Open lists and their “simplest forms”

No, this is not the much-anticipated essay on the possibility of open lists in New Zealand, but it does belong in the series on the MMP review.

Rather, this is a comment on the NZ Electoral Commission’s introduction to the issues on list types. While I find the issues pages at the MMP review website to be generally well done, the Commission does not get things quite right on the issues on open and closes lists. It says:

In contrast [to closed lists], open lists allow voters to express a preference for one or more candidates on the list and not just the party. Although the seats are still allocated among the parties based on their respective shares of the vote, voters may influence which candidates are elected to fill these seats. How much influence depends on the rules of the open list system. In its simplest form, voters have some ability to change the order of candidates set by a party on its list, but most candidates are elected in list order. More open systems allow voters to determine for themselves the rank order of candidates, and in some, voters can rank any of the candidates, regardless of party.

It seems very odd to me to refer to intermediate types of list as “the simplest form” of open list. Clearly, there are two basic principles that can define who is elected from a party list:

1. Party-given pre-election order, or

2. voter-given order.

These two mechanisms result in the two simplest methods: closed and open lists, respectively, as pure types. At one end of the continuum, there are no preference votes, and thus the party-given order fully determines who is elected (to whatever number of seats the list as a whole has won), and in what order. These are closed lists. At the other end of the continuum, only preference votes matter in determining who is elected, as there is no party-given order.

Any system that mixes these principles–entailing both a party-given order of candidates on the list and preference votes for one or more candidates within a list–is necessarily more complex. It must establish a threshold or quota, or some other means of determining when a given candidate has won sufficient preference votes to be elected in a different order than that established by the party, and which candidates will be elected based on their party-given order despite having relatively few preference votes.

I understand where the Electoral Commission is coming from here. A system in which “most candidates are elected in list order” while “voters have some ability to change the order of candidates set by a party on its list” is the smallest departure from the closed-list status quo. But it is certainly not as simple as a pure open list, where no quotas or thresholds are needed: the law simply establishes that voters may give one or more preference votes, and candidates are elected in the order of their preference votes obtained.

The last sentence is also a little misleading:

More open systems allow voters to determine for themselves the rank order of candidates, and in some, voters can rank any of the candidates, regardless of party.

Such systems are “more open”, in a sense, and regardless of whether they are more or less “simple”. However, this sentence introduces additional dimensions of variation in open or flexible lists, aside from the question of the balance between party-given order and preference votes:

1. Whether voters can give just one or multiple preference votes;

2. Whether, if multiple preference votes are allowed, these votes may be given to candidates across two or more different lists;

3. Whether voters may rank candidates on the list.

I am not aware of any system in which voters may give a ranked-choice vote (such as those in STV systems) within a party list. (Does anyone know of such cases?) This is what “voters may determine for themselves the rank order” implies to me, although maybe the Commission simply means that voters, collectively, determine the final ranking of the list–which is indeed the case under the “pure” (and “simple”!) open list. However, if I were reading this as a non-specialist who was at least aware of STV (which was an option in the referendum and is used in some New Zealand localities), I might understand it as an STV-style preference vote. This is, of course, a potential option in the review, although I would imagine a very unlikely one.

It should be noted that a weaker form of “ranking” is implied by allowing multiple preference votes with cumulation, as in Switzerland. Swiss voters may vote for candidates across lists (panachage), and may also give more than one vote to any given candidate. The Swiss voter has as many votes as there are seats in the district. Having as many votes as there are seats is also allowed in the relatively new list systems of Ecuador and Honduras, though without cumulation as an option. And then there are the–shall I say “simpler”?– multiple-preference vote systems that permit these votes to be cast only for candidates within one list, as in Peru and pre-1994 Italy.

Finally, the Electoral Commission includes the Japanese case of “best loser” provision under the heading of list types. I suppose that is technically correct, inasmuch as it is a means to affect the order of election of candidates from the proportionally allocated seats. However, I would prefer to keep these distinct from the question of having a preference vote for choice from among several candidates on a list presented by the party. Best-loser provisions within mixed-member systems do not give voters any choice of candidates, they merely provide for some mechanism of using the contests in the single-seat districts to affect the order of election in the proportional tier. Thus best-loser rules belong to the class of dual-candidacy provisions rather than list type per se (and, to be fair, the Electoral Commission does provide a link to its own discussion of dual candidacy).

I do not want to be overly critical of the Electoral Commission, which I think is doing an exemplary job of explaining what are fairly arcane issues to the public. Many more people will learn about options in electoral systems from the MMP review website than will ever do so from this blog. Nonetheless, I do not think either the cause of “improving MMP” or that of educating the public is especially well served by implying that flexible lists are the simplest way of giving voters a say in the order of election of candidates from party lists, or by suggesting that ranked-choice voting is a procedure used for party lists elsewhere. Open lists, where preference votes alone determine the order of the list, are much simpler–which is not to say “better”–than the quota and default-order provisions of flexible lists. EIther a ranked-choice system or a multiple-preference vote system that allows cross-list voting (panachage) is not only “more open” but also much more complex than the option of a single, decisive preference vote.

Whether any form of flexible or open list is desirable in an MMP system, or in New Zealand specifically, is a topic for another day.

6 thoughts on “Open lists and their “simplest forms”

  1. > ” a ranked-choice vote (such as those in STV systems) within a party list. (Does anyone know of such cases?)”

    1. ACT’s short-lived “modified d’Hondt” system, 1989-95.

    2. Don’t or didn’t Norway and Austria allow numerical ranking within lists (but counting by a Borda-style points system)?

  2. Their nomenclature can be found on many other places on the internet. “More open” lists are what you call – without a doubt more accurately – flexible lists. And to be fair, flexible lists are easier to understand than the openest (‘free’) lists as used for example in Switzerland, Luxembourg and local elections in some German states.

    • No, JD. What I call “flexible” lists they are calling the simplest departure from closed lists. “More open” lists is the term they are using for (pure or fully) open lists, as well as systems of multiple votes with or without panachage (“free lists”), etc.

      I think flexible (or semi-open) lists are a lot harder to understand than standard open-list systems. In fact, even I don’t really understand how flexible lists work sometimes!

  3. I stand corrected.

    How about a system where voters can rank a list’s candidates as much as they want, and all candidates left unranked are ranked according to the party’s ordering. The average placing is then calculated for each candidate, and seats are assigned from smallest to largest average ranking.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s