If you want party proportionality above all else, then D’Hondt is the simplest and most wisely used option. If you want party proportionality but at the same time to retain a link between MPs and constituents, then AMS is a reasonable, tried, and widely supported, compromise solution. But there are problems.
The constituency vote
AMS would continue to use FPTP in the election of its constituency MPs. The faults in FPTP may be deemed to matter less when used within AMS because the overall system is more or less party proportional, in a roundabout way. On the other hand, they may be deemed to matter more because they are more likely show up in AMS’s double sized constituencies and will be in no way corrected in the regional ballot in which it is the parties rather than the voters who choose the MPs.
Consider, for example, two specific problems.
(i) FPTP in double-sized constituenccies could make it twice as hard for an independent or a small party to gain a seat.
(ii) The ability of FPTP to elect a candidate on a minority of votes greatly increases the chance that a party will win more constituency seats than the regional D’Hondt would have allocated in total.
The bottom line is that FPTP is wrong in principle and AV is right in principle. AMS would be a very much better system if constituency candidates were elected using PV rather than FPTP.
The regional vote
There is also scope for PV in the party-list component of AMS.
Where there are more parties standing than there are seats available, as used to happen in the UK’s elections for the EU Parliament, it seems unacceptably undemocratic to disenfranchise supporters of parties who could not possibly win. Much fairer to eliminate those parties and to transfer the votes they had received to the second choices of those who voted for them. PV is the obvious way of doing that[1].
Moreover, when D’Hondt grants a party a second seat, the number of available seats decreases but the number of active parties does not. So at that point we should similarly eliminate the party who could not possibly win a further seat.
Furthemore, some countries’ D’Hondt-based systems disallow parties with less than a given threshold level of support (typically from 2% to 5%). If we were to enact such an exclusion[2] then the same PV-based elimination mechanism should apply.
A conclusion
Raw AMS is significantly undemocratic. AMS with PV is clearly much better.
So if it is AMS we want then the best plan would be
a. to introduce PV now, in time for the next election
b. to extend this as soon as possible thereafter to full AMS
– with PV rather than FPTP for the election of constituency candidates
– preferably with PV to shorten the list of parties to the number of seats available.
[1] Each party eliminated would of course be the one with the lowest current voting weight,
i.e. (Current votes for)÷(1+Number of seats won so far).
[2] Which cannot be recommended, even if it has the support of all the larger parties.