Friday, August 10, 2012

Democrats line up against Harris elections administrator

On the 'devil you know' theory. All these come from Whitmarsh's list.

Stan Merriman was first out of the gate yesterday afternoon.

Are our memories that short ?  We've had "expert" election administrators before., working under our County Clerk.   I won't name names unless you insist, but in the late 90's we had a guy alleged to be an expert who led our then County Clerk and both Dem and Repug county chairs down the primrose path to the Hart machines, without any paper trail and we've been stuck with them ever since, now with new "replacements" of this outdated, non-recount technology after the fire. Most of the rest of America has moved to other, more transparent technology with recount capability at least.  Later, when this administrator left to ultimately become an election association lobbyist, we brilliantly hired a guy fresh from the Broward County, Florida recount fiasco of 2000. He continued the advocacy for our black box voting technology and then moved on.  he moved on I think also to become a lobbyist.    So, our track record on these "experts" isn't so good, is it. At least having this position under an elected official gives  we the people the option to remove all incompetents from office. Including the "experts" who screw up.

Gerry Birnberg picked up an echo from John Behrman (who posts occasionally here).

I share Gerry’s reservations about an elections administrator: It is something we could come to regret a lot. But, that is not what Lane Lewis called for.

The Chairman’s position is much more astute, to the point, and practical. The phrase “forensic audit” reported in the Chronicle is not a felicitous phrase: a “forensic examination”, “election audit,” or “IT audit” are things needed at various times, but not the same thing. 

Behrman continued a bit more in high praise of Chairman Lewis. David Patronella fell in behind Merriman.

Stan is absolutely right. An appointed elections administrator is not the answer. In the 1980s Dallas County became the first county to get an appointed elections administrator. She in short time gained notoriety for short changing Democratic strongholds at election time. Officially nonpartisan, she owed her position not to the voters of Dallas County but to Republican officeholders and acted in their interesest. Minority and other Democratic legislators introduced several pieces of legislation to curtail her power some of which were enacted. I would hate to see us go down this path notwithstanding serious concerns with recent serious election problems in our county.

I just left all the typographical errors, sentence fragments, comma splicing, inappropriate capitalization and munged paragraphs in those excerpts because otherwise I would have had to type [sic] about a hundred times.

Several of these men have advanced degrees from institutions more noteworthy than Lamar University, so I suppose we can chalk some of it up to failing eyesight.

Meh. Anybody can make a mistake. Even me.

But nothing anybody has written yet -- not even Charles' skepticism -- convinces me I am wrong about the need for an appointed elections administrator for Harris County, and fast. As in an observatory capacity for November, and a supervisory one after January.

I wonder if Marc Campos is still with me? Guess we'll find out later this morning. I'll update here when he weighs in. In the meantime, let's allow Pokey Anderson to remind us what's at the root of the problem: "the electrons running Harris County elections".

At the risk of harping on something I've (cough cough) researched for years....

Harris County elections are run on non-transparent, all-electronic machines, driven by software that is by its nature non-transparent. Even software in use for years has bugs in it (constant Microsoft updates, anyone?), some important, some not. Software can be changed, by officials, by insiders, by hackers.

Are intrusions into critical computers difficult? Are they rare? 
1) In one year, the Pentagon logged more than 79,000 attempted intrusions; about 1,300 were successful, including the penetration of computers linked to the Army’s 101st and 82nd Airborne Divisions and the 4th Infantry Division. (2005)

2) "A government consultant, using computer programs easily found on the Internet, managed to crack the FBI's classified computer system and gain the passwords of 38,000 employees, including that of FBI Director Robert S. Mueller III." (2004, reported by the Washington Post)

3) Secret Service operating procedures, 100,000 Social Security numbers, and other "highly sensitive" national security information have gone missing from the National Archives. (2009, reported by Computer World).

4) A computer hacker got into the U.S. agency that guards the country's nuclear weapons stockpile and stole the personal records of at least 1,500 employees and contractors, a senior U.S. lawmaker said. (Reported by Reuters, 2006)

When you combine inherently non-transparent electronic machinery, subject to flaws or fraud, with how much money and power is at stake in an election, you are gambling your democracy.
Certainly, some election chiefs are better than others. But, after a certain basic level of competency, whether you have Mother Teresa or Jeff Skilling running your elections should NOT matter.

Then-County Clerk Beverly Kaufman's PR flack, David Beirne, told a meeting of League of Women Voters that they should not expect transparency in elections.
"They're faith-based elections," he sniffed.
No. It's not about faith. Elections are about transparency.

In 2003, I asked Beirne about the software, the guts of this stuff:
 "No one in our office has the expertise and background to be looking at the source code, the programming for the eSlate system. "
As for an audit trail, watch his language here:
"Right now what we do in the State of Texas and what's considered to be adequate in the state of Texas is that right now we can manufacture an audit trail any time after an election if it's necessary to do a manual recount. "

(In 2007, Beirne stopped working for the public and accepted a job working for the electronic voting machine organization, Election Technology Council. But, one could argue whether he ever was working for the public.)

The public should be able to tell if elections are being run fairly and accurately, by observing every step of the process. When it happens in a dark box, the public has no way to know. What if your bank told you your account had $50,000 at 10 pm, but only $30,000 the next morning, and you had made no transactions?

If the top election official "explains" losing 800 votes by blaming it on "garbage" phone lines, the public should be able to verify, without doubt, what the actual vote counts are.
You can't do it with the eSlate. Period.

Pokey nails it, and for their part Merriman (including above as well as in an op-ed in the Chronicle some years ago), Behrman (in continuing and official capacity), and I have all studied and written about this issue extensively ourselves.

I was on the conference call with Common Cause and Verified Voting yesterday which had as its topic election machine integrity; read the reports here and here. And be reminded that we all agree on at least this much: that neither partisan elected officials nor election officials appointed by partisans can really address the dilemmas we face in Harris County, Texas, and the nation.

But hey, an elections administrator is a beginning toward improving accountability. One I think we need. As with most of my political endeavors, I'm not concerned about being the minority view.

Update: Then again, maybe the County Clerk's office can just call their PR consultant, Hector Carreno, who also consults the Election Technology Council, and get this all *ahem* "papered" over.

Isn't it simply amazing how Carreno's fat fingers are in every single pie in the county?


Matt Bramanti said...

Hey Perry, got a question for you. A real question, no gotcha b/s.

What would a paper audit trail look like (or what did they used to look like)? Would it just be a machine-by-machine total, or a precinct-by-precinct total, or would it be down to very small detail, like Ballot 9653-2521 voted for Tom, Dick and Harry? Were the paper ballots retained, and for how long?

I think I've only ever voted by paper ballot once, and I don't remember whether the ballot had a serial number or anything.

I don't think that a paper-based election is inherently more secure than an electronic one. You can run a fair election or a crooked one either way. It's just a matter of defining, monitoring and auditing the controls.

PDiddie said...

To some it looks like the mail-in ballots currently in use. Those are bar-coded to individuals' personal data, which could provide evidence in a forensic audit.

To me it looks more like one of those old Scantron tests we used in college, with software that tallied instead of compared. These could be similarly coded to a voter's identity and provide a kind of "receipt" (as with e-Slate's log-in PINs) for audit purposes. I would prefer to see something with the voter's preferences; voter-verifiable evidence of their cast ballot.

The standard objection -- especially in the current environment -- is that all these printers and paper add expense.

Matt Bramanti said...

Oh wow, I didn't realize they'd be tied to a specific voter. Doesn't that raise concerns about a secret ballot?

PDiddie said...

Not for me, no.