Up: Issue 9 Previous: Paper 4

# Voting matters - Issue 9, May 1998

## Ordered List Selection

#### J Otten

Joseph Otten is the Green Party Policy Co-ordinator.

### Rationale

The electoral system to be used for the next European Elections requires ordered lists of candidates from each party. It was felt that the advice in the ERS booklet that If an order is desired, this is provided by the order of election (2.5) was inadequate - it would effectively lead to a First Past the Post contest for the top place on the list.

Were we to know in advance that we would win, say, n seats in a region, then it would be straightforward to use STV to select n candidates from the potential candidates and put them in the top n places in our list. If we don't know n in advance (which we don't!) then we can perform this operation for every possible n, i.e. from 1 up to the number of seats available in the region, and attempt to construct a list whose top n candidates are those victorious in the nth selection ballot. (There is really only 1 ballot - the division into n ballots is notional.)

This ideal solution fails when a candidate elected for one value of n is not elected for a larger n. In such cases either the STV result for a smaller n must constrain that for the larger (top-down) or vice versa (bottom up). Reasoning that the Green Party would be unlikely to win large numbers of seats in any region, we opted for top-down.

### Algorithm

Each count is conducted by ERS rules [1] with the following alterations. We start with the count for the first place (n=1) and work down.

#### 5.1.6 Calculate the quota

Divide the total valid vote by one more than the ordinal number of the count. Eg for the third count, divide the total valid vote by 4. If the result is not exact, round up to the nearest 0.01.

#### 5.2.5 Excluding Candidates

Do not exclude any candidate in one count if they have already been elected to the list in an earlier count. This may introduce distortions to the results of later counts, but is necessary to preserve the integrity of the earlier counts.

If a count is proceeding identically to an earlier count, and an exclusion by lot is required, then the result of the earlier lot should be taken as read. Otherwise the lot must be recast. (cf 5.6.3)

5.3.3 and 5.4.2 For the purpose of these rules (i.e. receiving transfers), a candidate elected in a previous count (not stage) should be treated as a continuing candidate for purpose of receiving transfers during the count, until they are deemed elected again.

#### 5.5.2 Completion of the Count

For the purpose of this rule, any candidate elected to the list in a previous count shall be deemed elected. Therefore the count may stop as soon as a single candidate is deemed elected, who was not elected in a previous count. In exceptional circumstances it is possible that two candidates, not previously elected may exceed the quota in the same stage. Only one may be elected. Resolve as follows (in order of priority):
1. If more than one value of papers is transferred during that stage, and only one candidate is elected as a result of the transfer of an earlier (i.e. higher valued) batch, then that candidate is deemed elected.
2. If both exceed the quota during the transfer of the same batch, then elect the one with the higher vote.
3. According to 5.6.2
4. By lot.

### Other deviations

My apologies to the Electoral Reform Society for these, but they do seem to be popular in some quarters.

Where regional parties have agreed to adopt gender balance constraints, then the usual constraint rules shall be used. This usually means excluding all the candidates of a particular sex at the beginning of an even-numbered count.

Each region was free to determine its own gender balance formula. For example one region might require a list of half men and half women with no constraints on position, and another region might require that the top two candidates were a man and a woman with no constraints on the other candidates. Whatever formula was chosen, this was applied within the system by excluding any ineligible candidates at the beginning of a round. Hence the top place on each list would be open to both sexes, and subsequent places would only be constrained in the event of an imbalance. Notably the London region decided not to impose a gender balance formula, and the top three candidates are all women.

On each ballot form there is a notional candidate called "Re-Open Nominations" (who is of indeterminate sex). If Re-Open Nominations is elected to the list, then there must be a fresh election for that place and lower places on the list. This is a distortion of STV which could be used by a majority to deny minority representation, although there is no evidence of this happening. STV, rightly in my view, omits this sort of negative voting, but it is popular in the real world outside public elections, such as in student unions.

### Conclusions

Although the justification for starting at the top of the list and working down, as opposed to starting at the bottom or even in the middle, is not particularly strong, this system is a reasonable solution to the question of seeking an ordered list. In particular it ensures that however MEPs are elected in any region from the party, they are as proportionally representative of the range of opinion in the party as their number allows.

### Reference

Up: Issue 9 Previous: Paper 4