Election Updates

The recent report about security flaws in the Diebold AccuVote TSX and TS6 touchscreen election systems are educational on several levels.

First, consider what it says about the certification process for voting machines. The problems associated with the Diebold machines, it would seem, should have been noted in the certification process. This is especially true with the hardware issues that have been raised.

Second, think about the bind Diebold--and its customers--are in because of the certification process. If Diebold fixes the problems that have been raised, it could create certification problems for its equipment. The current certification process is neither quick or easy, so making a fix to a voting system is not the same as Microsoft providing an instant update to a security flaw. There needs to be an expedited process created to address critical problems that arise.

Third, the Diebold system problem is a Diebold problem, not an electronic voting system problem. As recent studies by the Center for American Politics and Citizenship at the University of Maryland have shown, voting machines are very different and the experience voters have with these machines vary greatly. Just because Diebold machines have been found to have a problem does not mean all machines have a problem.

Fourth, this report illustrates the importance of states having effective and clear laws governing the security of voting equipment before and during elections. As a part of the work we are doing studying vote counts and recounts for the EAC, I have read every state's laws and regulations related to voting system security and there is great variation in the scope and level of detail across the states. A state like Georgia--which has the most detailed laws on security on the planet--is not likely to have problems with the Diebold system. In fact, Georgia's machines do not have any of the physical problems that have been reported because the state forced Diebold to modify the casings and create sealed covers over the screws and external ports. It will be interesting to see if other states follow this lead.

While former Secretary of State Kevin Shelley is long gone (and is no apparently working as some sort of consultant), the fallout from his administration of HAVA funds continues. The EAC is ordering that the state return over $500,000 in HAVA funds, and that another $2.5 million be returned to the state for reallocation to improve election administration in California.

Some details are in the Los Angeles Times this morning. The return of the HAVA funds is the outcome of state and federal audits that we wrote about months ago. The LA Times story offers a brief summary of the audits:

California has received roughly $360 million under that legislation, much of it given to counties for such work as voter education and purchase of new voting machines.

After a state audit found problems with how Shelley's office spent some of the money in 2004, the U.S. Elections Assistance Commission requested its own audit.

Based on that audit, which examined the expenditure of $8 million, the Federal Elections Commission has ordered the state to return $536,000 to the federal treasury, saying it was spent on things "clearly impermissible" under the federal law.

The $2.5 million in spending that may have met the purposes of the legislation but lacked proper documentation was ordered returned to the state election fund to be used to improve the administration of federal elections in California.

Federal auditors concluded that Shelley's office improperly extended a contract for a law firm in order to avoid competitive bidding and skirt cost restrictions in the contract. The law firm provided advice on dealing with the media in matters unrelated to improving the voting system.

Auditors also found it impossible to determine whether workers employed by Shelley's office for "voter outreach efforts" actually performed work in line with the Help America Vote Act because their work statements and invoices were so vague.

Hopefully this will end the troubled Shelley era, and that the $2.5 million will be well spent on efforts to really help election administration in California.

In a recent posting, Seattle area blogger Stefan Sharkansky weighs in against voting by mail, which is going to be adopted by King County, WA starting in 2007.

Here is what I posted on his blog:

Stefan,

You can find a fairly vigorous debate on the merits of voting by mail here: http://www.blueoregon.com/2006/04/democracy_begin.html

You promote a few urban myths about voting by mail. The reason it is less expensive is that there are no polling places. Your comparisons to Iowa and other localities where they have a mixed system are inaccurate.

Voting by mail will save King County money, compared to the current system where 75% of the voters cast their ballots absentee anyway.

Voting by mail may be an invitation to fraud--I look forward to your future posts on the topic. However, my own study of voting by mail in Oregon and absentee balloting nationwide (reported on here: http://www.sos.state.or.us/executive/policy-initiatives/vbm/execvbm.htm) found no evidence of fraud.

Finally, the record shows that absentee/early voters are actually counted more, not less, accurately than precinct day ballots, where poll workers are operating under severe time pressures.

There are reasons to oppose voting by mail, but the argument you make is mostly false.

I received an email message today noting that the ACE project has launched a new website, updating their previous Administration and Cost of Elections site. Here is the relevant info from the email:

Partners Launch New ACE Electoral Knowledge Network
The ACE Electoral Knowledge Network ( www.aceproject.org) — the next generation of the Administration and Cost of Elections Web site — is now live. With the launch of the Network, we make available new tools and resources for electoral practitioners around the world. The new Web site provides a ready environment for professionals to generate, share and apply knowledge on managing elections.

New features include:


An updated ACE Encyclopaedia that includes authoritative resources on electoral administration;
An interactive Comparative Data Section that allows you to compare data from different countries;
An online collection of Electoral Materials that includes reports, manuals and other hands-on materials from around the world that is searchable by region, document type or topic area;
An online newsletter (Elections Today) that provides the latest news on elections around the world;
Access to Regional and Country-level Resources; and
An Online Demonstration of several features that enable professionals to interact and build knowledge on elections administration.
Visit ACE today at www.aceproject.org . To receive additional information about the ACE Electoral Knowledge Network, please visit http://ace.at.org/registration .

County registrars and recorders have been reporting difficulties recruiting pollworkers for the upcoming June 6 primary election. In order to help alleviate the problem in Los Angeles County, the Board of Supervisors voted last night to increase the pay of pollworkers. Clerks will get a raise from $55 to $80, and inspectors will get a raise from $75 to $100. According to a story in our local paper, this will make the pay of pollworkers in Los Angeles County roughly equilivant to that paid to workers in other California counties:

The move will cost taxpayers an extra $625,000 annually but makes pollworkers' pay comparable to that in other large counties throughout the state.

By comparison, Alameda County offers pollworkers the highest pay in the state: $110 a day for clerks and $150 for inspectors. Ventura County pays clerks $70 and inspectors $100; San Bernardino County pays clerks $100 and inspectors $135.

While a nice raise, and hopefully a move that will make it easier for Los Angeles County to recruit pollworkers, given that these folks are working a long (14-hour) day, this is still not much money to serve as an inducement for pollworkers ...

Here is a link to a short but interesting piece by Marc Rotenberg, "Real ID, Real Trouble?", that appeared in the March 3, 2006 "Inside Risks" column of the journal CACM (Communications of the ACM). In the piece, Rotenberg argues that while there are many objections and questions that have arisen about the U.S. "Real ID" plan, his concern is with data security:

Several objections have been raised about the plan, including privacy and cost, but the most significant concern may be security. As Bruce Schneier has explained, ``The biggest risk of a national ID system is the database. Any national ID card assumes the existence of a national database, and that database can fail. Large databases of information always have errors and outdated information.'' Even if the identity documents are maintained in the states, problems are likely.

One example concerns the vulnerability of the state agencies that collect the personal information that is used to produce the license. In 2005, the burglary of a Las Vegas Department of Motor Vehicles put thousands of driver's license holders at risk for identity theft. The information of at least 8,738 license and ID card holders was taken during the break-in, and reports of identity theft have already surfaced. Another report uncovered 10 ``license-for-bribe'' schemes in state DMVs in 2004.

Of course, what applies to general identification systems like this, also may apply to voter identification systems ...

An interesting story in the Star Press, Muncie IN speculates on why one candidate did significantly better among absentee balloters than did another in a primary for county commissioner.

In precinct place voting, Larry Bledsoe beat Mike King by .4% (50.2 to 49.8). But among those voting absentee, Bledsoe won by 24%. Why the different margin among what was presumably party regulars?

Neither candidate reported that they had a GOTV effort directed at absentee voters, and the Democratic Party said it was neutral.

This is the first time I've seen reports of such a differential outcome among absentee voters and precinct voters in a primary--but I have to admit, there are extremely few examinations of early voting in primaries at all.

Research opportunity?

After controversy surrounding absentee balloting provisions for the April mayoral primary, early voting has begun for the May 20th runoff between Nagin and Landrieu.

Election officials note that, although there were threats of lawsuits following the April election, none have surfaced. It appears that the NAACP, other voting rights advocates, and state and parish officials are cooperating to ensure voter access.

New Orleans Picayune story

UPI Story

The Star Telegram reports today that the problems with voter registration that we wrote about recently in California are not isolated. Texas is having similar problems with its voter registration system. As the paper notes:

Dallas County Elections Administrator Bruce Sherbet was panicking. The new state voter-registration database was rejecting thousands of registered voters each day because the state could not match their names with those in driver's license and Social Security records.

Each time the state turned away voters, Sherbet's office sent them a letter saying they had been rejected.

In Tarrant and Denton counties, elections administrators had another problem. The state's computer system initially required driver's license numbers. If your number began with a zero, you got kicked off the voter rolls for insufficient digits.

Some of the problems have been corrected over the last month and the number of problem VR files being referred to counties have decreased. However, there are still logistical issues with how the state handles mismatched ballots. As the article reports,

One key difference between Texas and many other states is that Texas won't print the names of mismatched voters on poll lists. In some states, like Oregon, names of those who are rejected will still be printed on the rosters delivered to polling places, but they'll be flagged as requiring verification...

The critical issue is whether states will use provisional balloting and other methods to help improve the process of fixing voter registration as they transition to new VR technologies.

The Los Angeles Times ran this story yesterday, detailing the efforts by 56 House Republicans to let the bilingual balloting provisions of the Voting Rights Act expire.

The reasoning behind these efforts is:

The 56 lawmakers support the act, but say the language assistance to voters — provided throughout much of California — undermines national unity, increases the risk of election fraud, and puts an undue burden on state and local governments.

"We believe these ballot provisions encourage the linguistic division of our nation and contradict the 'melting pot' ideal that has made us the most successful multiethnic nation on Earth," the members said in a letter earlier this year.

However, as the story also notes, these legislators are unlikely to achieve their goal, due to electoral politics --- Republicans want to keep attracting Hispanic votes.

No doubt, the need in many election jurisdictions to provide balloting materials in English and other languages does add to the complexity of election administration. We've seen that time and again in our election observing efforts, especially in Southern California.

But as far as I know there isn't any evidence that bilingual ballots have themselves led to a heightened risk of election fraud in any of the jurisdictions that currently are required to offer them. If readers have citations to cases where bilingual ballots have themselves been utilized to commit election fraud, please pass them along.