Election Machinery

In some of my previous posts, I have investigated whether parallel testing can provide an adequate defense against fraudulent voting software. I concluded that parallel testing alone is an insufficient guarantee of election legitimacy, but I wanted to find out how many states even used parallel testing to verify their elections. I do not think parallel testing alone is enough, but it is surely at least a step in the right direction. So how widely is parallel testing actually used?

To find out, I first compiled a list of the 19 swing states that have been decided by 5% or less in the 2000, 2004, or 2008 elections. I then used voting machinery data from verifiedvoting.org to narrow the list to the 11 swing states that use electronic voting machines. Reliable data on actual election practices is hard to come by, so I called the board of elections in each of these states to ask about their parallel testing procedures. The results surprised me.

Of the six states that I was able to contact, five had never done any parallel testing of any sort. None had ever used parallel testing in a general election. The other states’ representatives either did not return multiple messages seeking information or refused to speak with me about parallel testing. Of the five states about which I was not able to obtain any official statements, my research indicates that only one of them has done parallel testing in the past, but this figure is still very preliminary.

For those who contend that we might catch fraudulent software with parallel testing of sufficient rigor, these numbers are sobering. In the last election, the vast majority of the states for which the result was in question did not use parallel testing at all.  In short, there is a long way to go before parallel testing can even be considered to be implemented, much less be effective.

Fortunately, many of the states that do not use parallel testing have implemented a voter-verified paper trail (VVPT) on their electronic voting machines. For example, the Voting Systems Director of North Carolina told me that no counties in his state use parallel testing. Instead, he described his state’s method of using a “real time audit trail,” a VVPT, along with a random manual recount of a statistically significant number of machines. Of the 11 swing states I examined, five (Florida, Georgia, Indiana, Pennsylvania, and Tennessee) do not use a VVPT in conjunction with their electronic voting machines (Tennessee plans to move to a precinct optical scan system by 2010).

It seems that many states are beginning to do something about voting security: according to verifiedvoting.org, legislation that requires an auditable paper trail has passed in 31 states.  These measures are positive steps — but parallel testing is nearly nonexistent in practice, and only some states use VVPT’s.  We still have a long way to go.

Earlier in the semester I read an article written by Michael Shamos concerning voting technology as an “engineering problem.” In his article, he cited a study done by Paul Herrnson of the University of Maryland titled “The Importance of Usability Testing of Voting Systems.” After expert reviews, numerous laboratory tests, and a large-scale field study of six different voting systems, Herrnson and his colleagues concluded that “changes could be made in all aspects of the process of voting,” and that additional research should be conducted. Throughout our discussions, we have examined security issues with election machinery as well as interface issues that voters may face while using a particular voting machine. Herrnson’s particular study highlights the “usability” of election machinery.

In Herrnson’s study six different voting systems were used in order to represent diversity of design and interface. Among his systems were the ES&S Model 100 (which uses a paper ballot and optical scanner), the Avante Vote Trakker, Diebold AccuVote-TS, and Zoomable prototype (touch screen machines), the Hart Intercivc eSlate (which uses a dial and buttons to move through the ballot), and the Nedap LibertyVote (which shows the entire ballot at once). Four different methods were used by Herrnson and his team which included a large scale experiment with over 1,540 participants, a laboratory experiment, reviews by computer-human interaction specialists, and “natural” experiments which were performed in Michigan and Florida.

The first thing that was studied was actually beginning the vote. Surprisingly some voters are simply not sure what to do when arriving at a machine. Some systems, like the Nedap and Zoomable systems, made it the responsibility of election officials to start the system for the voters. However, in other machines, like the Hart system, a 4-digit number was required. Although Herrnson cites this as simple, if it is a voter’s first time it can be slow and slightly confusing. In the Diebold machine, a card needed to be inserted. However, again Herrnson felt the process could be improved upon seeing that the card slot was not perfectly “reliable.” He cited as an example how the card slots of ATM machines worked better than those of the Diebold. To actually navigate the ballot, Herrnson described the interface of all the machines. On the Zoomable machine the full ballot can be seen. On the Diebold system, another touch screen model, a voter makes his choice for one particular office and then proceeds to the next screen to vote for the next office. A majority of voters felt that navigating the ballots of these two machines were easy enough even though some described the Zoomable’s layout as “overwhelming.” In comparison to the Diebold and Zoomable, the Avante touch screen received lower marks because voters were given less control in moving around the ballot.

In terms of correcting or changing the vote most people gave the ES&S system a poor rating. If a voter would like to change their choice it is suggested that they receive a new ballot. However, many people simply erase their choice and refill in another bubble. Many people lose confidence in whether or not their vote will be counted correctly after such erasures. Of the touch screen models, the Avante system was described as the “most challenging.” The voters must wait until the review stage of the voting process to make any changes.

Before casting a ballot, voters are supposed to have the opportunity to review their choices. Most voters found reviewing their choices on the touch screens to be a much simpler process in comparison to the optical scan ballot. This may be due to the fact that it is impossible to overvote on touch screens. It is also impossible to overvote on the Nedap system, but errors are more likely here then in the previous systems. If an undervote occurs a small text box appears to notify the voter. However, some people completely ignored the box or didn’t even notice it due to its very small size.

Both the optical scan ES&S system and the Avante Vote Trakker have voter verifiable paper trails. Surprisingly, however, having a paper trail did not add to the confidence of the voter. More people were confident that their vote was correctly counted on the Diebold and Zoomable machines then the optical scan and Avante machines.

Herrnson’s study also took into account voter accuracy, the need for assistance while voting, and the impact of the actual design of the ballot. As a part of some of his conclusions, he took into account the “digital divide” between younger and older citizens. Those who were more familiar with computers and technology (i.e. younger people) had fewer concerns with the machines, had greater satisfaction, and voted with fewer errors in comparison to older citizens or those who did not speak English as their primary language. Ultimately, Herrnson called for continued research on the issues of user interface with election machinery. Instead of focusing strictly on issues of security, research that includes how voters interact with the machines can be just as beneficial.

Upon reading Andrew Gumbel’s report on the Election Center’s inability to properly oversee the certification of election technology (Steal This Vote, pp. 255-7), I decided to visit the Election Center’s website myself (which, I might add, is filled with broken links and other features of poor web design) to try to understand their perspective on DREs and voter-verified paper audit trails. Sidetracked by the “news” column on the front page, I discovered Doug Lewis’ testimony against the  Bi-Partisan Electronic Voting Reform Act (BEVRA) of 2008. As the reading is quite dry, I included here a summary of his major points, followed by the major arguments of the opposing side, articulated by Senators Dianne Feinstein and Bob Bennett. While I don’t agree with the majority of Lewis’ specific contentions, I will argue at the conclusion of this post that BEVRA introduces more problems than solutions to the verification issue.

The Lewis Argument

After a few pages spent lauding the Senators in their effort to keep America’s elections fair by working across party lines in backing BEVRA, Doug Lewis articulates the following position: “We approach all bills with the basic guiding principle that elections in America are best handled, best managed, and best serve voters when they are left in the hands of state and local governments. The continuing push to have Federal election laws or Federal administration overtake much of the process is, in our view, the wrong approach.” Without explaining this belief, he continues with an assertion that auditing the system under which federal elections take place will only create an additional burden for local and state governments, especially (1) when people don’t want to pay for audits when the election is not close and (2) when they already have confidence in the system. Rather than voter-verified audits, he proposes that recounts be automatic if two candidates for office each receive similar percentages of the vote; in this system, votes will be counted again by machines, and only a substantial number by hand if the losing candidate believes there is “not sufficient evidence” that the recount was accurate. He then goes on to speak against other aspects of the bill, including a lack of funding for private enterprises that develop election technology.

The Feinstein/Bennett Press Release

The press release characterizes BEVRA as “an agreement on bipartisan legislation that would require electronic voting systems used in federal elections to produce an independent voter-verified record of each ballot cast.” The Senators claim that the bill will provide the greatly needed independent records to supplement voting technology, and also stricter security standards on new devices. Ballot designs, chain of custody practices, and poll worker training must also meet stricter standards on a state-by-state basis. Auditable records (”independent paper, electronic, audio, video, or pictorial record”) will be used with DREs, which can also be useful in the event of a recount. The bill also ensures that each state hold public auditing to make sure the totals are plausible and statistically valid. The press release also asserts that “the bill also includes a research grant program to help spur development and testing of new independent verification technologies [and] provisions to ensure military and overseas voters have their ballots counted.”

Why the bill is problematic

It seems counterintuitive at first to reject the bill based on what Feinstein and Bennett claim in their press release. After all, computer scientists like David Dill and Rebecca Mercuri want voter-verifiable audit trail, more testing on individual voting technology, and auditable/recountable independent records. And the opposing view expressed by Lewis is replete with faulty logic and misconceptions about the utility of audits (for all elections, not just close ones) and their purpose to restore confidence in the system in the face of increasingly risky technology.

But the text of the bill is, in fact, misleading in what it permits as an “independent” vote record, which could just as easily exist in an electronic form–more costly to implement and certainly less reliable and useful than a paper trail. But what is most disconcerting about the bill is what it leaves vague: whether the “independent record” has any legal status and whether the “independent record” will be enforced as the record to be used in the event of a recount or an audit (rather than just a database reprinting the records that it received from DREs in a given number of precincts). While the bill does stipulate that companies disclose machine source code to elections officials and the Election Assistance Commission with “adequate” time for them to review it prior to election (see pp. 12-13 of the PDF linked above), it leaves something called “commercial” information about their codes able to be kept confidential. Ideally, if an election is to be a public and reliable system, secrets about voting technology (and an ambiguous definition of the term “independent audit”) should not be a part of any bipartisan election voting reform legislation. While Doug Lewis was wrong in the aspects of the bill he chose to contest, he was absolutely right in his opposition of the bill and his praise of the two senators for at least trying to introduce stricter standards on machines and audit trails.

Alex P. Keaton, the young neo-conservative Republican, from the TV sitcom Family Ties has been disenfranchised based on prior felony conviction. Alex P. Keaton famously obtained a fake ID. Unlike other teenage high schoolers, Keaton obtained the fake ID to vote for Reagan in the upcoming Presidential election. Keaton must not have realized that using a fake ID is a felony offense in many states. Moreover, many states disenfranchise felons. Lucky for Keaton, Ohio (the setting of Family Ties), does not disenfranchise felons unless they are serving jail-time. Moreover, the use of a fake ID is not a felony there. But laws regarding voter disenfranchisement vary by state and their affect is appalling.

It is difficult to evaluate the reasoning behind felon disenfranchisement. Virginia used felon disenfranchisement as a way to prevent black men and women from voting. US Senator from Virginia famously said, “Discrimination, why that is exactly what we propose!” regarding the new policy on felon disenfranchisement. However, laws in Kentucky and Florida (the harshest prosecutors of felon disenfranchisement) can be traced 1792 and 1776 respectively. Roger Clegg, President of the Center for Equal Opportunity, summarized the prevailing wisdom on voter disenfranchisement from the 18th century until today; “If you aren’t going to follow the law, you can’t claim the right to make the law for others.” The term felony is used to describe severe offenses that merit to revocation of the right to vote. The problem is the gradual progression in many states to make less serious crimes felonies. The use of a fake ID is a felony in Florida, California, Virginia, New Jersey and New York, among other states.
The ACLU is pushing states to reform their policies on felony disenfranchisement. The ACLU is several lawsuits, accuse many state legislators of revoking the right to vote for minor offenses. Moreover, the process of disenfranchising voters is confusing and troublesome for the state. Many innocent citizens are disenfranchised because their names appear on felony blacklists. The number of innocent victims disenfranchised by administrative error is not well documented though. Currently, 48 states prevent inmates from voting (with the exception of Maine and Vermont). 35 states prevent felons from voting on parole and 30 prevent felons from voting on probation.

The facts mask the true effects of felony disenfranchisement. 13% of the black male population has been disenfranchised as a result of state laws. Moreover, if trends continue, 40% of black men will be disenfranchised at one time during their lives. State laws regarding felonies disenfranchise 5.3 million Americans. Florida, a swing state in the past three elections, has disenfranchised 960,000 voters who predominantly vote for Democratic Party.

Though Time magazine disagrees, asserting that felony disenfranchisement did not exist in the 18th century, the Department of Justice report on felony disenfranchisement asserts that the issue was a hallmark of many state constitutions. While I believe it is important to punish felons, I believe disenfranchising laws have become too severe. Although, I believe some people lose their right to vote after violating certain laws, using the term felony is too broad. A felony could be a murder, rape, kidnapping or on the opposite spectrum, the use of a fake ID or backdating stock options. So should Alex P. Keaton, Henry Samueli and Martha Stewart lose their vote because of minor felonies? No. Felony prosecutions have become too harsh, equating murders and backdating stock options. There must be different standards and reform in the system of felony disenfranchisement.

http://www.usdoj.gov/crt/restorevote/restorevote.htm
http://www.heywhynot.com/articles-felon-disenfranchisement.htm#nytimes
http://www.campusprogress.org/page/community/post/jr/C3V7
http://www.abc.state.va.us/facts/fakeid.html
http://www.sentencingproject.org/Admin/Documents/publications/fd_bs_fdlawsinus.pdf
http://www.time.com/time/nation/article/0,8599,1553510,00.html

Whether it was Chicago’s Mayor Richard J. Daley or the city’s notorious gangster Al Capone who said it first, “Vote Early and Vote Often” pretty well describes the technique by which Chicago’s Democratic machine maintained its political grip on the city for years.  In the late 19th Century, and well into the 20th, political machines used ballot manipulation devices to keep control of major cities and even entire states.  Louisiana and Georgia were at one time controlled, respectively, by the Long and the Talmadge machines.  There is disturbing evidence that voter fraud may be making a comeback in the 21st Century.

One of the most flagrant examples is to be found in the 2008 Presidential Election in the state of Ohio.  For a one-week period in early October, “Ohioans were permitted to register and vote on the same day.”  Under Ohio’s new early-voting law, first-time voters were casting ballots without meeting the state’s 30-day registration requirements.  Such requirements were made to allow time for verification of voter information such as age, address, citizenship and whether or not the voter is a felon and has lost their right to participate in an election.  No longer necessary in
Ohio! One morning you could register and vote in city “A”, then hop in your car and immediately register/vote in an adjacent city “B”, and then repeat until the polls closed. “Vote Early, Vote Often”, no identification required.

Such “one-stop shopping” — er — “voting” shatters every precaution that legislators have spent decades building into the voting process to ensure its integrity.  Early voting and the altering of the traditional election cycle have unfortunately increased the potential for fraud, especially in Ohio.  What security of voting integrity is possible when look-alike secret ballots are accepted, are non-traceable to a given voter, and the new voters have displayed no evidence of being US citizens, of being of minimum age or even documented residents of the state where they are casting ballots?  There is none.  So, vote early and vote often.  Integrity is not a requirement and no one is going to catch you.

Early voting was authorized in 34 states during the 2008 election cycle and perhaps 40 million voters took advantage of it in this election; it was an opportunity that possibly enabled significant numbers of new voters to participate.  However, the potential for wrongful voter participation has increased because of the shorter time periods to check voter credentials; creating a logistical nightmare.

Whether in the 19th, 20th or 21st Century, the rule of law and voting integrity is the foundation of our democratic process. We should not discard the hard “lessons learned” of our previous generations.  If we don’t want to go back to the days of Mayor Daley, Al Capone and machine politics, we need to keep the voting process honest.

With the increased use of technology in elections, some election officials see internet voting as the future of electronic elections.  In A report on the feasibility of Internet Voting 2000  from Secretary of State Bill Jones, he describes two phases of internet voting. Phase one is a recommendation to provide internet voting under the conditions of a traditional polling place. This transition phase is a chance to test the security of internet voting.  Phase two is a recommendation that allows voters to cast remote internet ballots from any place at any time. Phase two includes will include some form of voter identity verification to take the place of similar procedures at a traditional polling place. Jones admits that for now, “Internet Voting should be viewed only as a supplement to, not a replacement of, traditional paper-based voting.” However, with the implementation of his phases he eventually hopes to achieve his final phase of internet voting.

The question is why do we need internet voting?  Several people including Jones  believe that “The implementation of Internet voting would allow increased access to the voting process for millions of potential voters who do not regularly participate in our elections [and give them an] opportunity to cast their ballots at any time from any place via the Internet.”  For now the main population of voters that internet voting is meant to target are US citizens overseas.  According to Michelle Spitzer in her Florida Today, U.S. Sen. Bill Nelson “plans to present a bill […] that would allow federal funding to states and counties that provide Internet voting to U.S. citizens who are overseas.” Representative for Nelson, Dan McLaughlin sees technology as a means to increase voter participation.  Increasing participation could potentially extend not just to voters overseas but also to Americans in the US with limited mobility. 

The option of internet voting is seen as a convenience. Pat Hollarn, the supervisor of elections in Okaloosa County, Florida, was among the first to implement internet voting. According to Spitzer’s article , “ninety-three registered voters stationed near three U.S. bases overseas cast ballots [on] Nov. 4 at kiosk stations in Germany, Japan and the United Kingdom.” In response to this recent use of internet voting, Brevard County Supervisor of Elections Bert Childress praised the convenience of internet voting for both overseas voters and election officials.  However, I think there should be a difference between convenience and accessibility.  Yes, there should be measure to ensure that overseas voters can exercise their right to vote. Whether or not internet voting is more accessible or even more convenient is debatable.

One argument in favor of internet voting is the inconvenience and lack of decent security for absentee ballots.  Writer for Popular Mechanics Erik Sofge reports that “roughly 6 million overseas residents must register earlier than other voters” and voters worry that their absentee ballots could be mishandled. The security of internet voting used in the November 4^th election included “PCs with no hard drive, and various other components [were] either turned off or removed in advance to reduce security risks.”  Voters have no way of knowing that their absentee ballots were not subjected to fraud, but with internet voting, they can at least see a semblance of security.  Additionally, voters print out a record of their vote, and turn it in to kiosk workers. At some point however, those print outs must be returned to election officials which still leaves them open to the same fraud that mailing in an absentee ballot would. Compared to absentee ballots, internet voting seems like a more complicated way to vote. In addition to issues of mailing ballots overseas, the entire realm of the internet is open as a source of additional fraud.  Rather than using technology to make a problem more complicated, I think election officials should focus on making absentee ballots more secure and accessible for overseas voters. 

The secret ballot is the basis of our election system in the United States. Without the secret ballot, we could surrender our democracy to a doomsday of coercion, collusion, and corruption. Or our democracy could be safe without the secret ballot. While it is true that in the past the enforcement of the secret ballot safeguarded voters against coercion and persecution (by unions, employers, corrupt political parties, etc.), this does not mean that the secret ballot is necessary to the sanctity of elections today. HOWEVER, the secret ballot is a vital symbol of human rights and free speech in American society. But, if the secret ballot were all of sudden removed from the legal code, a corruption apocalypse would not ensue. Proof of this rests mainly on the fact that there are many relatively easy ways of defeating the “secret ballot” in the United States.

The numerous state legal codes requiring a secret ballot make designing voting machines a (sometimes very complicated) puzzle for companies. Scantegrity II, along with many other voting systems, becomes very complicated because it strictly requires that the ballot be secret (as do our laws). Maybe if the laws behind the secret ballot were somehow made more lax, voting machine companies could come up with a 100% accurate system for guaranteeing election results based on how voters voted. In a way, secret ballots almost “gave birth” to the unreliable and easily hacked “black-box” DRE machines. I am NOT advocating for a removal of the secret ballot, I am only arguing that today the secret ballot sometimes creates more obstacles than protection for voters.

There are many ways around the secret ballot for voters who want to give proof of their vote. I will talk about one main one this week, “pattern voting” and then conclude next week with more ways of defeating the secret ballot (note that most of these require a person “on the inside” working for the election staff).

1. Pattern Voting (PV)
Pattern voting is when voters vote in a specific sequence to distinguish their ballot from others’. Voters can do this with a DRE machines or paper (optical scan) ballots. Say a corrupt political party offers people money for their votes. The corrupt political party has a mole in the elections office who has access to the ballots. Actually, in Ohio the VVPAT printouts from the DRE voting machines are time stamped and can be viewed by the public, along with a numbered list of the voters for each precinct so a mole wouldn’t even be necessary there. The voters and corrupt party officials agree on a predetermined sequence of votes (maybe for the “less important” races) that distinguishes each voter.
Then using the mole, the party finds the ballots with these sequences and checks to see if they voted accordingly. Depending on whether or not they did, the voters would be compensated or punished. One can argue that there is a chance some uninvolved voter could have the same sequence of votes. This is unlikely considering how big most ballots are. Maryland’s 2006 ballot had a total of 197,271,752,498,675,712,000 vote sequences, so it is feasible to design a system where a large number of voters were essentially given an “ID number” based on their vote sequences.
It is interesting to note that DREs without paper audit trails are actually somewhat safer than ones with VVPAT when it comes to protection against pattern voting.

Next week’s blog post will continue proving the claim that the secret ballot can be easily defeated.

“Scantegrity is a security enhancement for any optical scan voting system. It lets voters verify that their ballots were correctly recorded and counted, but doesn’t change how voters mark their ballots” (IEEE security & Privacy, May 2008). Scantegrity was originally created by David Chaum, who also invented the concept of digital money. Scantegrity has many potential benefits. The system is compatible with all optical scan machines currently in use. It also uses a familiar voting procedure- the ballot is structurally similar to those used today. Most importantly, it provides “end to end verifiability of election integrity.” Voters are able to check to make sure their ballot was correctly included in the final tally, and also that the final tally was counted correctly.

The crux of the Scantegrity system is the ballot. When voting, instead of filling in an oval with pencil, a voter uses a special pen. This pen contains a chemical that reveals invisible ink. When filled in, a confirmation code is revealed. The voter then records this code, and is able to ensure his vote was correctly counted online. The chemical also darkens the circle, allowing it to be read by an optical scan machine. (See ballot picture below).

Click to Enlarge

The Scantegrity system is able to provide voter verifiability without revealing who the voter cast his ballot for through the use of the cryptographic commit protocol (hereafter abbreviated CCP). A CCP is a cryptographic device that allows a user to encrypt information so that it is unreadable to others, and then later prove that the encrypted information hasn’t been changed. CCPs are widely trusted and used daily in internet commerce. They are based on one way functions, or functions that are difficult (if not impossible) to find inverses of. The functions must also have a one to one correspondence between x and y. In other words, given any x you can find one and only one y, but it is nearly impossible to find the corresponding x if only given the y.

A simple use of the CCP would be to play rocks, paper, scissors over the phone. Suppose Bill and Alex want to play rocks, paper, scissors over the phone. Without a device to encode the information, revealed their move last would always win. A CCP would allow each player to encode his result and then share the hash, or encoded statement, with the other player. After both hashes had been received, the players would then reveal their original move and re-encode it using a CCP. If the hash of the stated move matches the original hash, the player has not lied. Neither player would be able to change their choice after giving their hash, as a different choice would result in a changed hash.

Scantegrity uses a CCP to encrypt both the confirmation code revealed when a voter fills in a ballot and the reference table used to match revealed codes to votes for a candidate. The CCP, when combined with an auditing process, allows voters to be sure their vote was counted correctly.

E-voting machines are a uniquely interesting concept; it’s a polarizing concept among those “in the know”. On one hand, you have people proclaiming that it is the future or voting, and on the other, you have people denouncing e-voting as an awful, opaque process with no accountability. But, I think most people are somewhere at a happy medium somewhere in between those two extremes. Unfortunately, I believe that this mass of people should develop a polarized opinion. Sounds weird, right? Well I have my reasons. Working at the polls during the past Presidential election was an enlightening experience and I learned a couple of things about e-voting that I had not know before:

1. Superficially it appears that e-Voting is very efficient and time saving. After all, e-voting is high tech, and technology makes everything better, right? Wrong. Unfortunately, e-voting is not always efficient and time saving. When lines got really long at the precinct that I was working at, I found that more it was because people took a long time to read the instructions on how to use the eSlate machines. And, to make matters worse, we only had three eSlates for people to use. So, a lot of people just started asking for paper ballots. The genius of the paper ballot is that it is very cheaply scalable. In fact, it’s  practically free. All one need’s to do is grab a nearby table and put up cardboard separators and instantly many more people can start voting, and alleviate long lines. E-voting machines on the other hand, cost thousands of dollars. So, in order to scale up, the government has to pay out thousands of dollars to accomodate more voters and/or create a faster flow for voters waiting in lines at precincts.

2.  Current technology for e-machines is awful. Basic security is terrible and some of them are really easy to hack. Why should we use technology when the technology is clearly not very good yet? Also, the user interface/usability is not always very good, and sometimes downright amusing. In San Mateo County, we use the eSlates, which use these weird dials to select items on the ballot. In fact, one of the voters who came in to vote in my precinct asked me, “Are we using the toys again?” I kindly asked the gentleman what he was talking about and he said, “You know, the Star Trek device with all the colored buttons and dials.” I looked over at one of the machines, and unfortunately he was right. The eSlates reminded me of a giant toy for toddlers. And, the two us wouldn’t be the only ones who agree that these machines aren’t designed very well.

Now, you may be asking, why would I want more people to be polarized on this issue? Well, if a lot of people decided they really hated e-Voting, then we would be able to go back to the good old days of using paper voting. Paper voting provides a clear audit trail, is cheap, paper voting booths are easily scalable, and they’re easy to use so lines would move quicker. And, if everybody decided they loved e-voting, then maybe the market for e-voting machines would grow and lead to major improvements. The technology/security behind the e-voting machines would improve and usability studies, while doing R&D, can be used to determine what works best for voters. All in all, my belief is that if people develop a stronger opinion on the issue, good things will happen

I used to have a moderate stance

Picture this: A single gray island stands alone, surrounded by a sea of red and blue.

No, I’m not talking about an Impressionist painting, or a peculiar travel destination.  I’m talking about the electoral map of the 2008 election - poetically perhaps, but a picture is worth a thousand words, and the electoral map is no exception - a convenient visual summary of how are nation voted in the election.  This time around, blue represents success and red disappointment.  And, in this Historical election (how many times have you heard that by now)  a single blue dot in a red state represents a little bit of history for the electoral college system.

Obama’s victory in Nebraska’s Second Congressional District not only earned him the first electoral vote a Democrat has gotten from the state in 44 years, but will also be remembered as the first time that Nebraska has ever split their electoral vote. Both Nebraska and Maine have enacted laws that allow for the electoral vote to be split between candidates. The way Nebraska’s law works, courtesy of MSNBC:

If a candidate wins the most votes in one of Nebraska’s three congressional districts, then he would get one of the state’s five electoral votes, even if he gets fewer votes statewide than his opponent.

This is the first time that the law has come into play since it was enacted in 1991. Though Ross Perot came close to winning an electoral vote in Maine (where the law is similar to Nebraska’s), the state has yet to perform a split, despite the fact that the law has been around since 1969.

This explains the blue dot in the red state - but what about that gray island? Perhaps its because election day turned out to be such a tidy affair, but the media has not seemed to express any amazement or interest in the fact that Missouri has yet to declare an official winner. Politico, in a recent article, seems to be the only one asking the question “What’s the matter with Missouri?” The article doesn’t take long to answer its questions - Missouri’s vote was very close this year, so its come down to provisional ballots - which take a while to count. However, by the time you read this blog post, the deadline will have passed; the counties were required to submit their results to the states by today.

So, very soon, that gray island will take on a color (almost certain to be red), and the only anomaly on the map will be that one blue dot in Nebraska. Missouri, its been fun watching you try to avoid that coat of red paint as long as you possible could. Was it because you didn’t want your status as a bellwether state to be compromised? It’s okay - we all make mistakes sometimes, and you’ve certainly kept the map interesting since. Don’t beat yourself up about it, better luck next time, and do try to get all those provisional ballots counted.

[edit: title 11/19/08]