Recommendations

managing it in government, business & communities
Chapter 10 - Concerns and Solutions on Electronic Voting Systems Adoption
Managing IT in Government, Business & Communities
by Gerry Gingrich (ed) 
Idea Group Publishing 2003
Brought to you by Team-Fly

When the practical experimentations with EVS continue on, researchers keep on studying EVS in their attempt to improve the process, address the consequences, and determine its impact on our political system. In this section, we will examine solutions emerging from this effort and give our recommendations for a smooth evolution to EVS. We believe the rapid development of technology will eventually lead to wide acceptance and adoption of EVS.

Security and Privacy

Privacy is perhaps the foremost reason for security. In this section, we will provide the recommendations for privacy and security following a voting procedure, including registration, ballot casting, and tallying.

Recommendation A: Enhance privacy in voter's identification check, such as using biometric systems combined with password or credit card. Research is required to determine which method would be most effective for improving privacy and security.

Verification of voter identification is critical to ensure that only eligible persons vote, and that each only has a single vote. Thus for EVS usage, before election, randomly created passwords would be assigned to each eligible voter and sent out by mail. Password coupled with biometric security systems, such as thumbprint or eye scanner, could be the effective combination necessary to enhance voter identification for EVS. Thus, some special devices, such as a laser-reader or thumb-scanner, would be required in order to process registration checks. Currently, technology companies are collaborating closely with biological institutions to produce biometric identification systems, some of which have already been used as safeguard management systems in residential buildings.

MasterCard is also taken as an effective alternative to improve security in electronic voting. European countries walk of the rest of the world in this respect, especially Northern European countries with established advanced telecommunication infrastructures. In 2001, Finland conducted a joint project titled Electronic Identification in Finnish Higher Education to investigate possibilities for implementing a smart card-based electronic identification system in institutions of higher education (Linden, Linna, Kivilompolo and Kanner, 2002). The satisfied results boosted the national implementation of this system. All Finnish citizens and foreign permanent residents with reliable identification are qualified to get an electronic identification card containing the holder's photograph, issued by a local police department. With a unique e-number embedded in the microchip, it is a secure network key for all government and many private sector on-line services. This e-card can assist the government in reliably identifying the user before voting. This verification process significantly reduces the possibility of fraud. In addition, the card can be used as an official travel document for Finnish citizens in 19 European countries (Population Registration Center, 2002). In the future, the government even believes that identification checks can be processed from a mobile device such as a cellular phone installed with a special chip (Fineid, 2002). This successful example from the EU may be a good reference for EVS implementation in the U.S.

Recommendation B: Enhance the security during the election process at three key levels: physical infrastructure, network infrastructure, and data protections.

During the election process, between ballot transfer and storage, and between system operation and maintenance, security still constitutes the most important EVS concern. Computer scientists have focused on developing security protocols for the voting process. In this respect, it is thought that security can be enhanced at three levels: physical, network, and database (EzGov, 2000). Physical security involves optimal server infrastructure design and maintenance for reliable data storage service. At this level, hardware malfunction, power loss, and backup plans should be carefully thought out to best prepare for unexpected emergencies (EzGov, 2000). Also, strict system auditing should be institutionalized internally to ensure hardware security. In particular, access to central computers should be guarded by rigid personal identification validation and biometrical systems.

At the network level, a firewall is the first level of protection from external intrusion. Usually, skillful computer hackers can access the server, damage the system, and change data just via IP address. Especially in voting, a robust firewall is essential to protect the rights of voters and prevent possible fraud. Network companies, such as Nortel Network, TriNet, and HP, are actively producing reliable network products. In addition to protecting central voting systems in the government, special security software and anti-virus software would need to be provided by the government Web site for citizens to download onto their own computer before online elections.

In the context of voting, privacy safeguards are intended to ensure that there is no way to link a vote to the voter who created it, and EVS security ensures that only eligible people can vote and that they vote only once. Currently, encryption is perhaps the best choice in this respect.

Computer scientists, in recent years, have continuously developed and improved security protocol. As early as 1981, public key cryptography was created as the first cryptographic voting protocol (Chaum, 1981). In 1991, the "one agency protocol" was published, which guarantees the secure process of ID tag distribution for voter eligibility (Nurmi, Salomaa, and Santean, 1991; Riera, Rifa, and Borrell, 2000). One year later, the more complicated "two agency protocol" was developed and widely utilized, which separates validation and tally processes to enhance security (Fujioka, Okamoto, and Ohta, 1993). In addition, other protocols related to vote-tags are also in development to improve the accuracy and verifiability of the tally (Riera, Rifa, and Borrell, 2000). In past EVS pilot experiments, no security-related problems occurred. So it is conceivable that EVS based on present security protocols would effectively guarantee the security.

The third level is referred to as data security and has associated with it two categories: transmission and storage. As to data transmission, personal and sensitive information, like candidate selection, need encryption processes. Data can be divided into several packages and encrypted before exchange (EzGov, 2000). At the time of this writing, a 128-bit Secure Socket Layer (SSL) is the highest level of security standard over the Web. Until now, it still plays a satisfactory role in protecting data transmission.

With respect to data storage, in order to prevent potential fraud, data should be separated and stored in different files, even on several servers (EzGov, 2000). "Sand Boxing" is a frequently used encryption approach, which separates user name, password and other data into various files. The powerful database system itself can also provide effective defense, such as central/distributed storage, authenticity approval, and multi-authority allocation. Presently, large database vendors actively provide reliable and secure databases. Especially the database software giant, Oracle, never misses a beat to enhance its database management systems. Complex management mechanism and special operating language strengthen the security of their products. Figure 1 illustrates the network connection for Internet-based voting systems.


Figure 1: Network Frame on Internet-based EVS

In order to ensure that EVS is reliable enough to provide voting services, it is necessary to conduct open testing in practice. In particular, federal and local governments should test consistency in a small-scope and non-critical election. Only after extensive testing and monitoring can the vulnerability of EVS be detected and valuable improvements be made. Also, during these pilot studies, voters would have more knowledge about electronic voting, and be better prepared for later usage in formal, larger elections.

Recommendation C: Develop and issue standard and effective laws and policies to protect privacy and security regulations.

Not only should security and privacy be enhanced from the technology side, but also through policy development. Legal preparation is always the initial step before taking any dramatic change, especially in political environments. In general, government policy on voting security guarantees will enhance people's confidence in electronic voting. As a critical political activity that represents the liberty of the whole of society, elections have gained significant attention from the government. For traditional on-site voting, federal and local governments have issued numerous laws to regulate voting activities with severe restrictions, especially for privacy and security issues (Done, 2002). Similarly, in order to create a safe and ordered environment for electronic voting, it is the right time for government to publish corresponding laws about privacy and security in the digital context.

As early as 1999, the Clinton administration has required agencies to put forms online for the top 500 government services, as well as keep privacy policies updated on all the federal agents' Web sites (Hiller and B lagner, 2002). Actually, when the Clinton administration recommended electronic communications and passed the Government Paperwork Elimination Act (GPEA) (Hiller and B langer, 2002), an impending problem became how to effectively protect information processing in online transactions. By referencing certain laws and policies for online security management in the private sector, government can enhance the Computer Matching Act, as well as develop associated standard, precise, and clear privacy statements specific for electronic voting procedures.

Recommendation D: Provide an extended education for citizens to gain trust in governmental policies on privacy and security protection.

Finally, education would play a critical role to help citizens gain more knowledge about online elections. Hiller and B lagner (2002) argue that education on privacy protection and security technology can enhance individual belief in technology effectiveness during electronic voting. In Arizona's online election, an extensive educational outreach program was launched two months before the election (Done, 2002). Detailed information about candidates and the whole voting processes, including when, where, and how to vote online or on-site, were announced repeatedly to the public. These efforts were found to have contributed greatly to the increase in voter participation. As a result, roughly 45 percent of all voters chose to vote online. It proves that effective education on mechanisms and operating processes of EVS can assist the government in gaining voters' trust in this voting revolution, in turn increasing voters' participation by using EVS.

Accessibility

Recommendation A: Research is needed to investigate the real impact of EVS accessibility on election results, as well as the factors associated with accessibility, such as race, income, education and age.

As to the concern about "digital divide," some evidence indicates that demographic differences in technology know-how may be more apparent than real. An alternative perspective is that the racial and ethnic divide may be a transitory phenomenon that will vanish as computer prices continue to drop (Westen, 2000). According to a recent Forrester Research Survey, the percentage of Asian and Latino Americans using computers at home, school, and work is more than that of white Americans (Irvine, 2000). Indeed, race, income, and education are factors in predicting turnout in conventionally held elections as well (Powell and Bingham, 1986; Verba and Nie, 1972; Verba, Sclozman, Brady, and Nie, 1993; Squire, Wolfinger, and Glass, 1987). Therefore, the problem is to determine whether such impact on elections really exists when EVSs are put into practice. In one empirical study on EVS implementation, Solop (2000a) investigated the Democratic primary election in Arizona and found that "racial ravine" did not significantly impact election outcome in an Internet voting environment. Recently, the authors conducted a survey to investigate the factors impacting EVS adoption among hundreds of part-time M.B.A. students in a southern state. The findings confirm the above arguments. This particular group of co-eds from the younger generation, comfortable with computers, did not regard accessibility as a problem for them when using EVS. So more research is needed to examine the true implications of EVS accessibility according to different demographic factors.

Recommendation B: State and local government should continue to experiment and implement multiple EVS channels simultaneously, as well as develop modern and inexpensive devices to alleviate any accessibility concerns. The kiosk voting and on-site voting can be adopted at the same time to reduce accessibility imbalances.

Moreover, an EVS is not confined solely to an Internet-based voting platform. It also includes telephone and touch screen systems as well. In the case of telephone-based voting systems, voter turnout should be greater than with Internet-based voting systems as the percentage of households with telephones is vastly greater than the percentage with computers. Compared with Internet voting systems, telephone-based systems do not require special skills or additional investment. Thus, telephone-based voting systems might be a good option for those lacking exposure to or experience with computers. In most states, such as Louisiana and Pennsylvania, touch screen EVS has already been widely implemented. Though most of them still stay at the level of on-site voting, at least voters are familiar with this mechanism of EVS, which is a good preparation for future remote voting.

Moreover, it is necessary to try the different mechanisms of EVS to see which ones are most effective in increasing participation. In Liverpool, England, Sheffield City Council tried many sorts of odd voting methods in various locations to improve the participation in upcoming national elections (The Economist, 2002). Such methods included the option to vote by mobile phone or the Internet and installing an electronic voting kiosk in close proximity to the World Snooker Championships to assist local fans in voting when they watch sports. It is believed that remote electronic voting could motivate voters' interest and increase voting turnout back from 30 percent (The Economist, 2002).

Besides, more and more new electronic devices are increasingly invented to make the election more convenient. For example, now small mobile devices are developed to facilitate online voting and mobile commerce. University of Wisconsin-Madison has developed EZ access, a trace technology, to apply to various mechanisms of EVS and facilitate the aged and disabled voters in voting (Law and Vanderheiden, 2000). This technology has been practically used in voting services.

The transition to EVS is not an overnight endeavor and requires the support of all voters. It will take years to achieve the final goal. During this slow but smooth transition, in addition to remote voting by different mechanisms of EVS, kiosk and traditional poll-site voting can also be taken as voting access alternatives. Voters need time to know EVS and gain accessibility to it, while they still hold the voting rights by adopting traditional voting methods.

Recommendation C: The government should develop special laws in response to voting process changes in order to prevent possible accessibility discrimination.

As EVS has changed the voting process, special laws are also needed to manage and guarantee citizens' voting rights. The Voting Rights Act, first published in 1965 and amended in 1970 and 1982, prohibits changes in voting procedures to result in racial discrimination. It should be further amended to prevent discrimination related to race, income and education due to accessibility of EVS. The Department of Justice would be responsible for this legal enhancement. Similar to the laws for privacy and security, legal protection for equal accessibility is an essential promise for all the citizens to use EVS. Only after addressing these legal issues can individuals put more trust in EVS.

Recommendation D: Organizations and all levels of government should provide training courses for citizens, especially minority and low-income residents, to increase their ability to use EVS.

A fundamental way to minimize the differences in computer access and skill is through training and education. In April 2000, President Clinton visited several communities that lacked Internet connectivity, which illustrated the willingness of the U.S. government and industry to get low-income communities connected to the Internet (Hunt, 2000; The Council for Excellence in Government, 2000). Funding for educational technology has increased by over 3,000 percent from $23 million in 1994 to $766 million in 2000 (The Council for Excellence in Government, 2000). During Arizona's 2000 online election, a government-sponsored group took laptops equipped with Internet connections to those who lacked Web access during that period (Done, 2002). Moreover, some simple training courses before an election in "technology-poor" communities could effectively increase voter participation. The Florida state government took the necessary steps to implement electronic voting systems (Washington Post, 2002). The government set aside enough money for efforts to teach voters how to use EVS. Besides, if possible, during the move to EVS government could consider providing mobile voting booths or installing computers at certain communities to offer voters with EVS options.

In fact, it is to be expected that with the passage of time, personal computer and Web-based technology will become more and more common. And it turns out to be true. According to a Census Bureau report released on September 6, 2001, (Associated Press, 2001), there has been a dramatic increase in the number of homes connected to the Internet in the U.S. within the past three years. Now over half of the 105 million people over the country have a computer at home (Associated Press, 2001. The differences among the persons with different incomes and races will significantly narrow down. More importantly, among the school-age kids (6 17), almost 90 percent have access to a computer. Those future voters have fewer problems in this aspect. Generally, citizen comfort levels with technology-based voting processes should rise as the role of computers in their daily lives increases. So we can optimistically predict the wide hold and usage of computers and advanced technology. Concerns about accessibility may subside in the near future.

Brought to you by Team-Fly


Managing IT in Government, Business & Communities
Managing IT in Government, Business & Communities
ISBN: 1931777403
EAN: 2147483647
Year: 2003
Pages: 188

flylib.com © 2008-2017.
If you may any questions please contact us: flylib@qtcs.net