Wikidata:Property proposal/Voting system

voting system edit

Originally proposed at Wikidata:Property proposal/Generic

Done: voting system (P8196) (Talk and documentation)

Description	Voting system used for this election
Represents	electoral system (Q182985)
Data type	Item
Domain	item
Allowed values	subclass of (P279)electoral system (Q182985)
Example 1	September 2019 Israeli legislative election (Q64159775) voting system party-list proportional representation (Q31764)
Example 2	2016 United States presidential election (Q699872) voting system indirect election (Q877353)
Example 3	2016 United States Senate election in California (Q18356713) voting system direct election (Q1196727)

Motivation edit

Use to allow linking elections to their voting systems (direct/indirect, etc). Thanks, --DannyS712 (talk) 23:19, 5 August 2019 (UTC)[reply]

Notified participants of WikiProject elections --DannyS712 (talk) 03:11, 6 August 2019 (UTC)[reply]

Discussion edit

Support --Trade (talk) 20:10, 10 August 2019 (UTC)[reply]
tend to   Weak oppose, seems redundant with
⟨ Israeli legislative election (Q24659965)      ⟩ subclass of (P279) ⟨ party-list proportional representation (Q31764)      ⟩
and
⟨ September 2019 Israeli legislative election (Q64159775)      ⟩ instance of (P31) ⟨ Israeli legislative election (Q24659965)      ⟩
. author  TomT0m / talk page 12:22, 12 August 2019 (UTC)[reply]
  Comment Anyway, allowed values should be « instance of voting system », not « subclass of voting system », or a specific election like would be an instance of « voting system ». And
⟨ 2016 United States presidential election (Q699872)      ⟩ instance of (P31) ⟨ voting system ⟩
does not make any sense. author  TomT0m / talk page 12:28, 12 August 2019 (UTC)[reply]
Except that
⟨ Israeli legislative election (Q24659965)      ⟩ subclass of (P279) ⟨ party-list proportional representation (Q31764)      ⟩
doesn't make sense - its not a subtype of the election system it uses. And I meant that the value of the property should be a subclass of a voting system. I'm not sure I understand what you mean in the second comment - its not
⟨ 2016 United States presidential election (Q699872)      ⟩ instance of (P31) ⟨ voting system ⟩
, but rather 2016 United States presidential election (Q699872) voting system indirect election (Q877353). --DannyS712 (talk) 12:55, 12 August 2019 (UTC)[reply]
@DannyS712: It may make perfect sense … if you consider the french label, which is
scrutin de liste proportionnel
, in english « proportional list voting ». Any proportional list voting is a voting, so we have
⟨ proportional list voting ⟩ subclass of (P279) ⟨ voting ⟩
, any israely legislative election is a proportional list voting, so we have
⟨ israely legislative election ⟩ subclass of (P279) ⟨ proportional list voting ⟩
, and September 2019 Israeli legislative election (Q64159775) is an instance of everything. If you choose that model you certainly not have to express in all elections items like September 2019 Israeli legislative election (Q64159775) which type of election they are because they are already classified as election of that type, with less redundancy.
⟨ 2016 United States presidential election (Q699872)      ⟩ instance of (P31) ⟨ voting system ⟩
then makes explicit how we classify elections, these are the subclasses we are looking for if we want the voting system, because there is other criteria with which we classify elections (for example any israeli legislative election is a national election, which make « israeli legislative election » a subclass of « national election »).

But I must admit I did not look in details to the english label and english definition of this item. If a voting system is a set of rule, it make perfect sense to link an election to a voting system. This set of rules defines a process, that is instanciated at each time this type of election type.

Essentially, thinking about this I’d try to define things like this.

An election is a process in which some people are chosen for some offices or a decision is taken, (referendum),so
⟨ election ⟩ subclass of (P279) ⟨ process ⟩
.

An election follows some rules, defined usually by a law : main regulatory text (P92). In a sense the rules entirely defines the process of the elections. The rules can be instanciated many times, for example the rules for the american president election are instanciated many times. In that sense the USA presidential election is … a voting system. If you see en:voting system it redirects to « electoral system », which is defined exactly like that, this is consistent. So it’s enough to say that the Trump election is a USA presidential election to link it to its voting system.

On the other hand, party-list proportional representation (Q31764), if you see the english article, is defined as a voting system … family, which changes everything. You would have to rename this property « voting system family » to be consistent, according to the examples. Then you are looking for instances of « voting system family » You don’t hav . So I think to make all this consistent, you would actually have 2016 United States presidential election (Q699872)voting system familyindirect election (Q877353). But I think all this is essentally the same analysis as before, so this is redundant and not very useful as well, we can do better without this property at all … author  TomT0m / talk page 09:44, 13 August 2019 (UTC)[reply]
SupportStudiesWorld (talk) 00:10, 13 August 2019 (UTC)[reply]
Support Need to document also the relation with main regulatory text (P92) (perhaps a costraint that there should be both or one implies the other?) --Sabas88 (talk) 07:18, 13 August 2019 (UTC)[reply]
Actually the regulatory text is tight to the election type, for example any israely legislative election (on a certain period of time) has the same rules. The statement should be made of items such as « israeli legislative election » to avoid redundancy, not on any election instances. It’s easy by a query or in lua to find the text if you know it’s an instance of « USA presidential election » then. author TomT0m / talk page 09:44, 13 August 2019 (UTC)[reply]
~~Oppose~~ This seems like a good idea, but I don't think the current samples (and definitions) are up to it. It would be good to have items that actually describe the voting system used in a given election. ~~Maybe "allowed values" should be instances, not classes.~~ --- Jura 14:10, 20 August 2019 (UTC)[reply]
- Support Maybe we can start with the above and gradually increase the quality (and granularity) of items being used. I don't think it helps mixing them with P31 values in general. Allowed values would probably be classes even if a specific system would only be used in a single election. --- Jura 16:55, 16 September 2019 (UTC)[reply]
Oppose. The format and available values need refining, and I'm still not sure whether this should be stored at the level of individual elections or types of elections. --Yair rand (talk) 22:45, 29 August 2019 (UTC)[reply]
Support @Yair rand: In 2018 voters in Maine switched to using instant-runoff voting (Q1491219) for elections to statewide office (including for US Senators and Representatives); it is appropriate, in such an example, to make clear (should items for specific elections in that state be created) that the elections prior to 2018 used first-past-the-post voting (Q5557375) and the ones afterward used instant-runoff voting (Q1491219). Mahir256 (talk) 04:48, 12 November 2019 (UTC)[reply]
- Yes, but that can be expressed by start/end date qualifiers. --Yair rand (talk) 06:22, 12 November 2019 (UTC)[reply]
Support Needed for infoboxes in biographies. Pmt (talk) 15:01, 3 April 2020 (UTC)[reply]
@DannyS712, Trade, TomT0m, StudiesWorld, Sabas88, Jura1: Yair rand, Mahir256, Pmt: Done --Tinker Bell ★ ♥ 21:56, 5 May 2020 (UTC)[reply]