[EM] The Global Fight For Electoral Justice: A Primer

[Erik Moeller]

James,

Once again, I am impressed by the deep knowledge of electoral reform
issues by people on this list, as is also evident from the data you
shared in your other message. Thank you, and I really hope the
Electoral Reform efforts in the UK are successful! A few comments:

> The single vote version of AMS (MMP) is not to be recommended because the
> measure of overall support for the various parties can be seriously distorted by
> tactical voting in the single-member districts, usually in an attempt to unseat the
> incumbent, but sometimes to keep out an "even worse" challenger.

I think of the district candidate as the party's chosen representative
of its platform to the local voters. He or she is the "chief
campaigner" in that district, and  so the party's overall success rate
in the district is up to them. If parties select poor representatives,
why shouldn't they be penalized for it? If they see an opening because
the major party put up a mediocre bureaucrat to campaign for them, why
shouldn't they benefit from taking that opportunity?

In Baden-Wuerttemberg the PR lists are derived from the plurality
results (albeit from absolute numbers in unequal constituencies, which
is obviously problematic). That means that your vote increases your
chosen candidate's chance of being awarded a seat, regardless of
whether they win at the district-level. All of this seems to
incentivize vigorous local campaigning between candidates, exactly
what we'd ideally want to see in a vibrant democracy.

At the same time, voters aren't overloaded with multiple choices for
each party. The party has made their pick of who they want to put
forward. Yet, voters get to scrutinize every candidate who gets a
seat. There are no closed lists.

Studies I've seen of the Baden-Wuerttemberg elections don't seem to
suggest that any major distortions are a real problem in practice. It
achieves comparable proportionality to the federal elections per [1].
The authors of [1] do note room for improvement in the implementation,
such as the issue of constituency size, but can't find fault with the
idea of a single-vote system. They also recognize the potential
democratic benefits of deriving party lists from the "single-member
district" (sometimes now actually multi-member) results.

> And if you are going to have effective ranking WITHIN each party list, why not just use STV-PR?

Per the above, I'm skeptical of the value of exposing intra-party
competition, as long as there are finer gradations of choices overall,
and we award seats based overall success. That's because voters have
real and meaningful choices in such a setup. For example, if the
Greens put up a candidate I really hate in my district, the step of
voting for the Left Party is a relatively smaller step, than voting
Conservative instead of Labour, so it's a real and meaningful
alternative.

That said, as with any system, the devil is in the details and we
really don't have many examples of this system in the real world. I
think it's an interesting option to consider for all FPTP countries
because of its relative simplicity and similarity to FPTP from the
voter's perspective. Provided one is comfortable with changes to the
total number of seats, it allows for similarly small districts as
before, and the voting system is dead easy.

I don't have a fundamental problem with STV -- it seems to work well
enough for Ireland -- and hope you are successful in promoting it! You
probably are familiar with this, but just in case -- these are some of
the arguments that were put up in British Columbia to defeat the 2009
STV referendum (which failed rather dramatically, unlike its earlier
instantiation): http://nostv.org/faq.html

I think ideally a referendum should be structured like the New Zealand
or Prince Edward Island one, giving voters both an opportunity to vote
for or against change, and a choice of 2-3 systems proposed by an
expert commission with a clear mandate (e.g.," find us a system that
achieves proportionality per a Gallagher index of 5 or less, while
avoiding fragmentation into more than 4-5 parties at the national
level").

Sincerely,
Erik

[1] http://www.tandfonline.com/doi/abs/10.1080/0964400032000242707?journalCode=fgrp20