On August 12, while I was vacationing, I came across a story about a judge in Maryland ruling that early voting was not constitutional in Maryland. (A better version of the story can be found in the Baltimore Sun). As the story notes:

An Anne Arundel County circuit judge ruled yesterday that the General Assembly’s plan to allow voters to cast ballots in the week before Election Day is unconstitutional, a victory for the governor, who has made opposition to early voting a central issue in the fall campaign. Judge Ronald A. Silkworth ruled that the General Assembly’s plan violated the clear language of the Maryland Constitution, which says that citizens may vote in their local election districts on a specific day, the Tuesday after the first Monday in November.

Here is the actual text from the Maryland Constitution that is referred to, which is in Article 17, Section 2:

Except for a special election that may be authorized to fill a vacancy in a County Council under Article XI-A, Section 3 of the Constitution, elections by qualified voters for State and county officers shall be held on the Tuesday next after the first Monday of November, in the year nineteen hundred and twenty-six, and on the same day in every fourth year thereafter.

However, Article 1, Second 3 of the Maryland Constitution reads:

The General Assembly of Maryland shall have power to provide by suitable enactment for voting by qualified voters of the State of Maryland who are absent at the time of any election in which they are entitled to vote and for voting by other qualified voters who are unable to vote personally and for the manner in which and the time and place at which such absent voters may vote, and for the canvass and return of their votes.

Interestingly, Article 17, Section 9 provides what would seem to be a tie-breaker, stating:

Now the bizarreness of Article 17, Section 2 is monumental for obvious reasons, such as we don’t have quadrennial elections, we have biennial ones! Moreover, read literally–or to use the language the judge did, the plain text reading–Article 17, Section 2 should also ban all absentee voting, because those individuals are not voting a ballot on election day any more than an early voter is voting on election day. (And note, Article 17, Section 9 would invalidate the Article 1 allowance of absentee voting, read literally).

This ruling has been appealed. It will be interesting how it turns out.

Another study was released this afternoon, also examining the recent primary in Cuyahoga County. This study was conducted by Election Science Institute, and is titled “DRE Analysis for May 2006 Primary, Cuyahoga County, Ohio.” This report, coming in at a at well over 200 pages, adds more detailed data and analysis to what is one of the most closely studied elections in recent history. Thad and I assisted in components of the analysis of this report, as did VTP colleagues Jonathan Katz and Rod Kiewiet.

Despite the title of the report, it actually examines the complete election administration practice in Cuyahoga’s recent primary, and issues a number of important findings (quoted directly from the executive summary):

After three months of exhaustive research, empirical evidence supports the key definitive finding: the machines’ four sources of vote totals – VVPAT individual ballots, VVPAT summary, election archive, and memory cards – did not agree with one another.

The vast majority of voters surveyed were pleased with their experience with the new system, liked touch screen voting and had confidence that their votes would be recorded correctly.

Improved training, both practical and procedural, is likely to minimize incidents experienced on Election Day.

Incident reports were widespread but concentrated, with 9% of precincts reporting 10 or more incidents. The most commonly reported incidents were voter registration issues (30.1%), election administration issues (22.6%), problems related to voting machines (16.2%) and issues involving booth workers (9.1%).

New strategies for voting machine allocation are needed to minimize voter wait time and distribute it equally across all locations.

VVPAT’s were missing, missing information and the tally of the individual ballots did not always match the VVPAT summary printed at the end of Election Day.

Discrepancies were found across vote counts stored on different mediums across the election system.

The current election system, if left unchanged, contains significant threats to inventory control of mission critical election assets, error-free vote tabulation, and tabulation transparency. One likely result is diminished public confidence in a close election.

As in the previous report that was done regarding this election, there is much worth detailed examination in both studies. Hopefully other election jurisdictions will undertake such detailed audits of their election administration practices in the future.

This report is picking up some attention in the press this evening; for example there is a summary story in the Washington Post as well as other media sources.

I think Mike Alvarez blogged about this article, covered here in the Arizona Republic, that shows where you vote may influence how you vote.

A set of links on the Maryland decision:

MSNBC

CBS News

ABC News

For those of you who wonder why people like Mike and I–and John Fortier and Norm Ornstein at AEI–care so much about disaster management and being prepared for terrorism and the like, check out this link. It has video of the aftermath of a ceiling collapse that ruined the voting equipment in a New Mexico County. As the story states:

A building where Bernalillo County voting machines are kept couldn’t withstand the rain. The building’s roof started to leak last night soaking insulation until it fell through the ceiling tiles. Water poured down onto computers. Absentee ballots from the primary election, which have already been counted, were ruined. The county clerk’s office is now scrambling to find a new facility and test voting machines. “We have some voting machines that got wet,” County Clerk Mary Herrera said. “We’re going to have to test them to make sure that they’re functioning efficiently to use for the election.” The next election–the Albuquerque Public Schools bond election–is scheduled for Sept. 19. The general election for state and local offices is Nov. 7.

I met with a colleague this week, someone interested in beginning new research on voting technologies. He posed a question to me, which I thought I’d post here and see what sort of feedback I received. He wondered what are the five most important questions that election administrators should ask when acquiring a new voting technology. Or, as he posed the question:

With respect to a voting technology, what are the five most important questions that you need to know before acquiring a new voting system?

Of course, this is posed to election administrators, but I’d be interested in getting feedback from any readers. Please send email to alvarez.research@gmail.com.

There’s a good article that came out this week in Technology Review, which provides and approachable discussion of the “scratch-and-vote” method developed by MIT VTP colleagues Ron Rivest and Ben Adida. Here’s a short snip from the article describing the method:

The S&V approach makes this auditing process secure because it allows a ballot paper to be checked without having to involve an election official (who in theory could be corrupt and tamper with a ballot). When applied to the Prêt-à-Voter scheme, S&V adds a scratch surface on the side bearing the candidates’ names, while the order of the candidates’ names is encoded cryptographically beneath the tick boxes. “This scratch surface is exactly like a lottery card,” says Adida.

To check that a ballot paper hasn’t been rigged, the voter simply scratches off the surface to reveal a number that can be combined with a number corresponding to the order of the names and a publicly available encryption key. In theory, voters could use cryptographic software at the poll to perform these operations; but in practice, trusted third-party organizations could provide a means for voters to check their ballot papers. If the codes match, the “audit” ballot is legitimate, and it should be okay to vote with the other ballot.

It’s exciting to see research and development efforts like these, because they do represent the future of voting technology. However, as Michael Shamos points out towards the end of the Technology Review article, the trick will be to get public and political acceptance of these novel approaches to balloting.

Recently a lengthy (398 page) report was issued regarding the spring primary election in Cuyahoga County, Ohio. That election has been closely studied, and as part of the scrutiny of that election, many issues have been discussed about the voting equipment, pollworker training, procedural problems, and voter confusion. Here’s the public report that has been issued (link from the Advancement Project). There is of course a lot here to digest, and above the problems that have been reported and discussed since the report was issued, so I’m not going to go into more detail on those points.

This report, in my mind, really points out the necessity that election officials work to undertake such comprehensive analyses of their election administration practices. The material released so far from the Cuyahoga study shows just how much data can be collected — and how it might be studied — to provide feedback to election officials as they seek to improve their administrative practices. For example, in the appendix of this report there are tables that provide preliminary analyses of incident reports filed by polling place workers and by election technicians. What a gold mine those datasets will be for administrators and researchers in the future!

Secondly, there are many important recommendations that are made in this report, in particular calls for better pre-election testing, procedural reform, and better training of pollworkers and election officials in future elections. No doubt, the comprehensive audit of the spring primary in Cuyahoga will provide much data for future audits in Cuyahoga, as well as baseline data for similar audits in other jurisdictions in the future.

Just a first reaction from the survey:

1) Probably most important–it looks like the EAC is going to distribute this to state election directors, and most of the questions are worded “statewide and by county/local jurisdiction.” I wonder who will be actually doing the legwork (and paying the cost) of completing the survey.

2) The EAC is clearer about what they mean by “early voting” than they were in 2004 (from the instructions: Early voting refers to any voting that occurred prior to November 7, 2006, for which there were no eligibilty requirements. For example, the voter did not have to attest that he/she would be absent from the votÎng jurÎsdÎctÎon on the day of the election.”). This will still result in all ballots from Oregon being reported as “at the polls” even though we don’t have polls in this state. All the ballots are “counted” on election day, though.

3) Question 35 is going to be an administrative headache. I suspect that most election jurisdictions are not set up to record the reasons why provisional ballots were rejected, even the EAC is asking for that information.

4) Same point with Question 39. How many jurisdictions record the actual reason that an absentee ballot was rejected? Is this information reported to the state?

The EAC has released a draft version of their 2006 Election Administration and Voting Survey (previously called the “Election Day Survey”).

It is now open for a 60 day public comment period, and I encourage all of our readers to take some time to look at the survey and send along thoughts to the EAC.

The 2004 version of the study (results reported here) had some flaws–the survey was distributed too close after a hotly contested election, when many officials had neither the time, or in many cases the information, needed to complete the survey. Some of the questions were worded imprecisely, or were intepreted unevenly, resulting in confusing survey results. And most worrisome, there are simply many large gaps in the dataset.

But as a continuing resource to monitor the march of election reform in the United States, the EAC survey is invaluable.

Mike and Thad have already done their part (read the report here). I recently received funding from the Carnegie Corporation of New York to supplement and extend the 2004 study (more on that soon). I hope the policy making and academic community can help the EAC make the survey as effective as possible.