From Citizendium
Jump to navigation Jump to search
Citizendium Ombudsman
Decisions | Referrals | Appeals | Guidelines | Archive
How to Edit
Getting Started Organization Technical Help
Policies Content Policy
Welcome Page

Report of election observer

Having been declared elected Ombudsman as an unopposed candidate, I was asked to act as election observer for the Citizendium elections of October 2010.

The elections to the Management Council, the Editorial Council, and for the positions of Managing Editor were conducted fairly in a manner consistent with the Charter, and in a manner that I believe to be consistent with generally accepted democratic principles. The ballot was secret, with details known to the fewest number necessary for rigorous independent oversight. The results of the elections were determined by two constables working independently, but in open sight of the other constables and myself as election observer. I independently checked the results.

The elections were complex, because of the multiple shared candidacies. The ballots provided a very large amount of information, giving approval votes and listing candidates in order of preference, and most of this detail proved essential for establishing that the outcomes fairly reflected voting preferences. Simpler ballot procedures would probably have led to uncertain outcomes requiring run-off elections, possibly involving several elections again with multiple candidacies.

The candidates for multiple offices had expressed preferences between the different offices for which they stood, and this was essential in establishing the final outcome. Some complexities that arose had been foreseen, others had not.

To give an example of the type of complexity that can arise with this type of election, I give an anonymised example with details altered; something similar arose.

Candidate A is running in elections X and Y, but has not stated a preference for which office he prefers. Both are close; if A had expressed a preference for X, then he would be elected to X ahead of candidate B who then takes his elected place in Y. If A had expressed a preference for Y, then this allows candidate B to be elected to X, which he prefers. However if candidate A’s preference is not known, then B is elected to Y ahead of him, and as a consequence of the withdrawal of B from X, candidate C is now elected to X ahead of A. So if A expresses a preference for either X or Y he will be elected to the body he prefers; but if his preference is not known he will be elected to neither.

The outcomes of all elections were considered according to the results that would have been achieved using different systems of determining the outcomes. We noted that the outcomes using some methods were not robust, in that for example with transferable vote systems, the outcome depended on the order of elimination of unelected candidates.

In the event, the method applied was simple: the election for managing editor was determined by simple majority of votes cast, and the winner’s name removed from the other election in which he had stood.

For the other three elections, the initial outcomes were determined by the order of candidates as given by the number of votes cast for the number of vacancies in that election; i.e. for a three candidate election, only the first three preference votes were counted. Then, for dually elected candidates their preferences were taken into account, and the next on the list from which they had withdrawn was declared elected.

These outcomes were then checked for robustness to see if the outcomes would have been the same if the withdrawn candidates’ names had been removed from the ballots. They were.

Finally, we noted that all elected candidates had received approval votes from an absolute majority of voters in that election.


  • For future elections, a simpler process can be applied, but it should not be so simple that there is too great a risk of tied results.
  • I suggest that in future elections, voters should be asked to list as many names as there are vacancies, in order of preference. I suggest that the results be calculated on the total numbers of votes cast for each candidates, but that preferences be used to separate tied results.
  • It should be noted that an election for 3 vacancies plus one reserve is effectively an election for 4 vacancies. It therefore needs to be decided whether to ask for a list of 3 names or for one of 4 names.
  • It should also be noted that the tenure of a reserve is not defined. Each election might therefore be seen as an election for fresh reserves.

Gareth Leng 09:24, 2 November 2010 (UTC)

Election Process

See CZ:Ombudsman/Election process