Spreadsheet of the week & related topical stuff

1. Paul Lockett, who 'blogs here has set up a fine spreadsheet which tells you how many seats each party gets for a given number of votes in a multi-member constituency under the d'Hondt system, i.e. the system used at last week's elections to the EU Parliament. (Rest of paragraph deleted as Paul has now honed it to perfection - see comments).
---------------------
2. On that note, I hereby declare last week's Fun Online Poll closed. The result was a dead-heat between those who preferred single-member consitituencies and those who preferred multi-member constituencies (the options '1' and '5' having been the product of two earlier elimination rounds).
---------------------
3. Bearing in mind the smokescreen that our Prime Minister intends to put up today on the subject of voting reform, and having taken everybody's objections to Proportional Representation into account, how about this for a plan, which I posted at Jock's earlier:

I rather like the idea of multi-member constituencies ('MMC') with one-man-one-vote for one individual candidate, you start by merging (say) three existing constituencies areas into one larger constituency (but retain the sub-constituencies, as we will see later).

The three candidates with the biggest individual shares win a seat. If a party is confident it will get a huge share of the vote it is free to put up two or more candidates*, but most smaller parties would put up a single candidate and hope for the best. That keeps the ballot slip nice and short.

Then the candidates have to sort out between themselves who will represent which geographical sub-constituency, on the basis that the one with most personal votes has first dibs, so that there continues to be a direct one-to-one link from a voter to 'his' MP.

* Of course, if one party has two candidates, it will make it very interesting as to how they co-operate or compete - it may help the party to have them share the vote as equally as possible, but if they do that, they increase the chance that both fail as well as the chance that both succeed ...
--------------------
4. As ever, here's a worked example: let's imagine the existing, neighbouring parliamentary constituencies of Epping Forest; Harlow; and Brentwood and Ongar formed an MMC. We can merge the 2005 votes achieved by each individual candidate from here, sort individual candidates in decreasing order of votes and we get the following table.

The asterisks denote those who won a seat under FPTP and are sitting MPs - as you see, assuming no change in voter or party behaviour, under my system, it would have made no difference - Laing, Pickles and Rammell would choose to stand for their original sub-constituency and that would be the end of that:
As an aside, if you put the parties' total votes (as below) into Paul's d'Hondt spreadsheet, you also end up with 2 Tories and 1 Labour MP:
But people don't like party lists, but this doesn't really matter as under my suggested system (which is not based on party lists), behaviour would change and the outcome would be much the same:

The Lib Dems would spot that if they had only put up one candidate, and he had received all Lib Dem votes, they would have won the first seat, with the Tories winning the other two (and Mr Rammell of Labour would have been squeezed out); Labour know this, so they would also only submit one candidate, and win the second seat. This would take a seat off the Tories, so they would only field two candidates and somehow try to ensure (by each of two candidates only campaigning in half the constituency, for example), that each of their candidates got more votes than the single Labour candidate. This might work for them, it might not, that's just luck on the day.

Of course, there would be no end of wrangling over how many sub-constituencies are to be merged each time, and which ones with which - smaller parties would prefer more members and larger parties fewer members, but let's not forget that even the Tories and Labour are 'minority' parties in some areas, so I guess that the outcome wouldn't be too unfair - in any event, it has got to be better than FPTP.
----------------------
5. By the way - this week's Fun Online Poll asks "Will the LibLabConsensus delay the next UK General Election until after the Irish have voted on The Lisbon Treaty again?", (which they hope will produce a 'Yes' - I think it will still be 'No', as it happens), thus saving Dave The Chameleon the embarrassment of having to have a retrospective referendum in the UK.

Vote here or use the widget in the sidebar.