NJ Election Discrepancies Worse Than Previously Thought, Contradict Sequoia's Explanation
I wrote previously about discrepancies in the vote totals reported by Sequoia AVC Advantage voting machines in New Jersey’s presidential primary election, and the incomplete explanation offered by Sequoia, the voting machine vendor. I published copies of the “summary tapes” printed by nine voting machines in Union County that showed discrepancies; all of them were consistent with Sequoia’s explanation of what went wrong.
This week we obtained six new summary tapes, from machines in Bergen and Gloucester counties. Two of these new tapes contradict Sequoia’s explanation and show more serious discrepancies that we saw before.
Before we dig into the details, let’s review some background. At the end of Election Day, each Sequoia AVC Advantage voting machine prints a “summary tape” (or “results report”) that lists (among other things) the number of votes cast for each candidate on that machine, and the total voter turnout (number of votes cast) in each party. In the Super Tuesday primary, a few dozen machines in New Jersey showed discrepancies in which the number of votes recorded for candidates in one party exceeded the voter turnout in that party. For example, the vote totals section of a tape might show 61 total votes for Republican candidates, while the turnout section of the same tape shows only 60 Republican voters.
Sequoia’s explanation was that in certain circumstances, a voter would be allowed to vote in one party while being recorded in the other party’s turnout. (“It has been observed that the ‘Option Switch’ or Party Turnout Totals section of the Results Report may be misreported whereby turnout associated with the party or option switch choice is misallocated. In every instance, however, the total turnout, or the sum of the turnout allocation, is accurate.”) Sequoia’s memo points to a technical flaw that might cause this kind of misallocation.
The nine summary tapes I had previously were all consistent with Sequoia’s explanation. Though the total votes exceeded the turnout in one party, the votes were less than the turnout in the other party, so that the discrepancy could have been caused by misallocating turnout as Sequoia described. For example, a tape from Hillside showed 61 Republican votes cast by 60 voters, and 361 Democratic votes cast by 362 voters, for a total of 422 votes cast by 422 voters. Based on these nine tapes, Sequoia’s explanation, though incomplete, could have been correct.
But look at one of the new tapes, from Englewood Cliffs, District 4, in Bergen County. Here’s a relevant part of the tape:
The Republican vote totals are Giuliani 1, Paul 1, Romney 6, McCain 14, for a total of 22. The Democratic totals are Obama 33, Edwards 2, Clinton 49, for a total of 84. That comes to 106 total votes across the two parties.
The turnout section (or “Option Switch Totals”) shows 22 Republican voters and 83 Democratic voters, for a total of 105.
This is not only wrong – 106 votes cast by 105 voters – but it’s also inconsistent with Sequoia’s explanation. Sequoia says that all of the voters show up in the turnout section, but a few might show up in the wrong party’s turnout. (“In every instance, however, the total turnout, or the sum of the turnout allocation, is accurate.”) That’s not what we see here, so Sequoia’s explanation must be incorrect.
And that’s not all. Each machine has a “public counter” that keeps track of how many votes were cast on the machine in the current election. The public counter, which is found on virtually all voting machines, is one of the important safeguards ensuring that votes are not cast improperly. Here’s the top of the same tape, showing the public counter as 105.
The public counter is important enough that the poll workers actually sign a statement at the bottom of the tape, attesting to the value of the public counter. Here’s the signed statement from the same tape:
The public counter says 105, even though 106 votes were reported. That’s a big problem.
Another of the new tapes, this one from West Deptford in Gloucester County, shows a similar discrepancy, with 167 total votes, a total turnout of 166, and public counter showing 166.
How many more New Jersey tapes show errors? What’s wrong with Sequoia’s explanation? What really happened? We don’t know the answers to any of these questions.
Isn’t it time for a truly independent investigation?
UPDATE (April 11): The New Jersey Secretary of State, along with the two affected counties, are now saying that I am misreading the two tapes discussed here. In particular, they are now saying that the tape image included above shows 48 votes for Hillary Clinton, not 49. They’re also saying now that the West Deptford tape shows two votes for Ron Paul, not three.
It’s worth noting that the counties originally read the tapes as I did. When I sent an open records request for tapes showing discrepancies, they sent these tapes – which they would not have done had they read the tapes as they now do. Also, the Englewood Cliffs tape pictured above shows hand-written numbers that must have been written by a county official (they were on the tapes before they were copied and sent to us), showing 84 votes for Democratic candidates, consistent with the county’s original reading of the tape (but not its new reading).
In short, the Secretary of State talked to the counties, and then the counties changed their minds about how to read the tapes.
So: were the counties right before, or are they right now? Decide for yourself – here are the tapes: Englewood Cliffs, West Deptford.
UPDATE (April 14): Regardless of what these two tapes show, plenty of other tapes from the Feb. 5 primary show discrepancies that the state and counties are not disputing. These other discrepancies are consistent with Sequoia’s explanation (though that explanation is incomplete and more investigation is needed to tell whether it is correct). Thus far we have images of at least thirty such tapes.
Comments
“Isn’t it time for a truly independent investigation?”
Given that these errors may have effected the result of elections, which I imagine will not be rerun, and cannot be fixed, one might argue that its too late for a truly independent investigation. Sigh.
Based on their previous explanation of obviously very poor if not malicious engineering, my guess is as follows.
Using the example with the Republican Party as option switch 6 and the Democrat Party as option switch 12, the poll worker presses button 11 by mistake. The red light next to button number 11 lights up (or doesn’t light up) and the operator panel display will show nothing. The poll worker then notices the mistake presses option switch 12 to activate the Democrat contests. The red light stays lit next to option switch 11 but the display now says DEM. Now the poll worker presses the green “Activate†button. The red light stays lit next to button number 11, and the operator panel display now says DEM and the ballot in the voting booth will activate the Democrat party contests. As the option switch 11 is the option switch recorded, however since it is unassigned it is not even totaled. The poll workers further confound the issue by being too busy to make sure that the “Public Counter†number matches the number of people that actually signed their registers to when they came to vote, after all it was very close, only 1 off anyone could miscount by one and not think twice when the “machine†says otherwise.
Now, that would be my guess, but the point being, there still needs to be an investigation.
I hope that post could be understood, I copy and pasted their original explanation example, and tried to edit it to what seems plausible given their already poor engineering design confession and the new examples.
What surprises me about all of this is just how few people are voting. With only 100-200 votes being cast, why do we need machines to count them? It seems to me that you could cast paper ballots, count them by hand, and still get the results much cheaper and hardly any slower, and without all of these nasty questions about reliability and fraud.
One more thing…
Reading the “statement” for the “Election Officers” It appears that the Public Counter and the Protective Counter are only measures in place to verify the voting machine was not tampered with between elections. My guess is the poll workers did not even count the actual voters that signed in to vote.
Tying things together. . .
“The Security Mindset and ‘Harmless Failures’” post from a few days ago plus this and other posts about Sequoia’s voting problems. From the security mindset, here is how I could see using this information to my advantage. (Assuming my guesses were right before).
I sign up to be a poll worker, I want to accuse Dems of “stuffing the ballot box” so ever time it is my turn to assist a voter and the voter is DEM I push the unassigned option switch first, then push the other, yada yada. Then at the close of the polls the numbers will be far enough off that it will be noticeable. The Dems will have far more “votes” than what the machine claims to have counted, this I have reason now to accuse Dems of “stuffing the ballot box” or tampering with the machines anyway. In the case of the primaries where they are separate contests might not make a difference, but if there are big discrepancies someone will notice, and accusations will fly.
Ooooooh that sounds fun
too bad I don’t live where these machines are used.
Not just an independent investigation, but top to bottom of all the source code, all the hardware, everything. Once you have a machine that gets number of voters not equal to number of votes, all of your confidence (such as it was) in the recorded results being causally related to what the voters tried to do pretty much goes out the window.
Tangent: is it a good idea to publish an image of the precinct workers signatures here?
Given that the votes are off by 1, could be a problem of initialization in counter variables in the software? Of course, for something so important, it needs to be resolved.
I would say the importance of the investigation is in figuring out how machines that behave like this made it to the front lines. Anybody can build a crappy voting machine. It is the state’s job to verify that it works.
From SAVElections Monterey County … Read Sequoia Source code review on the CA SOS website. ..Debra .Bowen decerified the Sequoia Edge II DRE touchscreen machines in Aug. 2007. Due to the outraged county clerks, in spite of the horrendous security vulnerabilities in the machines,in Oct.2007 she reapproved use of the Edge II with the condition that only 1 could be used per precinct. Each machine is now required to have a 100 % manual tally audit during the canvass to assess the accuracy of the DRE machine vote tally.
Unfortunately Bowen permits the elections staff to upload the tally’s from the DRE memory cartridges on election night. …before the manual tally is conducted. Our observation of the 100% manual tally confirmed that not all DRE’s were closed at the polls, poll workers failed to sign and date the summary tapes when they did close the machines, summary closing tapes were not always posted at the precinct which CA code requires. I was not permitted access to the VVPAT machine tally at the time the manual audit was conducted.
Ca code allows only 5 days after the statement of the vote is delivered to request a State paid recount.The recount order under CAEcode 15640 reads, The recount order will only be issued when “there is probable cause to believe…. Errors or failures , whether electronic, mechanical or otherwise, in the safekeeping, handling,tallying,counting,recording or certification of the ballots or votes cast, sufficient to make it likely that the result of the election was affected as to the successful candidate or proposition or tie holders, or sufficient to cast substantial doubt on the substantial accuracy of the results without regard to affecting any result.”
We have just received the precinct reconciliation reports from the Feb. 5th Presidential primary.Yes, it took more than 5 days to retrieve them from the elections department. In addition a significant number of reports show the total number of voters who have signed the roster do not match the number of paper ballots plus DRE votes cast at the polls……Plus there are other important data omissions on the reports related to accounting for reconciliation of how many paper ballots were delivered to the precinct… etc.
In addition to your demand for an independent analysis of the machines
I highly recommend that you review the number of signatures on the precinct roster.
We have found it impossible to provide meaningful citizen oversight to assure accuracy of our DRE vote tallies as well as the paper ballot tallies processed by our Sequoia 400-C scanners. We are demanding to hand count all paper ballots at the precinct and encouraging all voters to vote at the polls. Mike, yes hand counting ballots at the polls is the most secure way to vote. I hope you will learn from our experience we are definitely learning from yours.
Mike Ash: At some point, any sufficiently established paradigm transgresses to the status of compulsive obsession (at least for a period). Automation and computerization have crossed that point.
What Valery from SAVE Our Vote in Monterey, CA reports has been replicated throughout the country every time citizens try to work within the procedures and election law to check the results reported by electronic forms of voting.
The citizens of this country do not control our elections, and have a valid basis for questioning the outcome of both state and federal elections. Our elections are controlled by election officials and vendors who control the process, whatever audits might be conducted, the chain of custody of all records, the very technology that records and counts, any recounts, the access of citizens to oversight or the ability to examine records in a timely fashion, and in lobbying legislatures to prevent meaningful reforms.
A massive and pervasive fraud has and is being perpetuated upon our country which is criminal. But the law gives the legal presumption that the election officials are doing their job, and it is the citizens who must prove otherwise. While all evidence is controlled by the very entities that need to be investigated. Sequoia’s reaction to New Jersey’s desire for Professor Felton to examine the machines are a prime example of how scared these vendors are of scrutiny. Their hiding behind the wall of claimed proprietary information is another form of the abuse of the national security claim by those in government who do not want their actions examined.
We can have no confidence that the “representatives” of the people were truly chosen by them.
My question is that if so many discrepancies in Sequoia’s e-voting machines have been identified, then why can’t straight off changes be made? Do they hold some type of monopoly in their industry? It seems ridiculous that there aren’t any other alternatives available. If investigating Sequoia proves fruitless, then it’s up to the states to follow through with an aggressive plan to get this situation resolved immediately, if not sooner!
“Worse? How could it get any worse? Jehovah, Jehovah! *crack!*”
I’m sorry, but that’s the immediate thought that came to mind when reading the headline. We now return you to your regularly-scheduled indignation.
(honestly, there’s nothing I can add – I’m dumbfounded that this level of quality is somehow approved to determine our country’s leaders)
Go read Sequoia’s explanation. Under their scenario, voters were presented with THE WRONG CHOICES for PRESIDENT of the UNITED STATES. Specifically, Democratic party members were presented with Republican party choices or vice versa – and (this is important) – the error was not detected by the voter. This should be reason enough to disqualify the equipment!
I also note the use of the term “Democrat party” in Sequoia’s explanation – I’ll reserve my own comments on this slur, instead referring the reader to en.wikipedia.org/wiki/Democrat_Party_(phrase)
These tapes are clear an convincing evidence that the Sequoia voting system has produced incorrect counts of votes.
Given this evidence, there is only one thing to do:
CERTIFICATION OF THAT SYSTEM MUST BE REVOKED.
Given clear and convincing evidence that the vendor of that voting system has attempted to obstruct the audit process, there is only one thing to do:
CERTIFICATION OF THAT VENDOR MUST BE REVOKED.
We know from previous elections that some partisans are willing to rig elections if there is any possible way to do so. The only reasonable response is to make it completely impossible. The only way to do that is by a rigorous process that includes multiple control, complete transparency, and thorough audits. Computerized voting systems are inherently single-controlled, completely opaque, and cannot be audited in any meaningful way. Any one of those attributes should be enough to disqualify these systems.
Shoulder to shoulder, we must throw all bums out.
Has anybody compared the COST of buying and maintaining one of these machines on an annual basis against the cost of 300 printed ballots and four dozen pencils?
Am waiting to see what kind retraction Dr. Felten posts…
Much easier to cast aspersions, than to eat one’s words.
Concerned Citizen:
What do you think I should retract, and why?
In 1946 in Athens, TN, Citizen-soldiers returning from the Pacific theater faced a similar situation. They addressed the problem a bit more directly than just whining on blogs.
www.constitution.org/mil/tn/batathen.htm
I’m proud to say that my next-door neighbor, Bill White, led that action. Perhaps it’s time for history to repeat itself.
Re: the update…
Sounds like they are trying to bend reality to some less controversial form.
Looking at the larger images, it is pretty clear that the Ron Paul number was mis-struck by whatever crappy printer is used in the device. Yet, it still looks very much like a mis-struck 3 and not a 2 given the middle portion and the bottom left portion. The 2s just don’t have that much ink in the lower left quadrant on the rest of the page.
Same goes for the Clinton discrepancy.
I’m no expert, but that is what it looks like to me. I would highly recommend having an expert pass judgement.
Looking at the larger image from Englewood Cliffs, I think the Secretary of State is correct. The apparent “9″ in the Clinton total very clearly has an indentation on the right side at the base of the upper loop. This indentation is characteristic of the digit 8 when printed on the tape, but the one other occurrence of the digit 9 has no such indentation, but rather a flat right side. So I’d say the S of S is correct and this tape has no discrepancy, rather the 8 just did not print completely. On the other tape, the Ron Paul digit is completely illegible and could easily be either a 2 or 3. Since the one is hard to explain while the other makes perfect sense, Bayesian reasoning forces us to conclude that it is a 2.
The real point here is that it shows the bankruptcy of the claim that a paper voting trail will fix these problems! Instead, as I have argued in the past, it will often open up more controversy. Paper is far from a perfect medium and is prone to defects, errors, mishandling, corruption, and a variety of other problems. These kinds of apparent discrepancies in the paper trail just lead to argument, suspicion, and lack of confidence in the voting process. Look at what happened in Florida 2000, a perfect example of the problems with paper, which put us on our present course.
Computer scientists, so painfully aware of the limitations and difficulties in their own field of expertise, seem to be blind to the fact that errors crop up in all areas of human endeavor. Their seeming faith in the magical incorruptibility of a paper trail leads them to the vain hope that adding paper records to voting machines will free them of their imperfections. Hopefully these examples will throw cold water on these unrealistic dreams and return some sanity and balance to the consideration of the security properties of all aspects of voting systems.
I sure am glad I stopped by here first. I was just about to shoot off an angry email to the New Jersey SOS and a few others. I intended to “cc” it here when I saw Hal’s message above. I deeply hate to say it, but I think Hal is right. Upon much closer inspection, it does appear the number INTENDED to have been printed next to Clinton’s name was an 8, and not a 9.
However, what WAS printed was NOT an 8. It was something that looks just like a 9. It was only after zooming in to 200% and holding my laptop up to my face that I had to conclude that “the intent of the printer” 8. LOL… I feel like that guy in Florida in 2000 with the punch card up to his face.
So what does this mean? It means that the electronics inside the machine were “thinking” 8, which would be correct, but that the printer did NOT accurately print an 8. It matters little what the intent of the machine was, if what it actually reported differs from what it intended. The system as a whole failed.
I still take exception with the New Jersey SOS’s press release. The least they could have done would be to have included an explanation and admit that it was an honest misreading on Ed’s part, and a misreading repeated by thousands of other reasonable people who also viewed the image. Anyone viewing that would see a 9. It’s only after close, non-standard analysis that one can come up with an 8 for that ink blob.
I would consider the faxed tape copy not to be readable with certainty. Fax machines just don’t have adequate resolution for this sort of thing.
As for the other, I would agree that the so-called “nine” is a misprinted eight. The determining factor is not the ‘indent’ on the right edge, but rather the fact that the flag on the nine slopes two pixels up from the bottom while the eight does not.
I’m curious why numbers are so often printed or displayed in typefaces that don’t allow for unambiguous reading even under bad conditions. A similar problem exists on television sports broadcasts, where scores are often shown in fonts where the “6″, “8″, and “9″ can all render similarly.
If I had my druthers, I would have the government strongly encourage the following standard:
The left ends of a numeral three should have a minimal inward curve–much flatter than an eight.
The numerals six and nine should start angling inward immediately.
The numeral five should have a significantly taller round section than the six or eight.
The numeral zero should be conspicuously elliptical at the ends (as opposed to oval).
If a set of digits is well designed, it should be possible to identify digits with certainty even when they are largely obscured or misprinted. Such a set might not be as aesthetically pleasing as Helvetica, but would be much more reliably decipherable under adverse conditions.
Incidentally, the Interstate signs use a pretty good set of digits, though the “3″ could be mistaken for a partially-obliterated “8″.
I have to add to my last post, as I’m sitting here giggling and can’t help it. IF that is an 8, and I now think it is (a mutant 8, but intentionally an 8, nonetheless), then that means Sequoia , first of all, also thought it was a 9 like everyone else including the election officials, and second, published an explanation of how it printed 9, even though it didn’t! It’s like the husband who thinks his wife knows he cheated and wracks his brain for a convoluted alibi, spews it out to her, and then discovers she had no such suspicion. I am truly LMAO.
Maybe Hart can make more sense of it after they take over Sequoia later this week.
Supercat: I don’t know if that image was faxed or not, but I have seen many of those reports in their original form and Ed’s image is a fair representation of them. Those machines use thermal printers that never look good under the best of circumstances. Why they are allowed to use thermal printers for official records that are required to be held as evidence at a later time is beyond me. All you have to do is leave the paper in a hot location for a while (a hot warehouse, a car trunk) and the entire paper turns black. It’s a good thing that most major elections are held in November when it’s cold. Our law makers don’t understand and don’t care.