>> In this section, we will discuss technology and the administration of elections. Now, elections are the central and the core drama in democratic societies. Enlarge representative democracies it's impossible to think of democratic governance without elections. There's no election, well, you've got no democracy. The United States has an extremely decentralized system of elections. The US Constitution gives authority to states to administer elections, including elections at the federal level for the Office of the President and members of the House and Senate. And then states further delegate this authority over conducting elections to local jurisdictions, to cities, and counties, and districts. So in the United States, there are roughly 10,000 election jurisdictions. This means that different ballot designs, different technologies in the voting booth, and different rules for polling stations, and different oversight of the entire process characterize an election in the US. You could call this a decentralized system, or you could call it a patchwork design of 10,000 systems for administering election. One consequence is that it means carrying out a reform of the decentralized system can't be done easily because nothing filters through a decentralized system very quickly. That includes trying to integrate new technologies into the mix. The US also relies on volunteers, rather than on paid public officials for holding elections. And more than half of the volunteers in poll stations are over the age of 60. They're retired citizens for the most part. These are precisely the people most at risk for contracting Covid-19. So creating the conditions for a safe and healthy election, a pandemic election, will be very difficult. Hence, the emphasis this year on voting by mail. The decentralized system we have in the US brings with it a perhaps unexpected benefit as well. It makes our system very difficult to hack or to manipulate from the outside. We all can remember Russian attempts to interfere in our election in 2016. Well, with 10,000 different electoral jurisdictions with different technologies and no master network hacking into and changing a vote count on a widespread basis is very difficult. Our decentralized system makes our own electoral administration more robust, less fragile in a certain way. Well, we faced an unprecedented challenge in administering an election in the midst of a pandemic with widespread social unrest, active efforts by outsiders to interfere, and the sitting president who is seeking without any factual basis to support him, to undermine trust in core elements of the administration of the election itself. We will now feature our discussion with two dedicated colleagues working in this landscape to help us better understand how US elections are conducted and what role technology plays in them. Our guests are Tiana Epps-Johnson, Founder and Executive Director at the Center for Technology and Civic Life, which is doing groundbreaking work to make US elections more inclusive and more secure. And by Stanford Law Professor Nate Persily, who co-directs the Stanford Cyber Policy Center and is a former Senior Research Director of the Bi-partisan Presidential Commission on Election Administration. >> With all the technological innovation and new possibilities, why can't Americans vote online yet? How do you weigh high-tech versus low-tech as a balance when it comes to good and right respecting elections? >> When I think about that question, I think that there are a few different places to start. I think that I could go down the path of talking about the different vulnerabilities around technology and why we're not quite there with online voting. But I think I want to start with introduction targeted but the structural, and to the deeply decentralized nature of elections as was described. In a country where rather than having a centralized election authority, we have 8-10,000 different independent entities that have some autonomous authority over administering elections. It means that when it comes to technology, there are election jurisdictions like Los Angeles County that is at the forefront of innovation, working with companies like IDEO to develop the first open source election software for folks to vote on. Then there are counties that I work with across the country that have few thousand registered voters and don't have so much as a website with basic voting information on it. That is not an experience that is few and far between, 94 or so percent of the population is covered by an election jurisdiction that has a basic website, but the US is large and we have about a thousand counties that don't even have that basic information online. When we think about questions about the election, just as in the US, I think one of the big challenges is that the system of federalism leads towards a disinclination of collective vision about what our elections, space and infrastructure should look like or could look like. I think with the exception of things like the Presidential Commission on election administration, but Nate worked closely on. There hasn't been much of this North star for what technology should look like in US elections in a collective effort to move many election jurisdictions in that direction in a concerted way. I think that there is a question about whether technology is up to the task? But I think more this isn't organizing challenge of having 8,000 different systems, and working to try to move 8,000-10,000 different entities in a similar direction towards a type of technology or election system, tech based or otherwise that works for all of us. >> Before Nate gets into the conversation, if I could just ask you Tiana, the decentralized nature of the system, yes. Should we be working toward an online opportunity to vote? If I understood my recheck correctly, after canvassing the state of play, the Netherlands doesn't have anything like the decentralized system we have, and they chose to do paper balloting. >> Maybe it shouldn't be a goal at all that we strive towards online voting. [OVERLAPPING] What's your view about that? >> Yeah. My view is that I'm relatively agnostic about the method by which folks vote, as long as it's one that is secure and that it works for everyone. I don't think about this so much as a question about whether or not we should choose to move to an online system versus a paper system. I think that we should have ample options because we have the tools and technology and know how to make those available to people to allow folks to vote and the method by which they feel the most comfortable. I think that if we could get to a point where we're able to provide an online voting system that would work effectively and securely for people, that that should be one option that we make available. But I don't think that we're there yet, although we've seen some interesting pilots over the last decade or so. >> Let me just deal head-on with the security issues, with Internet voting, and then move on. Which is that, we need a voting system that voters trust and we need a voting system that candidates trust. I worked on the National Academy of Sciences panel that was dealing with the future of voting technology. The question that Ron Rivest, a computer security guru at MIT who developed the security protocols for banking and other things said, he says, "We need a system in which the loser is confident that in fact, he or she lost, and you need to be able to demonstrate that the votes were cast and counted in an accurate way." We are not at the stage yet with Internet voting where you can have that level of confidence. There are countries that have moved in that direction; Estonia, our colleague, Toomas Ilves in FSI, when he was president, he moved them in that direction. But we in the US, imagine given that we don't even have competence right now, but even just today, the President said that he might not concede this election, that's when we're basically working with paper ballots. Imagine what would happen if we just had to trust the system, so we're many years away. The question is is it secure enough that the loser will admit and concede the race having trusted the system? That's the way we should think about it, and we're just not there yet. People say, "Well look, we trust hundreds of millions of transactions over our ATMs. Why can't we do it with voting?" The answer is that there are things that we have in place in financial transactions that we can't replicate in the polling place. For example, you can't get a receipt that says how you voted so that you might then show someone so that they could buy your vote or something like that. There are also the risks of systemic failure if some would happen on online voting system, are such that it would be difficult to recover. >> I think what I would add is that also, we think about the analogy to the financial system. With online banking transactions, there are transactions where there is fraud all of the time in those systems, but those are insured and those losses aren't beared by the person that's the account holder. We can't have any error or margin like that. There's no insurance for having an accurate election that folks trust, and so the threshold of how accurate a online voting system needs to be is at a higher level than what we see even in something that is secure and as well used as the way that we do financial transaction. >> Let me give a little bit of history here on what's happened in the area of voting technology over the last 20 years. In the 2000 election, which was determined in many ways by the Palm Beach butterfly ballot and punch-card ballots in Florida, we went through a soul searching as a country about voting technology. The Palm Beach butterfly ballot machine, maybe I'll even show you, I have it next to me, is used paper and you would punch through a paper, then you would read the holes in the paper and that would be the vote. Because so many of the votes were indiscernible in 2000, and exceeded even the margin of victory, we made a massive investment as a country in new voting technology as part of the Help America Vote Act, which followed that. Billions of dollars then went into electronic voting machines, many of which did not have a paper trail and so then, there was concern that there was hacking, or tampering, or dysfunctionality in these electronic voting machines. Then, we shifted from those in the last few years to what we'll call voter verified paper audit trail machines. So either you vote on paper or you end up voting electronically, but there is a paper record of your ballot. That is to deal with the problem of recounts and to make sure that you're not just trusting the machines. Right now roughly, maybe China has even better stats in this, I think it's 85 percent of voters are going to end up voting on paper in one way or another. There will be a paper ballot. It's probably even higher than that now given the male balloting rates this year. What happens now is that in the event of a recount, we have something that we can look to that can be scanned both by machines as well as read by individuals. Almost all of the votes in the ballot ground states are going to have paper as their main backup. I should say one thing just speaking to the huge audience in the Netherlands who's watching this, which is that the United States holds more elections for more offices than any other place in the world. We have more types of ballots than any other place in the world. Our ballots are extremely complicated and by that, technological challenges are greater than they are in many other parts of the world. So we can't just have the same ballot that's given out to everybody in the United States and then they mark with a red pen, there's just a lot of variety and complexity there. >> Tiana, do you want to add something? >> Yeah. I think that I would use another example. Nick talked about this in the context of voting machines, which is one way to think about the role of technology, high or low tech in security in the voting process. I'd like to offer the example of drop boxes, which is a different way of thinking about how we use considerations of election security and mix both low tech and high tech options. One of the available ways for people to return their ballot they receive in the mail, and many states is to return them to a ballot drop box that looks very similar to a mailbox, where election teams regularly collect those ballots and then bring them to the place to process them and eventually count them. When we're thinking about a drop box, there's physical security. Drop boxes are often bolted to the ground and have cameras, and they are fireproof, and there's all this physical security. Then, there is this really nitty-gritty logistical pieces in the background. There are specific driving routes for the two-person teams that pick up each of the ballots every day. When they come to that drop box, there is some form of seal, but then, there is a code that they have to write down and record. They have to record the time. Sometimes these folks have GPS on them to really see that they are at the place at the time that they said that they are there. Sometimes there is a process of actually weighing the ballots to make sure when they're picked up and then brought to the processing facility, that nothing untoward has happened. Every time, if you look at any single piece of the process of how elections are administered, once you look down the rabbit hole of the logistics behind how these things are managed, there are these really rigorous step-by-step processes, well-documented processes that election departments have in place for everything from the voting equipment to the ballot drop boxes, to the paper itself, and how you handle things like paper ballots.
