the Additional Member System (AMS). This system combines elements of First Past The Post and a form of proportional representation using the d’Hondt formula.

Voters cast two votes: one for a Constituency Assembly Member and one for the independent candidate or political party they would most like to see represented in the Assembly on a London-wide basis.

Constituency Members each represent one of London’s 14 Assembly constituencies. They are elected by the First Past The Post system, where the candidate with the most votes in each constituency is elected. If there is a tie lots are drawn by the constituency returning officer.

If all Assembly Members were elected in this way, some independents or parties whose votes were spread right across London might not win any individual seats. All the people who had voted for the candidates who were unsuccessful in the constituencies would have no representation in the Assembly, making it unrepresentative of London as a whole.

So, voters cast a second vote for an independent candidate or party. These votes are counted and then the number of constituency members is topped up with 11 additional London-wide members, using a modified d’Hondt formula.
The d'Hondt formula operates as above. However, the modification adds one very important exception which isn't given very much publicity at all. From Fact Sheet 6 (PDF file), Parties must gain a minimum of 5% of the votes to be included in the contest. In short, if your party gets 4.9% of the vote, you don't get a single seat in the Assembly via the top-up method. Given that there are 25 seats in the assembly in total and the intent is (roughly) for each seat to represent one twenty-fifth of the votes cast, it seems most unfair that a party getting more than one twenty-fifth of the votes cast cannot earn one twenty-fifth of the seats.

Now this may sound like a terribly technical quibble. To an extent, it is. However, let's look at the actual results from the 2000 election and see how the top-up seats were applied in practice.

                        Cons    Labour     LibDem     Green   Seat goes to:
Constituency seats won     8         6          0         0
Round One    1st sum     8+1=9     6+1=7      0+1=1     0+1=1
             2nd sum 481,053/9 502,874/7  245,555/1 183,910/1
             Total    53,450    71,839    245,555   183,910   Lib Dem
Round Two    1st sum                          1+1=2
             2nd sum 481,053/9 502,874/7  245,555/2 183,910/1
             Total    53,450    71,839    122,778   183,910   Green
Round Three  1st sum                                    1+1=2
             2nd sum 481,053/9 502,874/7  245,555/2 183,910/2
             Total    53,450    71,839    122,778    91,955   Lib Dem
Round Four   1st sum                          2+1=3
             2nd sum 481,053/9 502,874/7  245,555/3 183,910/2
             Total    53,450    71,839     81,852    91,955   Green
Round Five   1st sum                                    2+1=3
             2nd sum 481,053/9 502,874/7  245,555/3 183,910/3
             Total    53,450    71,839     81,852    61,303   Lib Dem
Round Six    1st sum                          3+1=4
             2nd sum 481,053/9 502,874/7  245,555/4 183,910/3
             Total    53,450    71,839     61,389    61,303   Labour
Round Seven  1st sum               7+1=8
             2nd sum 481,053/9 502,874/8  245,555/4 183,910/3
             Total    53,450    62,859     61,389    61,303   Labour
Round Eight  1st sum               8+1=9
             2nd sum 481,053/9 502,874/9  245,555/4 183,910/3
             Total    53,450    55,875     61,389    61,303   Lib Dem
Round Nine   1st sum                          4+1=5
             2nd sum 481,053/9 502,874/9  245,555/5 183,910/3
             Total    53,450    55,875     49,111    61,303   Green
Round Ten    1st sum                                    3+1=4
             2nd sum 481,053/9 502,874/9  245,555/5 183,910/4
             Total    53,450    55,875     49,111    45,978   Labour
Round Eleven 1st sum               9+1=10
             2nd sum 481,053/9 502,874/10 245,555/5 183,910/4
             Total    53,450    50,287     49,111    45,978   Cons
Top-up seats won           1         3          4         3
Total seats won            9         9          4         3


However, let's look at the same data again, this time without the stipulation that a party needs to get 5% of the vote to be eligible to win a top-up seat. For the sake of screen width only, let's change that 5% to 2½%. Using the numbers of top-up votes cast quoted by the BBC:

                                                             Christian    British
                                                              People's   National
                        Cons    Labour     LibDem     Green   Alliance      Party   Seat goes to:
Constituency seats won     8         6          0         0          0          0
Round One    1st sum     8+1=9     6+1=7      0+1=1     0+1=1      0+1=1      0+1=1
             2nd sum 481,053/9 502,874/7  245,555/1 183,910/1   55,192/1   47,760/1 
             Total    53,450    71,839    245,555   183,910     55,192     47,760   Lib Dem
Round Two    1st sum                          1+1=2
             2nd sum 481,053/9 502,874/7  245,555/2 183,910/1   55,192/1   47,760/1
             Total    53,450    71,839    122,778   183,910     55,192     47,760   Green
Round Three  1st sum                                    1+1=2
             2nd sum 481,053/9 502,874/7  245,555/2 183,910/2   55,192/1   47,760/1
             Total    53,450    71,839    122,778    91,955     55,192     47,760   Lib Dem
Round Four   1st sum                          2+1=3
             2nd sum 481,053/9 502,874/7  245,555/3 183,910/2   55,192/1   47,760/1
             Total    53,450    71,839     81,852    91,955     55,192     47,760   Green
Round Five   1st sum                                    2+1=3
             2nd sum 481,053/9 502,874/7  245,555/3 183,910/3   55,192/1   47,760/1
             Total    53,450    71,839     81,852    61,303     55,192     47,760   Lib Dem
Round Six    1st sum                          3+1=4
             2nd sum 481,053/9 502,874/7  245,555/4 183,910/3   55,192/1   47,760/1
             Total    53,450    71,839     61,389    61,303     55,192     47,760   Labour
Round Seven  1st sum               7+1=8
             2nd sum 481,053/9 502,874/8  245,555/4 183,910/3   55,192/1   47,760/1
             Total    53,450    62,859     61,389    61,303     55,192     47,760   Labour
Round Eight  1st sum               8+1=9
             2nd sum 481,053/9 502,874/9  245,555/4 183,910/3   55,192/1   47,760/1
             Total    53,450    55,875     61,389    61,303     55,192     47,760   Lib Dem
Round Nine   1st sum                          4+1=5
             2nd sum 481,053/9 502,874/9  245,555/5 183,910/3   55,192/1   47,760/1
             Total    53,450    55,875     49,111    61,303     55,192     47,760   Green
Round Ten    1st sum                                    3+1=4
             2nd sum 481,053/9 502,874/9  245,555/5 183,910/4   55,192/1   47,760/1
             Total    53,450    55,875     49,111    45,978     55,192     47,760   Labour
Round Eleven 1st sum               9+1=10
             2nd sum 481,053/9 502,874/10 245,555/5 183,910/4   55,192/1   47,760/1
             Total    53,450    50,287     49,111    45,978     55,192     47,760   Christian People's Alliance
Top-up seats won           0         3          4         3          1          0
Total seats won            8         9          4         3          1          0


You see? It makes a difference in practice as well as in theory. I thought it would only actually make a difference in artifically-constructed degenerate cases which rely on far more parties getting small proportions of the vote, but it actually makes a difference in the real-world case of the 2000 election. The 55,192 top-up votes cast for the Christian People's Alliance - fully 3.3% of the total - were completely ignored. If they had received, say, 82,000 top-up votes, or 4.9% of the top-up votes cast, then they still wouldn't have won a top-up seat, despite being eligible for one as early as round four. Evidently all votes are equal, but some are more equal than others.

There's an argument that the requirement for a minimum 5% of the vote is to guard against extreme parties getting any influence. While I have little love for any overtly religious party, whatever the denomination, and by far less love still for the far-right BNP, it seems positively antidemocratic to me to be prepared to discard up to 4.9% of all the top-up votes cast for each small party - indeed, in 2003, about 15% of all the votes. The CPA deserve a seat on the London Assembly fair and square; I would argue that a London Assembly with 8 Conservative, 9 Labour, 4 Lib Dem, 3 Green and 1 CPA would be more accurately representative of the votes cast than the actual figures of 9 C, 9 L, 4 LD and 3 Green. (For the record, there would have had to have been fifteen seats awarded by the top-up method, or 29 seats in total, before the BNP would have stood to win one.)

What I would advocate for fairer representation would be to scrap this minimum requirement. By way of demonstration that this works, the European elections use a pure d'Hondt formula; there is no requirement for a minimum proportion of the vote in order to be eligible to win a seat within a region. Indeed, this is exactly how the UK Independence Party managed to win themselves three seats in Europe. (On the other hand, the regions are a little smaller - no larger than eleven representatives per region.) It would be interesting to see how different the results would have been from a single Great British constituency electing 84 representatives.

I can't recall hearing anything about this at the time, but I'm sure people must have pointed it out - not least the Christian People's Alliance themselves. I have to say I was pretty shocked when I found this out. Am I the only one?


4/10 indeed. Now that's geekery.

  • LJMaps meme? A poor man's GeoURL. Let's have automatic but overwritable userinfo-quoted-location-to-GeoURL conversion instead as a generalised solution and make our pretty maps based on that instead.

  • Sneaking in just under the wire, a wonderful thirtieth to daweaver! I hope he won't mind me praising the way he wrote a 39,000 word autobiography for keeps in a month, which (word count emphasis aside) is at least a para-NaNoWriMo acheivement in my view. It's his to share (or not) as he will, but it's one of the most ambitious pieces of writing, whether fiction or otherwise, I've seen on LiveJournal yet and fulfils its considerable promise remarkably well.
    Current Mood: shockedshocked
  • (17 comments | Leave a comment)

    Comments:


    [User Picture]
    From:amuzulo
    Date:November 2nd, 2003 03:01 am (UTC)
    (Link)
    Wow, looking at the top 10, my ex-girlfriend (the one I met at a board games convention) regularly participates in #1 LARP's, #6 Vampirism and #9 RPG's. She actually almost convinced me to join a LARP too. Scary! I still miss her though. Oh well. We still hang out occasionally though... I should email her. hehe :)
    [User Picture]
    From:karen2205
    Date:November 2nd, 2003 04:36 am (UTC)
    (Link)
    If you want to see my entry and can't, then email/shout/comment....whatever, and I'll add you to my Friends' list.

    /me loves voting system geekery;-) I have my own objections to using closed list systems at all, but some other time.....

    And I will come and argue the merits of state ownership, vs. private ownership at some stage. Unfortunately LPC work calls right now.
    [User Picture]
    From:jiggery_pokery
    Date:November 2nd, 2003 12:51 pm (UTC)
    (Link)
    I look forward to it! You are likely to find that my opinions are considerably more pragmatic and less idealistic than my biases, as an unfortunate consequence of the real world. Nice to dream, though.

    Have to admit that I don't really very know much about electoral reform other than that proposed by the Electoral Reform Society. It would be very useful for me to see a number of other perspectives on the matter and see whether I still think the same way in the face of opposition.
    [User Picture]
    From:hedwig_snowy
    Date:November 2nd, 2003 09:15 am (UTC)
    (Link)
    You'll always get 10 out of 10 from me.... :-) So, did you calculate whether American hobbies are more OR less geeky than British hobbies? I haven't seen any 65-yr-old men here collecting garden gnomes in at least a month...... :)
    [User Picture]
    From:jiggery_pokery
    Date:November 2nd, 2003 12:59 pm (UTC)
    (Link)
    So, did you calculate whether American hobbies are more OR less geeky than British hobbies?

    Interesting question, but quite a different one. There are doubtless hobbies which would be seen as geeky in the USA but mainstream in the UK and vice versa; for instance, there is a very large (mid-five digits of attendees?) annual festival at (I believe) the East of England Showground in Peterborough each year which celebrates all things American in the UK. I would be amused and relieved to be made aware of the existence of a counterpart event in the US.

    I also think it's possible to identify geekiness and perception as a function of hobby size - I would estimate all of the hobbies listed, except possibly the furry fandom, to have 105-107 self-identified participants in the US alone, so a 103 size hobby like the game show fandom - and that's a generous estimate - doesn't really make the scale.

    It's not all about size, though. Compare the Renaissance Faire, the Society for Creative Anachronism and Live-Action Role-Playing for geekiness. It's arguable, particularly if you are one of the people for whom the differences matter more than the similarities, which is the geekiest, but I have my own take.

    As ever, I do not necessarily regard geeky as a bad thing.
    [User Picture]
    From:addedentry
    Date:November 15th, 2003 02:03 pm (UTC)
    (Link)
    Do you know the name of the pro-American festival? I should love to attend.

    In principle, anyway; in practice, monster trucks and stetsons would be a turnoff.
    [User Picture]
    From:jiggery_pokery
    Date:November 15th, 2003 02:15 pm (UTC)
    (Link)
    Americana International.

    I was right about there being 40,000 visitors over the weekend, though that may well be ten thousand on each of four days. I was wrong but close about the venue: not the East of England Showground in Peterborough, but the County Showground in Newark. (In the East Midlands, at least.)

    The original concept behind the event was to combine an American car show with live music entertainment. The first show was based in a pub car park in Bramcote, Nottingham back in 1980 which attracted 80 cars, a few hundred visitors and included two bands.

    During the next few years, the venue for the show was Wollaton Park until 1985 when the event moved to Newark Showground in order to accomodate the increasing numbers of exibitors and visitors. Apart from the 1995 show at Donnington Park, the event has remained at this excellent location.

    The event attracts around 40,000 visitors a year and hosts around 5,000 American motors. There have been some fantastic live music acts perform over the years including about 75 U.S. acts over the last 10 years.


    You should love to attend, but you wouldn't.
    From:themightyuser
    Date:November 2nd, 2003 10:18 am (UTC)
    (Link)
    I'm pretty sure "Played Real Life Quidditch" has got to be high on the "geek-o-meter"... and I'm pretty sure "Was Head Judge and Organizer at a Real Life Quidditch Tournament" has got to be much higher. =D
    [User Picture]
    From:jiggery_pokery
    Date:November 2nd, 2003 01:15 pm (UTC)
    (Link)
    I say roll on version 3. (Or there's always the FAPQ.)
    From:daweaver
    Date:November 2nd, 2003 11:07 am (UTC)

    You know what's coming...

    (Link)
    in the anteanteantepenultimate paragraph

    Surely "in the avantpreantepenultimate paragraph..." Did Gordon Burns announcing the second-to-last head of The Krypton Factor as "the preantepenultimate programme of the series" pass you by?

    British MEPs are elected by a system of proportional representation called the Regional List system, using the d’Hondt formula.

    A picky git writes: It's a bastardised version of the d'Hondt formula. Normally, divisors would increase by 2, so we'd be dealing with 1/1, 1/3, 1/5, 1/7 etc. Using these larger divisors tilts the balance towards the big parties. Plus ça change.

    For the 2004 election, London is likely to elect nine MEPs.

    For those who missed the Electoral Commission's announcement on Friday, the UK loses nine seats next year, so that the European Parliament doesn't grow too massive with the New Ten coming on board.

    The new seat distribution

    Region               Now  Next  Change Electorate
    East Midlands         6    6      0     3,223,412
    Eastern               8    7      -1    4,132,682
    London               10    9      -1    5,060,066
    North East            4    3      -1    1,921,576
    North West           10    9      -1    5,195,366
    South East           11   10      -1    6,086,113
    South West & Gib      7    7      0     3,830,811
    West Midlands         8    7      -1    4,011,017
    Yorkshire & Humber    7    6      -1    3,746,989
    Wales                 5    4      -1    2,226,999
    Scotland              8    7      -1    3,892,646
    Northern Ireland      3    3      0     1,072,000
    Total                78   87      -9   44,399,677
    


    It would be interesting to see how different the results would have been from a single Great British constituency electing 84 representatives.

    You know what's coming next, don't you. Information tabulated by David Boothroyd

    The parties gaining over 5000 votes nationally:

    Conservative			3,578,219	36
    Labour				2,803,821	29
    Liberal Democrat		1,266,549	 9
    UK Independence			696,057		 3
    Green Party			625,378 [1]	 2
    Scottish National Party		268,528		 3
    Plaid Cymru			185,235		 2
    Pro-Euro Conservative		138,097
    British National Party		102,647
    Socialist Labour Party		86,749
    Liberal Party			59,239
    Scottish Socialist Party	39,720
    MEP Independent Labour		36,849 [2]
    Liberal				33,812
    Alternative Labour List		26,963
    Natural Law Party		20,329
    Socialist Alliance		7,203
    


    [1] Includes "Scottish Green Party"
    [2] Stood in West Midlands only. Boothroyd does not record the candidate's name.

    Using the 1-2-3 divisors, the results are as follows:

    Conservative			32 (-4)
    Labour				26 (-3)
    Liberal Democrat		11 (+2)
    UK Independence			6  (+3)
    Green Party			5  (+3)
    Scottish National Party		2  (-1)
    Plaid Cymru			1  (-1)
    Pro-Euro Conservative		1  (+1)
    British National Party		0
    


    Using the 1-3-5 divisors, we get these results:

    Conservative			30 (+6)
    Labour				24 (-5)
    Liberal Democrat		11 (+2)
    UK Independence			6  (+3)
    Green Party			5  (+3)
    Scottish National Party		2  (-1)
    Plaid Cymru			2  (nc)
    Pro-Euro Conservative		1  (+1)
    British National Party		1  (+1)
    Socialist Labour Party		1  (+1)
    Liberal Party			0
    


    The Liberals miss the 84th seat by 427 votes.
    [User Picture]
    From:jiggery_pokery
    Date:November 2nd, 2003 12:47 pm (UTC)

    I hoped for nothing less!

    (Link)
    Did Gordon Burns announcing the second-to-last head of The Krypton Factor as "the preantepenultimate programme of the series" pass you by?

    <h1>BLUFF</h1>

    The price of a lost point would be well worth the convincing video evidence to confirm the truth. :-)

    Many thanks for this. I shall make you are sent a copy of my forthcoming book, "How to get other people to write your LiveJournal for you". (It will, naturally, be a blank exercise book and a free pencil.) Seriously, glad you found it interesting.

    GB one-region EU vote: well, I think this clearly demonstrates why the suggestion is a poor one. When some parties only stand in certain regions, such as the Welsh and Scottish nationalists, they deserve to have their votes counted only in that region.

    I'd be curious to know whether the Cornish factionalists are putting up any sort of a presence in SW & Gibraltar. I was also amused to encounter the concept of "South West and Gibraltar" for the first time - there's a cute piece in an online newspaper from Gibraltar about it. Interesting to see that the Isle of Man is not represented at all and not too bothered about it, either.

    Very interesting point about the different divisor schemes - the Wikipedia entry on the subject separates the 1-2-3-4 divisior d'Hondt method as being distinct from the 1-3-5-7 Sainte-Laguë method. However, better still, in the Electoral Commission's report (36k PDF), they explicitly recommended the 1-3-5-7 Sainte-Laguë method over the 1-2-3-4 D'Hondt method even without being asked to consider it. Hurrah!

    I find this all fascinating, but I also recognise that I'm actually quite uninformed about electoral reform and will tend to parrot the Electoral Reform Society's recommendations fairly verbatim without knowing the reasoning behind them. It would be interesting to know whether there is much debate within the ERS about the various methods that exist or whether the opinions are clearly held with little disagreement. By analogy, is there an Electoral Reform Society Reform Society?
    From:daweaver
    Date:November 3rd, 2003 10:23 am (UTC)

    Re: I hoped for nothing less!

    (Link)
    A major correction: the 1-3-5 divisor gives only 83 seats, not 84. Therefore, please chalk up the Liberal party with the final seat.

    BLUFF

    Is the correct answer; though Mr Burns often referred to the last heat as the antepenultimate show, he never went one prefix further.

    GB one-region EU vote: well, I think this clearly demonstrates why the suggestion is a poor one. When some parties only stand in certain regions, such as the Welsh and Scottish nationalists, they deserve to have their votes counted only in that region.

    And yet... under one system, the Welsh nationalists would lose nothing, the SNP one, and both parties can only pick up votes from expat Welsh / Scots in the rest of the mainland. I suggest that the SNP would gain the extra votes for a third seat.

    I'd be curious to know whether the Cornish factionalists are putting up any sort of a presence in SW & Gibraltar.

    They didn't last time, and they'd need everyone in Cornwall to vote for them to stand much chance of getting a seat.

    in the Electoral Commission's report (36k PDF), they explicitly recommended the 1-3-5-7 Sainte-Laguë method over the 1-2-3-4 D'Hondt method even without being asked to consider it. Hurrah!

    Good for them. 1-2-3 favours big parties. I was going to make some comment about the choice of divisor being a contentious issue, as it does skew the results towards the large (or small) parties. Some have suggested using fractional divisors, such as 1-2.7-4.4-6.1, but that smacks of tilting the system in favour of one party, and probably reading too much into one set of results.

    The 1-3-5 divisor will tend to produce a result that converges rapidly to the ratio of total votes. Add up the parties in descending order, until we reach the party that does not receive half a quota. Here, that's the top eleven parties, and there were 9,810,519 votes cast. Divide that by 84 to get 116791.8929 votes per seat. Then divide each party's vote by that quota figure, to give:
    Party			Quotas	Sainte-Laguë seats
    Conservative		30.63	30
    Labour			24.00	24
    Liberal Democrat	10.84	11
    UK Independence		5.959	6
    Green Party		5.354	5
    Scottish National Party	2.299	2
    Plaid Cymru		1.586	2
    Pro-Euro Conservative	1.182	1
    British National Party	0.878	1
    Socialist Labour Party	0.742	1
    Liberal Party		0.507	1
    

    The quota figure is the target we'd be aiming for, with the least unsigned total difference. Here, we've a total difference of 2.953 quotas. To minimise the total difference, the Liberals must give their one seat to the Conservatives, with Hague's team having 31 members, and Meadowcroft's none.

    Compare with the 1-2-3 method, which gives the Tories 1.37 seats too many, and Labour exactly 2 extra seats. The BNP and SLP scored most of a quota, but get no seats.

    The actual result, of course, will vary wildly - here it's 23 quotas out, but we're no longer comparing apples with apples.

    Incidentally, this is the reason why I suggest it's vital to vote as much with your heart as your head at the Euro election. Give the vote to the party that comes closest to your ideals, whatever that might be. There is strength in diversity.
    [User Picture]
    From:jiggery_pokery
    Date:November 3rd, 2003 02:28 pm (UTC)

    Re: I hoped for nothing less!

    (Link)
    When I was looking this up originally, I found a "voting methods" mailing list where someone was railing against d'Hondt as leading to corruption (!!?) and proposing the alternative method, IIRC:

    1) Count up the number of votes per party.
    2) Divide the number of votes per party by the number of parties. Round down. Award each party that many seats.
    3) There will be at least one party left over unless the numbers are suspiciously convenient. Award the remaining seats in order of size of the remainder.

    Hence:
    Party			Quotas	First pass seats
    Conservative		30.63	30
    Labour			24.00	24 (unless actual # = 23.9951, etc.)
    Liberal Democrat	10.84	10
    UK Independence		5.959	5
    Green Party		5.354	5
    Scottish National Party	2.299	2
    Plaid Cymru		1.586	1
    Pro-Euro Conservative	1.182	1
    British National Party	0.878	0
    Socialist Labour Party	0.742	0
    Liberal Party		0.507	0
    Seats to award                  6
    The six biggest remainders are 0.959, 0.878, 0.84, 0.742, 0.63 and 0.586, so the final totals are:
    Party			Quotas	Final seats
    Conservative		30.63	30 + 1 = 31
    Labour			24.00	24
    Liberal Democrat	10.84	10 + 1 = 11
    UK Independence		5.959	5 + 1 = 6
    Green Party		5.354	5
    Scottish National Party	2.299	2
    Plaid Cymru		1.586	1 + 1 = 2
    Pro-Euro Conservative	1.182	1
    British National Party	0.878	0 + 1 = 1
    Socialist Labour Party	0.742	0 + 1 = 1
    Liberal Party		0.507	0
    ...which is identical to your least unsigned total difference method. I'm trying to work out whether this method is guaranteed to produce the least unsigned total difference or not and cannot prove or disprove this, but I suspect so. Conclusion: all much of a muchness, really, but it really would make a difference to the Liberal Party. My gut feeling is that I think I actually prefer the "least unsigned total difference" method to the Sainte-Lague method for ease of understanding, despite how darn elegant, fun and just plain geeky the Sainte Lague method actually is.

    On the Cornwall issue, the Tees Valley is made up of five unitary authorities known as borough councils, Stockton, Middlesbrough, Redcar-and-Cleveland, Darlington and Hartlepool. Stockton is situated on both sides of the River Tees, with Thornaby being that part of the Stockton area which is south of the Tees. (Middlesbrough is south of the Tees; Middlesbrough and Thornaby are delimited by the A19 in practice, if possibly not in legislation.) All 8 of the 55 wards in Thornaby were won by representatives of the Thornaby Independent Association, who want to establish a separate Thornaby Town Council as distinct from the Stockton one. (They probably want to campaign for Thornaby to be listed as distinct from Stockton-on-Tees in the phone book once more, too.) Yikes. Despite my biases, I have only a bad feeling about the efficiency of this one.
    From:daweaver
    Date:November 4th, 2003 10:43 am (UTC)

    Re: I hoped for nothing less!

    (Link)
    1) Count up the number of votes per party. 2) Divide the number of votes per party by the number of parties. Round down. Award each party that many seats. 3) There will be at least one party left over unless the numbers are suspiciously convenient. Award the remaining seats in order of size of the remainder.

    This is going to give the same result as my proposed method, and it's probably easier to code and explain. The objective of both methods is to minimise the wasted quota, and allocating the extra seats on the basis of the largest partial quota clearly meets that objective.

    In true maths professor style, I leave the formal proof as an exercise for the reader (;
    From:2ndavemusic
    Date:November 2nd, 2003 11:35 am (UTC)
    (Link)
    Interesting analysis! But why do I always get the feeling I'm not really smart enough to read your posts? :)

    [User Picture]
    From:jiggery_pokery
    Date:November 3rd, 2003 02:58 pm (UTC)
    (Link)
    It's a different world... :-)

    One of the many nice things about LiveJournal is that it lets you get bite-sized samples of many different worlds and existences.
    [User Picture]
    From:addedentry
    Date:November 15th, 2003 02:17 pm (UTC)
    (Link)
    There's an argument that the requirement for a minimum 5% of the vote is to guard against extreme parties getting any influence.

    There is, and it's of a variety that irritates me. If it's considered undesirable for certain (especially antidemocratic) views to be represented, then legislate to prevent candidates expressing those views from standing. Trying to fix it without being explicit about it not only denies unobjectionable minorities a hearing but leads to trouble when the racists do score over 5%.

    A similar argument is made in defence of first-past-the-post: that it ensures strong government by marginalising nth parties for n>3. Nonsense, of course, because the electorate's preferences can't be represented by a single dimension. Remember that the Liberals used to be the natural second party of government; and enjoy this three-way chart [scroll to end].

    Finally, the violent and messy disestablishment will be arranged by the C of E, at the same time as the monarchy destroys itself. That said, both institutions' historical resilience suggests it will be decades rather than years.

    > Go to Top
    LiveJournal.com