Liechtenstein's election system is something like open-list PR.
Its party-list component is unusually limited - votes do not go to an overall party list but only to candidates running under that party label in the district.
The two districts produce local represention of a sort, although each district covers half the country, the country is small and each district has one main city.
whereas NZ's MMP combines district FPTP with overall list PR. The voter can vote for candidate in the local district and can also place a party vote to be used to effect overall distribution party-wise. But other than the party vote the local vote cannot leave the district. This is unfortunate as in many cases, in the district, the groups that support candidates other than the leading candidate are totally disregarded, even though there are other groups of the same party in the adjoining districts that are also disregarded. Together they might have enough numbers to elect someone to represent them.
Catherine Helen Spence when she envisioned STV back in 1861, pictured a mash-up of candidate voting in district with candidate voting outside the district (no party lists were used).
A flexible voting system I am toying with is a mash-up of candidate voting in the district (under STV) with voters able to mark a preference for a "cascade" vote working down candiates of a party. Like in Catherine Helen Spence's system and like in NZ's MMP, a vote could leave the district.
But the vote could only be used in the end just to elect one person. so tht princile of STV is still preserved.
In STV-plus ("Monto-Spence Voting"), voter only has one vote (but can mark back-up preferences).
Depending on the voter's priorities, the ballot could be marked in a variety of ways:
-voter could mark preferences just for district candidates in which case the vote would not leave the district - it would be used to elect one of the district candidates or not at all;
-voter could mark preferences for candidates just in one party in which case the vote would be used to elect one of those candidates, whether within the district or outside the district, or not at all,
-voter could mark preference for candidates of party A in the district (which would allow the vote to go to electing a local candidate of that party if possible), then the "cascade" vote for A party (which, if the vote is not yet used, would allow the vote to be used to elect any candidate of that party anywhere if at all possible), and then could also mark preferences for other candidates of the district or for a "cascade vote" of another party (giving chance for vote to be used in that way, if it had not yet been used).
Altogether allowing votes to leave the district would better ensure that the vote is used to elect a candidate or a party that the voter prefers.
=========
the idea of STV-plus would be for votes to be able to transfer out of districts if that is voter choice instead of either having to cross party lines to be used or to be declared exhausted.
I just found a colourful expression of this goal in a publication from the 1890s:
"the hopeless minorities currently cooped up in political slave pens will be liberated; their votes will bear directly upon the final result, thereby offering the greatest incentive for all men to vote"*
the writer is speaking to benefits of STV in at-large city elections versus use of FPTP in single-member wards. I share the concern and would stretch it to loosen up district boundaries themselves.
at-large is possible in city elections but seldom possible in higher levels of government as STV's DM historically has maxed out at 21.
so to avoid voters "being cooped up in political slave pens" I am saying vote should be allowed to leave the district if that is the will of voter.
a novel and I admit perhaps overly-creative innovation....
the "party vote cascade" option would still be structured as candidate-based voting.
instructions to election officials are implied to mean, if vote would otherwise be wasted, to send it to the most-popular candidate of the preferred party that is still in the running -- anywhere (within Scotland or within a province of Canada, like that)
a bit of a blank cheque but in line with how it seems many voters do vote - for a candidate of the party of choice, irrespective of the candidate's name, etc.
but there would be no party lists prepared by party brass, and no allocation of seats to parties per se.
so the party cascade option may be considered a form of party list voting but is actually still candidate-based, the same as STV.
Locally voter has option to show support for one candidate in particular as first preference before marking second preference for "party cascade," in which case might stay in district or go elsewhere to elect a popular candidate of the chosen party.
*a longer excerpt from the 1894 publication can be found in the "Old Books" Montopedia blog
===========
Here's other remarks on STV-plus.
my porous-district "STV-plus" system has the advantage that each voter has just one vote and each vote would only be used once (or not at all in a few cases)
Malta's additional member top-up uses STV in districts and then allocates seats to make the final result even more fair. in particular cases.
I have not thought of it as way to liberate the "minorities cooped up in their political slave pens"
but yes anything that creates more proportionality and fairness or stops a wrong-winner election is good for voters in the end.
But if that kind of top-up was used to effect the same fairness as the STV-plus system, by allocating seats to parties if they are overall under-represenated based on first preferences, I would be afraid of unfairness and/or manipulation happening if a voter's vote could count for a secondary preference (under normal STV method) and then also as first preference to be counted for the top-up.
(the kind of unfairness possible or seen to be possible under MMP is discussed here:
A gameable electoral system? The Additional Member System in Scotland - The Constitution Society (consoc.org.uk)
The double-benefit of a vote is seen in New Zealand when a person votes one way for the local member and a different way for the party vote, sometimes electing both
while another voter (just a few percentages) see none of their votes used to elect anyone.
(in NEW Zealand, many do not see their local vote used, same as under any FPTP system, but most do see their party vote used to help elect a party representative.)
a MMP system that uses just one vote -- Denmark -- does not suffer that problem to same extent -- an effective local vote is valued so does not leave scope for much or any extra rep through top up.
or a rule that any vote that was used to elect a local member would not be used for party vote would also prevent that double benefit.
But such special rule is not needed in STV-plus because each voter has just one vote and that vote can be used just once or not at all (for the unfortunate few).
======================
Here's a basic description of the STV-plus system, the Monto-Spence system
I wish to combine list PR voting of a sort with normal STV.
It may be an un-necessary adaptation of STV, which already is much more fair than FPTP but I think it addresses some disadvantages of any district voting. and also allays some criticisms of STV for not being strictly party proportional (not that that is big concern to me).
STV-Plus ("Monto-Spence Voting") would allow transfers out of the district.
say voter when marking No. 1 preference has choice of marking preference for a candidate in the district or for a party "cascade" preference.
Both types of preferences would be on the same ballot.
(thus voter would one vote but can mark back-up preferences and by marking a preference for party cascade, could show preference for party-list-type use of the vote
instead of how under MMP or AMS there is one party vote and one local vote with no back-up preferences allowed)
under STV-Plus, multi-member districts would be used,
as well votes could leave the district.
(districts would not be guaranteed any set number of seats though)
(for example, likely voter would mark no. 1 for most preferred candidate of preferred party, then for No. 2 could mark party "cascade" for preferred party if does not care about FPTP-style "local representation" and does not have preference for any other candidate after No. 1.
then No. 3 could be again for a different local candidate or a different party "cascade" showing preference for second-most preferred party
and so on.
marking party "cascade" for preferred party would be sort of like party list voting.
like some versions of Gove method (Indirect STV),
the party cascade vote would go to the most-popular party candidate anywhere, then if it is surplus (so not needed), would "cascade" by going to second-most-popular candidate, then possibly to third most popular candidate, and so on,
transfers taking place in successive stages -- the use of successive stags would be same as normal STV.
with transfers crossing district borders as they might under party cascade choice, quotas arrived at by using total vote in district would not work.
A uniform quota would have to be used. (New York City used a uniform quota in the same way as this when it used STV from 937 to 1945. However in that system each borough had to get at least one seat. A minimum of one seat per district could be brought in under STV-plus but that would add an additional bit of complication.
Under NYC STV, the overall number of members fluctuated depending on voter turnout. Under STV-plus a set number of members would be elected overall, because to do otherwise seems to me to be even more complicated.)
And of course at the end, when the field of candidates across the jurisdiction (the province or the devolved state or whatever*)- when the field of candidates is thinned to the number of remaining open seats, the most-popular candidates, the ones that are the closest to quota, would be elected to fill the remaining seats.
*jurisdiction would be the outside limit of vote transfers - a unitary country or across a province in Canada, or across all of Scotland for the election of the Scottish Assembly, for example.
District would not be guaranteed any set number of seats.
But if a district has quota or more of votes and voters there give all their preferences to the local candidate, then a local member would be elected.
But if voters use party cascade and it happens that the most-popular candidates of those parties are outside the district, then the district may not get rep. but that too is choice of voters.
(Seats would be filled under "STV-plus" by putting all candidates in the jurisdiction into one big contest, with eliminations done step by step affecting the least-popular candidates overall, in successive stages.)
Voter would vote for local candidate(s) or for party cascade(s).
Voter would not have the choice of candidates outside the district but could aid candidate of the preferred party running outside district through marking preference for party "cascade".
Election officials would allocate seats judging votes overall.
More work for election officials but no more work for voters especially under optional-preferential voting.
Under AMS it might happen that a voter sees both of his votes used to elect someone while another sees none of his votes used to elect anyone.
Such problems and how some parties manipulate AMS voting is outlined in this article:
A gameable electoral system? The Additional Member System in Scotland - The Constitution Society (consoc.org.uk)
Under AMS, voter can only choose one candidate in local election and only one party with party vote, while he might have two or more favoured individual candidates and two or more favoured parties.
(In my "STV plus" system, just like under normal STV, the vote can only be used (in the end) for one choice, but under "STV plus" system the variety of back-up preferences would be wider than under normal STV.)
And under STV, voters are trapped in district so voter must choose to give back-up preference to a candidate of another party or to see vote possibly exhausted. In fact under full-preferential voting, voter must cross party lines or see the vote declared rejected.
But if voter could follow a party slate outside district, then voter would not face that problem.
Only parties that got or were due one or more seats in the last election ("due" being measured by the uniform quota) would have ability to use party cascade preference.
so the ballot would not be cluttered by multitude of parties having party cascade preference on offer.
Under AMS the party vote can be used to elect someone outside district but within region but can have no effect outside the region.
"STV plus" would see vote go anywhere in jurisdiction (province or country) to help a preferred party if that is the choice of the voter.
While using multi-member district means many voters would be satisfied in district as well.
Under AMS and FPTP, many voters are un-satisfied in district, and votes are wasted even in regions,
And a party that is spread over all of Scotland for example is unable to get rep. that might otherwise come to it.
Under "STV-plus" votes could travel farther and so help the preferred party anywhere if that is will of voter, without as much necessity to mark cross-party transfers
Under AMS
is there true feeling for local districts (micro-local district)?
is there true feeling even for regions?
is there reason to stop party voting at region's border?
Under STV
is there reason not to have party voting?
is there any reason why a vote has to be cooped up in a district?
(Australia in some states uses above-the-line party voting but there voter has choice of either party voting or marking many candidates below the line. and votes cannot leave district.)
STV-plus would give voter the right to mix and match candidate-based district voting with party support and would allow back-up preferences in both contexts. and would allow vote to leave district if that is will of voter.
STV-plus, same as many STV systems, would not use any single-winner districts except where necessary in hinterlands.
Would a system that had city-scale multi-member district coupled with overall rep. through out-of-district vote transfers, and both local-candidate voting and party voting have real benefits?
I believe so.
================================
Comments