Election Updates

My posting last week on Rhode Island's efforts to implement an electronic voter registration process in DMV or AAA locations spawned some email responses. One, from electionline.org's Doug Chapin, pointed me to Arizona's EZ Voter Registration process, which is described in this FAQ. This system is similar to the Rhode Island process, in that it appears to be part of the DMV process ("ServiceArizona"). As the FAQ describes it:

ServiceArizona is an easy way for Arizona residents to take care of Motor Vehicle Division (MVD) business, any time of the day or night, from the comfort of their home or office. Currently, Arizona residents can use ServiceArizona for online Vehicle Registration Renewals, Personalized/Specialty Plates, Fleet Registration Renewals, Address Changes, ordering duplicate Driver Licenses or Identification Cards with or without an address change, Driver License Reinstatement, obtaining a Restricted Use 3-Day Permit, viewing available Plate Credit, Duplicate Vehicle Registration, obtaining a 30-Day General Use Customer Permit, Plate Refund, Vehicle Sold Notice, and for Fleet Application.

And for voter registration!

The major difference relative to the Rhode Island system, is that the Arizona process seems to be one that can be conducted over any Internet connection, as long as the person attempting to register has an Arizona driver license, state ID card, or has already somehow filed a digitized signature with the DMV office.

One of the more interesting claims about the Arizona system is that they are "hard at work on an ongoing basis to keep ServiceArizona one of the most secure websites on the Internet" (on the FAQ page referenced above). The security features mentioned include:

• "All confidential records are kept locked inside an "electronic vault" protected by IBM's most advanced security technologies and among the most comprehensive security programs and services in the industry...the IBM Secureway (tm) line of products.
  
• IBM uses the industry's standard bearer for electronic commerce to encrypt your credit card number before it ever leaves your computer. This means that your credit card number is kept confidential within the system during and after your use of the system.
  
• Professional IBM security specialists have trained IBM's employees on security procedures. These procedures are checked and double-checked by both the Arizona Motor Vehicle Division and IBM's security experts on a regular basis.
  
• IBM conducts unannounced audits of every security checkpoint up to 50 times a year. This includes a rigorous inspection of IBM technology, procedures, training and personnel."
  

From this it sounds pretty clear that the vendor for this system is IBM.

Thanks, Doug, for pointing this system out to me. Given that Arizona has been providing NVRA implementation data for the most recent NVRA study, we will have some data with which to evaluate this process. The comparison between this system and the Rhode Island process will provide an interesting study to determine the effectiveness of these two different models of electronic voter registration. As these two systems appear to have very different system architectures, an analysis of those architectures for usability and security would also be very interesting to undertake.

We'll keep looking at how states are moving to implement the HAVA statewide voter registration mandates, especially how and whether states are allowing individual citizens or voters to have interactive access to the new databases. Feel free to send along email if you know of other state efforts in this area.

Many people have written off Internet voting. After the Department of Defense cancelled SERVE and the critiques of electronic voting became quite intense, little thought was given to this form of voting technology.

However, the Internet voting environment in Europe is much more dynamic than in the United States. Right now, Internet voting is being done nationwide in Estonia. Articles in EurActiv.com and the Associated Press discuss the system and its functionalities. The Estonian system is just one part of a larger e-government system that provides all residents with a digital signature on a key card that is used for a person to authenticate themselves (in conjunction with a PIN) for a wide array of government services, like filing taxes. The government in Estonia has been quite supportive of expanded e-government, obviously, and with 1.3 million voters, they are at a scale that allows them to conduct full trials of e-voting that would be equivalent to testing it in a medium to small state here in the U.S.

As the Financial Mirror describes the system:

The proposed e-voting process is relatively simple: you sit at a computer, slide your ID card through an electronic reader which calls up a special election website showing candidates in your constituency, and then select one with the click of a mouse. To confirm your choice, you must enter a PIN number - and that's it. The biggest difference compared to a paper ballot is that the voter can think over his decision and change it a countless number of times before the polls close. Alternatively, he or she can go into a polling station and fill out a traditional ballot slip - thereby canceling the e-vote.

Everything is not all wine and roses in Europe with Internet voting. In the UK, Internet voting trials scheduled for 2006 are being postponed because of concerns about fraud that arose in all-mail-paper balloting this year and last, and because small-scale trials have been expensive in the past.

Still, the Estonia experience may raise interesting issues about Internet voting. The success of the Estonians may lead other smaller European nations to follow suit and move to the e-government, e-voting bandwagon.

Every once in a while, I run across a simple --- but cool --- innovation in election administration. The one for today is a neat tool located front and center on the San Mateo County (California) County Clerk/Recorder website, and it allows an absentee voter to track and confirm her absentee ballot request. This tool also is supposed to confirm when your absentee ballot was received by the election officials. I say "supposed to" here because as I'm not a registered voter in San Mateo County, I've not tried it, but I trust that it works!

This is a cool innovation because absentee voting can sometimes be a mysterious process for voters, as they do not necessarily know if their request for an absentee ballot was received --- nor if their returned ballot was received and included in the final tabulation. As Betsy Sinclair, Thad and I have written in our paper "Whose Absentee Votes Are Counted: The Variety and Use of Absentee Ballots in California", there are many absentee ballots that are never returned to election officials, and some of those that are returned are not included in final tabulation for a variety of reasons. New and innovative tools like this one offered by San Mateo County may help voters keep track of their absentee ballot, and may help empower them to act if it seems that their ballot request was not accepted or if for some reason their ballot was not returned to the election office.

On Wednesday, Rhode Island unveiled a new electronic voter registration system, which integrates the state's "Motor Voter" procedures into their new statewide voter registration system.

At this point, precise details of the new electronic voter registration system are a bit sparse, but according to a press release from the Rhode Island Secretary of State's office, issued yesterday, the new system has the following functionality:

Motor Voter e-Registration is available to Rhode Islanders using services at any of the states eight DMV locations and also to AAA members at club locations. When a citizen is getting a license or ID for the first time, renewing a license, or changing their name or address DMV employees must ask citizens if they would like to register to vote as required by state law. The process takes minutes and is similar to a touch-pad used at supermarkets, banks or convenience stores. Citizens must still answer questions about their eligibility and swear an affirmation, just as they would have under the old system.

...Motor Voter e-Registration simplifies the voter registration process at the DMV by allowing Rhode Islanders to confirm information on the touchpad and sign the voter registration electronically. In the past, citizens needed to fill out a paper form, which was then delivered to the Rhode Island Board of Elections and forwarded to local boards of canvassers where it was manually entered into the voter registration system resulting in a longer registration processing period.

Apparently what then happens is that the voter registration application information is sent to a local election board through the state's "Central Voter Registration System" (CVRS), where it is then either approved or rejected.

As far as I am aware this is the first statewide voter registatration system that allows potential voters to interact with a computerized statewide voter registration database, using hardware and connections in the offices of a state agency (and apparently a private third-party organization, AAA). This new approach to implementing the provisions of "Motor Voter" could resolve some of the potential problems with voter registration applications that arise from state agencies and third parties, by speeding up the process, providing voters real-time response, and by helping to reduce common errors that can prevent such registration applications from being accepted or which can introduce errors into voter registration records.

Exactly how the new connections between the DMV and AAA offices have been implemented is not clear from the information that I've been able to find. According to testimony before the EAC during hearings in Boston earlier this year, Rhode Island Secretary of State Matthew A. Brown described the CVRS in detail, noting that it runs on a "private, high-speed state government network, RINET-MUNI", described as "a secure and efficient system." Some additional details of how the CVRS operates using the RINET-MUNI system are given in this FAQ. But from these descriptions, it does sound as if Rhode Island has taken a major step towards implementation of electronic voter registration.

Unfortunately, it might be quite difficult to eventually evaluate the effectiveness of this electronic voter registration system. As I was looking into Rhode Island's voter registration process, I checked the recent EAC publication "The Impact of the National Voter Registration Act of 1993 ..." for data on Rhode Island voter registration procedures (especially the data on how many registrations they process using "Motor Voter" procedures). Unfortunately, Rhode Island seems to have reported little to no data for the purposes of this report to the EAC, meaning that unless someone can get data from Rhode Island on NVRA implementation in 2004 through other means, there may not be any reliable baseline information from which to analyze how this new electronic registration system influenced some dimensions of the voter registration process in Rhode Island. This is just another example of how hard it can sometimes be to analyze election administration practices, as we just lack basic data altogether too often.

This was just put out by Doug Chapin in his October 13, 2005 "electionline Weekly" report:

EAC Announces Intent to Award Millions for HAVA-Related Research
By Doug Chapin
electionline.org

Late last month, the United States Election Assistance Commission (EAC) announced its intent to award a series of research contracts totaling nearly $4 million. The research program - and the federal funds it commits - is designed to begin addressing the EAC's research responsibilities under the Help America Vote Act.

At press time, contracts were still being finalized and thus the information has not yet been posted on the EAC website. However, electionline.org has learned that the proposed awards include the following:

* $681,400 to the American Institute of Graphic Artists (AIGA)'s Design for Democracy program. The project will focus attention on better design and so-called "human factors" research for ballot materials, signage, and a new National Voter Registration Act form;

* $650,000 to the National Academy of Sciences for "voting registration database technology assistance";

* $378,310 to the International Foundation for Electoral Systems (IFES) for recruitment, retention and training of pollworkers other than college students and $346,615 to the Center for Election Integrity at Cleveland State University for similar research regarding college student pollworkers;

* $352,000 to the National Association of State Election Directors (NASED), former Johnson County, KS election commissioner Connie Schmidt and Dr. Brit Williams, consultant to Kennesaw State University's Center for Election Systems to prepare proposed "election management guidelines";

* $300,212 to the Council for Excellence in Government to conduct "voter roundtables";

* $285,226 to publius.org for "public portal access research";

* $257,350 to the University of Utah's Thad Hall for research on recounts and state standards on what constitutes a vote, plus best practices regarding the same (NOTE: electionline.org is a subcontractor to the University of Utah on this project);

* $224,737 to the Center for Governmental Responsibility at the University of Florida's Levin College of Law for a legal clearinghouse, including a web portal for recent and pending election cases around the country;

* $217,105 to InfoVoter Technologies for "voter hotline research". [See the April 7, 2005 edition of electionline Weekly for information on InfoVoter's role in the 866-MYVOTE1 project in 2004]; and

* $110,000 to the Century Foundation's Tova Wang and election legal consultant Job Serebrov for research into voter fraud and voter intimidation.

Some of the figures in this story were drawn from the October 3, 2005 issue of Election Administration Reports published by Richard Smolka.

Thanks! - DMCj

Thanks for the update, Doug Jr.!

There continues to be materials added to the NIST "Threats to Voting Systems" website, material that we hope to dig into here in the near future. One important new addition are audio recordings of the workshop sessions. Interesting to listen to, though the quality leaves a bit to be desired and they have simply posted three enormous mp3 files (instead of breaking them into specific speakers, for instance). I'm sure that these mp3 files will become favorites of our friends who like to listen to mp3 recordings about voting technology and election reform, though I suspect many will wish that that NIST (after all, they are the "National Institutes of Standards and TECHNOLOGY") could have provided better audio --- and some video to go along with it!

Also, some of the speaker presentation slides are also available, which are handy if one wants to listen and watch at the same time.

Let's say you are an election official, and one of the following scenarios happens:

• You find out days before a major general election that snafus have either prevented or substantially delayed the mailing of more than 400,000 absentee ballots.
  
• You find out after the close of voting that a storage device used to record electronic ballots did not have the storage capacity that you assumed it did, resulting in the loss of 4,500 votes.
  
• Polling places open as expected on election day, but a major terrorist attack causes a massive disruption of the election in a large part of your jurisdiction.
  
• An bitterly contested election you administer is decided by only a few hundred votes, and the candidate that you announce to be the loser claims that fraud and malfeasance in your office cost him the election.
  

Each of these scenarios have happened in recent years (King County, Washington; Carteret County, North Carolina; New York City; Compton, California). And in each case, election officials scrambled to deal with the crisis, controversy, and public scrutiny.

A recent study from Donald P. Moynihan, "Leveraging Collaborative Networks in Infrequent Emergency Situations", provides election officials with some important lessons for how they might establish networks --- or utilize existing networks --- to deal with infrequent but serious crises. Moynihan analyzes a particular case, an outbreak of "Exotic Newcastle Disease" in California and Western states, and studies how a variety of federal, state, and local government agencies teamed up with some private organizations to produce an effective response to the outbreak of this poultry disease. The last major outbreak of this disease had occurred decades before, and technology had changed so dramatically, that the relevant stakeholders had little precedent to draw upon in devising a rapid and effective response strategy. But core to the response, as Moynihan argues, was the development of a task force, that itself facilitated connections between responders through existing networks and worked to forge new connections between responders who had previously not been networked together.

Moynihan's report concludes with a series of lessons learned and important recommendations. The five recommendations are well worth consideration by election officials as they develop emergency response and threat assessment strategies:

1. Pre-plan --- but expect to plan some more once the emergency occurs.
   
2. Identify the resources needed to deal with the emergency and match them with the competencies of organizations.
   
3. Create trust where you can; find alternatives where you can't.
   
4. Take advantage of technology innovation to dramatically improve emergency network coordination and efficacy.
   
5. Establish, formalize, and communicate basic procedures that familiarize workers with their tasks.
   

These are solid recommendations that election officials might adopt for emergency or crisis situation contingency plans, if they have such a plan. And if you don't have an emergency or crisis plan, then you might consider developing one soon!

On Friday, October 21st, Common Cause is holding the 2005 Southwest Election Reform Conference in Estes Park, Colorado. The conference agenda includes an opening plenary session, where I will be one of the speakers, followed by a set of workshops on policy, grassroots activity, and collaboration. After a networking lunch, there is going to be a set of meetings regarding how the states in the Southwest can move forward with election reforms in their individual states.

I think that the conference has the potential to accomplish quite a bit for moving election reform forward in the Southwest, especially since it is designed to promote an open and free dialogue about the key issues. If you are going to be in Colorado next week, Estes Park could be the place to be!

A critical question for the evaluation of election reform proposals is to assess their potential effect on outcome variables, and in the case of voter registration reform efforts (a subject of much recent debate, especially in the wake of the recent and controversial Carter-Baker report) one of the most important outcome variables is voter turnout. Thus, the first question regarding voter registration reform is how much might reforms increase voter turnout?

Most of the available research on this subject. studying the potential impact of voter registration reforms across states, asserts that making voter registration easier can increase voter turnout by as much as 7 to 10 percent. The previous research, whether based on studies of actual election returns or based on survey studies, generally tends to estimate increases in turnout of about this magnitude from significant efforts to reduce the barriers of voter registration.

In a recent study (published electronically in Political Analysis), MIT colleagues Stephen Ansolabehere and David M. Konisky offer a very different approach to estimating the potential effects of voter registration reform. Instead of looking at the possible effects of registration reform across states, Ansolabhere and Konisky look at the actual effects experienced by the states of New York and Ohio when those states moved to impose uniform voter registration requirements statewide, which occurred in New York in 1965 and Ohio in 1977. The question then becomes estimating how much voter turnout fell in New York and Ohio counties that did not require voter registration, after the imposition of voter registration requirements, controlling for other attributes of the counties that might be expected to influence turnout.

Ansolabehere and Konisky compile a dataset of county-level data, ranging from 1956 through 2000. When they look at the county-level data in both states across this span of time, they find that in presidential election years counties in New York with voter registration had average turnout rates that were 13 percentage points lower than counties without voter registration; in Ohio, the difference was 10 percentage points between counties with and without registration (page 9). Looking at non-presidential election years, the respective differences in New York were 10 percentage points, and in Ohio 5 percentage points. These estimates are in rough agreement with the previous literature.

But when Ansolabehere and Konisky turn to estimate the effect of imposition of voter registration by looking at the change in turnout in counties that did not have voter registration requirements relative to the counties that had voter registration, and then imposing some control variables in a multivariate statistical analysis (a "fixed effects panel regression" for election geeks), they estimate a slightly lower potential effect. They find that the effect of imposing uniform voter registration requirements led to a decline of turnout of about 5 percentage points in Ohio and 3 percentage points in New York. Yes, the effects of voter registration imposition are statistically significant and are in the direction that we expect, but these estimates are, of course, somewhat lower than those that have been found in other previous studies.

This is an interesting study, with a new take on how to examine the possible effects of voter registration reforms. This study raises some questions, though, that we ought to consider. One question is how representative Ohio and New York are, and whether we ought to generalize to the nation (or even other states) from the analysis of only these two states. My take here is that some additional research as to the same sort of effect in other states, especially in other regions of the county, would be very helpful. Second, I also wonder about the timing of these changes in Ohio and New York, and whether they tell us a great deal about the potential for voter registration reform today; after all, these changes in both states occurred a long time ago, and in a vastly different context than exists today. A third question is whether the "fixed effects panel regression" approach is ideal for this type of "natural experiment"; this "natural experiment" that Ansolabehere and Konisky have found might be a excellent place to explore the "propensity score matching" estimation procedures that have been the subject of much discussion recently in social science and statistics (two of our graduate students at Caltech have been working on applying these procedures to similar problems, see Delia Grigg's poster presentation on voting technology and residual vote rates and Betsy Sinclair's award-winning poster presentation on the ballot order effect; more on both of these innovative new research projects later.)

As I flew from San Diego to Salt Lake City on Sunday and read the Los Angeles Times, I found a great story that made me feel better about any issues we have with elections in the United States. Haiti Election Grows More Distant as Preparations Move Sluggishly blared the headline. How bad are things in Haiti? Here is an excerpt.

The council has yet to publish a list of approved candidates even though registration closed two weeks ago. Friday's planned lottery for ballot positions has been canceled. The candidate list is a necessary first step, to be followed by three weeks for legal challenges, another three weeks for contractors to print ballots and two weeks to get the materials distributed to polling places. A similar period would be needed ahead of second-round voting, which would push the run-off and outcome well beyond Feb. 7.

"Do the math," said Gerard Le Chevallier, the U.N. election coordinator who has overseen contentious ballots the world over. "We can't compress this critical path of eight weeks any more than we already have."

And you think there are problems with voter IDs in the US? Check this out....

A biometric voter card initiative also has been delayed. More than 3 million of the estimated 4.25 million eligible voters in Haiti have signed up for the cards. But less than 1% of the cards, which would be the first national ID cards in Haiti, have been produced. Until recently, violence in the teeming slums also prevented registration workers from opening sites for hundreds of thousands of the poorest Haitians.

Election officials in the U.S. get it right 99 percent of the time, but as we all know, there are things that could be better and things that will likely improve between now and 2008. But it is nice to know that, regardless of how bad an election is in the U.S., it could be worse. We could be in Haiti.