ACORN

[Jester]

Almost correct. In itself, your quoted statement simply states a fact and based only on that statement my inference would not be valid. However, taken in the context of the discussion and especially as a reply to kandrathe's claim, I felt that that inference was justified. Since these discussions are not as formal as a mathematical proof, at least some of the meaning can come from the context. Indeed, must come from the context, otherwise the verbosity required for rigor would lead to rigor mortis.:)

I don't know ACORN internally. I don't work for them, I've never made a study of their methods, I didn't interview hundreds of their employees. Indeed, if you'd asked me months ago what ACORN was, I'd probably have told you it's the seed of an oak tree. I am in no position to make strict, factual claims about negatives: I cannot say, with any certainty, that ACORN is not committing fraud. However, that is trivial. I cannot say you weren't one of the Apollo astronauts either, but I have no reason to believe so.

However, the remainder of the argument, that registration fraud is a necessary prerequisite to voter fraud, still stands, even if not applicable as a rebuttal to your statement.

Sure. Maybe there should be an investigation into the matter. Of course, there have been some, and now even the FBI is going to have a look. The Republicans have been pushing hard to unearth 'voter fraud', which they either believe is a huge problem, or want other people to believe that it is. Given how long that pressure has been going, how little it has turned up in the way of substantive problems, and how none of it has managed to implicate ACORN in anything worse than imperfect hiring, I'm not sure Bayes would be counselling me to revise my probability estimates upwards at this point.

New evidence would change the map. I'll be waiting when it gets here.

However, consider the following:
1) ACORN personnel have been involved in registration fraud (acknowledged even by ACORN).
2) Voter fraud exists (pretty much common knowledge)
3) Most (non-computer based) voter fraud requires registration fraud to exist (my, perhaps not applicable to you, but still un-rebutted argument).

1 is not only acknowledged by ACORN, but they appear to be actively combatting it. Now, there is always the possibility that this is a ruse, to hide actual fraud, but we have no evidence for that proposition either, yet. 2 is obvious, although the apparently trivial magnitude of the problem makes me wonder why this has become an event mentioned offhandedly in poli-sci journals, let alone in live debate by a major candidate, unless that candidate is using it as a political weapon, and not as an actual problem to be solved. 3 is true. However, we cannot demonstrate the proportion of faulty voter registration to voter fraud, and all indications are that it is a low-to-negligible proportion.

To conclude, from these facts, that "We do not have reason to suspect . . ." causes me to wonder just when your suspicion is aroused. Let's say we were speaking of a sinking boat. Would you be suspicious of the integrity of the hull when there was water in the bilges? When the gunwales were awash? Or not till you were sinking for the last time?

Well, let's say the FBI finds something, a document that didn't get shredded, an informant willing to come forward, some district where ACORN is extremely active and fraudulent votes turn up far more often than an equivalent district where they are not active. Then I'd be suspicious. Right now, what we have are a bunch of accusations from people whose accumulated trust with me is a negative function (I'm slightly less likely to believe the sky is blue if Karl Rove tells me it's true) and very little else, except an increasing track record of uncovering nothing else.

I think I just heard Bayes turn over in his grave ;) Russell's outlook is all very good for mathematical discussions, but in real life it is seldom (Kant would say 'never') that we have complete information. We must draw conclusions from incomplete information and not become paralyzed by partial ignorance. We gather what information we can, we adjust our assessment as well as we can based on this information, and at some point, for practical purposes, we determine that something is 'true' or 'false' even if the mental probability we've arrived at isn't the perfect 1 or 0 we would like.

Yes, but I am not asking for certainty. I am asking for some evidence which would move the probability up from being trivial (it's higher than Russell's teapot, you're right, that was a misleading analogy), to something I would be concerned about. For practical purposes, though *not* in a deductive sense, I am looking at this as thought it is false. The likelihood of it being true, from our current set of knowledge, seems very low. The likelihood of it being both true and a big deal seems even lower.

Because, of course, the FBI is an organization that is completely unbiased and has never railroaded any group in its history? You see, your requirement for absolute proof is a never ending regression.

You are honestly reading what I'm saying as a request for absolute, unassailable, ironclad proof? I'm asking for anything that would even cause me to *suspect* that these accusations are true, beyond mere circumstance. Specifically, I have to have reason to believe that the hypothesis of widespread fraud or conspiracy to commit fraud is more likely than the hypothesis that it's not happening, and right now, that's not a battle ACORN is losing.

And, of course, the FBI railroads people. Their say-so would not be absolute proof. However, if they themselves want to be taken seriously, they need evidence. If this really goes into Oliver Stone land, I'm sure they could simply fool me into believing it. Contrary to what you seem to be reading from me, I do not claim to have some impervious filter that siphons out all incorrect information. I can be fooled, and easily with the right tricks, but I`m not yet fool enough to believe political accusations without some evidence.

Now, if an organization is tied to behavior that is repeated, is it wrong to suspect that that behavior is endemic to the organization? And if it is endemic, is it further wrong to suspect that that behavior is condoned, even encouraged by that organization?

Could you imagine how badly every organization would come out in such an analysis... frequency and intent are what is relevant here. You cannot simply take any repeated behaviour on the part of employees and call it endemic, unless you think almost everything people do is endemic to almost every organization.

Also, remember the unique prohibitions ACORN is working under here. They cannot enact internal quality control to catch these things in a face-saving manner. The registrations, once filled out, are going to be sent in, as a legal requirement. Their worst employees, the mistakes, the glitches, the liars and cheats, the whole lot, their work cannot be purged prior to public release. Other organizations do not labour under these conditions, and might look quite sketchy if they did. ACORN has taken steps to single out suspected fraudulent registrations, to fire employees caught doing this kind of thing, and to cooperate fully with investigations into these irregularities. That would, at face value, make one suspect that the behaviour is neither condoned nor encouraged. One can of course suspect what one wishes, such as that this cooperation is a sham, but it does not seem to me that, in this case, there is reason to suspect any malfeasance.

The second count you are wrong on is a bit more subtle. This discussion is relative to the statement "There is no evidence that these voter registration issues have led to voter fraud." The question of the corruption of ACORN is a separate issue. ACORN could be as pure as the driven snow (right, ever read what is found in snow in industrial areas?), and all the registration fraud the work of evil underlings who've taken advantage of this pure and naive organization. That still does not save this final statement from being completely out of place in this post. It is a total non sequitor and an emotional red herring.

Sure. That`s a plausible hypothesis. But it lives in the infinite land of plausible hypotheses until some actual malfeasance is shown. So far, what`s been dredged up painstakingly at the behest of the Bush administration is downright feeble, lazy people filling out bogus cards, voters voting fraudulently because they are merely confused about the process. If you want to posit evil underlings, I`m afraid you`re in jar-of-palladium territory for me at this point. Until we know more, I think that last statement stands just fine in the context of this thread, and is neither emotional, a non-sequitur, nor a red herring.

-Jester