Did the Ohio recount amount to a criminal conspiracy?

[eomer]

Several of us got started discussing this subject on another thread where it was off-topic. Anyone who wants to continue - let's continue it here.

Specifically, OTOH, I wanted to respond to your post:

I think "extraordinary" is not synonymous with "illegal" in my mind. Doug Jones suggests (as I recall) that many basically honest election officials take shortcuts -- which may well be illegal -- to avoid what they regard as pointless busy work.

I'm not quite sure how I would define "extraordinary." But I might be more impressed by 5 examples of "extraordinary" (to indicate a conspiracy -- for lesser purposes, of course one might do) than by 50 examples of "illegal."

I swear I am not trying to bust your semantic chops here. The underlying issue is whether the behavior of election officials seems bizarre under a benign interpretation. Certainly that should take account of patterns. Some of the reports kinda cry out for follow-up -- what's with those Clermont ballots, anyway?

I agree that it is difficult sometimes to choose which word to use to characterize some of these acts. However, I would argue that 'illegal' is the correct word in the cases I have in mind.

One example (hypothetical for now) is a worker who is counting votes manually. That worker is required to count each individual vote according to whatever standard is specified by law. If the worker intentionally counts the vote in a way that is not the same result as implied by the legal standard then an illegal act has occurred, no matter what the motivation is. And it only takes one vote intentionally miscounted to say that illegality has occurred. Surely we must all agree that this is the case?

Here's another (hypothetical) example I would call illegal. Let's say the worker has a deck of punch cards for a precinct that has been pre-selected for the 3% recount. The worker goes through the deck and "cleans up" the chad issues by removing any hanging chads. This would constitute illegal tampering with ballots.

Hopefully we agree that the two hypothetical acts I described are illegal.

Now, did these hypothetical acts occur in the Ohio recount? I say there is no other explanation for the results than to conclude that these (or something similarly illegal) occurred. How else is it possible that in every county but one the 3% manual recount and machine recount matched?

For all its faults, the 2000 Florida recount looks golden compared with Ohio 2004. Remember the video of workers examining punch cards with magnifying glasses to determine each individual vote according to whatever standard they were told was the legal standard? To accept the Ohio recount as a true (and therefore legal) count of the votes we would have to believe there was not one single chad in all of the 3% precincts in all of the punch card counties that hung and therefore would have been counted differently by the machine than by hand. In addition we would have to believe that there was not one single ballot in all of the 3% precincts in all of the op-scan counties that would have been counted differently by the machine than by hand. This is simply not possible. In fact, I find it incredible even at the single county level. Before the Ohio recount got going I posted a number of times saying that there would surely be a full hand recount in almost every county because there would almost always be at least one ballot that would count differently by machine than by hand. The result that there was only one county that admitted a difference and performed a full recount is a result that is simply not credible.

How about the conspiracy part? Well, the same totally incredible result occurred in every county but one. The only reasonable explanation for this fact is some type of communication directed at many or all of the counties suggesting the approach of forcing the 3% counts to match.

I had said previously I would go back and look at the observer reports and cite evidence. I haven't done that here because when I thought about it I realized that the basis of my belief that there was a criminal conspiracy does not really rest on eye witness reports -- it comes from the sheer impossibility of the result if the count had been true and the fact that the same approach was used in all but one county. It's sort of like having the dead body and the murder weapon lying on the sidewalk but no eye witnesses to the crime. It is clear a crime was committed. What we can't prove so far is whodunit (although we have at least one prime suspect who had motive and opportunity).

By the way, what I am explaining here is why I believe what I do. The level of proof I used is whatever my internal thought process felt comfortable with. But I also believe that the circumstantial evidence I point to would be sufficient to trigger a criminal investigation if the rule of law actually applied to Ohio elections.

[OnTheOtherHand]

1. very reasonable -- now, on illegal, conspiracy, and extraordinary...

I think it's pretty darn obvious, as you say, that a bunch of one-off errors that (in my understanding as in yours) should have triggered county-wide recounts weren't allowed to do that. I agree that "illegal" is the correct word. I tried not to imply that "illegal" was an overstatement.

Doug Jones points out:

Routine error is fairly common in elections, no matter the technology used. Too frequently, these are not reported, even when state law requires it, simply because it is inconvenient. Otherwise honest election officials sometimes delete files, shred paper, conduct undocumented recounts and otherwise commit technically illegal acts simply because it is less work than properly documenting what is obviously an isolated and unimportant incident.

http://www.cs.uiowa.edu/~jones/voting/aaas2005.shtml

He perhaps should have put "obviously" in scare quotes, because routine "technical" illegality could serve to conceal malicious actions.

I don't think any sort of conspiracy, or imputation of "extraordinary" behavior, is required to understand why basically honest BoE officials wouldn't order a countywide recount based on one-off errors, in a race where no quantity of one-off errors would suffice to alter the outcome. Any sensible BoE official would be reluctant to order a recount in that circumstance, so I see no need for coordination.

Of course I am not vouching that all the BoE officials were basically honest, or that none were involved in a conspiracy or extraordinary behavior, or that no fraud occurred. I am just trying to articulate my response to this particular line of argument. I am more interested in discerning intent than legality per se.

[Febble]

2. Well at the very least

it suggests that if you are going to have transparent elections, the procedures need to be do-able. It is no good mandating random selection of precincts and full hand recount if there is a single error, if the result is going to be that harassed but fundamentally honest election officials try to get round the rule.

So the first thing perhaps that needs to be done is devise do-able recount procedures that can be scrupulously adhered tom, which would probably involve some standard of acceptable error.

My understanding of the San Diego story was that the mandatory recount came out absolutely accurate, but the random recount had a few insignificant deviations.

If all recounts are going to have a few insignifant deviations (all our hand-counted recounts do) then "significant" needs to be legally defined.

[eomer]

4. Okay, but...

we also have to remember that these "insignificant deviations" were coupled with the hand-picking of precincts. The combination of these two tactics would allow someone who knew which precincts were fraudulent to keep the fraud from being uncovered.

[Febble]

5. Oh, I wasn't arguing (in this instance)

that Ohio wasn't (or was) fraudulent. I was just saying, in my PollyAnnish way, that if we are thinking in terms of how the system needs to be reformed, one reform that needs to be done is to design audit procedures that are do-able - and can thus be reasonably (and scrupulously) enforced. Un-enforceable procedures are a gift to anyone who wants to smuggle fraud under the cover of plausible deniability, because anyone can just claim to be saving tax-payers' money by ignoring "obviously" irrelevant "errors".

That's what I was going on about in the other thread. At best, many BoE's seem lax, and perhaps they are lax because the rules aren't realistic. But once you have a culture of laxity then it is easy to abuse the system.

I didn't offer a view on whether or not the system was abused in this instance!

[eomer]

6. Okay, then I agree (in this instance)

Your point about do-ability of an audit seems right to me.

And when we are creating a more do-able audit protocol we probably need to raise the size of the sample from 3% to something higher or whatever modifications to the design are appropriate so it gives us a reasonable assurance. (I'm sure that's what you have in mind so this is not disagreement).

[Febble]

8. Yes

And given that fraud occurs at the level of the precinct, there ought to be a minimum percentage of precincts audited, as well as a minimum percentage of ballots. Maybe precinct could be stratified by size, and a minimum number randomly selected from each stratum.

Thinking about the UK system, and units of analysis, one difference that may be relevant is that our polling stations are small - we don't have any kind of separate precinct system - each polling station serves a portion of a constituency And our constituencies serve about 20,000-40,000 voters- so a lot less than one of your counties. Cities will be divided into several constituencies. No counting is done at the polling station - sealed ballot boxes are transported to the constituency count and each constituency conducts its own count. If there is a recount, the whole constituency is recounted, which is a lot more than recounting a precinct, but a lot less than recounting a large county. So once the ballot boxes are emptied (and they really are simply tipped out onto the counting tables), we only have one unit of analysis.

And of course whoever wins the constituency becomes Member of Parliament for that constituency.

[eomer]

3. Point taken, but...

I don't think we have the luxury of waiting until we can find evidence of motive. If we do then we are totally vulnerable to a plausible deniability scheme. And we know that plausible deniability is a favorite tactic of politicians.

So I say screw motive. Democracy is dead and lying on the sidewalk. Prosecute the people in Ohio who intentionally miscounted the recount. If they did it just because they were lazy then put their lazy asses in jail.

Edit to add: the reason I feel strongly about this is that we were entitled under the law to a recount that told us something. What we got was a recount that effectively told us nothing. If we are willing to accept this level of irregularity without raising a stink then we may as well all go back to watching CNN talk about runaway brides.

[OnTheOtherHand]

9. yes! no! auuggghhh!

A system that encourages basically honest people to act indistinguishably from crooks is an intolerable system.

(I don't think the Ohio system is quite that bad -- or perhaps I should say, bad in quite that way. I think we got observably worse behavior from Blackwell, who wasn't even breaking the law for the most part, than we got from the BoE folks who may well have been breaking the law. But we don't need to parse the degrees of intolerability too finely.)

If you want to lock up all the BoE folks and put their lazy asses in jail, then I think you may be missing part of the point. The time and money you are committing is coming out of someone's hide, and quite possibly at the cost of something else that we really, really want BoEs to do.

So, the reason I don't say "screw motive" is that understanding motives may provide us with really crucial insight into how to get the results we want in the future. One can't design effective policies without thinking about the people who have to implement them.

I don't really think that the Ohio recount told us nothing -- I think it appreciably reduces the plausibility of Lone Ranger tabulator hacks, without actual conspiracy (at least after the fact). If there were 10% discrepancies in 20% of precincts, or 3% discrepancies in 70% of precincts, that should have triggered some honest BoE official's suspicions somewhere. But of course we can't assume that the BoE officials are honest. (And some people believe the conduct of the recount is conclusive evidence to the contrary, although I'm not convinced.) Again, I agree with you that that outcome isn't satisfactory. We shouldn't be left to cross our fingers that the tampering with ballot security and so forth was generally benign.

[eomer]

11. The main thing (and maybe the only thing) that I really, really want

BoEs to do is to faithfully and precisely apply the procedures specified by law. I don't want the question of whether or not they have suspicions to play any part in it. The BoEs, if they have suspicions, should notify law enforcement, the courts and the public but otherwise still precisely follow the lawful procedures until told to do otherwise by a court order.

When I say the recount told us nothing, I mean nothing we can count on. Once we see the level of irregularities that occurred here then we can no longer safely assume anything. If the BoEs did not feel bound to follow the lawful procedures then we really don't know all the ways in which they deviated. If it were the case that the recount was safe, reliable and compliant; produced insignificant error rates; and then those insignificant error rates were fudged over but that was the only irregularity -- then I would agree that the recount had told us something. But unfortunately the recount was much more messed up than that. The precincts were pre-selected and the ballots in the selected precincts were apparently manipulated and tampered with. Under those conditions you cannot be sure of any results that came out of it.

With regard to the time and money commitment (not sure whether you mean to have done the recount lawfully or to have thrown the lazy asses in jail but, either way...) I can think of a couple other ways we could save billions of dollars here and there and would get those ways done first. Especially in Ohio there are some pretty famous cases of wasting (or, more accurately, stealing) the taxpayers money that we could address first. The last place I would be thrifty is anything that would jeopardize the reliability of our elections.

The important point that I'm not sure we agree on is that to me the actions of the BoEs and their workers were an outrage. To willfully deviate from the lawful procedure for determining the election result when you know that the whole country and in fact the whole world are watching and waiting for the result and consider it of paramount importance -- I mean what would it take to get these folks to take their job seriously? If they don't get the seriousness of their job then we need to educate them about it.

And finally I should clarify that I really don't want to throw any lazy asses in jail. I do want to throw any conniving, thieving asses in jail. Maybe a criminal investigation and the possibility of jail time for the lazy ones would help catch the thieving ones.

[OnTheOtherHand]

12. I think we agree on more than we disagree on, but--

please forgive me for not reiterating every point where I agree with you 90-100%. It's a good discipline, but I just don't have time!

I confess that I am not generally outraged by the actions of the BoEs and their workers. I think the formulation about "willfully deviat(ing) from the lawful procedure" is not very helpful. I'm prepared to be outraged with BoEs, but not on that particular basis. Some of the BoE staff who (apparently) violated laws seem to have been taking their responsibilities a lot more seriously than Blackwell was when he was within the law (I had written "following the law," but that seemed too generous).

But I agree that apparent tampering with some of the recount precincts renders the recounts in those counties totally useless. I guess I can't honestly say that I am outraged about it, but at best I think it was a terrible idea.

From a wonkish standpoint, when a bunch of BoE people seem to be behaving in terribly irresponsible ways, and when there is legitimate doubt about their motivations in doing so, it is time to seek more information IMHO.

[eomer]

13. I agree that we agree on more than we disagree, and in particular...

I agree that "{f}rom a wonkish standpoint, when a bunch of BoE people seem to be behaving in terribly irresponsible ways, and when there is legitimate doubt about their motivations in doing so, it is time to seek more information IMHO."

[eomer]

14. Our sad history of ulterior motives.

One more thing on why I look at this the way I do.

Our nation has a very sad history of voter disenfranchisement and a large part of that history has involved variations of one basic trick -- a mechanism that has an innocent explanation but an ulterior motive.

This history makes me especially wary of "innocent explanations". The bottom line in Ohio was that the intent of the recount provision was entirely thwarted. I'm going to assume the ulterior motive until it is proved otherwise. Historically I am going to be right.

[OnTheOtherHand]

15. excellent point; my take on wariness and assumptions is a bit different

I think that we approach these issues from a somewhat distinct point of view because I've had many different forms of government failure dinned into me. So I share your wariness of innocent explanations -- I certainly don't assume that any innocent explanation is probably true, although by DU standards it must sometimes seem that I do -- but I wouldn't "assume the ulterior motive," either.

Since we both want to know more, it shouldn't matter so much what we are assuming in the meantime, as long as we aren't hell-bent on locking up the BoE folks regardless.

[eomer]

16. Fair enough.

As long as we are going to get to the bottom of it sometime in the near future then I agree it doesn't matter much what we assume in the meantime.

At present there is a lawsuit looking into it. Let's both hope we can have another discussion soon that is based on knowing exactly what happened.

[garybeck]

7. the first count was a criminal conspiracy. the recount was too. and

the 05 election was as well. it's politics as usual in Ohio.

there are only 2 ways the republicans can win anything in Ohio:

1) very low turnout

2) criminal conspiracy

most of Ohio is working class. if you talk to people who have been watching their elections for years, they agree on one thing - the higher the turnout, the more democrats win. every time. When there is high turnout and the republican wins, you can bet it was stolen. they have the means, motive, and opportunity.

[OnTheOtherHand]

10. but... but...

There are many, many Ohio Democrats who have been watching Ohio elections for years and aren't convinced that the 2004 or 2005 elections were stolen. I don't understand the point of asserting otherwise.

I don't know what to do with the assertion that "most of Ohio is working class" -- obviously it will depend on definitions. Ohio's median income is close to the national average, and its poverty rate is somewhat lower, so we can take that either way.

It may be true that "more" Democrats win when turnout is higher -- I am wondering how to measure that -- how about we count the number of members in the Ohio House and see whether it is consistently more Democratic after presidential elections? Anyway, I'm not sure that matters much. If you look back to the 80s, the highest turnout was in 1984, when Reagan just clobbered Mondale (59-40). And since 1980, the Dem presidential candidate who did best in Ohio was Clinton in 1996, which had the lowest turnout. I can't imagine how a political observer could look at the turnout figures and say that Kerry obviously won Ohio.

[garybeck]

30. what?

no offense, but this statement is ludicrous:

"There are many, many Ohio Democrats who have been watching Ohio elections for years and aren't convinced that the 2004 or 2005 elections were stolen. I don't understand the point of asserting otherwise."

there are many many people in Ohio who know it was stolen. they are working hard to repair the election system there. to say that there is no point in saying it was stolen is strange to me.

if you don't like my point about the turnout, that's fine. there is plenty of other evidence about it being stolen.

I don't care what your "many many Ohio Democrats" think. If many democrats don't think it's stolen, it's because the democratic leadership is failing them. it's because the media is failing them. they only place they can find out what is really happening are handful of websites that focus on the issue. They are misinformed if they think that 04 and 05 were legit. the mountain of evidence is undeniable.

[OnTheOtherHand]

46. sorry I missed your response

such as it was.

Lots of Ohio Democrats "know" it was stolen, and lots of Ohio Democrats "know" it wasn't stolen, and lots of Ohio Democrats aren't sure. You are welcome to believe that anyone who disagrees with you is misinformed. You may find that you have a hard time learning anything that way.

[philb]

17. Yes, outright fruad was found in the recount effort

but nothing was done about it
http://www.flcv.com/greenrc.htm
http://www.flcv.com/mahonts.html & other counties
(deliberate miscounts and undercounts such as in Cuyahoga)

but more of the swing was due to manipulation of registrations, process, absentees, provisionals, etc.
http://www.flcv.com/ohiosum.html
http://www.flcv.com/summary.html Ohio

[kiwi_expat]

18. Under/over votes were not counted in the manual recounts.

"How else is it possible that in every county but one the 3% manual recount and machine recount matched?" -eomer

The counts didn't match EXACTLY in most Ohio counties. A lower standard was used: Good Enough for Government Work.

However, the main reason that the manual recounts matched the machine counts, as well as they did, is that under/over votes were not re-examined. I assume that not attempting to count the under/over votes in a manual recount is illegal.

IMO that was the essence of the "illegal conspiracy" among the BoEs and Blackwell.

[OnTheOtherHand]

19. "under/over votes were not re-examined"

I'm not sure I understand what you are saying here. I understand that disqualified ballots would be segregated from other ballots, but why would under- and over-votes not be examined?

Excerpt from Hamilton:

Over votes and under votes were immediately counted and double-checked by another worker (generally of the other party), then set aside, as were ballots for the presidential candidates other than Kerry or Bush (which tended to have no more than 5-10 votes, if that, per precinct).

Ballot stacks for Mr. Kerry and Mr. Bush were then counted separately two times, with Republican and Democratic workers trading off their stacks to confirm the count.

http://www.iwantmyvote.com/recount/ohio_reports/counties/hamilton.php

Was that unusual?

Now, the Washington Co report describes a case where the voter had apparently voted for Peroutka(?), crossed out that vote, and voted for Bush. The machine kicked this out as an overvote; the BoE members voted unanimously (over the observer's objections) that it should be counted as a Bush vote; and I think they stuck a white sticker over the Peroutka misvote so that the subsequent machine recount would math the hand recount. If the folks conducting the hand recount reach a different determination than the machines, then something has to give.

And the recount standard promulgated by Blackwell simply says that "if the computer count does not match the hand count," all ballots must be hand-counted. I don't know whether that was a deliberate preemptive attempt to trash the recount. If so, it was a damn subtle one. But it puts the recounters in an awkward position: are they supposed to assess clear voter intent (and if so, by what standard?), or are they supposed to pretend they are scanners? (For punch cards at least there were some chad guidelines.)

Recounting to assess the basic accuracy of the machine count just isn't the same as recounting to assess voter intent. It's an inherent problem. And actually, I know from my own adventures in scan recognition that it is a bit of an art form to look at a piece of paper as if you are a scanner (or actually a scanner plus a decision algorithm).

These comments don't settle anything about the conduct of the Ohio recount; I'm just exploring issues that I think should be sorted out for future recounts. A very vague, blanket statement of zero tolerance for discrepancies is actually a recipe for disaster, IMHO.

[kiwi_expat]

20. I should have said "under and over votes were evaluated using the same...

criteria that were used in the original machine count - to avoid triggering a full manual recount". Effectively, the under/over votes have not been re-examined or counted.

As for some of the BoEs settling for a less than perfect match, I believe the Cuyahoga observers' reports would confirm that. Please correct me if I'm wrong and I will find another example.

Perhaps Blackwell was not responsible for the under/over votes not being re-examined/counted. Maybe it was such an obvious way to avoid a full manual recount that the BoEs needed no prodding.

So, if 88 counties decide INDEPENDENTLY to avoid counting the under/over votes, is it not a conspiracy?

[OnTheOtherHand]

23. there sure is a lot to juggle, isn't there?

First, let me suggest -- or, rather, develop -- a distinction between two rationales for a recount: voter intent, as distinct from election integrity (i.e., identifying widespread fraud in the vote count). In an extremely close election, I would support a full manual recount to best determine voter intent. (But I would have to do more work in order to set a criterion for "extremely close." Florida 2000 would certainly count! Not everyone would support a full manual recount in this circumstance -- some would say that the outcome is basically a coin toss, and there isn't much point in throwing lots of resources into the count -- but hey, I have to be a purist about something. Of course to be serious about that, we will have to be a lot stricter about ballot security.) In a not-so-close election such as Ohio 2004, where a main concern was whether the count had been somehow hacked, I think it is reasonable to conduct a recount that uses the same criteria as in the original machine count. Even then there should be some slight wiggle room because, alas, even machine standards aren't quite deterministic; microscopic differences in alignment or scan contrast will yield different results -- and of course human beings don't have an inherent capacity to determine whether a stray mark will trigger a Mark Sense decision rule.

Alternatively, we could conduct a partial recount that uses voter-intent criteria, in order to determine whether for some reason one candidate is being disadvantaged by facially neutral application of the machine criteria. But then IMHO we would certainly need some criterion for deciding what level of discrepancy should trigger a full recount. If we are going to require a full manual recount whenever a 3% count finds any discrepancy between the machine count and a voter-intent standard, then we might as well just require full hand counts from the get-go.

--OK, now, I also agree that some counties clearly set aside the letter of Blackwell's directive, acknowledged that their hand counts did not match the machine counts, and nonetheless dispensed with a full recount -- while others went through other contortions to make the counts match. I think, all else equal, we are better off in the counties that OPENLY FLOUTED THE DIRECTIVE, accepting minor and inconsequential count disparities, than in the counties that did all sorts of weird secretive stuff to make the counts match. This situation sucks, to use a political science term of art. I don't mean to imply that BoE folks in the contortion counties were actually covering for vote-count fraud -- in general, I doubt it -- but we don't know.

If 88 counties decide independently to avoid counting over- and undervotes, it is not a "conspiracy." Should we care? Well, in Ohio 2004 -- unlike in Florida 2000 -- I see no reason to believe that finer inspection of over- and undervotes, in itself, would have altered the outcome of the election. But a recount standard that corrupts every BoE that attempts to apply it 'reasonably' is a really, really, really bad idea.

If one starts from the premise that there should have been a full manual voter-intent recount in Ohio, then one is bound to end up in a different place than I do about 2004. But that still leaves the question of what to do in the future.

[eomer]

26. My premise is that the rules should be precise and known in advance

and that the job of the BoEs is to faithfully follow them. In this case that leads to the conclusion that there should have been a full manual recount in Ohio but it is not by starting with that premise. It is by starting with a premise of the rule of law and ending up with the application of the law requiring a full manual recount.

During a recount is not the time to consider whether the rules are wise or misguided. It is a time to apply them precisely. If there is some ambiguity (which there wasn't in this case for the most part) then there will be a need to consider what the intent was. But anytime the rules are clear (like the punch card chad rules and the precinct selection rules and the number of times you count by hand and machine and what you do if they don't match) then the rules should be followed.

Once you abandon this approach then all hell will break loose (as it has).

[OnTheOtherHand]

31. OK, that is a fair distinction

I stated my point very poorly: I didn't mean to imply that you "start(ed) from the premise" that there should be a full manual recount, and derived your conclusions from it. I do think, and meant to say, that -- whatever your own personal views -- the more suspicious one is of the initial Ohio count, the more likely one is to be outraged by the failure to conduct a full manual recount. Republicans on average perceived a much greater threat to the "rule of law" from Bill Clinton's actions than Democrats did; Democrats on average perceive a much greater threat from Plamegate than Republicans do.

I certainly am not trying to argue for a general principle that election officials should ignore the rules. But I would be lying if I pretended that I believed all deviations from the rules were awful. If memory serves, one Ohio county BoE openly flouted one of Blackwell's directives on -- I don't quite recall -- perhaps whether to count provisional ballots cast in the wrong precinct at the right polling place, or whether to discard registration forms where the date of birth wasn't written on the envelope, or something fairly arbitrary like that, although I don't think it was the 80# paper rule. Reasonable people could disagree about whether it was a good idea for the BoE to disregard a precise SoS directive, but I can tell you for a fact that I for one was a lot madder at Blackwell than at the BoE.

My impression is that strictly applying the recount rule would indeed have led to a full manual recount, and would do so in practically every case in which a party requested a recount. I just think that's a pretty bad rule, on its face. Why even require a 3% random recount if the inevitable result is to require a full recount? That's nutty.

You seem to be mostly mad at the irresponsible BoE people who didn't follow the rule. I am madder at the rule. However, one can believe both that the rule is bad, and that it was irresponsible of BoE folks not to follow it. From a policy standpoint, it doesn't matter much how irresponsible we think the BoE folks were.

Personally, I think that BoEs should have carefully secured their ballots, then chosen precincts to recount at random or even by the observers' choice -- but at any rate in the open and not by prearrangement. That done, if the counts were pretty close, I honestly would not give a damn if they eschewed full manual recounts in cases where minor discrepancies did not bring the basic integrity of the count into question. But we should give BoEs as little discretion as possible in that respect. I think the rule should be rewritten. If we want to require a full manual recount, we should just do that, thus removing the perverse incentive for counties to tamper with the 3% recounts. If we think that is overkill, at least in some circumstances, then we should specify the circumstances more carefully.

[eomer]

35. We need to start over, to go back to basics.

By we I don't mean you and I, I mean the country. We need to go to a much simpler system that has very little need or possibility for discretion anywhere in the process. The political climate in the country is such that any discretion anywhere is just too dangerous and too provocative. What is at stake (life or death for people all around the world) is just too important.

The POTUS for the 2000 to 2004 term ended up being given to the losing candidate because of the discretion that existed in Florida. The SCOTUS seized on the indeterminate state of the system and used it as an excuse to make a blatantly political decision. Discretion in the system took the choice away from the people and handed it over to the Supreme Court.

There are other threads currently discussing the paper ballots, hand counted solution so we shouldn't really wander into that here. But I will just say that there is much less need or possibility for discretion in such a system as long as we don't find some stupid way to complicate it too much. I am all in favor of it.

[OnTheOtherHand]

37. be careful what you wish for

"The political climate in the country is such that any discretion anywhere is just too dangerous and too provocative."

Weirdly enough, in a slightly different context, that could have a radically different meaning than you intend. But setting that aside, I think any political scientist would feel honor-bound to tell you that trying to rule out "any discretion anywhere" is a non-starter. It's sort of like the doctrine of constitutional originalism, which doesn't really bind even those who profess to believe in it (as your SCOTUS example arguably underscores).

That said, I agree with the general direction of your comment. Simple standard procedures that everyone understands and can support help to rein in the arbitrary abuse of discretion. Simplicity is beneficial because the more rules one has to impose, the less chance that (1) they will make any objective sense or (2) the people who are supposed to follow them will understand them. Of course it isn't always possible or desirable to impose simplicity, but it's a good design criterion.

[eomer]

47. I agree, but...

what I meant was any discretion in counting the votes. I'm in favor of a system that is simple enough that there isn't any question (or very little if any) about what it means to count the votes.

I agree with your points about discretion in general but believe in the narrow context of counting the votes we can and should almost totally eliminate it. The problem is that some actors have realized that lots of discretion is good when they are the ones exercising it. They seized on it in both 2000 and 2004 to hijack the recounts.

[OnTheOtherHand]

48. I think we agree, then, at least on the future

A recount standard that isn't followed is obviously no good. And a recount standard that allows -- by rule, or in practice -- preselection of the precincts to be recounted is no good, either.

[kiwi_expat]

33. The original concern was determining if Kerry had won Ohio.

"In a not-so-close election such as Ohio 2004, where a main concern was whether the count had been somehow hacked, I think it is reasonable to conduct a recount that uses the same criteria as in the original machine count." -OTOH

I was one of the thousands of people who funded the recount with small donations. I clung to the hope that Kerry might actually win the election on the recount. And even though the recount occurred after the Ohio vote was certified, it was not too late to influence the Electoral College decision. (A very long shot, of course.)

The hope was that the under/over votes ("spoiled votes"), when inspected for voter intent in the 3% manual recount, would force a full manual recount in virtually all of the counties. And that once the under/over votes and the provisionals were counted, Kerry would be within 20,000 votes of a victory. That an extra 20,000 Kerry net votes might be found in a full manual recount seemed quite plausible. Greg Palast obviously thought so.

However, by the time the recount actually occurred, there had been so many rejected provisionals that a Kerry win on a recount seemed unlikely unless serious tabulator fraud had occurred. However, given the lack of random precinct selection for the 3% recount, a full manual recount might have been necessary to detect such fraud.

[kiwi_expat]

50. Maybe the manual recount should be in three phases:

The first-phase would be a 3% random manual recount to verify the machine counts. The machine-designated under/over votes would be EXCLUDED from the first-phase count.

The second-phase would be the full manual recount of any county in which the machine and manual counts did not match (still excluding the machine-designated under/over votes).

Then, after the candidate counts had been adjusted, if it was POSSIBLE for the under/over votes to change the winner, ALL of the state's under/over votes should be manually checked for voter intent.

* * *

The above 3-phase manual recount system is probably the most rational, but I personally would prefer to keep the current Ohio system, if the BoEs could be REQUIRED to examine under/over votes for voter intent - and thus trigger a full manual recount in most counties. Who knows what the full manual recounts might uncover.

[eomer]

28. I just read the full Cuyahoga observers' reports twice and can't find

where it says either way whether that the hand and machine counts matched or they didn't. If you can find it please copy/paste the paragraph so I can find it too.

Aside from that question, Cuyahoga is a great example of the kind of irregularities I am talking about. The precincts were selected in advance and the observers saw evidence of the cards having been manipulated before the recount. The cards appeared to have been sorted. This should not have been the case if the cards had been machine counted in the original count and then put away and not touched again until being opened up for the recount.

So even if they found an insignificant discrepancy in the public part of the process and didn't bother to finagle it away in front of our eyes, there is the appearance that they may have finagled away an original discrepancy during the earlier non-public part of the process but were somewhat less than competent at finagling and therefore had to first finagle and then fudge (if in fact they did fudge, which as I said I don't see the report of).

[eomer]

21. It is my impression also that under and over votes were counted.

OTOH, you mentioned in another thread that you agree with me more often than it would seem since we usually seem to be posting only the points on which we disagree. Funny that at the same time I was thinking the same thing. So let me seize this opportunity to agree with everything you just said.
:toast:

To not count the under and over votes would make the entire recount process a totally silly exercise (not that it wasn't that anyway but, you know, even more of one). I do believe they were counted in every county and would appreciate if kiwi could point to a report or something that tells us that was not the case.

The example you give of the overvote and the process they followed in Washington County is a great example of the kind of thing that threw the intent of the recount provision out the window. If you follow a process such as that, how can the hand and machine recounts ever not match? The process they followed in 87 out of 88 counties, in general terms, was to finagle with the count however many times and however many ways it took to get the hand and machine counts to match. That is why it was a sham process and a total violation of both the intent and the letter of the recount provision.

BTW, feel free to disagree with any of the above - we don't want to get too carried away with this agreement thing.

[kiwi_expat]

22. As I said in post #20...

The under and over votes were evaluated in the manual recount using the same criteria as in the machine count. The effect was that the under/over votes were not counted.

Here is an example from an Allen county observer report:

"The significant major discrepancy that I noted was the rejection of a ballot that was marked in ink. According to my understanding of the law, even though this ballot cannot be counted by the machine, it should be hand counted. Mr. Cunningham disagreed and the ballot was not counted."

Yes, I agree that not counting the under/over votes made the entire recount process a totally silly exercise. Or tragic.

[eomer]

24. Silly and tragic at the same time.

Thanks, it sounds like we are using different words to say the same thing.

[OnTheOtherHand]

25. I hope that's root beer -- I had better not get "loaded" too early

I think I was right to agree with kiwi a moment ago that actually, some counties didn't even finagle with the counts to get them to match -- they just said in effect, 'They don't quite match, but so what?'

I don't exactly agree that it was a sham process; if the BoEs were being basically honest, then it would have uncovered massive hacks. But the terms of the recount rule encouraged even basically honest BoEs to act corrupt, which is pretty much a worst-case scenario. In many people's minds, it completely defeated the purpose of the recount, by intensifying rather than resolving doubts about the integrity of the vote count. That's an appalling outcome even if the vote count was clean.

[kiwi_expat]

27. Ashland county recount appears to include 66 under votes.

"The vote recount produced these tally changes:
Bush/Cheney 16,211 (up 40 votes)
Kerry/Edwards — 8,577 (up 22 votes)
Badnarik/Campagna — 71 (up 1 vote)
Peroutka/Baldwin — 122 (no change)
Write-ins — 45 (up 3 votes)
Ashland County uses optical scan"

So, I was wrong, some counties did count the under/over votes.

But despite the fact that the Ashland manual recount was thus significantly different from the machine count, a full manual recount was not performed.

[eomer]

29. Yes, I think Ashland is a county that chose to fudge rather than finagle,

at least in the public part of the process. But like Cuyahoga, this process apparently had both a secret part of the process and then a public part.

It looks like you are right that in the public part they ended up with a difference between the 3% machine and hand counts but didn't do a full recount anyway (which is what I am calling a fudge).

Notice in the observer report that there were serious concerns about ballot security. The ballots were stored in a room that appeared to be used also as a lunch room and/or a meeting room.

So, once again, how would we know they didn't finagle first in secret but were then forced to fudge later? BTW, the fudging was necessary only because the observers challenged the standard being used to count overvotes. It would be a reasonable explanation to believe that they finagled the votes in private but then had to execute Plan B (the fudge) due to challenges by the observers.

Aside from the total lack of ballot security, the other evidence that would point to secret finagling is that the precincts were pre-selected.

I don't want to be on a witch hunt here but, golly, the number of irregularities in each of the examples we've talked about so far would, I think, have any reasonable person asking "what is going on here?"

[kiwi_expat]

32. I have to retract my accusation of a conspiracy re. not counting spoilage.

And the DIVERSITY of the irregularities among the various counties is breathtaking.

But I still think we can make a case that Blackwell and the BoEs conspired to avoid full manual recounts - even if all Blackwell did was to look the other way while the BoEs did what was in their own interests.


{p.s. I should have said "over votes" not "under votes" in post#27. I should probably not post at all at 3:00 in the morning.}

[eomer]

34. I agree there was diversity in the irregularities. But...

... there wasn't diversity when you look at it in more general terms.

87 out of 88 counties decided to take illegal measures to avoid a full recount. There is no diversity here.

And most if not all of those 87 counties decided on the same general tactic -- that of forcing the manual count to come out the same as the machine count. (You pointed out Ashland County as one that did not force the counts to be the same but instead allowed them to be different and then ignored the difference. That is true, but Ashland's original plan was to force the two counts to be the same. They applied that approach until a challenge by observers pressured them into taking voter intent into account.) There is not much, if any, diversity here.

Almost all of the 87 counties decided to not select the 3% precincts randomly. The specific non-random method varied from county to county but the decision to not do it randomly was almost universal. I believe, but would have to check to be sure, that almost all counties did the non-random selection in advance and in secret.

The similarities are to me what is breathtaking. I just can't believe that what happened here was the result of decisions made by each county independently. There has to have been someone directing, coaching, or at least planting ideas for this to have come out the way it did.

If there was someone directing, coaching or planting ideas then there was a criminal conspiracy. The ideas being planted were criminal.

[kiwi_expat]

36. I don't think Warren county forced the counts to be the same.

"Because the 3% sample was properly reconciled, the remainder of the ballots were processed through the computer and the results were tabulated electronically per the recount procedures. The *final* result tabulated was the same for *ballots cast*, but differed by 3 votes for ballots *counted* from the *certified* vote count. The final vote total *counted* was: Badnarik 193 (no change), Bush 68,037 (+2), Kerry 24,044 (+1), Peroutka 144 (no change)."
-Warren county recount observer


Champaign county doesn't appear to have forced their counts to be the same either. Three overvotes were counted as valid votes in the recount.

[eomer]

40. Sorry, I have to disagree. Champaign County is the poster child for how

the cheat was done.

First off, the observer report (http://www.iwantmyvote.com/recount/ohio_reports/counties/champaign.php) says that "{a}lso present was Dwayne A. Rapp, Vice President of Triad Government Systems, Inc. (and son of President.)"

Check out this excerpt from the executive summary of the Conyers Report:

The voting computer company Triad has essentially admitted that it engaged in a course of behavior during the recount in numerous counties to provide "cheat sheets" to those counting the ballots. The cheat sheets informed election officials how many votes they should find for each candidate, and how many over and under votes they should calculate to match the machine count. In that way, they could avoid doing a full county-wide hand recount mandated by state law.

http://www.truthout.org/docs_05/010605Y.shtml

And check out the more detailed version in the full Conyers Report:

First, during an interview, film maker Lynda Byrket asked Barbian, “you were just trying
to help them so that they wouldn’t have to do a full recount of the county, to try to avoid that?”
Mr. Barbian answered, “Right.” She further inquired: “did any of your counties have to do a full
recount?” Mr. Barbian replied, “Not that I’m aware of.”

Second, it appears that Mr. Barbian’s activities were not the actions of a rogue computer
programmer but the official policy of Triad. Rapp explained during a Hocking County Board of
Elections meeting:

“The purpose was to train people on how to conduct their jobs and to help them
identify problems when they conducted the recount. If they could not hand count
the ballots correctly, they would know what they needed to look for in that hand
count.”³⁸⁹

Barbian noted that he had “provided reports so they could review the
information on their own.”³⁹⁰

As one observer asked,

“Why do you feel it was necessary to point out to a team counting ballots the
number of overvotes and undervotes when the purpose of the team is to in fact
locate those votes and judge them?”³⁹¹

Barbian’s response was,

“...it’s just human error. The machine count is right...We’re trying to give them as
much information to help them out.”³⁹²

http://truthout.org/Conyersreport.pdf

The Green Party observer also notes (in addition to the presence of the Triad VP who is also the son of the Triad President) that the Deputy Directory had a spreadsheet up on his computer during the count. This computer screen was apparently visible and close enough to read because the observer noted the name of the spreadsheet.

Now how about the count -- did it match or not? I believe that they did force it to match and that the way they accomplished that was by the technique coached by the Triad VP. They ran the cards for the pre-selected precinct through the machine in advance and recorded the totals they needed to match. They made sure they understood the algorithm used by the machine and then they used that same algorithm when performing the hand count. They posted the totals that the hand count needed to match on the computer sitting nearby so they could doublecheck that they had come up with the matching totals during the hand counting.

The information in the report that talks about 3 "overvotes" being counted (which is what you pointed to) is saying that those 3 votes were counted both in the hand count and in the machine count. Apparently the machine was programmed to count a vote if there was a punch in one and only one column that was for a valid candidate. It was set to ignore totally a punch in any column that was not for a valid candidate (in other words, a punch in a column not associated with a candidate would not disqualify an otherwise valid vote). This sounds like a typical algorithm for a counting machine.

Other facts from the observer report that support my interpretation are that the precinct was pre-selected, that the deck had already been removed from secure storage and that no chads were observed during the hand count. These facts point to pre-counting and possibly chad-cleaning of the deck to make sure the machine count was predictable (no chads would fall off during handling).

So, far from being a counter example to my theory, I believe Champaign County is a perfect example of how it was done. It also provides evidence (or, I should say, outright admission) that the process used was not an independent decision by the local BoE but rather was a coordinated approach that was orchestrated (at least in this and other Triad counties) by an election vendor with strong Republican ties.



edit: minor wording

[kiwi_expat]

43. No matter how dirty the Champaign recount was, I still think that...

the manual recount counted 3 of the 4 overvotes rejected in the machine count.

Salem North had presidential assignments of:
Kerry 2
Nader 4
Peroutka 6
Badnarik 8
Bush 10

Here are the 4 overvotes:

1 and 2 counted as "2" 1 was absentee
2 and 4 counted as "2" because "4" was for Nader
6 and 10 overvote
9 and 10 counted as "10" not absentee


My question to you is how would the machine have decided to count the second overvote as "2" 'because "4" was for Nader'??

[eomer]

44. The machine was set to ignore the Nader column because

Nader was not legally a candidate on the ballot.

Probably you know this but I had to google it to refresh my memory so here is how that came about. What happened was that Nader on the ballot in Ohio was an issue being disputed. Apparently the printed materials in many counties (in this case the cards that were installed on the face of the punch card voting machines) had been printed with Nader on them because it was thought at print time that he was on the ballot. Then at the last minute Nader lost the court battle and was ruled not on the ballot in Ohio (http://www.command-post.org/2004/2_archives/cat_nader.html). By that time it was too late to reprint the materials.

So Nader showed up physically on the voting machine but when counting they set the machine to treat that column like any other column that was not associated with a candidate.

So I still assert that the count you are pointing to is the count for both the 3% machine count and the 3% hand count and that they forced them to match.

The relevant part of the report is:

We retested the reader/counter successfully. The Republican observer declared himself satisfied with everything and asked if there was agreement, which I gave. Then I asked about counting the rest of the precincts by machine and found that others had assumed my agreement with the hand count included agreement that no more counting was needed. After a bit more discussion we began the machine count which took about an hour.

which I admit has some ambiguity but I interpret the statement that they "retested the reader/counter successfully" to mean that the counts matched.

Other things that point to my assertion being true are:

1. the algorithm they used to count the hand count appears to be a machine counting algorithm,
   
2. Triad admitted they coached multiple counties to do it this way, including the technique of taking the machine counting algorithm and using it as the standard for hand counting and
   
3. there is no mention of any debate about whether to do a full hand recount due to them not matching. It seems from the part I excerpted that they all took for granted (edit: at the moment just after the 3% count was finished) that there would not be a full hand recount. This would have been obvious to all parties only if the 3% counts had matched.
   

[kiwi_expat]

38. At least two counties involved the observers in the precinct selection.

"Our observer reported that Brown only had a certain number of precincts which weren't too large for the three percent sample, so the BoE put these in a hat and had the observers pull them out until the number equaled at least three percent. He thought this was very fair and would meet the requirements for 'random' precinct selection."
- Greens' Brown county recount report

And Lake county let the observers choose which machines(/precincts) to recount.

[eomer]

41. Yes, that is why I said almost all rather than all counties (n/t).

[kiwi_expat]

42. That is why I said "at least 2" rather than just "2".

I only checked 9 counties before I got 2 examples. Please don't make me check ALL the reports.

I'm happy to leave it at "most" ("most counties selected the precincts in advance"), if you like. :-)

[eomer]

45. I'm happy to leave it at "most" also.

And I do appreciate the time you're taking checking the reports (and that you are pushing me to check them also).

Sorry if the way I posted that answer sounded snarky. Sometimes my online voice comes out that way so I should tell you that I appreciate your engaging in this discussion with me and don't want it to get personal.

B-)

[kiwi_expat]

49. Cheers! (n/t)

[philb]

39. There was also a problem of not allowing check of poll voter signatures,

and number votes counted versis number of valid voters


there clearly were a huge number of deliberate and systematic "undervotes"

people whose votes those in control didn't want counted

[Time for change]

51. I believe it was a conspiracy

I think that the following "extra-ordinary" events are relevant to that point:

1. The Warren County lockdown. Locking reporters out of the building while the "counting" was done was unprecedented, and it was excused by an incredible tale that turned out to be a lie. We don't know what went on in that building. We do know that, according to many people working in Ohio, Ohio was seen as almost certainly going for Kerry. Then rather suddenly, everything turned around. We have to assume that went on in that building was a conspiracy until it is investigated and we are offered a better explanation.

2. The witholding of machines from Dem. precincts in Franklin County. Even IF we buy the excuse that the allocation of machines was based on previous voting behavior, anyone with a modicum of sense of responsibility for running the Ohio election should have made sure that voting machines were supplied when the long lines were first identified.

3. The electronic vote switching in Mahoning County: All reports favored Bush (i.e., Kerry to Bush swithcing). I believe that was a conspiracy.

4. The great disparity between the NY Times reports of voter registration in Ohio vs. the official SOS figures. Here are some posts I have made on that subject:

http://www.democraticunderground.com/discuss/duboard.php?az=show_topic&forum=132&topic_id=1974460

http://www.democraticunderground.com/discuss/duboard.php?az=show_topic&forum=132&topic_id=2237782

This had to have been a conspiracy IMO.

5. The recount: So many counties either had technicians come and tamper with machines to ensure a match, or failed to select random precincts, or actually found non-matches, and then failed to do a recount anyhow. The pattern was so pervasive, it had to be a conspiracy IMO.

There's lots more, but I think that's enough.