Category Archives: Voter fraud

New Mexico SOS Turns Over 64,000 Registered Voters for Potential Fraud Investigation

The New Mexico Secretary of State Dianna Duran has provided 64,000 registered voter names to the New Mexico State Police for fraud examination. This represents about 5.3% of all registered voters and about 7.7% of voters in the 2010 election.  No details were given on how these voters were identified, but that represents an enormous number of records to work through.  No process was identified for how the New Mexico State Police would examine the records or how long it would take to review them.  More information about this story can be found here.

The article mentions the possibility of administrative errors.  This is very likely.  My own experience with working with voter registration files is that they are somewhat dirty.  By that I mean the file consists of missing data, incorrect data, duplicate entries, etc (also see an excellent report on the quality of voter registration files by Stephen Ansolabehere and Eitan Hersh).

What is the problem? A major part of the problem is that the data collection and entry process leads to data entry errors.  Individuals register to vote in their own handwriting and then administrative staff interpret and enter the information into the electronic system. Obviously this is a process rife with numerous points for potential error.

One possible solution to this is to move to some sort of electronic registration, perhaps self-registration on-line and/or augmenting information in voter registration files with information from other state and federal databases to obtain more correct information.  See for example the op ed piece by Mike and Dean Logan.

As the investigation progresses, it is worthwhile for us to understand the process.  One thing we can learn from the process is where the problems exist and how to fix them.  This is an opportunity for us to collect important data on administrative problems with the voter registration process.

Survey on the Performance of American Elections Data Available

As part of my pre-Thanksgiving clean-up, I have finally gotten around to posting the data sets and documentation for three surveys my colleagues and I did in 2007 and 2008 to gauge the quality of American elections. The studies were funded by Pew, as part of their Make Voting Work Initiative, along with the late, great JEHT Foundation and AARP (for the Nov. ’08 study). The studies were conducted in November 2007 (gubernatorial races in KY, LA, and MS), February 2008 (15 Super Tuesday states), and November 2008 (all 50 states). Lots of questions about how well elections were run, from the perspective of voters, plus some questions about why non-voters didn’t vote.

The data are all on the MIT dSpace site:

One feature of these datasets is that we did parallel administrations using the Internet and telephone (random digit dialing), so people interested in how these two survey modes differ should find things of interest to them there.

California man charged with election fraud

This is from a press release from the CA Secretary of State office:

Felony charges were filed and an arrest warrant issued Monday for an Orange County man suspected of committing voter registration and election fraud, Secretary of State Debra Bowen announced today.

An investigation by the Secretary of State’s Election Fraud Investigation Unit revealed that Nativo Lopez of Santa Ana leased office space in the Boyle Heights area of Los Angeles and allegedly registered to vote at that commercial address although he lived with his family in Orange County. Lopez also allegedly cast an illegal ballot from Los Angeles in the 2008 Presidential Primary Election.

On June 22, the Public Integrity Unit of the Los Angeles County District Attorney’s office charged Lopez with four felonies: fraudulent voter registration, fraudulent document filing, perjury, and fraudulent voting. A warrant was issued for his arrest and bail was set at $10,000.

Under state law, registering where you are not entitled to vote is a felony punishable by up to three years in prison; fraudulent voting is also a felony punishable by up to three years in prison.

Voter Confidence 2: By Mode and By Vote Choice

I wrote a few days ago about some data we collected as part of the 2008 Cooperative Congressional Election Study (CCES), testing assumptions about “voter confidence,” a survey respondent’s perception about whether their vote will be counted accurately or not.

We wanted to test for two other patterns that have been reported in past work by Alvarez, Hall, and Llewellyn. First, are their mode effects–do voters who cast ballots at the precinct place express higher levels of confidence than those who cast a by-mail ballot or early ballot? Second, are voters who cast a ballot for the loser (in this case, John McCain) more suspicious about the validity of the count?

As shown in the graphic below, we were able to replicate “loser’s regret”: McCain voters were substantially less confident than were Obama voters. However, we found no statistically significant decline in voter confidence among those who cast an absentee ballot (although the level who express “great deal” does drop). Now that I look at this figure, I realize that I need to re-run the statistics lumping together early in person and precinct place voters. More in a few minutes …

National Academies of Science voter registration workshop

I’m off to Atlanta — the site of yesterday’s Senate runoff election — for two days of discussion regarding voter registration systems as part of the ongoing National Academy study of Statewide VR systems. So if you happen to be in the area, Thursday’s open sessions look pretty interesting. Unfortunately I can’t find the URL for the workshop agenda on this slow 2G connection I have, but I’ll dig it up later when I have a faster connection.

This all assumes that I get to Atlanta today; we’ve been sitting here in the plane at LAX for 45 minutes waiting for the pilot who is supposedly stuck in traffic (would they wait for a passenger who didn’t plan ahead and got stuck in traffic?).

UPDATE:  here’s the current URL.

Governments and Companies Struggle With Data Security

There’s an interesting story in the Washington Post, “Companies Struggle to Keep Data Safe.” The lead paragraph in the story notes: “A staggering 94 percent of companies admit that they are powerless to prevent confidential data from leaving their company by e-mail, according to a new study from Mimecast.”

But it’s not just corporations that are having trouble with data security, especially data leaks through email:

“Most leaks occur via e-mail,” confirmed James Blake, Mimecast’s chief product strategist. “Two thirds of data leaks occur via e-mail.” He highlighted an Infowatch survey, which said that 95 percent of leaks are accidental. “I would go along with that figure,” he said. “From what I have seen most leaks are accidental.”

Yet e-mail leaks are nothing new. Back in May this year, the Conservative party accidentally e-mailed the voting intentions of 8,000 voters in the Crewe and Nantwhich by election, to a journalist at a local radio station. It was thought that the automated completion of an e-mail address was to blame for the mistake.

Government agencies — including election officials — might want to take a look at this article and some of the information it reports regarding data security issues.

The survey was conducted by Mimecast. Here’s the interesting content from the press release:

LONDON 29 July, 2008 – An independent survey commissioned by email management company Mimecast has revealed that an alarming 94% of companies are powerless to stop confidential information from leaving their organisation by email. The survey revealed that just 6% of all respondents were confident that anyone attempting to send confidential company information by email out of the organisation would be prevented from doing so.

The independent survey, conducted amongst a sample of 125 IT managers, revealed that 32% of companies would not even be aware that confidential information had been leaked so would be unable to take steps to minimise the damage or track down the source of the information. However, 62% would be able to retrospectively identify the email leak once the information had been sent, but confessed to being unable to prevent its disclosure.

According to Dr James Blake, security expert at Mimecast, “The picture revealed by this survey points to fundamental security issues with protecting not only a company’s own data but also customer data like patient records or credit card numbers.” He adds, “With the blurring of boundaries between company employees and external consultants, contractors, outsourcers and other third parties, it is now much more difficult to ensure the appropriate flow of information outside the organisation. Especially since the majority of employees are now knowledge workers with access to significant amounts of confidential data.”

According to Bob Tarzey, security analyst at Quo Circa, “These figures do not surprise me – on the whole employees are not sending stuff out maliciously, but through carelessness or lack of fore-thought. Education can help to some extent, but many employees are using communications tools all day, every day and mistakes will happen, so having checks in place makes sense. Affordability of available technology to tackle the problem is also a problem, as most businesses are unable to invest in the high end, on-premise Data Leak Prevention (DLP) products that large business can, so the availability of on-demand services like those offered by Mimecast to achieve the same end is welcome providing performance is not adversely affected.”

A few items from the newswire

Doug Chapin and I have been asked about this before–if you cast an early ballot and then die, is your ballot still counted?  I think we’ve agreed on a standard reply.  While everyone thinks this is an issue that can be dealt with using absentee ballots, what about early in person voting?  It seems to me that once you cast a ballot, that’s it.  (Of course, I’ve written before about the odd few states that have “do-over” provisions.

Am I the only one who does not see a problem with a private firm charging a small fee ($9.95) to process your voter registration?  Sure, it’s not necessary, but people pay for all kinds of things that they can get for free or low cost.

Enough about the professional skepticism–this story out of California does raise a point of concern.  If online voter registration is legalized, how do they check signatures on absentee ballots (as high as 40% in some California counties)?  The bill has a provision for capturing the digital signature from the DMV, but the Butte County recorder thinks these are badly out of date.  (Full disclosure: Pew and JEHT are funding a few studies of online voter registration systems.)