New York City’s decision to dispense with DREs and go with optical scan systems “with paper ballots” is a marginal contingent improvement, but is being celebrated as a Hallelujah moment by some. False confidence, though, is the very best place to grow a fraud.
What we CAN celebrate is going from zero chance of an election we can believe in (under touch screen DREs) to at least having a remote chance with optical scans, but only provided a whole series of unlikely and expensive events occur in the process of obtaining an optical scan recount to check the opscan machines. Without doing that proper recount (and verifying perfect chain of custody prior to the recount), because the opscans are every bit as hackable as touch screens and produce more false confidence that DREs (touch screens) do, we are left with little protection.
In fact, you might call optical scans (without hand counts on the first count that completely check the opscan machine) the “fool’s gold” standard for elections.
Though the New York resolution below is very nicely and wisely written to "make clear to all that elections must be observed by the people, and should not require "trust" in computer experts to say whether the votes were properly handled," and even though I very much agree with this frame and in some small way helped popularize it OPTICAL SCANNERS DO NOT ACHIEVE IT without
hand counts during the first count (as opposed to hand "recounts").
It says New York is voting for optical scanners for the above reason of observability, and it simply doesn't follow. The "Ban the DREs" movement gaining some steam right now also helps reinforce this fundamental misunderstanding, as does the existence of a paper ballot trail with optical scans.
But the reality is that in the overwhelming 90+% of all opscan elections, there is no human observation of the count. NONE. To say one has “observed” an opscan election is like watching a hi-speed photocopier copying 10,000 pages of SAT exam ovals and thinking your eye is somehow keeping up on a counting basis with the photocopier. There is simply no way for humans to verify an opscan election or observe it, if the machine is the only count.
In short, opscan is only IN THEORY observable by the public, and ONLY at the RECOUNT level. You may believe you were PRESENT at the counting, but you did not OBSERVE it and you can not VERIFY it.
The recount in most states is hard and expensive to come by, except in rare, close “automatic recount” jurisdictions. Having myself sued for recounts, such as the June 6, 2006 special congressional
election that was disputed in California's 50th Congressional District, I want to give a "witness" about that.
*In a DISPUTED ELECTION, i.e. when it really counts, you are going to have problems or are highly likely to have problems, getting to the paper ballots or ever getting a recount IN TIME to matter in the race.* Same goes with "audits." They *come too late in the process and the officials are very "proud" of their totals and resent the hell out of anybody questioning them (if innocent officials) and
are scared as hell at being caught (if guilty officials)* so in either case they are going to resist your recount or audit with every tactic from high recount fees to partial production of information to interposing every conceivable legal objection and delay.
Because of the above dynamics with officials involved in first counts, we see just this week two officials working in DEMOCRATIC COUNTIES in Ohio being sentenced for RIGGING THE RECOUNTS in the Ohio 2004 presidential election in Cuyahoga County (Cleveland, OH). Why Dems or Dems' appointees would do this is a mystery unless you understand the paragraph immediately above and that the urge to CYA and protect your fellow employees from hard work over Thanksgiving and/or
Christmas is much more powerful than the urge to accuracy or even the CRIMINAL LAW. Comprende vous? It most certainly does NOT take a bribe to get officials to resist the recount or audit. The first count – still an electronic count under all congressional bills in the 110th Congress so far – is the only count that REALLY counts in well over 99% of the nation’s races. And in many of the rest needing recounts, you’ve got to be rich to have rights.
Indeed, in the CA50 litigation, it is now the ides of March 2007 and we have yet to get our hands on a single opscan ballot from the June 6, 2006 congressional special election, but we are in the appellate courts trying and an opinion is expected in April.
Remember, the original Diebold hack by Harri Hursti was on an optical scan machine. They are every bit as vulnerable as DREs, plus the added feature of false confidence by the public.
The only way in which DREs are worse is that they never let you see your own ballot (but the opscans need not "see" your ballot properly during them electronic count so the effect is the same) and the DREs can never be recounted (while opscan recounts are rare, difficult, expensive, and feature
plenty of time for rigging, EXACTLY what happened in Ohio where they spent two freaking days rigging that recount to make sure it showed no problems!
RECOUNTS HAVE A LIMITED SCOPE, AS WELL: Even activists often think that recounts are more powerful than they are, but they only detect a smallish slice of possible error or fraud, only that slice of errors that occurs during the first count itself as a result of addition errors or scanning errors. *You can recount, for example, a stuffed ballot box a hundred times and still get the same stuffed result*. A recount matching the original count is only one of MANY tests one would have to do to pronounce an election "Fair". *So, activists should always correct, point out, jump on, or kindly redirect any person who says an election was 'fair' or "clean" because a recount matched the original
Having helped get rid of DREs in my own home county (and being left with optical scan) I know there is a slight improvement, presuming there has been a perfect chain of custody. But DREs are like the wolf, and optical scan is the wolf in sheep's clothing.