Are the Voters Running the State or is the State Running the Voters?

Live Connection with DPS Raises Question
As IBM Gets New Project Manager and State Frets Expectations

Ten Thousand Hours: Part Three
Weeks 14-21 of the Texas Voter Database Project

By Greg Moses

When Texas pays $4 million to license the eRegistry election management software from Hart InterCivic, will it be buying a product that the state has helped to develop? And when the voter management system is hooked up to live records from law enforcement, will the voters be running the state or will the state be running the voters? These questions we ask after going through facts found in project documents for the Texas voter database and election management project, weeks 14-21.

Weeks 14-21 continue to show evidence of intensive knowledge harvesting as developers of the proprietary eRegistry software place so many demands on the expertise of state employees that the employees register a complaint about their ability to keep up. And the contract assumption that Hart InterCivic actually has a COTS (Commercial Off The Shelf) product to offer the state comes under question when the first release of the software arrives without installation instructions or a users manual.

These weeks also see a change in managers for Team IBM as weekly reports begin to display more detailed summaries of contractor activities.

And the running issue of what it means to privatize functions of public technology finds another prime case study when Hart InterCivic representatives express concern that their March 3 demonstration of software to a focus group of county election officials may result in “disclosures”. The same focus group also raises concerns for the state project manager that “county expectations” also need managing. By the end of the period the state manager is asking for administrative help so that he can be more free to discuss the project more frequently with “external stakeholders”.

As much of the technical activity for this period revolves around live data connections between the Secretary of State (SOS) and Department of Public Safety (DPS), we can also raise questions about a trend in voter management toward live interfaces with law enforcement. Is voter management now a subset of law enforcement? As update information is freely swapped between SOS and DPS, we wonder. Are the voters running the state or is the state running the voters?

Let’s try this answer. The high tech frontier in election management is creating a voting population as a class of administrative privilege that will be more and more pre-screened and qualified to vote. Against this trend a clear “human rights” response is needed: “let all the people vote, period.”

* * *

The IBM weekly status report for Feb. 5-11 has interesting features. For one thing, it is mislabeled week thirteen (it should be week fourteen) which serves as consolation against small-minded perfectionists everywhere who utter dicta about how things really work in the big time.

But more importantly, the document sets a new standard for this $12 million project to build a privatized voter database and election management system for Texas. It serves as documentary evidence that something has changed at the project, and although it bears the name of the initial project manager for Team IBM, the report’s style, format, and presentation hint that the new manager is already doing the old manager’s paperwork.

In the week-fourteen report we glimpse for the first time the full range of activities being conducted by at least ten project components: security team, data migration team, data mart team, voter information programmer, DBA, lead architect, interface team, GIS team, test team, and application team.

The security team for example puts together a meeting between staff from the Secretary of State (SOS) and the Texas Department of Public Safety (DPS) where they work out the difficulties of making secure database connections between the voter registration system and DPS. As the reader may recall from previous articles in this series, the security team for the project will then take the knowledge gained from this meeting between state workers and turn it into a privatized computer solution that can be sold back to the state with service contracts attached.

The same kind of knowledge harvesting is going on with the project DBA (database administrator) who is meeting with “state employees and the data mart team to discuss reporting requirements.” In a handy term for wordplay, the DBA is working up “demoralized” tables that can be used by eRegistry software, a proprietary product of Hart InterCivic. Once the DBA gets clear about the data warehouse design, eRegistry will be “reverse engineered.” According to the contract with the state, “reverse engineering” of eRegistry is something that only Hart is allowed to do.

And this raises an interesting question for intellectual property buffs. If eRegistry is reverse engineered during a public contract with Texas to whom do the rights to the reverse engineered product belong? Is this another case where public money, public expertise, and public functions are all contributing to privatized value of a product in the marketplace? I don’t have the answer, but I think it’s an interesting question to ask.

Knowledge harvesting among state workers has created such “multiple and parallel demands on SOS resources” that state workers say they can hardly keep up with everything that Hart needs them to do. Meanwhile, the state is letting Team IBM use its screen reader to run tests with.

Also busy down at DPS is the interface team figuring out how to make the connections for live checks of drivers license and social security numbers, so that numbers placed on voter registration applications can be checked against the DPS database. Again, fresh questions come up for public inquiry. What all is attached to these numbers down at the DPS? What events can be triggered by the “live checks” that come in, other than a simple yes or no from DPS to SOS that a number does or does not belong to a name?

In addition to “live checks” of numbers, the interface team is also working on getting “signature images” from DPS that can be matched to signatures on voter registration cards. Remember you sign your drivers license, and that signature image is stored on a DPS computer. For some reason which is probably not very permanent, the DPS declines to offer up its signature images — I mean YOUR signature images — to the SOS.

By week fifteen it’s official, there is a new manager on board for Team IBM.

The state project manager during week sixteen reports that he has begun working with the new project manager for Team IBM who “is getting his arms around the project.” We don’t yet know what the new IBM manager had to say for weeks sixteen and seventeen (those reports were not included) but by week eighteen, the project plan is in version 84, and the list of project issues for Team IBM has grown in three weeks time from eight to 23.

Two of the new project issues in week 18 revolve around a 25-member county focus group that meets in Austin on March 3 (week 17). SOS has assembled the focus group in an effort to sell more counties on the project. But Hart registers “disclosure concerns” about all these folks seeing its eRegistry prototype, and the state project manager worries about “Counties Expectation Management.” Belonging to the focus group places public election officials in the dicey position of previewing plans for election management (a very public issue) and proprietary software (a very private property) at one and the same time. What does Hart fear they will talk about? What does the state project manager fear they have come to expect from this project? If we put the two worries together, may we infer that all did not go well on March 3?

March 3 turns out to be an unlucky day for another reason. Someone
from Team IBM copies the entire project folder and then “inadvertently” places the duplicate back onto the public workspace. For the next two weeks, updates to the project will be split between the two folders creating “confusion, lost comment documents, and process flow breakdown.” When the glitch is discovered about March 18, the state project manager will have to spend time bringing the two files back together.

Week 19 is a crucial week for the project. It is the deadline for Hart InterCivic to deliver its first COTS software, the Commercial Off-The-Shelf product that the state will eventually pay $4 million to use. On the due date for release of Hart One, the state receives a letter from Hart with a URL and password. On March 17, Hart presents a “walkthrough” of the software. But documentation was yet to be found. On the face of things, what Hart delivered does not yet look like a COTS.

Hart follow-up items appear in the report for week 20. The eRegistry configuration guide is reviewed on March 21. And on March 28, notice is received from the escrow service that deposits Hart software (actually this is a week 21 event).

During week 21 the external interface team meets with IBM and SOS to talk about tech submit issues and directory structure. External interfaces for the project have grown from an initial 10 to 19 since the project started. Does this mean that the number of databases hooked up to the election system has nearly doubled in scope?

According to plans contained in IBM reports, it looks like the week of April 18 will be eventful. There is a County Focus Group scheduled for April 19 to review the first Hart release of mid-March. On April 20 the SOS staff will be tasked to define their business rules in ways that are pertinent to the logic of eRegistry. And on April 22 Jeff Osborn will be in town.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s