[EM] APR (7): Steve’s 7th dialogue with Juho (Steve)

[Juho Laatu]

> On 12 Dec 2014, at 16:55, steve bosworth <stevebosworth at hotmail.com> wrote:

> J: The key difference in my description above was that there would be no primary. The right of a grouping to participate would be decided by other means.
> S:  Would you be willing to spell out these “other means” and explain how they would be as effective as the ones offered by APR?

I think I already mentioned some. One could for example require a new grouping to collect 10000 names of supporters. If it does so, it will get the right to nominate a list of candidtes. APR has two phases, first to get to the primary (or maybe anyone could nominate any number of candidtes??) and then to the actual election. The approach that I mentioned above would be simpler (and simplicity may nean efficiency).

> S:  What do you mean by “internal proportionality”?

I mean political proportionality within the parties. For example the right wing of a party would get its proportional share of the seats. Basic open list approach does not guarantee this.

> J:  … and geographical proportionality more exact. I could use also the idea of representatives with different weights.
>  
> S:  Of course I would be happier with this but it seems that these additions from APR could undermine the “geographical proportionality” which you want.

APR does not support geographical proportionality, but it could be enhanced to do so, or some features of APR like weighted votes could be introduces in geographically proportional systems.

> S:  Yes, in practice, traditions can slow progress but it would help me to understand your values better if you could specify the long term ideal electoral system for your country toward which you would like your country to move as fast as possible.

What I wrote describes pretty well how I would like to change the system (in the next reform within few years) even if traditions would not slow the progress. This approach includes some tendency to keep the system understandable to the current voters and avoiding the risks of too radical changes at one go. One can not change the population, so one has to serve it, taking into account what its thinking patterns are.

> J:  Better political and geographical proprtionality can be achieved in many ways, with or without APR. For example by allowing also small groupings to win seats (seats allocated at national instead of regional level, and no cutoffs in some countries), and to improve political propotionality within the parties using ranked votes (in STV style).
>  
> S:  APR’s Primary provides an automatic mechanism by which “small groupings” can win seats, as well as determine where the “cutoffs”, if any, will be.  How would your preferred mechanisms avoid making these decisions arbitrarily?

Also other systems than APR can be proportional and allow small groupings to get representatives. APR could allow also groups that are considerably smaller than 1/n of a representative body with n seats, because of the weighted votes. This would be difficult for systems with equal weight representatives. They can be accurate though (not arbitrary), to the level of one representative, without any additional tricks.

> J:  Geographic proportionality may be quite exact at district level, but if those districts are large (as typical in multiparty countries), geographic proportionality within those large districts may be poor. Geographic proportionality is typically considered good because it supports local representatives that are close to the voters and they represent the values of those voters better that representatives that have no connection to that area.
> S:  How do you answer the argument that APR has the advantage of providing an objective test of the extent to which these “local representatives” are (or are not) “close to the voters and represent the values of those voters”, i.e. a test of what you report as “typically considered” by some unknown source, based or not based on evidence.

APR allows voters to elect local representatives. Traditional geographically proportional systems (including that of the US) guarantee proportional local representation. APR can thus not provide the same thing (the guarantee). One has to decide what one wants and pick the method based on that.

> S:  Please explain more fully, do you mean “revealed” voluntarily or even when citizens are determined to keep them secret?

If some voter announces her association with some opinion group, or if that association will be recorded in some closed or open registers, or if the voter's party specific ballot paper will be seen by others, then that voter's preferences are revealed. Some cases are voluntary, some mandatory, some could be coerced.

> J:  ... I believe the system [APR] also gives those voters that reveal their preferences somewhat easier or more efficient ways to influence the outcome.
>  
> S:  Do you mean this “ease and efficiency” is an advantage offered by APR as I do? Or that citizens who publicly join a particular association during the Primary will be more vulnerable to being bribed or coerced?  If the latter, please explain.

My understanding is that whatever way you vote in an APR primary, that vote will be recorded in a way that reveals voter's preferences. To my understanding in the actual APR election votes are secret. Voters of the primary can influence the outcome of the primary, and thereby also indirectly the outcome of the actual election. Those voters that will not vote in the primary (maybe because of privacy concerns) (or will not reveal their true preferences) will have less say in the outcome of the election as a whole.

The easiness part of my comment referred to the more difficult process of voting as a voter that has been automatically tied to a regional association (that may have many kind of candidates instead of a group of homogeneous candidates that are close to the voter in any case).

> J:  In countries that support full voter privacy, any steps in that direction would probably be quite unacceptable. …
>  
> S:  You seem not to have accepted my earlier explanations of the different exact ways that each APR voter can keep their real party preferences absolutely secret.  Please explain why you do not accept that explanation?

Note that systems that emphasize privacy typically make it impossible to the voters to reveal how they voted. Being able to hide how one voted is not enough to meet this criterion.

> J:  … I'm used to a system where the system intentionally makes it impossible to the voters even to prove (to outsiders) how they voted. …
>  
> S:  APR is one of these systems.

No, APR makes it possible to the voters to prove how they voted. They could prove that e.g. by showing their association specific ballot paper to other people.

> J:  …This is related to the risks of vote buying, coercion (= violent husband telling his wife "how the family will vote"), risk of revealing one's hidden opinions, risk of group pressure against minorities etc.
>  
> S:  Yes, and that is why I also favour APR’s methods.

But isn't it possible in APR to vote for some association and then show to one's husband the association specific ballot that one got?

> > > S: Do you not see that APR?s Primary would make it easier for new additional parties ?associations? also to field candidates and that this would allow them and their electors to benefit more, both qualitatively and from the exact proportionality offered by APR?
> > 
> J:   I think this can be achieved also without a primary. APR may have some nice ideas on how this process could work, but it is possible to achieve similar kind of results also with much simpler arrangements.
>  
> S:  Please give me details of these “much simpler arrangements” that you have in mind.

See my first comment above on how one could avoid the primaries and achieve "association formation" by other means.

> S:  I have not asked for a “generic answer” but feel that it would be helpful if you would describe your preferred alternatives to each of the following claimed advantages offered by APR:

Yes, I can at least comment the properties one by one (and I think I have already commented all of them at some level). I may not have an alternative module for APR in each specific point.

> J:  All systems have some problems, rounding errors and there may be tradeoffs when trying to meet many different requirements at the same time. It is very much ok to trade between all different requirements so that the overall quality of the system (taking also into account practical limitations like the political environment) will be maximized.
> 
> S:  Tradeoffs are only necessary if 2 or more relevant values conflict.  The only 2 that I see as potentially in conflict are:  A) the desire to maximize the representation of each citizen by the rep she most trusts, and B) the desire to have a legislature that is most likely to make laws based on evidence and rational thought.  Because “A” by itself could theoretical produce mathematically a dictator in the legislature, “B” has prompted APR’s 10% limit on the weighted votes that can be retained by any rep.  This means that any law would have to be agreed to by a group of reps who together have more than 50% of the weighted votes, i.e. at least 6 reps.
>  
> Do you see any additional values or yours that are potentially in conflict and which would require more tradeoffs in your preferred  electoral system?  If so, please explain.

Yes, there are many design decisions to make. One very typical is simplicity vs. complexity with some added features. Simplicity makes the system easy to understand and use to regular voters and thereby improves the democratic process.

> J: Working methods could certainly be improved also in APR, since I guess it is still pretty much a draft method.
> 
> S: Perhaps mistakenly, currently, I do not see any remaining flaws in the most recent “draft” of APR.  Please tell me about any that you see and suggest any corrections.

I think you can read all my comments either as doubts about if the design decisions are the best possible, or as more neutral questions wondering if the system could still be improved, or as questions to you on what kind of system you really want, and how those needs could best be met.

> J:  In my terminology, I think you are saying that APR's political proportionality could approximate the results of geographic proportionality (fully or to some extent) if the voters so wish. …
>  
> S: Yes.
>  
> J:  …That is different from implementing geographic proportionality (that guarantees each region its proportional share of the seats).
> 
> S:  I now understand that your “geographic proportionality” will be entirely determined by the number of citizens who happen to live in the relevant district.  Do you accept that this means that in some cases, some of these citizens will not be able to help elect any candidate that they like or trust.  Please explain why you want to deny these citizens this APR option.

If voters can not change the district that they vote in, it is possible that they do not trust any of the candidates of that district, but they could trust candidates of some other district. In my previous mail I explained how voters might be able to change their district. In addition it is possible to develop systems that both implement geographical proportionality and allow voters to vote candidtes of other districts (such systems are however probably unusual). I don't want to deny voters this option, but I might opt not to support it in many practical systems since it does not add very much, but may add complexity, or alternatively remove geographical proportionality (that may be wanted for some reason).

> J: Scientific method can be used in a very rational way. Same with any logical and sensible discussions. Political and social discussions often do not pay much attention to respecting such rigid practices.
> >
> S:  Do you give highest importance to “evidence and rational thought” in your own “political and social discussions”?

I try to respect the "rational thought" criterion (in all discussions where that approach makes sense). "Evidence" is just one rational tool among others.

> I want to assume that you do because you are freely engaging in our dialogue.  If you do value “reason” as defined below at the end of this post, we both want to discover an electoral system that maximizes the chances that rational laws will be made.  If so, I would very much appreciate it if you would detail your constructive criticisms of my above explanation of how APR would be such a system.

I think I have commented all points at some level. At some points I have given general answers to cover numerous points. Please indicate if there are some remaining key questions left.

> J: The scientific method and Wikipedia are good examples of areas where people have been able to expand such rational practices also to discussions of a group. Political discussions may include some individuals with rational thinking and argumentation, but as a whole, political debates are something else. The best strategy in political discussions (if one's target is to win the debate in question) usually is not that of rational argumentation.
>  
> S:  Winning is not the most important thing for me.  Instead, I hope we can work together to discover the most rational and practical electoral system in the above sense.  What do you think?

Yes, I think the whole EM list is based on rational discussions. People also defend their own viewpoints and their favourite methods sometimes a bit more than what rational discussion practices would require, but in general this list has good traditions in being open, constructive and rational.

Juho




-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.electorama.com/pipermail/election-methods-electorama.com/attachments/20141212/247dac97/attachment-0001.htm>