To the astonishment and disappointment of many, Secretary of State Bruce McPherson has chosen to re-certify both the Diebold AccuVote-OS optical scan and AccuVote-TSx touchscreen machines for use in California's election this year. This in spite of the fact that computer scientists commissioned to do a proscribed but nevertheless revealing report on Diebold's AccuBasic Intepreter determined that aspects of the software that runs both systems is prohibited under federal standards. According to the report the computer scientists, David Wagner, David Jefferson, and Matt Bishop, all members of California's Voting Systems Technology Assessment Advisory Board (VSTAAB), advised McPherson that the risks posed by the presence of prohibited "interpreted code" were "manageable" and could be "mitigated" by tightening security around Diebold's voting machines.

In December, McPherson, in response to the revelation that AccuBasic could be used to mask pre-stuffed memory cards as demonstrated in a test election in Leon County Florida, denied certification to Diebold's equipment pending a review the Independent Testing Authority (ITA) laboratory Ciber, Inc. - the same laboratory that had certified the systems in the first place. In a published statement McPherson said "[d]uring a thorough review of the application for the Diebold system currently  pending certification, we have determined that there is sufficient cause for additional  federal evaluation." In a letter to Diebold dated December 20, McPherson noted that "it is the Secretary of State’s position that the source code for the AccuBasic code on these cards, as well as for the AccuBasic interpreter that interprets this code, should have been federally reviewed," and requested that Diebold submit the source code relating to the AccuBasic code on the memory cards and the AccuBasic interpreter to the ITA for immediate evaluation and required this additional review before proceeding with further consideration of your  application for certification in California. He concluded that "[o]nce we have received a report from the federal ITA adequately analyzing this source code, in addition to the technical and  operational specifications relating to the memory card and interpreter, we will expeditiously proceed with our comprehensive review of your application."

In the end the ITA apparently wasn't expeditious enough for McPherson. In a letter to Diebold dated February 17, McPherson admitted that "[t]o date, we have not recieved a report from the federal ITA, nor have we received any information regarding the status of its review from the federal ITA." Apparently the "resource constraints at the ITA" precluded the review of the AccuBasic source that was somehow overlooked during the original certification process. No problem - the Secretary of State certified them anyway. After all, no one seriously thought the "Independent" Testing Authority would act in any way that was detrimental to a valued client like Diebold. This way they are spared the uncomfortable and onerous job of explaining why it doesn't matter that Diebold's software is in violation of federal standards.

While McPherson's decision to side with Diebold will no doubt have significant and well-deserved fallout in November, it seems that, in the short term, he made the politically expedient decision. With a special election for Duke Cunningham's seat scheduled for April and statewide primaries in June, McPherson chose to bow to pressure from county clerks eager to buy Diebold. As Ian Hoffman notes in an Oakland Tribune aritcle this morning:

The decision is likely to set off a buying spree for as many as 21 counties, more than a third of the state, as local elections officials rush to acquire one of only two voting systems approved for use in the 2006 elections. Registrars and clerks prefer having voting systems for at least six months before conducting a statewide primary like the one in June, partly because it is California's most complicated and error-prone type of election.

In his certification document for the Diebold systems, McPherson states that the approval was subject to several terms and conditions. Among those conditions, McPherson specifically states that voting systems "shall comply with all applicable state and federal statutes, regulations, rules and requirements, including, but not limited to, those voting system requirements set forth in the California Election Code and the Help America Vote Act of 2002 [HAVA]." He continued, "further, voting systems shall also comply with all applicable state and federal voting system guidelines, standards, regulations and requirements that derive authority from or are promulgated pursuant to and in furtherance of the California Election Code or [HAVA]…including but not limited to, the 2002 Voting System Standards/Guidelines, developed by the Federal election Commission and adopted by the Election Assistance Commission (EAC) and EAC Advisory 2005-04, dated July 20, 2005."

But just exactly how and when will Diebold's systems be brought into compliance with federal and state standards? Unlike Prof. Michael Shamos, who, after a cursory review of the Diebold systems, decided that AccuBasic was not prohibited by federal voting system standards and therefore was "okay" for use in Pennsylvania, the VSTAAB computer scientists, who took four weeks for their review, confirmed that "interpreted code in general is prohibited by the 2002 FEC Voluntary Voting System Standards, and also by the successor standard, the EAC’s Voluntary Voting System Guidelines due to take effect in two years. In order for the Diebold software architecture to be in compliance, it would appear that either the AccuBasic language and interpreter have to be removed, or the standard will have to be changed."

They go on to suggest that Diebold remove interpreters and interpreted code, suggesting that the architecture of the AccuVote-OS and the AccuVote-TSx could be changed so they do not contain any interpreter or use any kind of interpreted code, in order to bring the codebase into compliance with standards. But when? Before the special election in April? Before the primaries? And who is liable if they don't? Diebold? The counties that purchase Diebold equipment?

And is such a conversion even possible? As Joseph Holder of VoteTrustUSA's Voting Technology Task Force pointed out in a widely-circulated email:

"…as currently configured, neither the Diebold optical scan voting system nor the touchscreen voting systems are in compliance with the 2002 Voluntary Voting System Standards because they both contain interpreted code. The Standards do not contain any exceptions based upon mitigating circumstances or procedural safeguards. The procedural security measures recommended by the Advisory Board, and adopted by the Secretary of State, not only do not make the system comply with the standards, they are based entirely upon the presumption that there would be no opportunity nor incentive for any "insider" to take advantage of the known security vulnerabilities.

The Standards have not been changed, and under the new adopted standards the prohibition against interpreted code still exists. The only alternative is for the AccuBasic language and interpreter be removed. However, AccuBasic is fundamental to Diebold's software architecture. Remove it from the current Diebold voting systems that are deployed or approved, and they will not work."

"Voting system manufacturers and/or their agents shall assume full responsibility for any representations that a voting system complies with all applicable state and federal requirements as referenced above." If they falsely represent that their voting system does comply with the above requirements, they will be responsible for any costs for "upgrade, retrofit or replacement of any voting system..."

How could Diebold possibly claim to meet this condition? While historically Diebold has demonstrated little hesitance to misrepresent its election products, the current situation would seem to leave little plausible ground for Diebold. After all, the report that provided assurance to the Secretary of State, clearly states that interpreted code is prohibited by federal standards. Will Diebold now simply proclaim that intepreted code is no longer present in their software?

A last condition made by the Secreatry of State would appear to put counties in a financially risky position as well if they rely upon Diebold's assurances:

"Any voting system purchased with funds allocated by the Secretary of State's Office shall meet all applicable state and federal standards, including the 2002 Voting System Standards, etc.

It would appear that under this condition, if a voting system was found to not be in compliance with the requirements, the counties who have obtained funds under California Proposition 41 or federal funds under HAVA to pay for the voting systems would have to return them. They then would have to go after the vendor for either compensation or force them to make their voting systems into compliance.

Before the local election officials give a sigh of relief, it would be incumbent upon them to run all the documents released by the Secretary by their County Counsel. The conditions of approval may put them at greater risk legally and financially than a threat of a suit from the Department of Justice for non-compliance with HAVA. After all, the Diebold voting systems themselves are not in compliance. They should be more concerned with voter confidence and potential liability, than electoral expediency.