The Challenges of iVoting Implementation
Part 2
When we last left this discussion, I had laid out a basis for our interest in technologies just over the horizon or "ready next" and in particular the growing interest in smartphone voting. I am essentially carving up a technology backgrounder white paper for easy reading here, so its taken a little time to figure out how to do that, while also being distracted helping develop other topics. Today, I help us dive in with a survey of the primary challenge areas to "Pajama Voting" (I love that phrase).
To develop a complete, credible, and defensible iVoting solution, there are three (3) areas of voting where new technology must be adapted (and adopted):
- Ballot casting,
- Public evidence, and
- Election administration.
In general, for iVoting to be adopted by public officials the technology must (at least) have proof of practical operation, and it must be usable by local election officials.
In order for iVoting to be practical it must fit into the existing norms and laws of the federal and local governments. It must be considered “another channel” for voting in the same way that absentee voting or military voting are. It follows then, that iVoting must be able to exist alongside these other forms of voting. The Internet is already used in a limited capacity with regard to overseas and military voting, which suggests that iVoting has at least the potential to exist alongside other forms of voting.
But even if iVoting were able overcome the formidable technical challenges it faces and create a reliable and legitimate digital ballot return mechanism usable by local elections officials, would it be transformative?
Unfortunately, not.
First, even if most voters used digital ballots, we would continue to need other voting channels that exist today. Unfortunately, these channels lack legitimacy and validity. Therefore, in order to be transformative iVoting must provide strong evidence of legitimacy on every ballot, no matter which channel the ballot goes through.
Seriously, this is an important point. Many advocates accuse opponents of requiring a perfect system when no perfect system exists today. Not so. But, simply adding another channel of ballot casting as an incremental means is insufficient if it only adds administrative burden. No one (and certainly not the Institute) is advocating for iVoting, if and only if it results in the "perfect" voting system. Yet, to be worthy of resolving the non-technical challenges to adoption and deployment any iVoting solution should be as transformative as possible.
Thus, transformative means at least offering a solution that is not additionally burdensome on the overall administration of elections, and we believe any iVoting solution will have to also provide better assurances of validity and legitimacy to entice election officials to bear the risk of an additional channel burden.
One other point: "imagineers" visualize a day when smartphones are the only, and most (add your favorite adjective here) means of participating in an election. That day is very far away (and believe me, we like to think forward). You see, these types of digital transformations can take generations. Consider for instance, that despite the proliferation of smartphones there is still a considerable global installation of copper-line based telephones. Consider how many black-and-white televisions still exist in the world. The point is, digital transformation is not a "light-switch change" of environment--the transformation is evolutionary (albeit sometimes at a rapid pace). And so too, with elections, there will be, for a variety of reasons, multiple means of casting ballots. So, this transformation will also be evolutionary, but its capability must be transformative.
Then there are the crypto-currency analogies. Digital currency has similar requirements and uses a digital ledger. However, election ledgers have additional constraints such as the varying skill levels of IT staff, budget limits, and technical tasks. And to be sure, there are several other differences between the currency of finance and the "currency" (the ballot) of public elections (we'll deal with those in more detail elsewhere and another time).
Moreover, current (existing) election technologies also need improvements in order to support an additional (iVoting) channel. Today, the existing election infrastructure is unable to support advancements in digital ballot return and digital election evidence.
One example of these many inadequacies is in voter check-in. Voter check-in requires that the system identify that a voter is eligible and that only one ballot is counted for each eligible voter. In a paper-based in-person system this is simple enough as voters are only allowed to vote at one precinct and the poll workers are given a list of all eligible voters within the precinct. But it is easy to imagine a situation in which a voter first votes on their phone and then fills out a paper ballot in-person, essentially voting twice. An iVoting system would have to prevent this scenario.
American democracy requires election management and operations that monitor multiple channels of ballot casting (in-person, by mail, etc.). So, first we need to update the antique “railways” for the current U.S. election “train,” so that we can more efficiently, cost-effectively, and safely run current types of engines ("diesel" to carry this analogy further), while supporting the few remaining steam locomotives, and still enable new Maglev trains as well — depending on what each locality’s election organization has or needs.
So, in summary, the challenges to an iVoting solution are many, but in general, any solution must be transformative of the current ecosystem; it must not be unduly burdensome on the existing administration processes and systems, and any iVoting solution must offer better assurances of validity and legitimacy than what we have today, or election officials will reasonably ask, "Why bother?" If the answer is, "Because it will increase participation," that answer alone does not address the transformative issues of election administration (and participation is a seaprate aspect of this we will address in a later posting).
Finally, any iVoting solution is not going to suceed, if it fails to address the totality of the election process including ballot casting, public evidence, and election administration. I will more closely examine these elements in coming installments.
I look forward to your comments and the on-going conversation!
--Sergio