CVR's and Ballot Secrecy

I’ve seen a couple of remarks lately that the Cast Vote Records from our electronic voting systems can not reveal the identity of the voter. This statement would be true in a rationally run election system, but ours are not run on a rational basis. While I am not aware of any method to use the CVR alone to identify a voter, I am aware of methods in which the CVR in conjunction with other documents such as poll book check ins, and voter histories can tie a surprisingly large number of voters to their ballots. With some combinations of election equipment, it may be possible for all voters and be tied to their ballot.

We are currently getting out tails kicked in Texas over open records due to ballot privacy issues, and unfortunately my chief nemesis in this fight is Paxton of all people. I think I wrote an article on that a while back that can be found on these pages. I'll review two law suits that have been filed in Federal courts here surrounding these privacy issues.

The first case I’ll discuss is brought on because we vote using programs such as UCAVA, mail, and countywide voting programs. These programs are the culprit in damaging ballot secrecy and should be eliminated and or severely curtailed immediately. This law suit was brought by the Pubic Interest Legal Foundation and was filed against Harris County Texas who uses the Countywide voting program. It was filed later than the second one I’ll address, but the issue has caused many more problems for us.

Using myself for an example, I occasionally work as an election clerk and when I do, it is in a location away from my home. It is almost a certainty that I will be the only person from precinct 1504 to vote in that location, therefore my ballot and how I voted will be very easy to identify, since mine is the only ballot of that style to be voted there. It is only possible to vote out of precinct due to programs like the countywide voting program. This might seem that it is a pretty small problem, but the ladies that first discovered this demonstrated that they could identify the ballots of several thousand people in a county the size of Tarrant. They stopped short of getting the ballots so that they could actually see how a person voted, but they did take the exercise far enough to prove they could if they wanted to. As a humorous aside, one of those ladies did find that her husband voted differently than she did on an issue. I bet dinner that night was interesting.

These ladies wanted to use this information as a way to attack countywide voting, which is nothing but a way to obfuscate an elections accounting. It is used to support cheating, I’m quite sure. Thy asked for my input and while I saw it’s usefulness in that regard, I did caution that it would raise privacy issues. I was right, but I didn’t know how right. This has become a nightmare issue.

A short time later a second set of people were conducting a recount in a neighboring county. I’m not certain if they were familiar with what the first researchers found, or if they found the issue entirely on their own, but they too discovered that thousands of ballots could be identified because of out of precinct voting. They did go so far as to get some of those ballots, but were discreet in their use. My understanding is that most often they used them to show a given representative how insecure programs like county wide voting were by showing the representative their ballots. Then someone with a bone to pick got hold of the ballot of our out going state Republican Party Chairman and released it publicly. The stuff hit the fan. Biggly.

The response was to limit the public records that would be released and/or redact a goodly amount of information from the documents that were released violating several laws in the process. This has done significant harm to our ability to see what’s going on in our elections and it is coming at a time when we have some people finding very interesting things in our voter registration records. It has complicated their efforts to get the needed records.

Withholding the information does not solve the problem. Government functionaries can still access that information. While our guy with the political ax to grind used this to try and embarrass some one in the political realm, there are a number of ways that this information could be used by government officials to coerce people into certain actions. Ballot privacy is supposed to be total. The answer to the problem is to do away with countywide voting programs as they are the main culprit in causing this problem and a lawsuit has been filed against Harris county in Federal court in an attempt to do just that. It is seeking to have the Countywide voting program halted on privacy grounds.

Because of the limited number of UCOVA ballots and the way they come in to the elections centers, it is nearly impossible to provide privacy for a UCOVA ballot. At one time the federal website stated that fact, and some jurisdictions require a waiver if you vote in that manner. In smaller counties or in small precincts mail ballots may be difficult or impossible to maintain complete ballot secrecy.

The second lawsuit was brought by a researcher in Central Texas. It is an Federal court in Travis County. While I have not always been in agreement with this person, I think in this case she is likely right, and may have found a very large crack in the electronic voting systems. She alleges that particularly when ES&S poll books are used with ES&S tabulators, that she can identify the ballot of each in person voting in a given location. In this case, the poll book issues the ballot including the ballots electronic number, which is then either inserted into a BMD or filled out by hand and fed into a tabulator. She has claimed that she found a nexus that connects that voter to that ballot number. If she can tie the voter to the ballot number, the CVR and/or associated reports can fill in the blanks. She has submitted redacted evidence of this to the court and states in her suit that she will reveal the methodology in camera to the court. I have not personally seen the evidence, but this capability being available would answer many questions.

She has further stated that the capability may exist with other combinations of equipment. With the Hart system that we operate, I think it would be difficult, but there may be some data that we don’t normally look at that would make it possible. More research is needed.

This issue is caused because we are not using ballots preprinted with a serial number divorced and instead allowing the equipment to number the ballots. Keeping that data might be possible, but it would be difficult. The ballot serial number shouldn’t be present in the same system as the voters information.

Jeff O'Donnell, has spoken of weaponized ballot privacy and I think that’s exactly what it is. Democrats were complaining about the totally unauditable DRE voting equipment that was in wide use. Their complaints were valid and people clamored for paper. We wound up with an electronic, paper hybrid, but it would have been auditable. What is a bad guy to do? They came up with a brilliant idea that promised the auditability of paper but at the same time announced the countywide voting programs which were sold as a convenience of being able to vote anywhere. I reality it was a way to confuse the accounting of the elections like the DRE’s did while giving the appearance of connivance and audibility. The public bought it hook line and sinker. Along with advertising that the systems are auditable, they claim that the release of documents can violate ballot privacy and thus withhold documents that are needed to do complete audits.

Note that they didn’t tell you up front about the privacy issues that would be created with these programs, and in the process of withholding information, they are admitting that ballot secrecy can be pierced and that the information to so so is available even though it shouldn’t be. These folks didn’t tell anyone about the issue and when the public found it, they cut off the publics ability to find additional problems, and of course we were/are on the verge of several important discoveries. nice work if you can get it.