I have always vaguely assumed that the single transferable voting system would bring about the greatest degree of "proportionality" in elections. George Kerevan has given me a fuller
fuller explanation of the actual mechanics of STV. I hadn't realised that the STV system was not necessarily good news for the smaller parties:
Although Tommy Sheridan’s SSP was second in 19 Glasgow council seats on 1 May, its chances of getting more than the odd member under STV is small. Expect semi-permanent Labour and Lib Dem coalitions, as at Holyrood.
Kerevan then goes on to explain why local Labour party activists are so opposed to a new voting system that may harm its smaller rivals like the SSP:
It works like this: a new STV ward will amalgamate four existing wards. Suppose two of the original four are middle-class wards where voter turnout is high. And suppose the other two are ghetto housing schemes with a low voter turnout. Under STV, the extra weight given to those middle-class votes might overwhelm the low turnout in the housing schemes. This will differentially affect Labour, which wins seats with tiny votes in the ghettos while the Lib Dems and Tories win by useless big majorities elsewhere.
I had never realised before that differential turnouts were so beneficial to Labour. Presumably this is one of the factors explaining why the Labour party is able to win 90% of the seats on Glasgow City Council with less than 50% of the vote.