As I have been reading about the election in Wisconsin, I am reminded of a couple of things that Mike Alvarez and I have written — as well as a paper I have now with Charles Stewart — about elections.

First, elections are a process. There are certain steps that have to be followed — by voters, poll workers, election officials. These steps constitute a set of standard operating procedures (SOPs) and SOPs are what keep complex organizations engaged in complex activities from going awry. An aircraft carrier functions without incident because everyone on the ship follows SOPs for their job; together these SOPs integrate and a plane can launch without anyone getting killed.  Elections too have standard operating procedures.

Second, the events in Waukesha illustrate the problem I have with the narrowness of how many people view election audits.  There is a very large focus on audits as events that occur for checking vote counts and the best statistical method for doing the count.  However, the Waukesha problem illustrates that an audit is about ensuring that a PROCESS works, not that a bunch of numbers add up.

For example, audits are about having detailed and appropriate SOPs and training election workers on how to implement these SOPs.  As Atul Gawande would note, you have to have and then follow the checklist.  The checklist for Waukesha County should have included an array of activities, including counting and reporting the ballots.  The checklist should also have included having at least two people doing most of the activities on election night — the two person rule is auditing 101.

(Mike Alvarez, Lonna Atkeson, and I have an edited book on election audits coming out next year and this is one thing we emphasize).

Third, voters are sensitive to problems that occur in the voting process.  When voters don’t think that their votes will be counted correctly, there is some evidence that these events will affect a voter’s likelihood of voting in the next election.   Moreover, part of the issue with voter confidence comes when unexpected events occur.  When election results surprise, it can affect voter confidence.  In in person voting, when problems happen at the polls, voter confidence and voter’s attitudes about election officials plummet.

So the problem in Waukesha County is this — when their election officials:

1. don’t bother to have a set of standard operating procedures and
2. don’t bother to have a checklist, and
3. then do not follow the items that should have been on the checklist, and
4. ignore the most basic rules of auditing (the 2-person rule and having a checklist),
5. they should not be surprised when all hell breaks loose and people criticize them because, well, they deserve it.

…people around the world are as psychotic about knowing the results of the election 5 minutes after the polls close as they are in America. In Estonia last night, the National Electoral Committee’s website froze for over an hour and, of course, that is the talk of the election today. The interesting part was that the Electoral Committee used Twitter and Facebook to get election results out and, in the course of an hour, more than 3,000 people “friended” the Electoral Committee so that they would get the stream of election results. Sometimes, social media is a great backup plan!

Also, one nice thing about the huge rise in early and Internet voting in Estonia — almost half of the voters used those methods — is that vote results were known earlier than ever before because the precinct counts were that much easier, with fewer election day ballots to count at the polls.

Estonia had their elections for parliament Sunday and there were several interesting aspects to the event today.

• E-voting exploded as a percent of votes cast, with almost one-quarter of votes cast online.
• There was an invalid vote cast.  This has never happened before; in general the system is designed for there to be only valid votes — you cannot cast a blank ballot online — so there is currently an investigation into the cause of the invalid ballot.
• A single woman cast 553 online votes.  Only the last vote counted.  According to the presentation the government gave today, she voted 18 times a night sometimes.

So we are having the most bizarre election controversy here in Utah.  To appreciate the bizarreness of the case, I am going to tell you the end first and then bullet out the story.  Rep. Craig Franks — a legislator from Utah County — found out himself that he does not live in his own district and is about to be removed from the legislature.  Now here is the story.

• Franks wants to buy a new house.  Before he bought it, he consulted his County Clerk’s election boundaries and — voila! — the house he wanted was in his district.
• He files to be a candidate for re-election in 2010.  The County signs off on his candidacy forms.  The Lt. Governor (who is the state election official) signs off on his candidacy forms.
• He goes to vote on election day in the precinct in which he is assigned by the County.  He is on the ballot.  He votes for himself.
• He wins.  The County and the State certify his election.

Under the State Constitution you have to live in your district so they are kicking him out of his seat, even though, when he ran, he was told he did live in his district and was certified as living in his district.  It turns out that the county maps were not accurate.  The problem is the issue of language and annexation, as this story notes:

Sen. John Valentine, R-Orem, an attorney who is providing legal advice to Frank, said information he has uncovered from the hearings and report from the 2001 redistricting committee calls into question whether legislators actually intended all of Cedar Hills, including Frank’s residence, to be in House District 57.

“It appears the maps may be wrong,” Valentine said. “The interesting question is: Is Representative Frank’s property in an area that was in Cedar Hills’ boundary at the time the committee adopted the map, because the text says [the line] is the Cedar Hills boundary.”  The maps attached to the act setting the boundaries do not include Frank’s neighborhood.

When the Legislature was drawing the district maps, Cedar Hills was in the midst of annexing the land where Frank later built his home. If that annexation was completed by the time the maps were adopted, then the boundary — it could be argued — would have included Frank’s home, Valentine said. If he lived in the district the whole time, then he should continue to serve in the House. Frank lives in Valentine’s Senate district.

Mike, Lonna Atkeson, and I are writing a book and as part of the research for it I looked at data on residential mobility from the Census. The data are pretty amazing and potentially have very important implications for voter registration and election administration.  Here are the key points:

1. The number of people moving between states has declined dramatically over the past decade, from roughly 7.5 million per year from 1998 to 2005 to 4.8 million per year in 2007 – 2009.
2. Same county moves slowly increased over the 2000s, from 21 million in 2000 to almost 25 million in 2009.
3. Intrastate moves to a different county have also plummeted, from between 7.4 and 8 million moves in 2000-2007 to 6.3 million in 2008-2009.

For election administrators, this means that keeping election rolls clean within a county are likely to be more difficult, as people move about the county and that rules regarding registration portability within a county are critical.  Rules for provisional voting, for instance, are likely to be more important, as are efforts to update registrations through the change of address process.

The question of how provisional voting changed between 2006 and 2010 is one that should be studied after the 2010 election data are all in.  In addition, these changes in mobility suggest that thinking about voter registration modernization means thinking about how to best keep everyone accurately registered — short moves and large — and that modernization is even more important than ever.  Finally, it will be interesting how mobility changes or doesn’t change between 2010 and the 2012 elections.

The DC Elections are pretty low key but the process for same day registration is not overly streamlined. There are two forms plus the data entry that have to be completed. This process takes 15 or 20 minutes per voter. Most voters are willing to wait but some have left. More exciting I just saw a child topple the special ballot box.

Linda McMahon — co-owner of World Wrestling Entertainment — is on the ballot in the State of Connecticut for the United States Senate. I came across this video where her husband criticizes the Secretary of State of Connecticut for allowing local poll workers to determine of wearing WWE merchandise to the polls will be allowed or will be considered campaigning.

I am on the program committee for the CEDEM conference and this is their call for papers.

CALL FOR PAPERS

Conference for E-Democracy and Open Government
http://www.donau-uni.ac.at/cedem
DEADLINE for Submission: 1 December 2010

CeDEM11 presents the following Tracks:

E-PARTICPATION

with co-chairs:  Julia Glidden (21C Consulting, UK) and Jeremy Millard (Danish Technological Institute, DK)

OPEN GOVERNMENT

with co-chairs:  Philipp Müller (University of Salzburg, Business School, AT) and Axel Bruns (Queensland University of Technology, AUS)

E-VOTING

with co-chairs:  Melanie Volkamer (Technical University Darmstadt, GER) and Thad Hall (University of Utah, USA)

OPEN DATA and OPEN ACCESS

with co-chair:  Andy Williamson (Hansard Society, UK)

During the last 10 years, the world has focused on social media and the new forms of societal behaviour, including content generation, collaboration and sharing as well as network organisation. These behaviours and expectations, in particular transparency and access to data, new ways of interacting with government and democratic institutions will continue to develop, and profound changes in society are to be expected. Society has been confronted with Open Government, Open Data and Open Access. What have the experiences been so far? How do these impact society, democratic structures and organisations? What changes occur at citizen level? What are the implications for democracy, society, science and business?

DEADLINE for Submissions: 1 December 2010

We would like to invite individuals from academic, applied and practitioner backgrounds as well as public authorities, NGO/NPOs, education institutions and independent organisations to submit their research and project papers as well as workshop proposals related to the CeDEM11 Tracks.  We welcome different multidisciplinary and interdisciplinary approaches and disciplines including (but not limited to): law & legal studies, social sciences, computer sciences, political sciences, psychology, sociology, applied computer gaming and simulation, democratic theory, media and communication sciences.

Conference Chairs:     Peter Parycek (Danube University Krems, AT), Manuel J. Kripp (E-Voting.CC, AT), and Noella Edelmann (Danube University Krems, AT)

Confrence Website: http://www.donau-uni.ac.at/cedem
Join us on Facebook:  http://www.facebook.com/pages/CeDEM/120470071343471

So Slate has an explainer piece about if you have to spell Lisa Murkowski’s name correctly for it to count.  The thing the story totally misses — and what will prove to be controversial — is not the name part but the Alaska law governing filling in the oval on optical scan ballots.  Alaska law is clear on this point (15.15.360):

(10) In order to vote for a write-in candidate, the voter must write in the candidate’s name in the space provided and fill in the oval opposite the candidate’s name in accordance with (1) of this subsection.  [NOTE:  (1) of the subsection states — (1) A voter may mark a ballot only by filling in, making “X” marks, diagonal, horizontal, or vertical marks, solid marks, stars, circles, asterisks, checks, or plus signs that are clearly spaced in the oval opposite the name of the candidate, proposition, or question that the voter desires to designate.)

(11) A vote for a write-in candidate, other than a write-in vote for governor and lieutenant governor, shall be counted if the oval is filled in for that candidate and if the name, as it appears on the write-in declaration of candidacy, of the candidate or the last name of the candidate is written in the space provided.

In short, this is going to be fun.

So this is an op-ed of sorts for the blog, about the NY election.

Here we are 10 years after the 2000 election debacle and New York has managed to have an array of administrative problems in their primary elections.  Given how much we know about elections today – 10 years after the Florida 2000 elections – how did New York manage to botch things up so well?  The answers are simple but defy the easy analysis that many people generally give.

First – let’s get this argument out of the way – it isn’t the voting machines.  If you went online to read about the problems in New York, you could be forgiven for thinking that the City’s touchscreen voting machines had malfunctioned and stolen all the ballots.  In fact, the City used paper ballots that were electronically counted by scanner, which is a form of electronic voting but that also involves paper ballots.  This and other voting systems have been used across the United States without problem during the 2010 election cycle.

So if we cannot blame the machines, what was the problem?  The answer is easy:  people and processes.  Elections are about the interaction of people – voters and poll workers – who work through a process (the election rules and regulations) to implement a technology (the voted ballot that gets counted).  It is easy to blame the technology for failures to train the poll workers to implement the election effectively or blame the technology because the election board failed to work through how to make the process work effectively in the polls.

Reading about the problems that actually occurred in New York in the primary, they are clearly people and process problems.  When poll sites don’t have the scanners that they need at the opening of the polls, it is a simple planning problem.  Why didn’t the election board have the logistics for delivering scanners and delivering enough scanners to each precinct worked out more effectively?  The answer is clearly a lack of planning.  When poll workers struggle to get the scanners up and running, again, that is a training problem that falls at the feet of the city’s Election Board.

When, as the New York Times reported, “workers seemed flummoxed by procedures that accompanied the new equipment, especially for accepting ballots when the scanners did not function,” the implication is clear:  poll workers were not trained to implement the equipment and did not have any sort of easy to use step-by-step guide that walked them through how to set up the tabulators.

Training poll workers and providing them with effective aides for managing the election is not rocket science; cities and counties across the US have been improving their training and processes over the past decade.  And these training and processes matter.  Several studies have found that, when voters have good experiences at the polls, especially interacting with their poll worker, the voters are more likely to be satisfied with the electoral process and more confident that their vote was counted accurately.  When poll workers do not seem competent or voters encounter a problem voting, confidence plummets, and confidence matters.  Poorly run elections affect democracy directly; voters who are not confident in the electoral process are less likely to say that they plan to turn out to vote in the next election.

Fortunately, New York can fix the problems they encountered but it will require the Election Board to be proactive and rethink how they train poll workers and what materials they provide to the poll workers on Election Day.  Studies have found that poll workers find hands on training more effective – no more dry lectures to large rooms of poll workers! – and they need simple guides to walk them through key activities, like setting up and closing down the equipment.   The poll workers also need for the voting equipment to show up on time and for there to be clear procedures regarding how to handle ballots if problems arise.

It has become almost cliché to blame the voting equipment for problems after an election.  Doing this lets the election officials off the hook for failing to properly train the poll workers and for failing to put in place procedures for handling ballots, setting up the equipment, and handling problems.  The Election Board in New York needs to step up and take responsibility for their failures and work diligently to fix them.