News and Updates
Interview/Speech/Testimony

Internet Voting: Past, Present and Future

Published

As technology advances and more transactions become electronic, many wonder when voting will truly enter the digital age. There are many issues to consider when it comes to exercising the right to vote through a computer. This includes building trust in the process, preserving secrecy of the ballot and ensuring citizens are not coerced or intimidated when using technology to vote.

Ben Goldsmith, Election Technology Adviser and former IFES Chief of Party in Kosovo and Pakistan, has helped and advised a number of countries, most recently Norway, in implementing election technology. He answers some questions about the challenges, advantages and evolution of this practice.

To begin, how do you define Internet voting, and is there a difference between electronic voting and Internet voting?

The terminology used to describe voting and counting technologies is not authoritatively defined; similar phrases, like “e-voting,” are used inconsistently by different organizations and experts.

I define electronic voting as the use of electronic means to mark a ballot paper. Normally, this will be an electronic ballot paper that is then also counted electronically. A typical electronic voting machine in a polling station would work in this way, with the voting machine recording the ballot choices and producing results at the end of polling. However, there are some electronic voting solutions that produce a paper ballot, which can then be counted either by hand or electronically.

Internet voting is a form of electronic voting and involves casting a ballot through the Internet. Normally this involves logging on to a website through any computer with access to the Internet, including your home PC. Votes are not stored in the machine used to cast the vote, but on a central vote server that tallies votes at the end of polling. However, Internet voting can also be conducted using voting kiosks. These kiosks contain computers installed in public places for the purpose of Internet voting, but are not supervised by electoral officials. Computers can also be installed in the polling station to allow Internet voting.

When was Internet voting first explored? How many countries have tried Internet voting and how many use it regularly?

Internet voting was first used for binding political elections in 2000 in the U.S. in a pilot across several states targeting overseas voters. Since then, 13 more countries have used Internet voting. Two use Internet voting nationwide (Estonia and the United Arab Emirates); five use Internet voting in some parts of the country or for certain members of the electorate (Australia, Canada, France, Mexico and Switzerland); two have ongoing pilots (India and Norway); three have piloted Internet voting and decided not to continue its use (Finland, the UK and the U.S.); and two adopted Internet voting, but decided to discontinue it (Netherlands and Spain).

Countries that use Internet voting tend to target specific categories of voters – for example, expatriate voters, military personnel posted overseas, absent voters or voters with disabilities.

What have we learned from these countries’ experiences?

Despite the Internet playing an increasing role in the lives of people around the world, there has been  significant caution about the adoption of Internet voting. Some have argued that we trust the Internet for banking transactions and to make online purchases, so we should trust it to cast ballots. However, these online transactions have a level of transparency, which means you can check they are accurate. The need for ballot secrecy means transparency and auditability in Internet voting is difficult to implement.

This is illustrative of the biggest lesson we can learn about Internet voting: this system has much to offer, but it brings significant challenges that need to be properly addressed if it is to positively impact the electoral process. The auditability of the system is one of these challenges, and this inherent lack of transparency has implications for the trust placed in the system. Other significant challenges include: ensuring that voters can cast their votes over the Internet free of coercion and undue influence; properly identifying online voters; ensuring that only the registered voter can cast his or her ballot; protecting Internet voting from security breaches; and, protecting the secrecy of stored vote data.

How can the secrecy of the ballot be maintained, to ensure that coercion and intimidation of voters does not occur?

Casting a ballot in the polling station is meant to ensure the voter can freely express their voting preferences. As soon as you take the act of voting outside of the secure environment of the polling station, this privacy cannot be ensured. Some see this as unacceptable, and a violation of international electoral standards.

However, the comparison is often made between postal voting and Internet voting, with many countries accepting postal voting. Furthermore, with Internet voting additional measures can be provided to protect the secrecy of the ballot. Several countries now allow repeat voting over the Internet; voters are able to cast their votes online as many times as they wish. Only the last vote cast is counted. Another mechanism to protect against voter coercion is to allow Internet voters to cast a paper ballot in a polling station and to guarantee that any paper ballot cast will be counted, invalidating the voter’s Internet ballot. Both measures mean that anyone coercing a voter while casting a ballot over the Internet cannot be sure the coerced vote will be the one counted.

What kind of protection can be offered against hackers intending to manipulate the vote?

Encryption and digital signatures can be used to ensure only legitimate data is received by the vote server and that this data has not been altered while being transmitted through the Internet. While much can be done to protect against manipulation of data stored on the vote server through firewalls, intrusion detection and transaction logs, there is little that can be done to ensure that every PC is protected against malicious code that will change a voter’s ballot choices before they are submitted.

Ultimately, the best protection against this possible manipulation is to ensure Internet voting systems have what is called “end-to-end auditability.” This means that every stage of the process can be checked to guarantee that votes are cast as intended, votes are stored as cast and votes are counted as stored. With this level of auditability, any manipulation of votes could be identified. For Internet voting, this end-to-end auditability is complex to establish, but not impossible.

How can you ensure that it is the registered voter that is actually the one casting the ballot?

The best way to do this is to use established and secure online authentication mechanisms. Some countries already have such mechanisms, which are used for services such as online banking and tax returns. In Estonia, for example, Internet voting requires voters to authenticate themselves using their national ID card. This ID card has a chip that is read by a card reader attached to the computer; the voter’s identity is verified through their smart card PIN numbers. Because this card is used to access other government services, most voters would be reluctant to provide these authentication credentials to anyone else.

What are the advantages associated with Internet voting?

Despite the challenges discussed, Internet voting has a lot to offer. With increasing levels of access to the Internet, the most obvious benefit is the ease of access to the voting process. This is especially important in the context of declining turnout in elections in many democracies around the world. By decreasing barriers to participation, it is hoped Internet voting can slow or reverse this decline. Voters who are overseas, military personnel, emergency service personnel who work on Election Day and voters with disabilities can especially benefit from the increased access.

Internet voting can also eliminate significant logistical challenges associated with paper-based elections. The removal of these logistical challenges can reduce the costs of holding elections, although new costs will also arise.

Given the above discussion, should countries move towards the adoption of Internet voting?

There are many challenges to be dealt with in implementing Internet voting, but also many benefits that could be achieved. There is no right or wrong answer as to whether a country should adopt Internet voting. The advantages and disadvantages will need to be carefully weighed up, electoral stakeholders will need to be consulted and the specific electoral environment assessed to determine whether Internet voting is appropriate and feasible. It is also very important for countries which wish to adopt Internet voting to conduct pilot projects to ensure that their expectations of Internet voting and their anticipated benefits can be achieved, and the challenges can be properly dealt with.