Rev. 04 Feb 26
<numerous way around the world where nations have striven to provide proportionality while preseriving a link between MPs and geographic constituencies (“representative democracy”). AMS is the British version of that.>
AMS is a brave attempt to provide party proportionality while retaining constituency links. <Brief outline of AMS, and of its top-up proportion?> <See Wikipedia>
In Scotland, which MVM take as a model, this involves having 73 MSPs elected by FPTP in single-seat constuencies and then, to move towards overall proportionality, 56 (43.4%) further (“top-up”) MSPs appointed from regional party lists in 8 regional groupings of constituencies.
Greater London Assembly 25 members, 14 for single-member constituencies and 11 (44%) additional for whole region.
AMS is regarded as one of the principal options for the future UK electoral system.
Proportionality
But who wants proportionality, and why? The most successful party rarely complains of having a disproportionally large number of seats. The second of two parties will complain if they have an overall majority of votes but a minority of seats; but that is more likely to be attributable to gerrymandered constituencies than to the voting system itself.
Much clearer complaints will come from smaller parties, who generally have difficulty in obtaining any seats at all. So much so that even their active supporters can be disinclined to waste their votes on them. Which particular problem can be largely overcome by having preferential voting, which the Scots sadly fail to do. But that merely induces honest voting: it doesn’t give the small parties any seats. So a top-up proportion of (say) 10% would give them a fair chance. More than that (e.g. 15%-20% as recommended by Jenkins, or even 43% as in Scotland) is a possibly justifiable concession to the disgruntlement of the larger unsuccessful parties.
It can readily be agreed that there is a strong need to reduce blatant and offensive disproportionality. Which, simple logic says, inevitably means increasing proportionality. Which aim is too readily translated by the less successful political parties into a requirement to actually establish party proportionality. The purpose or benefit of which is not questioned, or even discussed.
Top up size
x
Constituency size
to keep the same number of constituencies, increase size by x/(1-x). So 50% top-up implies double , and 10% implies a modest 11% increase
Why FPTP
Why single member
Separate ballot
<Top-up by separate ballot – votes explicitly for parties rather than for individuals, mostly with party affiliations – and top-up proportion high enough to ensure high level of overall proportionality>
scope for party manoevres to cheat the system, e.g. “decoy lists” and fake independents
But why does the top-up require a separate ballot? The constituency ballot will have registered total regional support for each party, but may well have left some parties unfairly under-represented. What possible objection is there to allocating top-up seats on the basis of the regional totals of the first choice (top ranked) votes for each party?
<There would be no separate top-up ballot (as required by orthodox AMS) but just the totalling of first-choice votes across all constituencies in the region.
Regions
<It would be regionally based.[1] (England has 9 regions identified for statistical purposes, which were also used as constituencies for EU parliamentary elections. In addition, each of the 3 devolved nations would be a separate region.)>
To whom
<And the allocation would be to parties identified in the constituencies ballots
How many
<The allocation of top-up seats would be by the same D’Hondt-based calculation used in orthodox AMS, but taking into account seats won in the constituency elections, with the aim of yielding overall proportionality.>
Extra seats for each party determined by D’Hondt-based calculation exactly as for standard AMS. Based on first-choice votes: rough and ready but anything cleverer also dubious and much more difficult>
Overhang
Germany “give up”. Insult to voters and truning reasonable aversion to gross disproportionality into pathological obsession with proportionality at all costs.
Which candidates
Given that, the top-up seats can be allocated to specific individuals
a) from party lists independent from constituency ties,
b) from party rankings of their constituency candidates,
but rather more reasonably to <Each party’s top-up seats are allocated to the best performing losing candidates of that party> c) those unsuccessful constituency candidates whose votes were the highest proportion of e.g.
— (i) the total votes cast in that constituency, or
— (ii) the [first or final?] votes cast for the winning candidate in that constituency, or
— (iii) the total electorate in that constituency,
<or … best of all (final vote)÷(1+number of seats) {or (final vote)×2÷(1+number of seats)}
or, even better, those whose final vote is highest % of active votes in final round {may have reduced by voters running out of preferences} <best performing>
Party grouping
option to allow any group of parties to ask to be treated as a unit for top-up purposes. May give some hope to smalles parties, none of whom might qualify for any top-up seats. But though the group might collectively qualify for one or more top-up seats, those seats would be awarded to the highest performing losing candidates from parties in the group, regardless of which of those parties they belong to.
Treat independents as party
Similarly we might choose to automatically recognise all independent candidates collectively as a single group.
- England has 9 regions identified for statistical purposes, which were also used as constituencies for EU parliamentary elections. In addition, each of the 3 devolved nations would be a separate region. The two largest of these 12 regions are South East (91 constituencies) and London (75), and the two smallest are North East (27) and Northern Ireland (18). ↩︎