READER COMMENTS ON
"Did Doug Hoffman Actually Win the NY-23 Special Election? If So, Does It Actually Matter?"
(27 Responses so far...)
COMMENT #1 [Permalink]
...
Soul Rebel
said on 11/13/2009 @ 1:57 am PT...
COMMENT #2 [Permalink]
...
Shortbus
said on 11/13/2009 @ 3:39 am PT...
Nice piece Brad, reminding us how the Republicans argued the Cali voters had no say so in their own election yet go figure the New York Post had this in their rag just yesterday
If the count overturns the election, Owens could be removed from office.
Hmmmmmm?
COMMENT #3 [Permalink]
...
aaCharley
said on 11/13/2009 @ 4:00 am PT...
Since the Bilbray / Busby contest was not particularly close, the results should have been allowed to stand. Further, note that the election result was certified. Was there a certification sent along with Owen to the swearing in event? Your hypothecations are stretching credibility.
COMMENT #4 [Permalink]
...
Wachovia CD Rates
said on 11/13/2009 @ 7:15 am PT...
There is no certification provided to winner, so a big question marks on the credibility of victory. This seems to be a fun tale, but it is hard-core reality.
COMMENT #5 [Permalink]
...
Robert C. C.
said on 11/13/2009 @ 7:37 am PT...
as if precedent will stop them from making what they will out of this, even if it is only a disinfo campaign of how wronged they are.
COMMENT #6 [Permalink]
...
Big Dan
said on 11/13/2009 @ 9:20 am PT...
#3 AACharley: The Bilbray/Busby was 49%-45%, and the Owens/Hoffman was 49%-46%. And you said the Bilbray/Busby contest "was not particularly close".
Why do you think 49-45 isn't close...but 49-46 is???
Bilbray defeated Busby by four percentage points – 49 percent to 45 percent in the heavily Republican district
http://legacy.signonsand...060801-9999-1m1vote.html
COMMENT #7 [Permalink]
...
Big Dan
said on 11/13/2009 @ 9:24 am PT...
So: AACharley thinks 49-46 is close...and 49-45 isn't. Then questions someone else's credibility.
COMMENT #8 [Permalink]
...
Pman69
said on 11/13/2009 @ 10:23 am PT...
Looking at absolute vote difference, the NY-23 race difference in votes is nearly half that of CA-50 in 2006. You have to be careful looking only at percentages. Things like vote count, and rounding off the decimal portion of the percentage can cause the numbers to look very similar when they really are not.
Just as reference, I got these numbers from Wiki:
NY-23 in 2009:
136,855 total votes: 3026 vote difference
48.7-46.5 : 2.2%
CA-50 in 2006:
134,302 total votes: 5530 vote difference
49.5-45.3 : 4.2%
Now having looked at the numbers, back when the "unofficial" numbers were a more than 5000 vote difference, with 93% of the votes returned, hoffman conceded. Now that they actually got the vote counts from the machines that they did not have initially, the vote difference is right around 3k.
I wasn't able to find any information on CA-50 as to what the specific vote discrepency's were outside of there being a chance for an invalid vote total. Here we have an actual vote adjustment bringing the vote difference much close.
Personally, I think once someone concedes, that is it. It is over, but I just wanted to point out that perhaps things are a little different than CA-50, although what happened in CA-50 is outrageous.
COMMENT #9 [Permalink]
...
Brad Friedman
said on 11/13/2009 @ 10:30 am PT...
"I think once someone concedes, that is it. It is over"
Just, for the record, a concession has absolutely no legal baring on anything. It's a political nicety. The voters still decide who won and who lost. IF the votes are counted, that is, and IF the courts (and Republicans) allow those totals to be used.
COMMENT #10 [Permalink]
...
Pman69
said on 11/13/2009 @ 11:37 am PT...
"The voters still decide who won and who lost. IF the votes are counted, that is, and IF the courts (and the Republicans) allow those totals to be used."
I couldn't agree more, but I don't believe Republicans should be singled out as I believe that regardless of which party holds the house, we would have the same behavior if their candidate is the one that initially won. If the Democrats were to prevent the conservative from becoming the NY-23 representative, would you agree with their pulling the same stunt that the Republicans did back in '06? In my mind it is still wrong, and just as deplorable if they were to try the same thing.
I also should have said "...that should be it..." in my original statement. I was stating my opinion, and if a candidate is conceding the race to his or her opponent, they are giving up, or admitting that they lost. If they are not sure, they should not concede.
COMMENT #11 [Permalink]
...
Brad Friedman
said on 11/13/2009 @ 12:08 pm PT...
PMan69 said:
I believe that regardless of which party holds the house, we would have the same behavior if their candidate is the one that initially won.
While I appreciate the sentiment, that's not necessarily the case. Some 6 elections were contested by Democrats under the Federal Contested Elections Act in 2006. Among them, the race in which 18,000 votes were lost in Sarasota, in a race decided for the Republican by 369 votes. The Democrats who controlled the House by then, refused to overturn or even examine actual evidence in any of those cases. They were all unanimously dismissed by the House Admin Comm. It's hard to believe a Repub-led Admin Comm would have done same with all six contests had they been filed by Republicans.
If the Democrats were to prevent the conservative from becoming the NY-23 representative, would you agree with their pulling the same stunt that the Republicans did back in '06?
No. But for the mitigating factor that it's difficult to argue that a precedent set, and accepted, by the House (Republican-led at the time, but not objected to by Dems, to my knowledge) and the courts. As noted in the article, I disagreed with the argument, but it was accepted legally, so it would be difficult to stomach changing the rules, moving the goalposts to benefit Repubs, when the same folks did nothing to benefit Dems.
That said, if the voters of NY-23 preferred a different candidate than the one installed, and the ballots are there to prove it, I can't imagine my NOT being in favor of that person being seated, no matter how dumb Democrats have been in the past.
So it certainly presents, as also noted in the piece, a potentially interesting situation should it come to that!
The democratic (small "d") argument is a no-brainer. The political one, however, is less so.
COMMENT #12 [Permalink]
...
jjray
said on 11/13/2009 @ 1:03 pm PT...
Reply to Brad. "Just, for the record, a concession has absolutely no legal baring on anything. It's a political nicety."
I would agree this is true for the official count, i.e., the state must finish the vote count and declare a winner regardless of whether one candidate concedes on election night (before all the counting has been completed). If we are talking about vote recounts, a concession may have some effect depending on state law.
COMMENT #13 [Permalink]
...
Brad Friedman
said on 11/13/2009 @ 2:14 pm PT...
If you are correct, JJRay, I'm not sure which state law you're referring to. (Happy to be schooled on something I don't know about here, if you wish to cite an URL to support your assertion.)
COMMENT #14 [Permalink]
...
Big Dan
said on 11/13/2009 @ 4:49 pm PT...
COMMENT #8 [Permalink]
... Pman69 said on 11/13/2009 @ 10:23 am PT...
Looking at absolute vote difference, the NY-23 race difference in votes is nearly half that of CA-50 in 2006. You have to be careful looking only at percentages. Things like vote count, and rounding off the decimal portion of the percentage can cause the numbers to look very similar when they really are not.
Just as reference, I got these numbers from Wiki:
NY-23 in 2009:
136,855 total votes: 3026 vote difference
48.7-46.5 : 2.2%
CA-50 in 2006:
134,302 total votes: 5530 vote difference
49.5-45.3 : 4.2%
I still think that's close and similar. You think that's not similar? Here's what I think isn't close: 792,102 vs 3,000.
"Half" is a percentage, you say the difference was "half", that's 50%. That makes it seem not close when you say "50%" or "half".
I'll use the numbers you rounded up:
136,855 - 3026
134,302 - 5530
That's EXTREMELY similar! More similar than I originally thought.
Here: this isn't similar:
1,298,990 - 55,221
136,855 - 3,026
COMMENT #15 [Permalink]
...
Big Dan
said on 11/13/2009 @ 4:54 pm PT...
Are you two guys politicians or something? The two guys saying those numbers aren't similar?
COMMENT #16 [Permalink]
...
Ernest A. Canning
said on 11/13/2009 @ 6:59 pm PT...
The issue before the CA court in the Busby/Bilbray contest pertained exclusively to jurisdiction. The court did not rule that Bilbray won the election; merely that it was the House of Representatives and not the courts that has the exclusive jurisdiction to determine whether to remove and replace a member once the representative is sworn in.
The court's decision in the Bilbray case does not mean that Hoffman would have no remedy in the event that the still uncounted absentee ballots should put him over the top. His remedy would be to go before the House of Representatives, claiming that he should replace Owens as the duly elected representative.
I would note that if the recent Coleman/Franken election contest is indicative, it is highly unlikely that an absentee ballot universe of a mere 10,000, cast at the time it when it was still a three way race, in all probability, will not permit Hoffman to overcome a more than 3,000 vote gap.
COMMENT #17 [Permalink]
...
Lora
said on 11/14/2009 @ 9:19 am PT...
From Brad's article:
The bad news, however, for now-hopeful Republicans/Conservatives is that when the results of the Busby/Bilbray race were contested by a California voter, the Republican-leadership of the U.S. House Administration Committee sent a letter [PDF] to the CA judge in the case, arguing that neither CA voters nor the state courts had jurisdiction to contest the election any longer, once the candidate had already been sworn into the U.S. House.
From Ernest (#16):
The issue before the CA court in the Busby/Bilbray contest pertained exclusively to jurisdiction. The court did not rule that Bilbray won the election; merely that it was the House of Representatives and not the courts that has the exclusive jurisdiction to determine whether to remove and replace a member once the representative is sworn in.
The court's decision in the Bilbray case does not mean that Hoffman would have no remedy in the event that the still uncounted absentee ballots should put him over the top. His remedy would be to go before the House of Representatives, claiming that he should replace Owens as the duly elected representative.
I'm trying to put this together. I take it this means that theoretically Hoffman could petition the House if the final vote count shows him winning.
Now what if, after all the votes are counted Owens is still ahead, but there is evidence that suggests a corrupted count. Because Owens was sworn in, does that mean Hoffman is now S.O.L., like Busby?
Or is there anything Hoffman could do, and was there anything else Busby could have done?
"Don't concede until all votes are counted" seems to be the moral of this story.
COMMENT #18 [Permalink]
...
Brad Friedman
said on 11/14/2009 @ 2:48 pm PT...
Lora wrote:
I take it this means that theoretically Hoffman could petition the House if the final vote count shows him winning.
That is correct. According to the Republicans' 2006, court-approved precedent, the voters and courts of NY no longer have a say in the matter. Hoffman would have to convince the U.S. House to unseat Owens and seat him instead.
Now what if, after all the votes are counted Owens is still ahead, but there is evidence that suggests a corrupted count. Because Owens was sworn in, does that mean Hoffman is now S.O.L., like Busby?
If Hoffman fails to timely file his election contest under the Federal Contested Elections Act within 30 days, yes, he'd be S.O.L.
In CA in 2006, before the Repubs pulled their "Constitutional" stunt, no contest was filed under FCEA for, among other reasons, the fact that the state contest could not even begin until *after* the election was officially certified by the state of CA. That didn't happen until 22 days after the election (as I recall), and then the contest period begins. Yet, Bilbray was sworn in just days after the election itself.
Or is there anything Hoffman could do, and was there anything else Busby could have done?
Busby could have filed under FCEA within 30 days of Bilbray's seating, but she wasn't contesting the election --- the voters were (and I don't believe they may contest under FCEA, only the candidate). Of course, until ballots were actually counted, it was impossible to know who won or who lost, and we weren't allowed to file for a count any ballots until *after* the 22 day certification period. Until then, we had only Diebold to go by. After that, the San Diego Registrar, Don Haas (one of the nation's worst!) capriciously and arbitrarily decided to charge about $1/ballot for such a count --- even while next-door Orange County was charging something like .14/ballot for same hand-count.
That would have amounted to tens of thousands of dollars to the citizen contestant which, as you may imagine, wasn't available to them.
Hoffman's only recourse at this time would be to file under FCEA in Congress. How he does that, without having evidence he actually won the election, or that there is reason to dispute the count will be up to him and his GOP attorneys.
"Don't concede until all votes are counted" seems to be the moral of this story.
And the moral of this entire blog. One we've been trying to get across to candidates and the parties and the public for years...
COMMENT #19 [Permalink]
...
Cosimo diRondo
said on 11/14/2009 @ 7:25 pm PT...
A bit off-topic, but I was wondering if Brad has seen this:
http://web.mit.edu/newso.../2009/rivest-voting.html
Ron Rivest is the 'R' in 'RSA', the public-key cryptographic algorithm upon which virtually all of your online-shopping is built. I haven't seen Bruce Schneier comment on it yet, but he most certainly will (probably here: http://www.schneier.com/crypto-gram.html), and I would trust his opinion.
But do politicians have any interest in supporting a scientifically-endorsed tamper-proof voter-verifiable election system? Seems to curtail their options somewhat, and takes all the fun out of Election Tuesdays.
COMMENT #20 [Permalink]
...
Soul Rebel
said on 11/14/2009 @ 8:31 pm PT...
So funny a coincidence. I'm teaching a college-math-topic class to some very bright middle school kids, and we learned about RSA cryptography a couple of weeks ago. The proof of the algorithm was a little over their heads, but I enjoyed it. Mathematically, a very simple set of ideas using modular arithmetic. Yet in practicality the sheer size of the numbers involved render any kind of random guess-and-check search for prime factors impossible for even the fastest computers.
COMMENT #21 [Permalink]
...
Brad Friedman
said on 11/14/2009 @ 8:54 pm PT...
Cosimo -
Asked in another thread about the same thing, here's the answer I offered:
Since you mentioned it, BD, and since I haven't had much time to hit the latest "open source" onslaught of new e-voting schemes, and since someone else just emailed me about the same announcement, along with the question "evoting that's verifiable - what's the catch?", I might as well post what I sent via email in reply:
The catch is that it's not actually verifiable. It can also be gamed. But most notably, it's "verifiable" only by you, for your own vote. If *I* have a question about the election, *I* can't know if anybody elses vote was counted as cast.
Also, even if you trust the verifiability of your own vote (which, in general, misses the point of citizen-verifiable elections), we've ALL got to trust in rocket scientists and cryptographers that the cryptography is accurate and secure.
Using cryptography and transparency in the same sentence is pretty much an oxymoron, doncha think?
Brad
Of course, one of the developers of that particular voting scheme is also a BRAD BLOG reader, so I'm sure he'll jump in with his own points and/or rebuttal if inclined here.
And to your specific description of this scheme as a "scientifically-endorsed tamper-proof voter-verifiable election system", um, not only does such a claim sound familiar (Sequoia, Diebold, etc. said same about their systems when they were selling them), but it seems you must be one of the developers standing to make money to make such a claim that so fervently seems to misdirect voters from what actually matters to them and their ability to oversee their own elections.
COMMENT #22 [Permalink]
...
Soul Rebel
said on 11/14/2009 @ 10:05 pm PT...
I read the article provided by Cosimo. RSA encoding is to the best of my understanding crack-proof due to the size of the numbers involved and the need to identify specific prime number factors for the decoding process (known by the decoder, these primes provide an instant translation).
But the issue here does not appear to be the encryption. The issue appears to be "Does the matching of my bubble code to the online bubble code necessarily mean my vote was cast for the candidate whom I cast it for?" And the answer to that still appears to be NO. It is still a faith-based system, in that I have to have faith in the integrity of the people working the databases, the security of the databases, chain of command, access, the integrity of the ballots themselves, and so on.
Nothing doing here, back to square one. It isn't the mathematics of the system, the cryptography is fine. It's EVERYTHING ELSE.
COMMENT #23 [Permalink]
...
Agent 99
said on 11/14/2009 @ 10:37 pm PT...
And, no matter who we elect, or think we elected, they turn immediately into clones of those we voted, or thought we voted, out....
COMMENT #24 [Permalink]
...
Cosimo diRondo
said on 11/15/2009 @ 4:16 pm PT...
BF@21 said:
... it seems you must be one of the developers standing to make money to make such a claim that so fervently seems to misdirect voters from what actually matters to them and their ability to oversee their own elections.
Well, I NEVER! I had just read the linked article and was simply asking your opinion, as someone with the bigger picture on voting issues and procedures. As others have pointed out, the cryptography is not the issue - it's rock-solid. Rather, the evaluation needs to be made on the application which uses it (and the humans using the application), and cryptography is not necessarily mutually exclusive with transparency. Banks don't seem to have any confidence issues with cryptography. If only we valued democracy as much as we do shopping.
The details were sketchy in that article. But it seemed to me they were probably using a crytographic hash of the data to be able to verify it had not been altered. This of course does not make the overall system tamper-proof...but I certainly could not glean that from the details provided.
For the record, I am most certainly NOT involved as a developer or in any other sense. I do security software for a living, so it piqued my interest, that's all. I'll cut you some slack for all the good work that you do
COMMENT #25 [Permalink]
...
Brad Friedman
said on 11/15/2009 @ 11:51 pm PT...
Cosimo -
First, my enormous apologies for the incredibly inartful wording I used in that reply to you! Was trying to get out of the house for a longish roadtrip, and was flying through comments/replies *way* to fast! Sorry. I know who the developers of that product are, and I know they are not you. So again, very inartful wording there. Hope you'll forgive.
Was trying to suggest that only someone hoping to gain from such a product could use such a phrase as you did (as if an ad for the product). I did so poorly.
Now to the substance of your comment above...
You said, "Banks don't seem to have any confidence issues with cryptography."
Well, that's because these are apples and oranges issues on a number of levels. Banks can check and re-check the cryptography to *their* satisfaction. Voters cannot (unless each one of them wishes to employ their very own set of computer scientists to do so.) On top of that, banks don't have secret transactions, so there are ways that all parties involved in the transactions can track and/or verify the transaction, along with independent paperwork to verify their transaction in the case that somehow any encryption has gone wrong. All parties also have various forms of legal recourse if needed in such a case, and are able to track every step of the process --- along with the independent paperwork --- to make their case.
None of that exists in elections. The vote is cast, it's then gone, and it's up to the citizenry, as a collective, to be able to verify that ALL votes cast were counted as intended. Totally different ballgame.
And, as far as I can tell, the Scantegrity scheme still does not allow for any such collective verification that every vote was counted as cast.
COMMENT #26 [Permalink]
...
Cosimo diRondo
said on 11/16/2009 @ 9:59 am PT...
Thanks for the reply, Brad. I completely understand. (To sorta quote one of my favorite movies, which ironically also has an IT security tie-in:) I just figured that somebody had a case of the Mondays
Cryptography can work both ways. The little guys can also use it against The Man...for example, using open-source freeware like PGP to throw a little twist in Big Brother's illegal surveillance schemes. (Google Phil Zimmermann to see how the gubmint harassed him for creating the original PGP.)
I'm sure there are smarter people than I actually getting paid to think about applying cryptography to electronic voting. While this particular scheme may not satisfy all the requirements of open-government types, I feel that cryptography will play a role, for the better, in some as-yet undesigned (or undisclosed, possibly suppressed) methodology. Cryptography can be used to verify that data has not been altered (cryptographic hash), and to verify the source of the data (digital signature), even without necessarily encrypting the data itself, so that we can all see it.
COMMENT #27 [Permalink]
...
Soul Rebel
said on 11/16/2009 @ 11:47 pm PT...