As noted by Bancki in a previous thread on the Citizens Assembly on the electoral system in the Netherlands, the citizens now have a proposal.
It keeps the actual party-list PR in a nationwide constituency, but proposes two major changes:
1. The d’Hondt formula should be replaced by the simple quota & largest remainder formula because it is fairer to smaller parties.*
2. Voters should get more power in determining who will be elected from the party lists. Normally the top candidates are elected; lower-ranked candidates can only get elected if they obtain 1/4th of a simple quota (only 10% of MPs get elected this way).
The Citizens Assembly proposes that a voter can vote for the whole list or for one candidate. The seats are distributed “proportionally” between the top of the list and the most vote-getters: if party A gets 20 seats, 40% of A-voters vote for the entire list, and 60% vote for some candidate on the list, then the top 8 are eleted and the 12 other seats go to the 12 candidates with the highest personal score.
That would be an interesting way of attempting to split the difference between open and closed lists. Elsewhere, I have proposed intra-party d’Hondt, allocating seats based on the shares of the votes cast for the list (as a whole) or to specific candidates. Neither intra-party d’Hondt nor this Dutch proposed method has ever been used, to my knowledge. My quick expectation would be that this proposal would allow the election of candidates with smaller personal-vote shares than would intra-party d’Hondt. That may be precisely consistent with the citizens’ goals, but in many other jurisdictions, the intraparty fragmentation promoted by rules in which large numbers of seats are filled by simple rank in preference votes has produced considerable dissatisfaction.
Apparently, the citizens like fragmentation, on both the intraparty and interparty dimension. I have just discussed the intraparty dimension. Regarding the interparty dimension, the decision to change from d’Hondt divisors to simple quota and largest remainders (SQLR) also would favor fragmentation. The proposal favors SQLR because it is “fairer” to small parties–overly so, I would argue. It can allow a party or faction thereof to split off and present its own list, with the result that the separate lists of the formerly unified party can obtain more seats collectively with the same votes than they could have running on one list. For this reason, many countries (including the Netherlands, previously, and Colombia most recently) using PR have abandoned SQLR for d’Hondt. A threshold can help overcome this effect, but the existing Dutch threshold is very low (about 2/3 of 1% of the nationwide vote).
Bancki reports that he can find only a Dutch text of the full proposal. If anyone reading this also reads Dutch, please post in the comments any additional information you can glean from the report.