[EM] Anyone got a good analysis on limitations of approval andrange voting? (long)

[Juho]

On Nov 10, 2009, at 6:48 AM, Matthew Welland wrote:

> Thanks all for the discussion and pointers. I still can't concretely  
> conclude anything yet but here are some rambling and random thoughts  
> based on what was said and my prior experiences.
>
> Plurality
> Leads to two lowest common denominator parties which are not  
> accountable to the voters. This conclusion supported by real world  
> observation.
Use of single winner districts has this tendency in general. The other  
single winner methods below give some more space to third parties but  
if you want to get rid of favouring the major parties and get  
proportional representation of all the parties/interest groups then  
some proportional multi winner methods method could be used.

> Feels right to the non-critical mind, "one man, one vote"
> Very fast at the polls
> Approval
> Encourages participation of minor parties and thus should keep the  
> big guys paying attention to a wider base.
> Almost zero marginal implementation cost. Hanging chads count just  
> fine :)
> Understandable by anyone but feels wrong at first "not fair, you get  
> more than one vote".
> Apparently has a terrible flaw but no one seems to be able to  
> articulate it in layman terms. No real world experience available to  
> illustrate the problem. Here is where I need to learn more. Data  
> provided to date is unconvincing to me.
Here's one example. We have left and right wings with approximately  
50%-50% support. Left wing has two candidates (L1, L2). Most right  
wing voters approve only the right wing candidate (R). Some left wing  
candidates approve both leftist candidates but some approve only one  
of them. Right wing candidate wins.

In order to avoid this problem left wing might recommend all its  
supporters to approve both left wing candidates. If they do so left  
wing will not have the above mentioned problem of vote splitting but  
on the other hand L1 and L2 will get the same number of votes since  
nobody can indicate if L1 is better than L2 or the other way around.  
Choice between L1 and L2 is quite random since the decision will be  
left to those left wing voters who don't follow the recommended  
strategy and to those right wing voters who approve also one of the  
left wing candidates.

In this example the problem thus is that voters can not express at the  
same time both that left/right wing is better and that one of the  
candidates of that wing is better than some others of them. This  
problem may lead to interest within the left wing to nominate only one  
candidate (=no spoilers, no leftist third parties) and we might be  
close to the plurality related problems again. Approval may work in  
this case quite well as long as the third party is small and all its  
voters understand that they should approve also one of the major  
parties (left), but when the support of that third candidate grows  
things become more complicated.

> Does not meet the desire of some to be able to differentiate between  
> "I like", "I like a lot" etc.  (note: this seems like perfectionism  
> to me. Large numbers of  voters and opinions all over the bell curve  
> should make individual expression at the greater level of  
> granularity irrelevant.)
Out of the discussed methods Range is the only one that can express  
"like"/"like a lot". But it has its own problems (partly due to this  
property).

The example above tries to demonstrate that while large number of  
sincere Approval votes might statistically lead to a good result there  
is the risk that the votes will be not as well in balance (for  
strategy and candidate positioning related reasons). I think it is a  
general assumption that in Approval voters would not vote sincerely in  
th sense that they would approve those candidates that they consider  
"approvable" but they would follow strategy "approve part of the  
leaders (likely winners) and candidates that you prefer to them".

> Very fast at the polls. Pick yer favorites and head home for beer  
> and telly.
> Range
> Can break the vicious cycle of plurality
I didn't understand this. (If this is about the two party dominance I  
commented already above.)

> Not voting for someone at all can have a strong influence on  
> election outcome. This is very non-intuitive and would take some  
> getting used to.
> Allows for nuanced voting.
This is the benefit of Range. The related problem is that while this  
works well in non-competitive elections (e.g. polls, olympics with  
neutral judges) in competitive ones (e.g. political elections) voters  
have an incentive to exaggerate. This may lead to Approval-like  
behaviour where most voters give min and max points to most  
candidates. In that case Range would be very much like Approval.)

> Pain in the ass at the polls (relatively speaking). You can't safely  
> disregard the candidates you don't care about so you *have* to  
> assign everyone a ranking, possibly addressable by defaulting to  
> zero for all candidates? This is considered a feature and I agree it  
> has merit. But in reality it is a deal breaker for joe six pack and  
> co. (and for lazy sobs like me).
It is a good practice to consider unmarked candidates to be "raked  
last" in all methods. This is because otherwise the most unknown  
candidates would benefit (and might win) while known strong candidates  
could get lots ow negative (lower than default) votes. (This comment  
maybe covers also point 2 above.)

> IRV
> Demonstrably broken. 'nuff said.
Yes, IRV has its flaws. All the others above have their faults too, so  
I wouldn't say IRV is in its own class in this respect.

> Suite of complicated systems that strive to reach "Condorcet" ideals.
> No regular bloke would ever trust 'em because you can't explain how  
> they work in one or two sentences.
I don't think they are too complex. Most election methods are complex  
enough so that average voters can not describe the tabulation  
mechanisms in detail. It s enough if they understand the overall  
principle and have some trust on experts that tell them that the  
system works ok. (Condorcet methods may be complex in the sense that  
they require computers to be used in the vote counting process (while  
Approval and IRV may still be hand countable), but that's another  
story.)

> Technically superior to other systems.
In the sense that in most cases voters can quite safely express their  
sincere opinions in the ballots. Also in the sense that the best  
candidate will be elected (if you think the Condorcet winner should be  
elected).

The basic idea of Condorcet methods is quite easy to understand. If  
there is a candidate that would beat any of the others, then that  
candiate should be elected. In most elections there is such a  
candidate. If there is no such candidate then additional rules are  
needed to determine the winner. Voting is as in IRV, and that seems to  
be easy enough for the voters (already in use in many places).

> Not clear what problem with approval they would solve. Unless you  
> are a perfectionist and insist that individuals express nuances of  
> opinion...
See comments in Approval above. In Condorcet (and IRV) voters may  
quite reliably indicate their preference between all candidates. In  
Condorcet voters can not express as detailed nuances as in Range (but  
they can express maximum support to both A>B and B>C at the same time,  
and risk of loss of nuances due to exaggeration is smaller).

Condorcet methods have also their own strategic problems (e.g. burial)  
but the one can expect/hope that strategic voting would not be  
feasible and common in practical elections (not even competitive ones).


> Some time ago I put together a site (primitive and unfinished[i]) to  
> promote approval voting and in the process I spent a lot of time  
> trying different systems on the web and repeatedly testing my own  
> site. I noticed some interesting things from all that playing around.
> It was very uncomfortable to go back to plurality after trying other  
> systems. It "feels" unfair and broken.
> It was very tedious voting in any of the ranking systems.
> Approval felt boring but good.
> I have checked in on this list now and then and I admit I don't have  
> the time or skills to follow all the arguments but it strikes me  
> that approval  voting is good enough to break the deadlock, at least  
> in US politics and that it doesn't have any major flaws.

Yes, it is a step forward from Plurality (unless you want to keep the  
two party system). I think there are also better methods (more complex  
though, and longer / more difficult steps from the current state).  
Also PR (proportional representation) based methods are worth  
considering (not only single winner methods).

>  The very understandable desire to be able to articulate in a finer  
> grained way in your vote is perfectionism. With millions of voters,  
> for every person on the fence about a particular candidate there  
> will be some to either side who will essentially make or break the  
> vote. If you are on the fence, approve or disapprove, it won't matter.
> So, to re-frame my question. What is the fatal flaw with approval?  
> I'm not interested in subtle flaws that result in imperfect results.  
> I'm interested in flaws that result in big problems such as those we  
> see with plurality and IRV.

I tried to address shortly the Approval related problems above. Maybe  
there are no "fatal" problems but I'd expect some (maybe less than  
fatal) problems to emerge, at least after the third parties grow in  
size and become serious contenders of the current two giants.

If you think the dominance of two (or few) major parties and lack of  
representation of smaller groups is one of the "big problems" then  
also PR (and typically associated multi winner districts) could be  
considered. The improved single winner methods (when used in single  
winner districts to form a larger representative body) are one step in  
this direction but they will not offer fill proportionality. Maybe  
this is good enough or one step in the right direction, or maybe  
exactly what you want.

Just my 2 cents,
Juho


>
> [i] www.approvalvote.org
>
> ----
> Election-Methods mailing list - see http://electorama.com/em for  
> list info

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.electorama.com/pipermail/election-methods-electorama.com/attachments/20091110/a197c0a0/attachment-0004.htm>