The The Baltimore Sentinel reports on the continuing fight over early voting in Maryland.

The recent set of articles in the American Prospect typify the debate over voting by mail. By mail voting has some distinct advantages and some disadvantages. But let’s make sure we know what voting by mail does do, and what it does not do, before we advocate for its wholesale adoption.

Voting by mail is super convenient. Citizens can cast their ballots at a time and place of their own choosing. And there is little doubt that citizens feel good about voting by mail–numerous surveys in Oregon have shown overwhelming support for the method.

Election officials also like it. Don’t compare vote by mail to precinct place voting. In many states, such as California, more than a third of the ballots come in via the mail (absentee voting). In extreme cases, such as Washington state, more than 3/4 of ballots are absentee. (This is why many Washington counties–most notably King County, the state’s most populous–are going completely by-mail.)

In such “mixed” systems, where elections officials shoulder the costs not just of election day voting, but also widespread absentee balloting, it’s not surprising that they prefer to opt for one system. Post-election analyses, such as those produced by the Voting Technology Project at CalTech, show that absentee ballots are more likely to be counted, and counted more accurately, than ballots cast at a polling place.

So what’s the rub?

Articles like that by Robert Kuttner go beyond what we know about voting by mail to make grandiose, unsubstantiated, and often flat out wrong claims about voting by mail.

Let’s take the most commonly disseminated urban legend about voting by mail: that is enhances turnout. Kuttner makes the common fallacy of attributing all of Oregon’s turnout advantage to voting by mail: But the more deeply you explore the Oregon system, the better it looks. It costs less than half the traditional polling-place system, and has turnout 10.5 percentage points above the U.S. average.

This is, of course, completely silly. Oregon has always had higher turnout than the nation at large. One would think that Kuttner–listed as the editor at the Prospect–doesn’t read his own magazine, because the excellent article by Don Hamilton of the Portland Tribune shows the fallacy of the turnout claim (check out the chart–it tells the whole story).

By mail voting increases turnout only in lower profile, lower intensity contests (such as state and local elections). It does so by encouraging among regular voters. But if you are looking at voting by mail as a way to expand the electorate, or enhance participation among less empowered groups, there is no evidence that it will work.

Within about an hour after publishing our editorial on the looming problems associated with California’s statewide voter registry, I received from Kathay Feng (California Common Cause) a copy of an emergency regulatory change to California’s HAVA-mandated statewide voter registry. As Secretary of State McPherson’s cover letter explains, county registrars in California can “apply a driver’s license number found through Calvalidator to a voter record for a registrant, without the need to contact the registrant to confirm it, provided there is a single “exact match” for the registrant’s first name, last name and date of birth.”

This will still be a long and slow process for county registrars, though hopefully this will help. It is still possible that county registrars might be overwhelmed by the volume of voter registration checks that might still be required even with this emergency regulation. Only time will tell — we ought to get word from election officials in a couple of weeks as to whether this change reduces their workload and will insure a minimum of problems for newly registered voters in the June primary.

California’s Looming Electoral Nightmare

R. Michael Alvarez and Thad E. Hall

As we enter the 2006 election year, the media continues to focus on the travails associated with electronic voting equipment. But below the surface is a greater threat that has received little attention—the transition to statewide voter registration systems.

As a result, unless steps are taken now, tens of thousands of people may try to vote this spring in states like California, only to find that their names are not on the rolls and they cannot get a ballot.

California is currently the poster child for a poorly-functioning statewide voter registry. Recent reports have noted that at least 25% of new registrations and re-registrations filed since January 2006 have been rejected because the information on the application did not match data in other state databases or because critical information was missing.

Rejected applications are sent to the county election official, who must spend precious time and money contacting these people to determine whether they are really eligible to vote. In Los Angeles County, these problems have been particularly acute, where over 40% of applications this year being rejected.

As we get closer to California’s June primary, the problem will just get worse.

In a normal year, county registrars are flooded with absentee ballot requests just before the election, as well as with new registrations. Now they will be flooded with requests, new registrations, as well as rejected registrations. As a result, tens of thousands of rejected voter registration applications may not be processed and these people may be denied the right to vote.

The irony is that these statewide voter registration systems, mandated under the 2002 federal “Help America Vote Act” (HAVA), are supposed to solve problems with voter registration practices, not to create new ones. Statewide files are intended to produce more accurate voter lists and minimize the chances of voter disenfranchisement on Election Day. States developed these new statewide voter registration systems quickly to beat the January 2006 deadline set by HAVA. California modified an existing system to meet the HAVA requirements. Many states, like Colorado and Wisconsin, relied heavily upon corporate vendors to solve the problem, and several of these vendors have come under fire or have been fired.

We hope that these voter registration systems fare better than we expect. To ensure they work effectively, the problems with voter registration systems needs to be addressed head-on now, and in the long-term.

In the short term, given California’s troubled statewide voter registry, we need to insure that qualified voters are not denied the right to cast a ballot in upcoming elections.

First, we need to build some flexibility into the process before the June primary in California. If someone tries to register before the deadline, but their registration application is not processed because of a matching problem or missing information, that person should be entered into the voter rolls and they should be flagged as “provisional.”

If they request an absentee ballot or vote early, they should be allowed to cast a ballot. But they should also get a notice telling them about the problem with their voter registration application and be allowed to submit corrected information with their ballot. Election officials will have plenty of time after the June primary to figure out if they were eligible to vote.

Second, precinct workers throughout the state must be flexible and permissive in the June primary when it comes to voters whose names do not appear on the rolls in their precincts. Voters not on the rolls will vote provisionally. When they complete the information on the provisional ballot envelope, they will have re-registered to vote and provided the local election officials with the information they need to correct the registration. After the election, full-time election officials can verify voter eligibility.

Third, the Secretary of State should stop relying solely on computers to look for problematic voter registration applications. The State should conduct manual matches on rejected applications and try to fix the problems when they can, as many of the problems likely are the result of obvious data entry problems or inadequacies in the state’s system. It should not be the role of the counties to remedy problems that exist at the state level.

In the long term, we need federal standards and certification of statewide voter registries. Today statewide voter registries are being developed and used without any federal supervision, standards, or testing. This makes little sense. Voting equipment used for casting ballots undergoes state and federal testing that follows a set of voluntary federal voting systems standards. Without standards for statewide voter registration systems—not to mention a complete lack of federal standards, testing and certification—it is no wonder that we have voter registration systems that perform poorly.

The debate over vote by mail is heating up, with a few notable entries into this important discussion.

• The American Prospect features a special report, “The New Ballot Box”, available on its website.
• Mydd and DailyKos have links to the Prospect stories, along with some reader comments.
• A local Oregon blog, BlueOregon, has entered the fray. Many Oregon residents are tireless advocates of vote by mail.
• Finally, an Oregon group has started the Votebymailproject to advocate for this method of voting.

I’ll blog on these various efforts over the next week. I’d like to also tell you that there is high quality academic scholarship on the broader impact of voting by mail on campaigns and on voter decision making, but unfortunately, the main agency that funds such work told me two years ago that voting by mail was a non-issue!

Thad already wrote about his wonderful testimony regarding the vote count/recount project he is heading up for the EAC. The EAC has posted on their website testimony from the others who provided comments regarding this important topic.

Panelists included:

1. Panel 1: Perspectives from states:

   • Washington Secretary of State Sam Reed
   • Rosanna Bencoach, Manager, Policy Division, Virginia State Board of Elections
   • Jill LaVine, Registrar of Voters, Sacramento, CA

2. Panel 2: Research on Vote Counting and Recounting

   • Thad Hall, Assistant Professor, Department of Political Science, University of Utah
   • Doug Chapin, Directer, Electionline.org

Worthy reading.

Thanks to Mike and Thad for inviting me to participate in this blog.

For those of you not familiar with my work, I’ve been conducting research into American elections and public opinion for about 20 years. I’ve worked for two of the premier survey research organizations in the United States–the General Social Survey at the National Opinion Research Center, University of Chicago, where I obtained my BA; and the National Elections Study at the Center for Political Studies, University of Michigan, where I got a PhD.

I teach at Reed College in Portland, OR, where the sun is shining and we can still ski on a 200 inch base at Timberline!

In the past 18 months, I’ve become increasingly interested in the phenomenon of early voting, methods by which voters cast their ballots at places other than the precinct place and at times prior to election day. I’ll write more about early voting in coming days.

My blog, earlyvoting.blogspot.com, has been used to post news stories and news updates. I am going to move that work over to here, as well as use the opportunity given to me by Mike and Thad to post longer commentaries on research and policy development.

The final report for the Buenos Aires e-voting pilot project is now available. There is an Executive Summary available in English, while a more complete report is available in Spanish. I’ve only recently received these materials, and have not yet had a chance to digest them completely yet.

As readers know, I was invited to participate in the evaluation of this project; we are compiling a complete archive of all the materials from the pilot project on the VTP website.

What has been interesting since October 2005 has been the chance to get access to the data generated by this pilot project, and to initiate some research projects using this data. Hopefully in coming weeks there will be some working papers arising from the pilot project data, as there are some exciting findings that are emerging from the detailed analysis of the information collected as part of this pilot project.

Mike, Paul Gronke, and I are at the meeting at the State of Illinois Board of Elections. There is Commissioner here chewing a cigar, so we really are in Chicago! What is below represents the discussion that occurred at the meeting.

The elections commission chair noted that this was the first election in the state with early voting, with disability accessible voting systems in every precinct statewide, and the first in recent memory without punch cards. The equipment problems that occurred were not unique, this happens in every election. There were problems with polling places and problems with the wrong ballot styles. In actuality, there were fewer telephone calls to the state than they expected. Voter participation was down throughout the state. Most election administrators recognize that switching systems does not immediately solve all problems, but causes some new unique problems.

There were some interesting problems. Not surprisingly, there were printer issues scattered across the state. More interestingly, in some southern counties, DRE machines were not closed out in some precincts where the machines were not used because there were no voters with disabilities.

The state elections director reiterated that there were fewer equipment-related problems than they expected. Instead, the problems were routine and similar to the problems that generally occur. The voting system director’s report noted that the new voting equipment resulted in counting errors and some machines broke down and had to be replaced.

The underlying problem with the voting equipment was with the printers. There were some memory card problems and judges not being well trained and knowing what to do. They did not close them out well; they did not close down the machines and bring back the memory cards. When the memory card was problematic, they were able to use the paper trail to address the problem.

They discussed the fact that different voting machines performed differently in the election. For example, the Hart InterCivic machines had a problem with their printers and with some screen problems. There were also support problems from some of the vendors, where the technical support people were not completely knowledgeable and could not fully support county needs.

They then had testimony from three individuals. First, the President of Sequoia talked and he stated that the primary election in Chicago had unique challenges because of the use of multiple voting devices within every precinct. There were 19,000 voting devices introduced into the City and County and the primary problem that arose in the election was a delay in tabulation. All total, 95 percent of ballots and 88 percent of precincts were reported on election night. Sequoia will be providing additional training and making changes to their system between now and the 2006 election to improve user friendliness, especially in the area of the controllers. Chicago and Cook County were using leased machines in this election; they will be purchasing the newest line of machines for the 2006 general elections.

Second, the director of election in Cook County spoke. He noted that Cook County is the first jurisdiction nationally that has combined two distinct voting systems in a precinct and tabulated and reported the results in the precincts. He said that the voters liked the new system, no voters were disenfranchised by the transition, and there were no reports of fraud in the election. Again, the problem was that the results were reported later than expected. There are no problems in the election identified in the 5 percent recount or in the discovery recounts. The county will be hiring independent experts to examine the equipment and will also be placing a voting equipment manager in every precinct. In November, they will not share equipment across precincts within the same polling place and will have an expedited process in place to get results if transmission from precinct electronically is problematic. Shockingly, the county also wants more money from the state to improve the election process!

The City thanked the state for their work in certification and testing the equipment, which led to improvements in the systems. In 1975, people didn’t worry about elections but now elections are more complex. There is a huge bureaucracy and expanded elections. For example, early voting went from 1 site to 21 sites and 14,000 people voted early. They did have problems on Election Day and this was one of the most difficult elections they have had. They did get the votes counted accurately, but the transition to the new equipment was difficult.
Questions:

Were the problems in the election concentrated in certain parts of Chicago, especially the inner city?

The tabulation system works this way: poll workers take the memory pack from optical scan and DRE and combine them together. There were 900 precincts transmitted without a problem. Most of the problems occurred from two instances of sharing equipment in a consolidated precinct. First, there was the problem where there was shared touch screen equipment used in polling places with consolidated precincts. This created problems for poll workers who were unsure how to accumulate votes from one DRE into their own precincts. Second, there were problems associated with two precincts sharing one vote accumulator and transmission equipment system; here the problem was the one precinct had to wait (often up to one hour) for the other precinct to finish their work. Printing the tapes was very slow and it gave a large amount of data that was not needed for tabulation but for auditing purposes. The city did conduct hands-on training but the process for accumulating votes was problematic. There were jamming problems with the some equipment but they were able to feed all the ballots through. Simplifying the task of transmission will make things easier. They will not share equipment in November to address this problem.

The human error side of this is important and is exacerbated by people wanting data immediately on the election results. One state commissioner noted the following (I am paraphrasing here mostly):

election equipment problems occur whenever there are system transitions. The voting equipment is getting more sophisticated and the poll workers are getting less sophisticated. The manual for some of this equipment requires being an electrical engineer. You need a person in every precinct that understands the details and the equipment who are more permanent and know what to do and how to manage things. You have to get a hire quality of election judge and to pay them more. Do you need to vote from 6 am to 7 pm, especially in early voting? Even if you fix the equipment, you are still going to have human problems. For low pay they have to deal with rude voters and colleagues falling asleep on the job. They are hard working people but they can only do so much.

One final issue that was discussed was associated with the security of early voting ballots. The discussion centered on the fact that early voting centers were all located in municipal localities. However, voted ballots were stored in a banker’s box, sealed with tape and signed, in a secured place in the offices. The concern expressed centered on the chain of custody of the ballots, and whether this system achieves that custody.

This is our 300th essay on Election Updates, another milestone for us!

As I write this 300th essay, I’m happy to announce that after a long and hard negotiation this morning over soda and sandwiches, Paul Gronke and his Earlyvoting blog is going to be consolidated with our Election Updates blog. Paul is going to point his blog readers to Election Updates, and from here on out he’ll be writing about early voting in the U.S. and across the world from Election Updates! Thad, Melissa and I are excited by this consolidation, and we hope that our readers appreciate finding all of this material in one place.

More soon from the visit to Chicago. The Illinois Board of Elections meeting this morning was quite interesting, and of course we’ll have much more to say about MPSA in coming days.