Thanks Ashley!

That is indeed a key next step question.

Part of the answer is to be found by gaining & surfacing community traction through channels like OpenIDEO ;).

Another is to highlight & build on the current successful pilots mentioned in the concept above.

I.e., is Los Angeles (jurisdiction of challenge sponsor: L.A. County Registrar-Recorder/County Clerk ) ready to be the next accessibility trendsetter here, perhaps starting with a local election or referendum (especially since, for LA County, you can already look up real-time election vote counts, verify voter registration, & download sample ballots online)? Maybe LA could pilot through a school district vote, providing additional online voting access at school computer labs afterschool or on weekends. Will tech-savvy Silicon Valley (home of IDEO's Headquarters beat them to the punch)? Or perhaps Washington D.C., our nation's capital (and location of challenge sponsor Information Technology & Innovation Foundation), will step up to the plate?

I'm also hoping the rest of the OpenIDEO community can chip in here on how best to encourage adoption of online voting technology =). Everyone, any ideas?

Nice thoughts and questions, and yes! All of our challenge partners are excited to leverage our community's bright ideas and use them to inform their efforts going forward.

Another note: One thing that I think about when it comes to adoption and technology is network effects (http://en.wikipedia.org/wiki/Network_effect). If we can increase the number of districts using online voting, the technology will become more valuable to everyone....I'm going to have a think about this and will keep you posted :)

Just *bumping* this up, as I think Ashley posed a key next step question here.

I'd love the talented OpenIDEO community's help in coming up with ideas that address "How can we encourage election officials and other people to adopt these online voting systems?", given the important benefits it can bring, accessibility and otherwise.

Thanks Rachel. When I checked out your concept, after getting an email notification that you'd built on this concept, I literally "WOWed" out loud about your creativity =). I was excited to add a link to your new innovative concept from this one.

Keep up the good work too!

P.S. Hmm, seems in my excitement, I actually forgot to applaud your concept. Not to worry, now remedied ;)

Hi Shane,



thanks for raising your concerns about electronic voting machines, particularly from Ireland's experiences, as it show the huge importance of having a voting system that is trusted to be secure & reliable by the general populace.

The watchouts you bring up here are even more important, considering that the stalled voting system switch we are talking about in Ireland is to basic electronic voting (which is already widely used in other countries: for example, in the United States, the majority of accessible voting systems are actually already direct recording electronic voting machines, and almost 30% of the votes cast in 2004 were by direct recording electronic voting machines). Just to be clear, this concept is proposing online/mobile voting, which would be a form of electronic voting.

TAKEAWAY 1: switching to electronic voting machines (already used elsewhere, including in the US) stalled in Ireland, due to the lack of backup verification/audit mechanisms. This points to the huge importance of having a voting system (physical, electronic, online, or otherwise) that is trusted to be secure & reliable.



It seems that a key problem with the electronic voting machines in Ireland, is that they lacked the paper trail for an audit/verification system. Without this audit/verification system, I agree, I wouldn't trust this type of electronic voting either.

In the online/mobile voting proposal above, the audit/verification system is designed so that each vote generally creates at least 4 duplicate records (a printed paper receipt for the voter, a printed vote paper trail for the election organizer, a digital receipt for the voter, and at least 1 digital record of the vote--that can be easily duplicated multiple times). The concept allows for verification by the election organizer (you can count up paper ballots, just like in a regular paper ballot system), verification that their vote was accurately included by each voter, & verification of accurate tallies by independent election observers, while also preserving voter privacy. In other words, the backup paper trail & amount of verification possible for this online/mobile voting concept is actually more than is possible with a traditional paper ballot system.

TAKEAWAY 2: the proposed online/mobile voting concept actually provides more of a backup paper trail (at least 4 duplicate records, 2 physical), and more verification/audit possibilities (by election organization, voter, & independent observers), than a traditional paper ballot system.



You're also right Shane that an electronic system to protect user privacy relies on complex cryptography, which is hard for people to understand. However, I I don't think this obstacle is necessarily insurmountable. After all, we trust these same encryption algorithms in our online banking (both for password verification that ensures only you can access your account, while ensuring even the bank doesn't know your secret password; and for protecting your financial information as it's sent back & forth between you & your financial institution). Another example, most people today don't truly understand how airplanes, automobiles, & trains function, but we all generally trust these inventions to move us around without exploding & killing us.

TAKEAWAY 3: Trust in complex systems is not just based on complete knowledge & understanding. It's also based on other factors like history/experience, legitimacy given by others that you trust, etc. This is certainly a design opportunity to find ways to build trust/legitimacy in online/mobile/electronic voting, such as through education materials, and pilots in local elections & referendums, rather than immediate topdown national directives.



Finally, I agree with you Shane, that any new innovation, particularly one that isn't implemented properly, can cause problems, and even not achieve it's intended benefits, such as increased voting participation. That's why careful implementation and minimizing the risk through pilots is so important. And, when online voting is properly implemented, there's certainly the chance to increase voter turnout. For example, Markham, one of the 1st major Canadian municipalities to try online voting in 2003, increased turnout by 35% after implementation. Afterwards, 80 more Canadian municipalities have implemented online voting. And ~80% of Canadians in a survey said they want the option to vote online & would do so if it was available. In fact, Canada's chief electoral officer is seeking approval to further test online voting in an upcoming byelection.

TAKEAWAY 4: Online voting, when properly implemented, has already been shown to increase voter turnout, and gain trust, legitimacy, & popular support from the voting populace.



NOTE: Shane, I really appreciate your comment, which points to very important watchouts & legitimate concerns. Given Ireland's experience ( €55 million is a huge amount to spend, and unfortunately not achieve the desired goals! ), I certainly wouldn't recommend pushing down a top-down directive for some huge voting system change in Ireland right now. At the same time, other countries may be in a different situation where voting systems like electronic, online, mobile, & otherwise may make sense.

Interesting discussion.
I can see the inevitability of online and/or mobile voting in the USA just as we have largely moved away from (but not eliminated) imprint transactions on credit card (A month ago I did an imprint transaction for the first time in over 10 years).
There are two issues in this discussion I should like to revisit.

1 Separation of establishing entitlement to cast a ballot and the act of casting that ballot - together they form the process of voting.
In physical polling places this should be straightforward regardless of whether the ballot is cast on paper, through a machine or online terminal to which access is only available once entitlement is established. It is this 'obvious to all' separation that preserves the privacy and anonymity of the cast ballot.
With the banking transaction, unlike in the voting system, it is highly desirable that the transaction is correctly identified with the transactor (voter). Thus analogies with the banking system (including the telephone voting system used by Australian visually impaired voters) carry impressions of non-anonymity.Overcoming this perception is a necessary part of the education of potential online/mobile voters.
In mature democracies I suspect most people will not care if the "system" actually knows what choices an individual has made because beyond the collective result the "system" has no interest in following up the choices of individuals. On the other hand in emerging democracies and places where corruption is extensive the "system" may have a vested interest in knowing how an individual has voted and introducing online or mobile voting even in places like India probably posses as many problems as it solves.

One additional word of caution regarding online voting that is banking related is the issue of phishing. - again perhaps more concerning in emerging democracies.

2. The paper audit trail is an interesting issue.
I can see the slight benefit in providing it to allow physical recounts in close elections although malfunctions in printing are probably as likely, in the overall scheme of things, as are malfunctions in the electronic count on which the production of the paper trail is based in the first place. I presume that, as in this country, recounts occur largely for two reasons the margin of the result is sufficiently narrow to warrant reexamining the votes, or there is dispute in determining the intention of voters through the marking of the ballot paper.
Where a paper copy of the vote is produced it is a matter of judgement (or legislative edict) which is the true record and therefore where there are discrepancies between electronic counts and those based on the paper copy, which count is to be accepted as the correct result. Again in Australia in the physical polling places we maintain a count of the ballots made available, the ballots issued, the ballots remaining unissued and the total of votes cast. Usually these balance but occasionally we have to go through rubbish bins to locate ballots thrown away and very rarely the election official in charge will have to file a discreption report. I imagine but don't know if a similar paper record is produced following a telephone or online ballot here probably with similar questions about the necessity for it.

Finally, at least for Australian voting, I remain a fan of the use of paper based ballots for most situations (for one thing it provides gainful if low paid employment for some people). On the other hand I am in favour of using technology where it can assist those who have no access to voting or whose access causes discomfort or unsatisfactory voting experiences.

Thanks Paul for your continued adds! Lots of great points raised, and I'm enjoying the discussions =).

Unsurprisingly, I agree with you about the inevitability of online voting ;), though I'm certainly not going to pretend I have a crystal ball and can predict time lines. I think the best we can do is try to understand where society is moving, and initiate positive impact projects that take this and specific regional contexts into account.

To address the 2 issues you raised to revisit:




1) Agreed about the importance of providing separation between "identification/verification of voting eligibility" and the "actual vote choice made", and really, all the points you made here. In the traditional physical voting process in private booths (assuming no one tries to "hack" the system, such as directly through physical coercion or indirectly, say through hidden cameras), this separation is relatively complete & intuitively obvious. A digital voting system can have the same level of complete separation (as you mentioned can be, and actually is often currently, done with electronic voting in physical locations. For an online/mobile system, if desired, the same level of complete separation can be enforced (literally no stored linkage between your identification & your vote), though you're right that there's a design/education challenge to overcome given perceptions of non-anonymity. In terms of phishing this is currently a concern


  -Note: One thing I'd like to add here, is that there's a partial tradeoff between A) separation of voting eligibility/actual vote choice to ensure anonymity, and B) keeping enough records to maximize verifiability/auditability (that totals are accurate, all votes have been counted, all votes are legitimate, etc.).

To take the extreme case, the most anonymous vote would only keep track of vote totals...period (any other records would be continuously destroyed): even with anonymous paper ballots, there are many ways to try to triangulate someone's vote, from factors like knowing which polling place they went to, to narrowing down stack position in the ballot box to when they voted, to obtaining biometric information like fingerprints/DNA from the physical ballot =P. This would basically provide no verifiability/auditability though, you need to trust there were no weaknesses/corruptions in the system.

On the other extreme end, the most verifiable/auditable vote would be publicly recording your vote & associated identity in full view of others (like how many of our democratic legislative bodies function). Full accountability, no anonymity.

How this tradeoff is balanced is really a decision that each voting populace needs to make for itself. An interesting thing with advanced encryption is that it provides a way to decrease this tradeoff, and thus increase both desired goals. Through advanced one-way encryption, it's possible to reach a relatively high level of anonymity & verifiability/auditability at the same time.


2) In regards to paper trails, physical and digital, agreed that just like the systems that they backup, nothing is perfect, and there's always the possibilities of malfunctions. Yet, these redundant backups together, can create more resilience in the voting system results & legitimacy: e.g. it makes it harder for anyone to completely corrupt/rig the vote, increases confidence in the voting system & final vote tallies particularly in close elections (or in the cases of conflicting recounts, provides evidence for when a revote may be ruled necessary). Of course there are tradeoffs here too, so the exact amount of redundancy is another decision to be made, and likely to change with time, given how experienced/confident people are with a particular voting system.

Hmm...some of my comment seems to be cut off. Don't have time to rewrite. But just to quickly build on your mention of phishing. This is definitely a concern to be addressed, particularly in countries where "offline phishing" already occurs: i.e. fake poll locations, ballot boxes, & ballots setup specifically to rig elections.

This is a long and thoughtful discussion about a very interesting and well developed concept.
What I keep from the discussion are the following points:
1. the importance of understanding the development / implementation of online voting systems as any major new IS implementation and change management project and thus to include in the process communication, training and understanding the context of use.
2. the importance of anonymity in voting vs. the importance of identification in banking which might make the metaphor of banking confusing for some of us. In fact, when I read Shane's comment about this issue, I thought to myself: of course, this is crucial! I want my vote to be anonymous!
3. the conversation on online voting "vs." or "and" physical polls. Per many of the inspirations, the experience of voting as a materially and socially situated experience seems to be important in many contexts and important to keep.
4. the independence and increased accessibility that online voting (at your pace, with your own interface, etc.) might provide for people with disability.
5. when people don't have online access, providing access to this online interface in poll stations but also in other public spaces:
http://www.openideo.com/open/voting/concepting/vote-anywhere/
and http://www.openideo.com/open/voting/concepting/use-your-bank-or-doctors/
is crucial.
6. The importance of having a consistent interface as Whitney suggested it for other interfaces: http://www.openideo.com/open/voting/concepting/election-web-site-templates-consistent-access-everywhere/
is also important

Thanks Anne-Laure. Indeed a long & thoughtful discussion, probably made unnecessarily long due to my stream-of-consciousness rambling =P

Agreed with all 6 of your takeaways, and thanks for linking to some of the other concepts. You're absolutely right that they and this concept together are very complementary in concert to improve voting system accessibility!

Thanks for the follow-up question Shane.

I briefly mentioned the basics of the voter verification mechanism above, but I'll try to clarify & flesh this out a little more for you here.

The most important record for other records to backup & validate against in this case, is the primary digital record. After all, this is the 1st & main vote that's generally totaled to determine the results of an election.

When someone votes, they receive an easy to read/understand receipt (digital & printed) clearly detailing how they voted, along with a special code that they can use to lookup & verify that their vote is included correctly in the primary record (which can actually be publicly posted, for both individual voters, and for independent election observers/media to verify individual votes & totals).

If desired, there's even the possibility of having a live video feed of the election organizer's printed vote paper trail, to ensure that it matches up with your own receipts printed at home. Or, in the case of having a separate identity verification at a physical location (one of the options proposed by Rachel Happen, & incorporated in the concept above under the sub-section "*POTENTIAL FOR SEPARATING ACTS OF VOTING & ELIGIBLE VOTER IDENTITY VERIFICATION"), you can have all paper receipts printed out during physical verification for you to check. Both the virtual & physical options can allow you to 100% match all 4 duplicate records against each other to make sure they match. Personally, I think this is overkill, because the main important thing is to verify the primary voting record is correct as done in the previous paragraph, however this is certainly a possibility if needed.


One important thing to note to keep things in perspective, Voter Verified Audit Trails certainly can increase trust in, legitimacy of, & support for a voting system, and rightly so since you have redundancy checks in place if things go wrong. For individual voters though, some studies have found most of them comfortable/unconcerned enough with the voting system, that they don't bother to verify their votes even when a physical Voter Verified Paper Audit Trail is put right in front of them after they cast their vote (I'm guessing the studies I came across weren't conducted recently in Ireland, where I'd expect different results to surface).

As everything has a balance/tradeoffs, we also need to be wary about over-engineering solutions that might over-complicate ease-of-use, and skyrocket costs.

Hi Myles,

I focused the concept description on illustrating how online/mobile voting might be applied in the United States, given the jurisdiction of the challenge partner (L.A. County Registrar-Recorder/County Clerk) and the geographic focus of the challenge sponsor (The Information Technology and Innovation Foundation).

However, I absolutely agree that voting on mobile phones could be very useful in developing countries given widespread mobile use, as well as voting concerns on accessibility, privacy, safety, etc. As can be seen from some of my concepts/inspirations on previous challenges, I'm a big fan of smart & basic feature mobile phone applications for both the developing & developed world ;)

In terms of avoiding double voting from one phone, one way would be to only allow one vote from a particular phone # (which can be screened). Perhaps a better way (to allow people to correct a voting error, or if they change their mind, say due to breaking news of a corruption scandal), is to allow for each person/phone # to change their vote up to a certain time, and only have the last vote preference recorded.

As for implementing so that people are likely voting independently from their phones and not being coerced to do otherwise, below's one potential way to tackle this. People pre-register to vote with their phone #'s and are given an associated secret voting pin #/password & confirmation secret phrase (similar to how dial-in televoting is already working in some countries). This ensures that each phone # that votes is only associated with one registered voter (so someone couldn't just buy/obtain a bunch of mobiles/SIM cards to vote multiple times). This double confirmation of identity through an associated phone # AND pin (stored through one-way encryption on the database side), also helps guard against hacking/phreaking attempts.

People vote by texting their pin #, then vote choice, which is then confirmed with a reply text that contains their secret phrase. If a friend tries to vote for someone, the friend would not know the pin #. If the militia tries to coerce someone to vote/change their vote a certain way, that person can simply fake a vote by entering a fake pin #. A confirmation reply text would still be sent, proving to the militia that the person had voted, but it would contain the wrong secret phrase, discreetly confirming to the coerced individual that his/her vote is still intact.

One could also imagine this "texting" vote system implemented by "automated voice" to increase accessibility, though this would likely have larger end-voter cost implications in developing countries, unless subsidized by the government/carriers.

-Thanks for the comment and though provoking questions Myles =). The above are just off-the-cuff ideas, so I'd also love to hear your and other's thoughts on how to tackle the questions you raised.

Thanks Paul and great question!

At this point, apart from one local primary, I don't think any governmental elections were held 100% on the Internet, though questions have certainly been raised about marginalization due to internet access. Governmental online elections are at an early stage of development in the United States, though digitizing of the process in general is relatively far along.

Disenfranchisement has occurred in the past (mainly through inadequate access to physical polling places, and more recently the internet). Although internet access in the US is quite high, it's disproportionately underrepresented in certain populations including low-income & rural areas.

I believe the best solution given current conditions would be to offer online voting as an option, while still having the traditional pop-up physical polling places (that are increasingly using electronic voting & tabulation, which is quite compatible/complementary with remote online voting). Traveling mobile polling solutions as suggested in various inspirations/concepts could also increase enfranchisement.

Personally, I like the thought of public libraries & schools with computer labs as natural polling places.

Thanks for the reply and I agree with you on the enhancement of physical polling places. I have included a concept (based on the Australian experience of being able to use any DTMF phone) that references your concept as an option for those with adequate access to internet or smartphones.
As usual my dial-up access is limiting how much I can incorporate in my original concept publishing.