Evaluating Electronic Voting in the US
The following document is the product of a CHDS project that asked homeland security experts to evaluate how a better use of data could improve homeland security. Because of the nature of the projects, we are making an exception to the editorial guidelines of the collection in order to leave the citations in the way they were offered, instead of using hyperlinks as we normally request.
I. Introduction — Evaluating E-voting
By Donalyn Dela Cruz
On January 20, 2017, the U.S. will welcome its 45th president amid controversy and presidential speculation over Russia’s alleged influence on the presidential election. American Intelligence Officials confirmed Russian President Vladimir Putin “ordered an influence campaign in 2016 aimed at the U.S. presidential election.”[1] The conclusion by the U.S. intelligence community on Russia’s successful so-called attempt to influence the election may be different from hacking the election system, however, it is worthy to revisit the many ways in which Americans cast their votes on Election Day and how they are calculated.
During the 2016 election more than 125 million Americans voted. Analysts have argued this was the country’s lowest turnout in twenty years.[2] Some voters did not make it to the polls either from apathy or dissatisfaction with the process. Currently, Americans have limited options on how and when they can cast their votes. Every state determines its own options and voting equipment. Some states are making digital advancements. Understanding the magnitude of how local elections affect communities, online voting could potentially ease the burden of small local county elections departments. In California, the Monterey County Elections Department, for example, had nearly 60,000 ballots left uncounted at the close of the polls this past November 8.[3] With nearly 70 percent of the voters voting by mail, and a total of 137,000 ballots cast by the certification 30 days after the election, an enormous 43 percent of the total ballots cast were actually counted after Election Day.[4]
Many local elections offices are not adequately staffed to manage a timely outcome for results. This can lead to a nearly 30-day wait to find out who actually won in local races, many of which are determined by very narrow margins.[5] Online voting in a community like Monterey County could potentially solve the issues facing overwhelmed elections departments in smaller areas.
This paper evaluates the debate over electronic voting and the implications to big data. It examines what some countries are doing and explores the challenges associated with making online voting a prominent process for Americans to cast their vote. After analyzing the impacts of electronic voting, the paper will conclude with an analysis and recommendations.
II. A look at three cases of Online Voting
By Donalyn Dela Cruz
Electronic voting or Internet voting have taken place in dozens of countries.[6] Several European countries have conducted electronic voting trials, however some of those projects have been suspended.[7] This section profiles those who have conducted pilots of Internet voting technologies and have implemented e-voting with a degree of success. They include Australia, Estonia and Switzerland.
A. Australia
Australia has a compulsory electoral system.[8] The country has experimented with pilots of Internet voting technologies. In New South Wales (NSW), the Electoral Commission provides an opportunity for voters to cast their vote online or by telephone via the iVote® system. The online process of voting is used only for the NSW State by-elections and for the 2019 NSW State general election.[9]
NSW voters must register for iVote® online, which when done successfully, allows them to vote on electronic devices that have an Internet connection. Users are assured of their privacy during the registration process.[10] Registered voters are provided an iVote® number by SMS text, email, personal phone call or mailed to the registered postal address. When a voter completes their voting session, the iVote® system provides him/her with a 12-digit receipt, including a code that is stored in the system as part of their voter verification.[11] The 12-digit number is what voters provide when logging into the iVote® by Web. However, in 2011, there was a problem with the program that tarnished the vote count. “Other problems pertained to voter authentication, including a circumstance in which voters using truncated ID numbers (fewer digits than official ID numbers were required to have) were able to log in and vote.”[12]
The NSW Electoral Commission responded to criticism about security flaws and stated “All election contests counted by the NSWEC’s software where the error may have had an impact have been carefully analysed. In no other instance was the election result from 2012 found to be affected…The NSW public can be confident that the electronic count system used for elections is a vast improvement from manual counts and has drastically reduced the risk of counting errors and waiting times for election results.”[13] NSW has made improvements to its system and in 2015, there were more than 280,000 votes used through the iVote® system. One article on this system noted the following data:[14]
- The iVote system returned approximately 5% of the vote.
- The iVote system was taken up largely by those who were outside the state on polling day (91% of votes taken).
- 97% of users reported being fairly or very satisfied with their use of the iVote system
For the rest of the country, the voting system continues to evolve. The Australia Electoral Commission (AEC) conducts its elections in accordance with the Electoral Act. “If the legislation was to be changed to provide for voting, or aspects of voting, by electronic means, the AEC would be ready and able to implement this change.”[15]
B. Estonia
Since 2002, Estonia has been developing its Internet voting system. In 2005 the country launched the system for local government council elections, then in 2007 Estonians could cast their vote for parliamentary elections online.[16] Similar to Australia, voters who use the Internet voting system are provided voter authentication. This is derived from a citizen’s national ID number and now can also be determined by a mobile phone ID.[17] Estonia has also faced criticism regarding ensuring a secure voting system. In 2011, the Estonian government was sued “on the basis that it was possible for a virus to block submission of an Internet vote without the voter’s knowledge, and made a successful demonstration of such a vulnerability to the Court. Nonetheless, because there was no other mechanism to evaluate the reported result, the Court found no evidence that the reported result was inaccurate, and rejected the legal challenge.”[18] The government notes on its website, “Estonia takes the security of Internet voting very seriously. Voting over the Internet is as secure as ballot voting. A variety of technical, administrative, legal and other measures are used to safeguard the integrity of the system and most importantly the security and secrecy of the votes.”[19]
Internet voting is offered as an option for Estonian voters and it has proven to be popular. E-voters set a record of votes in 2015.[20]
C. Switzerland
The Swiss government has been working on its electronic voting system since 1998.[21] Switzerland’s voting system is a practice of direct democracy in which any citizen can challenge any law.[22] Switzerland has not gone nationwide with its electronic voting system. In 2015, Switzerland announced that the electronic voting system would be used by some Swiss living abroad along with about 90,000 citizens in Geneva and Neuchatel.[23] Part of the reason it has not gone national with e-voting is because of security concerns. One Swiss writer noted, “The Swiss authorities have opted for the gradual and carefully controlled introduction of electronic voting. E-voting still remains in a test phase, in which the number of voters who can access the electronic voting system is limited.”[24]
Users receive a code and a picture printed on his single use voting card to enable the server to identify him but also to enable him to identify himself to the server.[25] Swiss citizens seem very concerned about privacy and security when casting an e-vote. Yet, over time, the popularity of e-voting in the designated cantons is gaining momentum, and there is a “willingness to engage in e-voting projects.”[26]
These three examples show that while no country has perfected an online voting system, each are confident in moving forward in an electronic voting system. Despite security concerns, these countries are not deterred in their efforts to increase ways in which citizens can cast their vote.
III. Challenges of Online Voting
By Joseph Coulter
This segment will explore the challenges associated with moving from the current voting system to an online voting process throughout the United States, thus enhancing the e-government sphere. Most broadly, e-governance allows direct participation of the public in activities of the government. It is not simply about digital access to government information or the ability to pay taxes via web portals, but about how citizens relate with each other and their government. It allows citizens to communicate with each other and to participate in the democratic political process. As an extension of e-government, online voting has the possibility of improving efficiencies in vote tabulation, reducing the amount of time involved in the voting process, and increasing voter participation. However, the underlying assumptions that apply to most aspects of e-government theories do not apply to e-voting because of the severity of consequences if failure occurs- thus the idea of conducting elections for public office via online voting raises significant questions.[27][28]
A. Big Data Security and Data Integrity
When the Help America Vote Act (HAVA) passed in 2002, (which among other objectives intended to phase out the use of punch card ballot systems that led to problems during the 2000 U.S. election), many districts replaced their existing ballot systems with electronic voting systems.[29] While this was done with good intentions, the movement to electronic voting systems opened the process up to new risks associated with the security and reliability of those systems.[30]
Concerns regarding the dependability and trustworthiness of those electronic voting systems were backed by repeated demonstrations of fragility. In 2010, the District of Columbia tested an electronic voting system. During this public trial designed to test a proposed web-based voting system prior to deploying the system in the general election, a professor from the University of Michigan and his graduate students were able to take over the system from more than 500 miles away.[31] In 2016, Virginia decertified 3,000 WinVote touchscreen voting machines after learning about security problems plaguing the systems.[32] These problems included inadvertent access of the systems Wi-Fi network that disabled machines, and the “inexplicable” subtraction of votes cast in a school board election.[33]
The introduction of remote voting methods will only serve to exacerbate the issues that have been faced with the existing electronic voting methods, since access will potentially be available from any global location. A web-based voting system must remain available and uncompromised on an open network, and it has to serve voters connecting from untrusted clients while still maintaining both the integrity of the election results and the secrecy of voters’ choices.
Security researchers and academics have both urged election officials to resist adoption of electronic voting systems, worrying that conducting elections for public office over the internet raises security risks that surpass those experienced with physical machine based systems.[34] According to researchers, internet voting would be vulnerable to a variety of well-known cyber-attacks. Such attacks could even be launched by an agency or government beyond the reach of U.S. law and could cause privacy violations, vote-switching, or vote buying and selling.
Attempted attacks on voting processes did occur during the 2010 Internet voting pilot project in the District of Columbia. The University of Michigan team that infiltrated the system detected and blocked attempted intrusions by IP addresses based in Iran, India, and China.[35] Thus, the project demonstrated the danger of an organized attack by an enemy agency is not mere speculation. Of even greater concern, had the University of Michigan group not already entered the voting system, these attempts may have remained undetected.
An internet based voting system needs to be much more secure than a typical website or data portal used for government purposes. It must have data integrity, which provides protection against alteration in an unauthorized manner since the time it was created, transmitted, or stored by an authorized source. In this way, a secure voting system would maintain both the secrecy of voters’ choices and the integrity of election results. Ensuring the integrity of information requires being able to detect corruption or change to even a single bit of transferred or stored data.[36] This must all be completed while remaining available, and serving voters connecting from a variety of clients, both trusted and untrusted.[37] If web-based voting were to be enacted, ensuring that information has not been altered by unauthorized or unknown means would be among the greatest challenges faced.
In 2008, computer scientists from 32 universities and computer agencies across the nation, including Princeton University, Stanford University, and Johns Hopkins University issued a statement detailing challenges towards the creation of a nationwide secure web-based voting system.[38] According to this group, “there is ample reason to be skeptical of internet voting proposals.” Citing vulnerabilities in existing technologies, the group listed technical security challenges that must be overcome in order to have a secure and verifiable internet voting system. These challenges included:
· The prevention of malicious software, firmware, or hardware that can change, fabricate, or delete votes, or deceive the user in a variety of ways including leaking information about votes to support voter coercion, and preventing or discouraging voting;
· Disruption of denial of service attacks from networks of compromised computers, causing messages to be misrouted, as well as the prevention of other kinds of electronic attacks;
· Finding a process to prevent undetected changes to votes, not only by outsiders, but by insiders too (such as technicians, system administrators, and election officials) who have justifiable access to election software or data;
· The creation of mechanisms that provide a reliable voter-verified record of votes that is as effective for auditing as paper ballots, without conceding ballot secrecy;
· The need to design a system that is dependable and verifiable even though Internet-based attacks can be mounted by anyone from any place on the globe.
These security problems persist despite many different trials and efforts to solve the issues. Citing the risks associated with the manipulation of votes and election results and the risk of data interception and theft from servers, both concerns represented by the previously referenced scholars, the Homeland Security cyber-division presently recommends against the adoption of online voting for elections at any level of government.[39]
B. Verification of Users and Voter Anonymity
There is a fundamental conflict between keeping votes anonymous as is required by most states and voter verification in a web-based system. The majority of states (44) have constitutional provisions guaranteeing secrecy in voting, while the remaining states have statutory provisions which refer to secrecy in voting.[40] While existing paper ballot methods do this function very easily, leaving a document that can be tallied by vote counters without connection to an individual voter, electronic systems leave digital audit trails.[41] The relationship between voter identity and the cast vote should be private, and should not be able to be revealed by another party. The requirement for anonymity makes electronic voting different from other applications. The obscurity of information and data sources makes fraud easier since addition, deletion, or modification of anonymous votes can be hard to detect.[42]
As secrecy in the voting process is of vital importance, concerns regarding the secrecy of ballot procedures when casting a vote from a remote location, rather than the controlled environment of a polling booth should also be noted. It is not typically feasible in the home (from where an electronic vote is likely to be conducted) to achieve the level of isolation during the voting period that is possible at a supervised polling location. As noted in a review of electronic voting in Europe, while the home may be secure from outsiders, for many families a home is not a place that can be easily made private.[43] This is particularly true for households with limited space, which means that the socio-economically disadvantaged may also be politically disadvantaged from lack of domestic privacy.[44] When voting takes place in an unsupervised context, it is virtually impossible to ensure that it will be carried out in secret. That lack of secrecy constitutes a violation of the principles of freeness and fairness that govern U.S. elections.
Users of a web-based system, in this case the voter itself, will also need to have a secure way with which to enter the system and validate their relationship with the voting guidelines and their eligibility as users. In principle, this is not unlike the requirements of an e-commerce website which connect a user’s purchase with their payment method all accessed through a user-controlled login method. However, unlike errors and fraud associated with e-commerce where a consumer may eventually learn of unauthorized purchases made using their account (via funds being drafted from associated accounts), as a result of the requirements of anonymity and ballot secrecy in the voting process, a voter would be unlikely to learn if their vote was manipulated or intercepted before being cast on an Internet voting website.[45]
C. Jurisdictional Ownership
There are over 10,000 election jurisdictions that administer major federal elections at the county level or below.[46] In most states, elections are organized and run at the county level, however in some cases they are administered by townships or county subdivisions.[47] Oregon conducts elections entirely by mail, however provisions are made for voters to cast ballots at county election administrative offices.[48][49] In each of these jurisdictions constituents maintain different priorities. With each of the diverse ways that these jurisdictions and the state-run election commissions operate, finding a uniform method for organizing a web-based voter system will be a challenging prospect.
While there are presently uniform standards for certain aspects of the election process, the question that must be asked with regards to e-voting methodology is what does a regulatory model for voting systems look like? Although Section 301 of the Help America Vote Act (HAVA) does establish some baseline standards for voting systems procured with HAVA funds, the current Voluntary Voting System guidelines impose no federal requirement that voting systems used in local jurisdictions comply with a national standard.[50] This is because logistic factors affecting election processes in individual constituencies differ greatly. Any attempts to implement a nationally run internet based voting process will need to be applicable to each electoral district, while still being easily evaluated outside of the environment in which they operate.[51]
While being able to cast a ballot from a home computer or cellular phone might make voting more convenient, the security problems presented by an internet voting system present an undesirable risk to the integrity of an election. The consensus from computer experts all point at the security vulnerabilities and as yet unsolved technical requirements of such a system. When combined with other questions, such as verified access by voters to election systems and protection of their data, along with jurisdictional questions regarding the creation of guidelines and oversight, internet voting is a technology that has significant obstacles in its path before it is ready for real consideration. While not ready at the moment, these obstacles are not insurmountable and, with proper planning and coordination, internet voting could conceivably be available in the future.
IV. Impacts of Online Voting
By Micki Trost
This section will explore the impacts of moving to an online voting process within the United States. Although the definition of online voting includes email, fax and casting your ballot through web-based platforms, this proposal looks at moving to a web-based voting process through the internet on computers or personal electronic devices (such as smart phones and tablets.)[52] The section will explore three specific impacts related to transitioning to an online voting process. The first area of impact will determine if online voting increases or decreases the number of people participating in the election. The second area of impact will discuss the compare the costs associated with a new online process to the current system. Lastly, the area of providing a final count for election results will be reviewed. These three areas of impact will be used to determine if an online voting process should replace the current process used by the majority of elections within the United States.
Voters in the United States expect the election process to be transparent, accessible, and secure while protecting their right to a private vote. If these three conditions are met voters are more likely to trust and support the system. Currently, online voting is occurring for military members and during small test runs within the country. In 2016, the state of Utah ran an online election and in 2000 the state of Arizona tested the online process. The first step in online voting may be the ability for voters to register online. There are 38 states that allow voters to register online through email, text, or web services. [53] The Ohio Secretary of State Jon Husted is quoted as stating, “It [online registration] eliminates another excuse for not voting. Nobody can say it’s too hard. You don’t have to leave home to participate in Ohio democracy now.”[54] Let us take this frame of mind and apply it to the entire online process and examine the impacts of online vote casting. This section examines how government agencies can maintain the trust of voters through online voting.
A. Increasing the Number of Votes
The lure of online voting is to capture the participation of the 85 percent of American adults who use the internet.[55] Online voting is currently limited to unique situations such as military deployment or unique testing situations. According to a technical paper written by the U.S. Election Assistance Commission there are 11 United States online voting projects that have taken place and their results have been mixed related in increasing the number of ballots cast. In 2000, Alaska conducted an online vote for the Republican Party’s primary election. The results were “not a grand success,” according to Randy Ruedrich, Alaskan Republican Party’s chairman, due to the problems caused with internet connectivity. The Alaskan rural communities had a much smaller voting rate than the state average caused by internet connection issues.[56] In the same year Arizona allowed online voting from anywhere in the world using the internet for their 2000 Democratic Party Primary.[57] The election resulted in only a 10 percent participation rate by Arizona votes and 46 percent of those ballots completed online.[58] During the 2000 Arizona online elections questions were raised regarding the use of the internet by racial and ethnic categories. The initial argument was that only white, affluent voters would have access to online voting. A Pew Research Center study from June 2015 shows that “78 percent of blacks and 8 percent of Hispanics use the internet, compared with 85 percent of whites and 97 percent of English-speaking Asian Americans.”[59] In 2016 the state of Utah used online voting for their March 22nd primary elections and GOP caucus vote. [60] Utah provided online voting through a company called Smartmatic without security issues or any other voter related issues. The Smartmatic website states that 90 percent of Utah registered voters participated in the online vote and that participation was the highest among voters aged 56–65 years old.[61] The Smartmatic case study also shares that the average time spent voting by those participating in the Utah election was 73 seconds. This greatly reduces the amount of time needed to participate in an election and may result in a higher participation level. The country of Estonia implemented online voting in 2005 with an initial result of 1.9 percent cast by this method. That percentage has grown to 31 percent in 2014.[62] The statistics coming from Estonia show continued growth of online voting from year to year.
B. Remote Voting Increases Votes
The need for remote voting is becoming more common due to military deployments and the expansion of United States citizens to remote areas of the country. Overseas online voting is already occurring through email and faxing, however, this proposal is to allow votes to be cast through internet access. The switch on elections conducted entirely by email ballots has already occurred in Washington, Oregon, and Colorado to accommodate remote voting. In 2012 the U.S. Election Assistance Commission found 17 percent of ballots were cast through the mail.[63] The Commission pushes for embrace remote voting and treats “remote voting as a first-class capability rather than as a back-up system with second-class effectiveness, speed, security, and integrity.”[64] Returning to the 2016 Utah online election the case study reports that voters participated from 45 countries and from all continents.”[65] The information from this case study supports the idea that online voting allows a higher participation rate by remote voters and it also increases accessibility.
C. Increased Accessibility Increases Votes
The U.S. Election Assistance Commission also suggests in its 2011 technical paper that online voting will increase accessibility for the 20 percent of Americans living with disabilities. The Centers for Disease control and Prevention state that one in five adults have a disability and that the “most common function disability type was a mobility limitation.”[66] The ability to cast a ballot from any location and from their home with their own computer may allow for an increase in voting. Currently, under the 2002 Help America Vote Act polling places are required to provide at least one location for disabled voters. Online voting could help jurisdictions meet this requirement. A study sponsored by the U.S. Vote Foundation found that “In 2007, 20 percent of Americans with disabilities said they were unable to vote in a presidential or congressional election due to difficulty getting to the polls, or barriers at polling locations.”[67] Additionally, voters with disabilities will regain their privacy during the act of voting. The U.S. Vote Foundation found that these votes required assistance during the voting process when attempting to vote in-person resulting in a loss of privacy. A second consideration for accessibility is providing ballots in multiple languages. The federal government requires ballots to be printed in languages based on the “voting-age citizens who live in those communities and have limited English skills…”[68] The use of online voting will allow individuals to access ballots in multiple languages. Image 1, provided by Pew Research Center, shows the growing need for additional languages for ballots. Providing ballots in multiple languages to meet the needs of the voters increases accessibility and the potential for increased voting.
D. Cost of Elections
The U.S. Vote Foundation found that the average cost of elections in 2015 ranged from $2 to $20 per vote. The cost of online elections will be beneficial in the mid to long term, however, the initial set-up or transition expenses could be extreme. The U.S. Vote Foundation estimates that the costs may “range from approximately one and a half million to tens of millions of dollars.”[69] The current costs associated with elections includes the printing of ballots, staff compensation for election judges as well as permanent election office positions, equipment upkeep and purchasing, and other logistical expenses. The online voting will bring a reduction in staffing, polling places, and printing costs. However, new costs may be associated with fixing a malware or Distributed Denial of Service (DDoS) attack, annual license fees and the upfront equipment costs. An example of how local jurisdictions can transition to online voting is found on the Smartmatic website. The company handles the pre-election process including: legal framework analysis, polling location management, and voter administration.[70] The company then moves to election support that includes: staff management, observer accreditation, candidate administration, campaign monitoring, ballot production, training, logistics, voting, and consolidation or the final vote results. And then Smartmatic finishes with a formal wrap-up to review the election process with an after action review to improve for the next election. This company is not unique; however, by implementing an online election through these services a reduction in full-time staffing may be required as it is handled through this one contracted service.
An additional cost will come in the form of expenses to change the election laws themselves. The U.S. Vote Foundation states that “the U.S. legislators must change the legal framework of elections in nearly every jurisdiction that wishes to use internet voting.”[71] The Foundation stated the process could take 5–10 years to implement.
D. Election Results
A clear benefit of online vote casting is determining the election results. Online voting ensures votes are accurate and timely.[72] The 2016 official vote results came weeks after the November 8 presidential election. The November 20, 2016 Los Angeles Times headline reads “This Is Why it Takes So Long to Count Votes in California.”[73] An online election will produce immediate results of the votes cast.
Online voting can build upon the current trend in local jurisdictions adopting the process of online registrations. The results are mixed related to the impact of online voting in producing a higher number of ballots cast. The online voting process will allow for an increase in voting accessibility for remote voters including our U. S. military members. The increased accessibility will better serve individuals with special needs including mobility issues, disabilities, or those who are non-English speakers. The long-term impacts of online voting may lead to a reduction in costs; however, it will likely cause an initial increase in startup costs. A clear benefit will be the ability to produce near immediate election results.
Images
Image 1: [1] D’Vera Cohn, “More voters will have access to non-English ballots in the next election cycle.” Pew Research. (2016).
Image 2: Source: https://www.verifiedvoting.org/resources/internet-voting/
I. CONCLUSION
By Tiffany Brown
Election officials and security experts are legitimately concerned about risks posed by Internet voting systems, which, for the time being, are uniformly considered unusable in large-scale elections. However, the same stakeholders have much to gain by continuing to explore and test Internet voting in smaller elections. Implementing Internet voting in local elections can serve to identify and address security issues on a smaller scale, and they can provide valuable political and social insights for broader consideration. In this way, individual communities and states who are willing to participate in online voting early on lend themselves to a greater cause by beginning to help build a successful track record — something a new election system must have before it is considered for large-scale elections.[74]
Perhaps the day will never come where, as a nation, every US citizens has the ability to cast their votes online using a home computer or personal device. Amid current security concerns and increased, startling advancements in cybercrime, it is not beyond comprehension that the risks will eventually be deemed too great a risk for something so closely associated with public trust as large-scale elections; but until that time, internet voting remains a consideration that continues to be explored. Any real possibility of implementation for large-scale elections will indeed be the result of experts testing and developing smaller systems while continuing to weigh the benefits
Security, Cost & Public Trust
From a standpoint of transitioning the public into an online voting environment, online voting represents a natural stepping stone from the current electronic and internet-based election activities such as online registration and electronic voting via on-site voter kiosks. Americans are already conducting a number of e-commerce activities and using governmental sites that did not exist a decade ago. At the current pace, it seems likely that citizens will have a desire to add voting to the list of things they are able to accomplish using a personal device; however, as discussed extensively in this paper, cyber concerns regarding online voting systems are largely unassociated with the online and electronic innovations made to elections in recent years such as online registration and electronic ballot systems used in on-site voter kiosks. End-to-End Verifiable Internet Voting (E2E-VIV) systems have been launched with varying success for local elections but these commercial internet voting systems to date are fundamentally insecure.[75]
The greater fear stemming from using a voting system which cannot ensure the integrity of the key components of democratic elections — privacy of voting choices, anonymity of the voter and accurate results — is a loss of voter in confidence should the system fail. Any recommendations for implementing internet voting must not lose sight that voter confidence is extremely fragile. The idea that “voters trust is hard-won and easily lost”[76] cannot be overstated, however, in the course of working toward developing a larger system the public can trust, a degree of acceptable risk must be assumed by those overseeing the smaller systems which serve to help pioneer the effort.
If testing election systems can result in decreased voter confidence, then thoughtful approaches around implementation of those system tests will be of critical for success. In identifying a community to participate in internet elections, those with a historical propensity for innovative election processes (such as electronic voting or mail-in ballots) may be more open to the idea of internet voting than another group who still participates in a more traditional system such as one where voters must be physically present to cast a vote.
Oregon, the first state to implement a full mail-in ballot system by over a decade, did so by a public vote to change the election process, suggesting that there is value in seeking public opinion regarding any decision to change voting systems. An election can help with buy-in because voters are given a choice.In the course of getting to make that choice, they are given information about the option by way of a voters’ pamphlet., This allows the voter to have more information at their disposal to be able to educate themselves and formulate an understanding and opinion about it.
Aside from changing voting methods by way of elections, the same information provided in a voter pamphlet could be provided to citizens by way of a marketing strategy aimed at communities uninterested in voting on the issue, yet open to dialogue and community consideration in a less formal manner. Getting buy-in by way of a vote (or from a more tacit form of approval) does not mean that voters will readily accept system insecurities or failures as part of the process — nor should they — but the outreach and education provided by way of election or marketing campaign might result in a more forgiving public response if people understand and accept the risks (and the plans in place to address them) should security issues arise. Whether educating the public in an effort towards system orientation or providing an option to express their opinion in a vote, marketing and education are key components for successful implementation.
Some worry that security issues will render internet voting too costly to maintain, easily outweighing the benefits.[77] The anticipated costs of implementing internet voting fall into two categories: The initial costs to set up a system and the ongoing costs of then running and maintaining it. Citing E2E-VIV systems that still lack a level of security deemed necessary for large-scale elections coupled with anticipated costs associated with preventing and mitigating cybercrime moving forward, the price tag for implementing the new voting system is high but will plateau after system startup. In addition, the increase in costs stemming from new system startup is offset by a decrease in running elections at the status quo. Internet voting would remove a historical need for workers who can staff polling stations and count ballots, eventually decreasing election costs.
Speculation and yet unknown factors regarding internet voting costs suggest that jurisdictions will have no choice but to invest in costly systems, however, an early trend in internet voting of hiring election vendors to conduct operations creates more predictability around elections costs for interested jurisdictions. In the long run, large increases to election costs are not anticipated as a result of internet voting — in fact, some believe the opposite will be true and that election costs will decrease — but in the event that an increase does occur, a cost-benefit analysis to analyze the relationship between the new system and benefit that it provides to society.
A. Benefits
Used thus far in smaller, more informal elections throughout the US, there is not enough data yet to suggest that implementing internet voting will increase voter turnout, but it is probable. Online opportunities have increased voter registration and other e-commerce activities suggesting that an online option for a given activity has the potential to increase participation. Unfortunately, increased voter registrations do not necessarily translate into increased voter turnout further suggesting that simply getting registered is not the biggest obstacle for voters — casting a ballot is. For citizens living abroad, Internet voting provides a simpler solution to completing a ballot than previous methods.
Some have credited high voter turnout in Colorado with a mail-in system, but that is hard to determine conclusively since the state also has some of the highest marks from the Pew Charitable Trust suggesting that other strengths in their election process could be a factor.[78] Although there is some indication that mail-in ballots can boost turnout in low interest elections specifically, it is not widely held that mail-in ballots increase voter turnout but they make it easier for those who intended to vote anyway.[79]
That makes sense since mail-in ballots only provide a new, simplified way to receive materials and cast ballots while removing the necessity of leaving one’s home to vote. Critics of the mail-in ballot system contend that it takes away from the community process of meeting in a particular place on a set day to vote, something they believe is an integrally important part of the election process, now lost in certain communities to the mail-in system. If it is true that sense of community is an important feature of voting, then internet voting holds new promise of (re)building a voter community within which to participate.
Virtual communities abound on the internet for a variety of interests and election interests are no exception. Not only will those without previous access to voting be able, but they may also benefit from community resources the internet has to offer around elections. Social media, internet news, blogs and countless other resources continually offer new ways to inform, educate and (perhaps most importantly) inspire voters in ways never previously imagined. Mail-in voting changes the election environment in a significant way by allowing voters to educate themselves on the issues and cast a vote from home, but critics argue that it weakens the democratic process by taking an election-day meeting place out of the equation thus lessening voter engagement. In other words, the mail-in process made voting easier but lacked the physical engagement that some felt was important to the process. If one can accept that virtual engagement in society is rapidly replacing activities formerly characterized by physical engagement, and that a sense of community is an important piece of the voting system, then consider the vast opportunities which exist on the internet to engage voters.
It is not the technology that will change voting and increase voter turnout — it will be the education and outreach that accompanies it. Voters who begin turning to the internet for voting will not simply be using it as a vehicle to cast a ballot like Oregonians turned to the postal service in the late 80s. In that case, postal resources made the change possible but not in an interactive way: The materials were delivered and the ballots were picked up or dropped off. Consider instead, the internet serving as both ballot shepherd and community resource for finding information and like-minded individuals with which to discuss politics. Where citizens once met at the polls on one day, as a community, to take part in the democratic process, we’re already seeing how that environment could manifest itself online.
B. Community Impacts
Online voting could begin to address some impacts to communities that experience prolonged delay of election results. If online voting became a nationwide standard for election districts throughout the country, we might understand how online voting could prevent what happened in the 2016 presidential election; but a national approach is not the recommendation. On the contrary, the proposal is to start small, therefore an example more relevant to a local approach suggested with Monterey County Elections.
Online voting also creates additional opportunity for citizens with functional or access needs by providing them an opportunity to choose from a variety of personal platforms and the ability cast a vote without having to visit a specific location. Mail-in ballots also provide an opportunity to vote from home but the voter must commit additional resources (voter time and expense=burden hours) to get the ballot from their home to the election office. One challenge common to vulnerable populations is those without internet access, an issue addressed which can only be addressed by providing the internet to voters, something that could be accomplished in partnership with libraries or other public locations which already feature other e-commerce services such as libraries, clerk & election offices, employment offices, etc.
Internet voting could increase voter turnout, as noted in the case studies. Online voting may be more appealing to those who are looking for a convenient way to participate in elections. This will require target-marketing efforts to provide outreach and information for those who are reluctant because they don’t completely understand the process.
VI. RECOMMENDATION
By Tiffany Brown
At the outset, a recommendation to implement online voting seems to defy logical assessments and guidance agreed upon by security experts and elections officials who assert that conducting online elections is still too risky because current systems lack the desired level of security and vote reliability. They contend that, unless and until the security of current verifiable end-to-end voting systems can be guaranteed to protect the integrity of elections, the risk to election integrity is too high. Voter confidence is too precious and hard to recover when lost.
But that attitude is one- and short-sided, as it fails to take into consideration both the potential benefits of internet voting as well as the likelihood that public demand will result in internet voting become increasingly utilized; early adoption is important and key to future successes. The goal therefore should not be to keep online voting at bay but to find ways to move it forward and develop the technology, albeit at a cautious rate, all while remaining mindful of the delicate interrelationship of government and voters throughout the process. Successful efforts have occurred on small scales. In some cases, security risks and attempts to hack were identified, but those are experiences that serve to help experts better understand how to build a resilient system.
The success of the Oregon mail-in system adopted in 1987 is attributed to two factors: The first was the decision to change elections wholly, doing away with the previous system altogether and offering only the mail-in system (voters can either mail their ballots via US Mail or drop at designated ballot boxes). The second was that the decision to change the system was made by the voters via an election. The system was not vulnerable to issues around casting redundant votes because there was no alternative system, and the public was given a voice in deciding whether or not to change election laws. While this proposal neither suggests enacting law nor overhauling current systems, Oregon is a useful example of how voter engagement can be used to smooth a system transition.
Initially, online voting systems should be implemented in communities who have agreed by way of vote (or more tacitly) by way of immersing themselves in the marketing and public outreach education made available. In addition to explaining technologies and processes, outreach efforts should also speak to the realities of current system security, balancing information regarding vulnerability with the benefits of online voting. While reports have concluded that hacking efforts by the Russian government to affect the US election resulted in more subtle impacts outcomes than initially feared, the experience put a spotlight on the frailty of internet security and voter trust. Communities with an innovative election history or who have agreed to participate can provide a more engaged group with which to develop systems. It is only through continued efforts of communities willing to consider the benefits of internet voting that broader applications will be applied. Those smaller, local examples provide opportunities for system development, cybersecurity mitigation and increased public education.
[1] Sanger, David, “Putin Ordered ‘Influence Campaign; Aimed at U.S. Election, Report Says,” The New York Times, January 6, 2017, http://www.nytimes.com/2017/01/06/us/politics/russia-hack-report.html?_r=0
[2] Wallace, Gregory, “Voter turnout at 20-year low in 2016,” CNN, November 30, 2016, http://www.cnn.com/2016/11/11/politics/popular-vote-turnout-2016/
[3] Wu, A. (2016, November 9). Update: Election 2016 — Vote Counting Continues. The Salinas Californian, p. 1A.
[4] Elections Results. (2016, December 2). Retrieved January 5, 2017, from Monterey County Elections: http://www.montereycountyelections.us/Election%20Result.htm
[5] Wu, A. (2016, December 2). Villegas and Davis Win Council Seats After Tight Races. Retrieved January 5, 2017, from The Salinas Californian: http://www.thecalifornian.com/story/news/2016/12/02/villegas-and-davis-win-council-seats-after-tight-races/94845228/
[6] Verified Voting. Accessed January 1, 2017 https://www.verifiedvoting.org/internet-voting-outside-the-united-states/
[7] Samuel Jaberg, “Hacking fears jeopardise e-voting rollout,” September 2, 2015, http://www.swissinfo.ch/eng/voting-with-a-click_hacking-fears-jeopardise-e-voting-rollout/41635672
[8] Wikipedia. Electoral system of Australia. Accessed January 1, 2017.
[9] iVote. https://www.ivote.nsw.gov.au/
[10] iVote instructional video — Registration. https://www.ivote.nsw.gov.au/ivoteprocess.aspx
[11] iVote, “4.11 Authentication of Vote,” http://www.office.elections.nsw.gov.au/voting/ivote/procedures/technology_assisted_voting_approved_procedures_for_nsw_state_elections
[12] “Australia,” iVote, accessed January 1, 2017 https://www.verifiedvoting.org/internet-voting-outside-the-united-states/
[13] Electoral Commission NSW, “Response to academic research about NSW’s proportional representation electronic vote counting,” June 14, 2016, http://www.office.elections.nsw.gov.au/__data/assets/pdf_file/0008/218681/PRCC_statement.pdf
[14] Sam Campbell, The NSW iVote Project — Internet Based Voting in Australia, https://webrootsdemocracy.org/2015/12/06/the-nsw-ivote-project-internet-based-voting-in-australia/
[15] Australian Electoral Commission, “Voting within Australia — Frequently Asked Questions,” Accessed January 1, 2017. http://www.aec.gov.au/FAQs/Voting_Australia.htm#e-voting
[16] Estonia.eu, “Estonian Internet voting system,” Accessed January 1, 2017. http://estonia.eu/about-estonia/economy-a-it/e-voting.html
[17] European University Institute, “Internet voting a success in two European countries,” February 12, 2013, http://www.eui.eu/News/2013/02-12-InternetvotingasuccessintwoEuropeancountries.aspx
[18] Verified Voting, Estonia, Accessed January 1, 2017, https://www.verifiedvoting.org/internet-voting-outside-the-united-states/
[19] Estonia.eu, “Estonian Internet voting system,” Accessed January 1, 2017. http://estonia.eu/about-estonia/economy-a-it/e-voting.html
[20] The Baltic Course, “E-voters number posts a new record in Estonia,” Accessed January 1, 2017, http://www.sgo.com/media/article/the-baltic-course-e-voters-number-posts-a-new-record-in-estonia/
[21] Gerlarch, Jan, Gasser, Urs. “Internet & Democracy Case Studies Series: Three Case Studies from Switzerland: E-Voting.” Berkman Center Research Publication No 2009–03.1. March 2009. http://cyber.harvard.edu/sites/cyber.harvard.edu/files/Gerlach-Gasser_SwissCases_Evoting.pdf
[22] Wikipedia. Voting in Switzerland. Accessed January 2, 2017.
[23] Elections 2015, “Electronic voting in Switzerland,” Accessed January 1, 2017, https://www.ch.ch/en/elections2015/electronic-voting/electronic-voting-in-switzerland/
[24] Janani Krishnaswamy, “The Insider’s view of Swiss e-voting,” Government Gazette, September 28, 2015, http://governmentgazette.eu/?p=6562
[25] Ballotpedia. Electronic Voting as used in Switzerland. Accessed January 2, 2017, https://ballotpedia.org/Electronic_Voting_as_used_in_Switzerland
[26] Gerlarch, Jan, Gasser, Urs. P. 14.
[27] Fang, Zhiyuan. “E-government in digital era: concept, practice, and development.” International journal of the Computer, the Internet and management 10, no. 2 (2002): 5.
[28] Moynihan, Donald P. “Building Secure Elections: E‐Voting, Security, and Systems Theory.” Public Administration Review 64, no. 5 (2004): 515–528.
[29] “The Help America Vote Act Of 2002”. 2016. Justice.Gov. Accessed December 29 2016. https://www.justice.gov/crt/help-america-vote-act-2002.
[30] Barrett, Brian. 2016. “America’S Electronic Voting Machines Are Sitting Ducks”. WIRED. Accessed December 29 2016. https://www.wired.com/2016/08/americas-voting-machines-arent-ready-election/.
[31] Wofford, Ben, Glenn Thrush, Thomas Mills, and James Higdon. “How to Hack an Election in 7 Minutes.” POLITICO Magazine. Accessed December 24, 2016. http://www.politico.com/magazine/story/2016/08/2016-elections-russia-hack-how-to-hack-an-election-in-seven-minutes-214144.
[32] Zetter, Kim. 2016. “Virginia Finally Drops America’S ‘Worst Voting Machines’”. WIRED. Accessed December 29 2016. https://www.wired.com/2015/08/virginia-finally-drops-americas-worst-voting-machines/.
[33] Ibid.
[34] Waddell, Kaveh. 2016. “How Electronic Voting Could Undermine The Election”. The Atlantic. Accessed December 29 2016. http://www.theatlantic.com/technology/archive/2016/08/how-electronic-voting-could-undermine-the-election/497885/.
[35] Wolchok, Scott, Eric Wustrow, Dawn Isabel, and J. Alex Halderman. “Attacking the Washington, DC Internet voting system.” In International Conference on Financial Cryptography and Data Security, pp. 114–128. Springer Berlin Heidelberg, 2012.
[36] Patel, Dhiren R. Information Security: Theory and Practice. PHI Learning Pvt. Ltd., 2008.
[37] Wolchok, Scott, Eric Wustrow, Dawn Isabel, and J. Alex Halderman. “Attacking the Washington, DC Internet voting system.” In International Conference on Financial Cryptography and Data Security, pp. 114–128. Springer Berlin Heidelberg, 2012.
[38] 2016. Verifiedvoting.Org. Accessed December 29 2016. https://www.verifiedvoting.org/wp-content/uploads/2012/09/InternetVotingStatement.pdf.
[39] “More Than 30 States Offer Online Voting, But Experts Warn It Isn’T Secure”. 2016. Washington Post. Accessed December 31 2016. https://www.washingtonpost.com/news/post-nation/wp/2016/05/17/more-than-30-states-offer-online-voting-but-experts-warn-it-isnt-secure/?utm_term=.9c3739d5f59e.
[40] Fitzgerald, Caitriona, Pamela Smith, and Susannah Goodman. “The Secret Ballot At Risk: Recommendations for Protecting Democracy.” August 18, 2016. Accessed December 30, 2016. http://secretballotatrisk.org/Secret-Ballot-At-Risk.pdf.
[41] Kobie, Nicole. 2015. “Why Electronic Voting Isn’t Secure — But May Be Safe Enough”. The Guardian. Accessed December 30 2016. https://www.theguardian.com/technology/2015/mar/30/why-electronic-voting-is-not-secure.
[42] Cetinkaya, Orhan, and Deniz Cetinkaya “Anonymity in E-Voting Protocols.” Accessed December 30, 2016. https://pdfs.semanticscholar.org/6ffe/22e93a299bdcfa6789b155805c4f84b4db67.pdf
[43] Birch, Sarah, and Bob Watt. “Remote electronic voting: Free, fair and secret?.” The Political Quarterly 75, no. 1 (2004): 60–72.
[44] Ibid.
[45] “The Dangers Of Internet Voting”. 2015. The Heritage Foundation. Accessed December 30 2016. http://www.heritage.org/research/reports/2015/07/the-dangers-of-internet-voting.
[46] Fischer, Eric A. “Voting technologies in the United States: overview and issues for congress.” Congressional Research Service [Library of Congress], 2001
[47] Ibid.
[48] The County Administrative Offices of Oregon are included in the total count of jurisdictions.
[49] Brace, Kimball W., and Michael P. McDonald. “Final Report of the 2004 Election Day Survey.” Submitted to the US Election Assistance Commission (2005).
[50] “The Help America Vote Act Of 2002”. 2016. Justice.Gov. Accessed December 29 2016. https://www.justice.gov/crt/help-america-vote-act-2002.
[51] Alvarez, R. Michael, and Thad E. Hall. Electronic elections: The perils and promises of digital democracy. Princeton University Press, 2010.
[52] Verified Voting, “Internet Voting,” accessed on December 30, 2016, https://www.verifiedvoting.org/resources/internet-voting/.
[53] Craig Kanalley, “Voting Online: The Few Places You Can Cast Ballots With The Internet,” Huffington Post, (2012).
[54] Michael D. Pitman, “Ohio to start online voter registration on Jan. 1,” Journal-News, (2016).
[55] Andrew Perrin and Maeve Duggan, “Americans’ Internet Access: 2000–2015,” Pew Research Center, June 15, 2015, http://www.pewinternet.org/2015/06/26/americans-internet-access-2000-2015/
[56] U.S. Election Assistance Commission, “A Survey of Internet Voting,” Washington, DC, Sept. 14, 2011, 14.
[57] Ibid. P.15
[58] Ibid.
[59] Andrew Perrin and Maeve Duggan, “American’s Internet Access: 200–2015,” Pew Research Center, June 15, 2015, http://www.pewinternet.org/2015/06/26/americans-internet-access-2000-2015/.
[60] Rob Weber, “Media treatment of Online Voting: Double-Standard is the standard,” Cyber The Vote!, March 23, 2016.
[61] Smartmatic, “Utah-Republican Presidential Caucus 2016,” accessed online December 30, 2016, https://www.smartmatic.com/case-studies/article/utah-republican-presidential-caucus-2016/.
[62] Quinlan Mitchell, “Could Online Voting Increase Turnout?,” (2014).
[63] U.S. Election Assistance Commission, “A Survey of Internet Voting,” Washington, DC, Sept. 14, 2011, 13.
[64] Ibid.
[65] Smartmatic, “Utah-Republican Presidential Caucus 2016,” accessed online December 30, 2016, https://www.smartmatic.com/case-studies/article/utah-republican-presidential-caucus-2016/.
[66] Centers for Disease Control, “CDC: 53 million adults in the US live with a disability,” accessed on December 30, 2016, https://www.cdc.gov/media/releases/2015/p0730-us-disability.html .
[67] U.S. Vote Foundation, “The Future of Voting, End-to-End Verifiable Internet Voting, Specification and Feasibility Assessment Study,” July 2015, p. 17.
[68] D’Vera Cohn, “More voters will have access to non-English ballots in the next election cycle.” Pew Research. (2016).
[69] U.S. Vote Foundation, ““The Future of Voting, End-to-End Verifiable Internet Voting, Specification and Feasibility Assessment Study,” July 2015, p. 102.
[70] Smartmatic, accessed online December 30, 2016, https://www.smartmatic.com/voting/.
[71] Ibid.
[72] Peter Hayes, “Online Voting: Rewards and Risks,” Atlantic Council, 2014, p.5.
[73] John Myers, “This is why it takes so long to count votes in California,” Los Angeles Times, November 20, 2016.
[74] The Future of Voting: End-to-End Verifiable Internet Voting
[75] ibid
[76] ibid
[77] While the benefits will be discussed in full shortly, it is important to understand for the purpose of this argument that they are more compelling than a convenient alternative to current solutions. It is the opinion of the authors that mere convenience for the voter who will vote regardless of ease does not warrant the risks associated with online voting.
[78] Eason, Brian, “Colorado crushes national voter turnout figures, but mail ballots aren’t a huge factor,” Denver Post, November 20, 2016, accessed December 23, 2016.
[79] ibid
