Thursday, December 07, 2006

Digital Voting

Zippy Catholic has proposed a formalism for why non-negotiable issues (NNs) imply that people should be abstaining from elections left and right (except for the rather bizarre case when both candidates are equally wrong on all NNs). He proposes the following:

Define each candidate's morality word P to be his platform on all of the non-negotiables in play in the race. A "1" means he is wrong on that non-negotiable, a "0" means he is right. I may only vote for candidate A without making a proportionalist judgement if the logical AND of his platform Pa with all of the other candidates' platforms Pi is equal to his original platform Pa. If it isn't, that means that some other candidate is right on a non-negotiable for which my candidate is wrong: so (no negotiating now!) I may not vote for my candidate.

His idea of "negotiating" means as between issues, so that the bits of the NNs can't be compared among one another, but that's not what I had in mind (nor is it what CAA had in mind, I think). Zippy's definition means that unless two candidates are exactly the same on all the same issues, any comparison between bits would be illicit (they simply can't be valued as against each other), so I think that his conclusion is correct. But I don't agree with his formal concept of non-negotiability, and I'll appeal to the concept of bit significance (most significant bits to the left, in this case) to explain why not.

Let's take Zippy's formalism, but replace the morality WORD with a morality REGISTER made of two 5-bit words, with a most significant word (MSW) including 5 bits for non-negotiable issues and a least significant word (LSW) including 5 bits for other issues (negotiables). It looks like this:

xxxxxxxxxx
MSW LSW

Since 1 is a wrong, these words indicate a harm score for the candidate. The MSW has a score from 0-31, as does the LSW. To include my concept of great moral weight for these issues, I will weight the MSW by a factor of 64 (which will of course seems contrived, but it's selected to illustrate the principle). So the harm score for each candidate would be given by the following formula:

H = 64*MSW + LSW

where H is the harm inflicted by the candidate if elected.

My point is that if you are voting against a particular candidate, the good accomplished by your vote is the harm score of the person you voted against minus the unintended consequence of the harm score of the candidate you had to support in order to keep the other guy out of office. In the following, G(VA) means the good accomplished of voting against the candidate, H(VA) is the harm that would have been inflicted by the candidate voted against, and H(VF) is the unintended harm of the candidate that you voted for (NOTE: Grammar regarding prepositions has been sacrificed for easier subscripts).

G(VA) = H(VA) - H (VF)

Because of the weighting factor, a candidate who is wrong on any of the NNs is going to have a higher harm score than someone who is wrong even on all of the other issues. Formally, this is because the LSW can never have a value higher than 31, and any non-zero NN value is going to be at least 64. That doesn't mean that you can't choose between people with different scores in the MSW. In fact, in my formalism, you can rather easily do this by assigning the most serious issues to the most significant bits (say, abortion, stem cell research, cloning, euthanasia, and same-sex marriage, left to right). So long as your rationale is voting against a candidate in relative error on the non-negotiables, it's licit to vote for the other candidate. This is analogous to voting in favor of a more restrictive abortion law, even an unjust one, if a less restrictive law is already in place: the harm you avoid is greater than the unintended harm you obtain. But that rationale will never obtain in the case of someone wrong on even one NN as compared to the other candidate, because that difference on the NNs will always (based on the weighting factor) entail more harm than being wrong on all the negotiables combined. Therefore, there can never be a net avoidance of harm by voting for someone in relative error on the non-negotiables, just as there can never be a net avoidance of harm by voting for a more restrictive law to prevent the passage of a less restrictive abortion law. Both laws are wrong, but one clearly does more harm than the other, so voting against it avoids more harm.

Since my concept turns on inherent disproportion between the WORDS, it doesn't mean that bits in the MSW cannot be weighted and compared in comparing the relative harm inflicted by the candidates. It simply means that any degree of relative correctness on negotiables will be irrelevant as compared to relative error on the NNs.

9 comments:

CrimsonCatholic said...

Forgot to turn on the comments. Sorry.

Anonymous said...

I am having a hard time seeing how this says something other than that the non-negotiables are in fact, from the perspective of the voter (commenter Decker2003's contention), negotiable.

CA says:
In some political races, each candidate takes a wrong position on one or more issues involving non-negotiable moral principles. In such a case you may vote for the candidate who takes the fewest such positions [...]

CA states outright (without any explicit justification) that each of the NN's has equal weight, so that a voter can choose a candidate based on the numeric sum of how many the candidate supports irrespective of which ones in particular the candidate supports. That directly contradicts the bedrock moral principle that there is no such thing as licitly choosing the lesser of two evils.

CrimsonCatholic said...

CA states outright (without any explicit justification) that each of the NN's has equal weight, so that a voter can choose a candidate based on the numeric sum of how many the candidate supports irrespective of which ones in particular the candidate supports. That directly contradicts the bedrock moral principle that there is no such thing as licitly choosing the lesser of two evils.

Yeah, that wouldn't be justifiable. You have to evaluate the relative harm worked by having a candidate publicly proclaiming the views, and you can't accomplish that simply by tallying up who has the most hits on the list. Abortion, for example, strikes me as being obviously having more moral weight than same-sex marriage. It's entirely possible that someone with a defective view on euthanasia and same-sex marriage would inflict less harm through his views than a candidate who is wrong on abortion.

My formalism doesn't suffer from this defect, and neither does most of the CAA guide. Note the following statements:
"In those cases, citizens must vote in the way that will most limit the harm that would be done by the available candidates.
...
Where every candidate endorses positions contrary to non-negotiable principles, choose the candidate likely to do the least harm. If several are equal, evaluate them based on their views on other, lesser issues."

Other statements seem confused. For example, this one deals exclusively with voting record, without even dealing with the sin of public advocacy:
"Such a voting record is a clear signal that the candidate should not be chosen by a Catholic voter unless the other candidates have voting records even less in accord with these moral norms."

Others are just plain wrong, e.g.,
"In some political races, each candidate takes a wrong position on one or more issues involving non-negotiable moral principles. In such a case you may vote for the candidate who takes the fewest such positions or who seems least likely to be able to advance immoral legislation, or you may choose to vote for no one."

That last one was particularly odd, because the subsequent explanation was accurate. It even (rightly, in my view) left open the option that not voting at all could rightly be judged to promote the common good in particular instances. It appears to simply be a mistake: right reasoning, wrong answer. Judging it on the whole, I think the only systematic error in the whole CAA guide was failing to treat public advocacy (as opposed to voting) as a sin in and of itself. What you cited appears to be an isolated case of error, not a systematic flaw in the reasoning.

Anonymous said...

Judging it on the whole, I think the only systematic error in the whole CAA guide was failing to treat public advocacy (as opposed to voting) as a sin in and of itself. What you cited appears to be an isolated case of error, not a systematic flaw in the reasoning.

Well, other than the business about there being such a thing as non-negotiables, I suppose. And it is difficult for me to address the issue of systemic flaws, at this point, because the only system of reasoning I see is that various platforms have various moral issues in them that have to be analyzed and ranked rationally in order to make a decision how to vote in a particular race.

CrimsonCatholic said...

And it is difficult for me to address the issue of systemic flaws, at this point, because the only system of reasoning I see is that various platforms have various moral issues in them that have to be analyzed and ranked rationally in order to make a decision how to vote in a particular race.

Yes, and ultimately, that is the only rational way to decide how to vote. "Non-negotiables" is one heuristic to simplify that evaluation based on the real cases one would face, and so long as it is applied according to the rules I've outlined, I suppose I don't see what the drawback is.

I guess this is what I don't understand. It is well-understood that one can use certain licit means to prevent evil even if it has the unintended consequence of causing some other evil, assuming that the prevention of the evil is a good proportionate to preventing the other evil. In the context of voting, this is specifically endorsed in the case of voting for more restrictive abortion laws. But there is no reason why the prevention of different kinds of evil cannot be proportionately compared in the same way.

For example, if you work as a secretary at a hospital that performs abortions, your cooperation with acts of direct abortion is compared against your obligation to provide for yourself and your family. They are clearly different sorts of obligations, but they can still be compared. What I can't see is why voting would be the sort of thing that would be an exception. In fact, it seems that this way of thinking (i.e., choosing to prevent the greater evil on pain of accepting the lesser) is specifically endorsed by Evangelium Vitae. So what is it about candidates that makes the cases different? We know that the evil permitted by voting against someone is an unintended consequence of voting to prevent something else from EV. There are multiple sorts of prevented evils, but that's OK, because we make those comparisons in double effect analysis, if not routinely, then at least not uncommonly.

There's something I'm missing, but I'm not sure why "non-negotiables" can't function as a heuristic for determining the relative severity of harm pertaining to certain issues as opposed to others.

Anonymous said...

Non-negotiables" is one heuristic ...

How can a categorical claim be a heuristic? Is there such a thing as a categorical rule of thumb?

CrimsonCatholic said...

How can a categorical claim be a heuristic? Is there such a thing as a categorical rule of thumb?

If heuristics aren't, then "non-negotiable" isn't a categorical claim either. I suppose they're meta-catgorical; they categorize cases where you don't have to do the work of categorizing strictly.

Anonymous said...

..."non-negotiable" isn't a categorical claim either.

I think once we've conceded that as a premise of our discussion the discussion is Jabberwocky, and we may as well go have a beer.

CrimsonCatholic said...

That might well be the case anyway, but it would seem that if we can talk about heuristics, and heuristics aren't categorical claims, then we can talk about other things that aren't categorical claims either. However, speaking of categorical claims, I've left a comment at your place regarding categorizing the sort of act that voting is. I think that is an issue that needs to get resolved to avoid talking past one another.