Election Updates

According to a wire report in this morning's Los Angeles Times, the much-discussed plan to allow Mexican expatriates the right to vote seems to have attracted little interest from Mexicans living abroad. As of the Wednesday deadline for determination of the final number of absentee ballot requests, the Federal Electoral Institute claimed a total of 56,749 valid absentee ballot applications were received.

Each of these 56,749 will receive an absentee ballot sent out between April 15 and May 20.

We've written before about how hard the process was for Mexicans living abroad to request and obtain an absentee ballot, clearly the data now bears out the predictions by many (including us) that the difficult provisions will lead to a low abstentee voting rate.

The U.S. Public Policy Committee of the Association for Computing Machinery (ACM) released a sixty-page report recently, "Statewide Databases of Registered Voters: Study of Accuracy, Privacy, Usability, Security, and Reliability Issues." This report essentially breaks down the issue of statewide voter registration databases by five criteria: accuracy, privacy, usability, security and reliability. The report provides many recommendations and discussion of issues associated with each of the five issues, echoing recommendations that the Caltech/MIT Voting Technology Project made in July 2001 about voter registration systems.

Having others help focus attention on statewide voter registration databases is a great, as readers of Election Updates know I've been concerned about the lack of testing and certification of these databases for a long time (having discussed these concerns again most recently at the AEI-Brookings event, see my slides, I also talked about testing and certification of voter registration databases at the "Voting System Testing Summit" in November 2005). Also, in October 2005 I wrote for the NIST conference on voting system threats about security risks associated with statewide voter registration systems.

One issue this ACM report seems to not have mentioned (I haven't found mention of it yet after my first quick read of the report) is the important issue of interoperability, which was called for in the recent Carter-Baker commission report, is something many states are now discussing, and which Thad and I have recently written about in our study, "The Next Big Election Challenge."

InfoSENTRY, an information technology company that specializes in security and systems analysis, recently released some new survey data on opinions of Americans regarding confidence of electronic voting systems. Based on the news release put out recently by InfoSENTRY, the survey question they posed to a national probability sample of 1004 American adults was:

“Now I am going to read to you some methods people use to vote in elections for public officials and ballot issues throughout the United States. As I read each one, please tell me on a scale of 1 to 5, where 1 means very low trust and 5 means very high trust, how much you trust each voting method to produce confidential and accurate election results. [READ AND ROTATE STATEMENTS]
1. Going to a polling place and making your choices directly on a computer screen and having the computer count the results
2. Going to a polling place, marking your choices on a paper ballot, and having your ballot counted by a computer scanner
3. Getting your ballot in the mail, marking your choices on the paper ballot, and mailing your ballot back to be counted by a computer scanner
4. Using a computer at your home, office, or some other place of your choice to cast your ballot over the Internet”

The press release had the following information on the results of the survey:

The following table shows the “positive trust scores” for the four types of voting systems in each of InfoSENTRY’s last three national opinion surveys. InfoSENTRY calculated these positive trust scores by adding the results of respondents who selected a “4” or “5” on the five-point scale.

	2004 Positive	2005 Positive	2006 Positive
	Trust Score	Trust Score	Trust Score
DREs	68%	62%	68%
Polling Place Optical Scan	59%	64%	61%
Vote-by-Mail	31%	35%	38%
Internet Voting	32%	29%	30%

It is interesting to note that, based on the way this question was posed to respondents that opinion has been relatively stable regarding electronic voting and polling place optical scanning, with these survey respondents expressing a relatively high degree of trust in both these voting technologies since 2004. Second, it is also interesting to see, certainly between 2004 and 2006, a slight increase in the expressed level of confidence in vote-by-mail, perhaps a reflection of the increased availability of vote-by-mail methods across the nation, and their increased use in recent elections in many states.

It would be fascinating to see more detailed analysis of these survey data; the press release promises "detailed results and analysis of this survey and attitude trends ... on February 14", but I've not yet been able to find more detailed information than the press release. If and when it is available, I'll post it in a subsequent essay.

It seems that the Haitian election controversy that I've been following has been politically resolved, after negotiations between the interim government in Haiti and international officials led to what from the outside looks like a questionable solution: the compromise was to not include over 80,000 blank ballots in the vote count, which effectively gives Rene Preval slightly more than the 50% necessary to avoid a runoff presidential election. Here is a link to the Washington Post story about this compromise.

This morning's Los Angeles Times ran a story that covered the irregularities in more detail.

First, in the LA Times story it notes that: "On Wednesday, U.N. peacekeepers recovered thousands of marked ballots and other election materials from a landfill north of the capital."

Second, it also notes that an estimated 85,000 ballots with no presidential vote, and at least 125,000 additional votes were invalidated --- that is 210,000 votes through out of the tabulation of the 2.2 million votes cast, a 9.5% invalidation rate (note here that there is still some disagreement about the number of ballots with no presidential vote recorded, 80,000 or 85,000).

Third, deeper in the story there is a quote from an official from the United Nations, who "estimated that about 3% of the ballots cast seemed to have gone missing." Add that to the blank and invalidated rate, that indicates that 12.5% of ballots cast were either lost or thrown out.

Fourth, "David Wimhurst, spokesman for the U.N. peacekeeping mission charged with providing security for the election, noted that isolated incidents left nine polling places ransacked, possibly accounting for the dumped election materials."

Also, this morning Delia Bailey, one of our VTP graduate student researchers, pointed out to me that the Washington Post currently has some very disturbing photographs on one of their websites. If you navigate to the following link, and click on "More photos" below the picture of the girl checking the preliminary election results in Port-au-Prince, you'll get a new window of six photos. The first four are interesting photos of Haitian election administration. But the fifth photo (showing an election worker at a U.N. facility checking trash bags filled with paper ballots awaiting tabulation) and the sixth photo (showing bags and boxes full of paper ballots spilling out of a truck, coming into a U.N. facility before tabulation) are both very disturbing documentary of what can only be considered highly questionable and very lax election administration procedures.

Building on the work of the UK's Electoral Commission, the UK's Department for Constitutional Affairs (DCA) announced earlier this week a serious of election pilot projects to be conducted in sixteen localities in their May 2006 local elections. The complete description of the localities and their pilot projects is here.

Some of the pilot projects include:

• Early voting in Broxbourne, South Tyneside, Sunderland, Gateshead and Newcastle.
  
• Electronic checking of postal vote signatures in the Epping Forest District Council and the London Borough of Newham.
  
• Early voting at supermarkets in the London Borough of Lewisham.
  
• Mobile voting and "E-enabled, networked early voting (at) local military bases" in the London Borough of Newham.
  
• The use of a "networked electronic registrar" in Shrewsbury and Atcham Borough Council.
  
• "Early mobile voting for service voters at military bases in Westminster."
  

According to the material provided so far, it appears that the DCA and the Electoral Commission will work with these local authorities to make sure that these efforts are adequately evaluated. We look forward to seeking the results of these evaluation studies, as the previous projects undertaken by the Electoral Commission have been very valuable (as we have written about in the past on Election Updates).

Some of the background information regarding these pilot projects, including the project prospectus, are available from the Electoral Commission.

Today's Pasadena Star-News has the second installment of my opinion column, this one on California's need for redistricting reform.

This opinion piece follows some earlier essays on Election Updates on redistricting reform in California:

1. "Why did redistricting reform fail in California and Ohio in 2005? New survey research provides great data and a lot of insight." This essay discussed some wonderful survey data that looked at Proposition 77 in California, and why it failed to pass last fall.
   
2. "Competitive districts possible in California, new study concludes." This essay briefly discussed a thoughtful analysis by the IGS up at UC-Berkeley about competitive district plans in California.
   
3. My first opinion piece in this series on the need for general election reform in California. This piece argued for redistricting reform, in addition to reforms in campaign finance and the initiative process.
   

As today's opinion piece argues, California needs redistricting reform now --- and the current political conditions seem suitable for fixing our redistricting process.

Electoral chaos continued today in Haiti, with continued unrest and mounting anger among Haiti's electorate regarding delays, alleged irregularities in balloting and vote tabulation, and continually changing vote totals.

According to current reports of partial tabulation results,
a stunning 7 percent of all ballots counted so far have been invalidated. Other media reports repeat this 7 percent invalidation figure, and also have claimed that at least another 4 percent of ballots examined so far were blank.

When the dust settles, examining how many ballots were invalidated in this election will provide some important data for analysts to study. Hopefully these data will be available to the international research community.

Just to keep my promise, the AEI-Brookings folks have put on their Election Reform Project website the video and transcripts from their event last week.

Here are links to the materials:

1. Senator Obama's address, video.
   
2. Senator Obama's address, transcript.
   
3. Panels, transcript.
   

According to a story in Computerworld, the Bush budget proposal seeks to increase the funding for National Institute of Standards and Technology's core programs by 24%, from $431 million to $535 million. The increased funding in intended to help NIST in the area of information security, including their work on voting system security.

In the story, there is a brief quotation from Bruce Schneier:

Bruce Schneier, founder and CTO of Counterpane Internet Security Inc., said that while NIST does good work in areas such as in voting security, it has no enforcement capability and no way to ensure that its security developments are implemented. He said that he would like to see the federal government use its purchasing power to require companies to meet certain security standards. “We all benefit when a big player uses its purchasing power,” he said.

Schneier is right on both fronts: NIST has been doing good work, and has started out some useful research agendas on voting system security --- but there is much, much more that the federal government can do to insure that the results of such research get incorporated into the marketplace. It will be interesting to hear more about how the research results of NIST on voting systems and voting system security might best be implemented in the near future, whether that will be through a beefed-up standards program, through stronger federal testing and certification, or via some other mechanism.

As I discussed at last week's AEI-Brookings election reform event, election officials need to concentrate on threat assessment and mitigation strategies, and we need to fix the broken federal testing and certification process.

There was a fascinating story in this morning's Los Angeles Times, about a bizarre situation in Vernon (a small city in Los Angeles County).

Vernon is an industrial town, and to be honest, I wasn't aware that it had more than a handful of registered voters within the city limits. Turns out that there are less than sixty registered voters, in a city that has 93 residents, but somewhere around 44,000 people who work in the city limits.

The other interesting dimension to this story is that Vernon has a city council, but apparently there hasn't been a contested election since 1980; that's probably the reason that most political observers may not even know that there are elections in Vernon, or even registered voters.

But the plot recently thickened, as apparently recently eight people have moved into the city limits, registered to vote, with three then filing papers to run for city council. According to the story, the city has denied the validity of the residence taken up by these new registrants, and then invalidated their voter registration status. That latter action has allow the city to block the potential candidacies of the three hopeful city council challengers. One of the odd twists to this story is the alleged connection between a disbarred lawyer, who used to work for Albert Robles (a former city official from South Gate, who was at the center of a massive corruption scandal in that city), and these potential challengers to the Vernon political machine.

However, the Los Angeles County Registrar Recorder's Office has now informed the city that they do not have the authority to invalidate voter registrations. According to the LA Times story:

Deborah Wright, executive liaison to the county registrar, said the city had sent her copies of the three men's candidate nominations with a stamp that read "canceled."

But the city "doesn't have the authority to cancel voter registrations," Wright said. "They have no authority to do that at all."

Even if the building was illegal, she said, election law does not require voters to live in a legal residence. A 1985 court ruling held that even homeless people have a right to register to vote, she noted.

"A person could be squatting in an abandoned warehouse, and the city may not like it," she said. "But that doesn't affect their ability to register to vote."

Thus, according to this story, this sounds like old-time politics, involving the potential disenfranchisement of voters by a local political machine intent on retaining their political monopoly. We'll try to follow this story as events unfold. While the numbers of possibly disenfranchised voters are small, in a situation where there are really only a few handful of voters in the city's electorate, resolution of this situation could have a decisive effect in Vernon in the future.