*Update: neither of the proposals was successful. * On Monday, 16 May 2022, colleagues in Luxembourg will meet in a General Assembly to change the electoral rules for Luxembourg. This is not a new topic: Generation 2004 has considered for a long time that the current rules are flawed. We tried to change the rules in 2019 , but missed the required 2/3-majority (not by much, though). We have learned from this and have incorporated the feedback received into our new proposal .
The assembly will be held in a hybrid mode. This means that you can support our proposal even without leaving your office. Just stay present (virtually) in the meeting, so that you receive the voting link, and make sure cast your vote within the (relatively short!) timeframe.
As always, remember that an “abstain” vote has exactly the same effect as voting “against”. A 2/3 majority of those present in the meeting must vote “for” in order for any proposal to be adopted. So, if you connect to the meeting, please make sure to vote, too!
Why is a new system necessary?
We could give a long technical answer and go into details about the obscure distinction between “list” and “mixed” of votes in the current system, its majoritarian character, the double counting of “list” votes, the alphabetical order, and so on. But it is sufficient to look at the results of the last three elections to clearly see the problem:
|Period||Percentage of votes for the “winning” list||Number of resulting seats|
|2013 – 2016||36.2%||13 out of 20|
|2016 – 2019||25.5%||14 out of 20|
|2019 – 2022||40%||14 out of 20|
As you can see, the current system allows for a list with only 26% of the votes to have a 2/3 majority in the local staff committee. We are not aware of any other electoral system in the Commission that is so obviously biased. However, it gets worse: if you win 14 more percentage points like in 2019, you may not get a single seat more! Or, comparing with 2013, winning 11 more percentage points actually costs you a seat! Such a system actively discourages collaboration: why should you try to build consensus and convince more voters, if it does not bring you anything in terms of seats in the staff committee?
We have incorporated the feedback received in 2019 into our new proposal. Without going into technical details, we developed and updated our proposal to fulfil the following characteristics:
- Fully proportional, e.g. 25% of votes translate into (roughly) 25% of seats
- Smaller lists with less than 20 candidate couples are not punished (a difference to the other proposal on the table, as shown below)
- The order of list candidates on the ballot is left to the discretion of each list; there is no more forced alphabetical order (Luxembourg is the only Commission site where lists cannot choose the order of their candidates!)
- Electronic signatures are explicitly allowed (a Covid-19 lesson; but we consider this also necessary for colleagues who have many missions, e.g. the nuclear inspectors in DG ENER)
The other proposal by Ensemble Luxembourg
It maintains an alphabetical order of the candidates in a list. While this might seem like a small point, it has lead in the past to silly games where a candidate with a “good” family name was paired as a full member with an alternate member to form a couple that was high on the list. Once voted, the full member would vanish and the alternate member took over. This is unnecessary and not transparent: lists should have the freedom to put their “core” candidates at the top of the list – like in any other serious electoral system of which we are aware.
Punishment of small lists
The competing proposal punishes small lists, and makes it essentially impossible for individual candidate pairs (that is, those not affiliated with any list) to be elected – and yes, we are aware that Ensemble Luxembourg says that it does not, but just carry on reading and then decide for yourself.
An example: let’s say that there are 1000 votes, one individual candidate pair (P) and two full lists (L1 and L2, respectively), each with 20 candidate pairs. 400 voters make one cross for the individual candidate pair and no crosses on the two lists. 300 voters vote for all 20 candidates on list L1, the remaining 300 voters vote for all 20 candidates on list L2. We show the corresponding d’Hondt tables for the two proposal below, where green means that the corresponding candidate is elected, yellow means a tie and red means that the candidate is not elected.
|Divisor||Individual pair P||List L1||List L2|
|Basis for d’Hondt||400||300||300|
|1||400 (1.)||300 (2.)||300 (3.)|
|2||150 (4.)||150 (5.)|
|3||100 (6.)||100 (7.)|
|4||75 (8.)||75 (9.)|
|5||60 (10.)||60 (11.)|
|6||50 (12.)||50 (13.)|
|7||42.86 (14.)||42.86 (15.)|
|8||37.5 (16.)||37.5 (17.)|
|9||33.33 (18.)||33.33 (19.)|
|10||30 (20?)||30 (20?)|
As you can see, in the Generation 2004 proposal, the individual candidate pair is elected first; which is only fair, since this pair has received more votes than any other pair on any of the other two lists!
As a small detail, in our proposal the tie for the 20th seat would be resolved by random draw, which we consider the most fair – unlike giving priority to more senior and older candidates as the other proposal does.
|Divisor||Individual pair P||List L1||List L2|
|Basis for d’Hondt||400||6000||6000|
|1||400||6000 (1.)||6000 (2.)|
|2||3000 (3.)||3000 (4.)|
|3||2000 (5.)||2000 (6.)|
|4||1500 (7.)||1500 (8.)|
|5||1200 (9.)||1200 (10.)|
|6||1000 (11.)||1000 (12.)|
|7||857.14 (13.)||857.14 (14.)|
|8||750 (15.)||750 (16.)|
|9||666.67 (17.)||666.67 (18.)|
|10||600 (19.)||600 (20.)|
As you can see, the individual candidate pair is not elected under the other proposal, although they beat every other pair on both of the lists. This is an inherent weakness in the other proposal – and shows the strength of the Generation 2004 proposal.
A bit more history
In 2019, we managed to find consensus with five other trade unions and staff associations in Luxembourg on a joint proposal to change the electoral system. There were compromises, but the most important element – a fair, proportional counting method that would not penalize small lists – was already there.
We were unable to find the same consensus this time: the same counting method, proposed jointly in 2019, suddenly became a problem for the other organisations. We are therefore presenting our own proposal now. However, on the bright side, we could get rid of the compromises and table what we firmly believe is the best thought-through proposal tabled in the past decade.
And if you are wondering about the visual similarity: yes, the same Generation 2004 member, who was the main drafter behind the joint 2019 proposal, did most of the drafting on our current proposal, too. We take it as a recognition of our good work that Ensemble Luxembourg also used the 2019 joint proposal as a basis for their text.
As usual, do not hesitate to contact us  for any further queries.