Wednesday, May 28, 2014

E-voting scores another triumph in Belgium


Since 1991, Belgium has been experimenting with voting technologies to automate their elections. And last weekend, the stage was set for another e-voting triumph as millions cast their vote electronically in the election with the highest turnout in the EU. A staggering 90% of the eligible voters headed to the polls, while the rest of the union averaged a mere 43%.

During the elections, Belgium elected three parliaments, the regional, the Federal and the European. To do so, over 1,000 polling stations in Brussels and Wallonia were set up using technology developed by Jites. Also, more than 3,000 polling stations across 153 communes in Flanders and Brussels used technology and services provided by Smartmatic (over 17,400 voting machines). In addition, authorities hired Stésud to deploy a system for the digital transmission and recording of results which is called CODI.

All voting systems and software programs were certified by an independent advisory body, PWC (Price Waterhouse Coopers) on the following dates:
Jites: April 4, 2014
Smartmatic: April 4, 2014
CODI (Stésud): April 30, 2014

Due to a change in Belgium´s voting legislation, it was possible to publish partial results of electronic voting. According to the IBZ website, once the results of 10 centers within a district were available, they were reported. Also, thanks to the digital transmission system, the election results were announced earlier this time around.

Some interesting facts about the election:
  • The electoral roll comprised 8,000,458 voters. 7,871,504 residing in Belgium and the rest living abroad.
  • Total number of electronic polling stations: 4,482.
  • Total number of manual polling stations: 6,301.
  • Municipalities took important initiatives to improve accessibility for disabled voters, such as:
            - Provision of priority parking
            - Customized access to the polling station
            - Guidance for using the voting booth
  • Between 6.30 am and 9.30 am, while setting up the polling stations, there were 234 incidents requiring technical support. In 2012 -for the same time interval- there were 463.
  • At 4.00 p.m. all electronic polling stations were closed. The traditional polling stations closed an hour earlier to start the manual count.
  • A red pencil was used for manual voting. Red is considered by authorities to facilitate reading, and deter electoral fraud. 
  • Paper ballots were produced using green energy (wind energy). A high quality paper that incorporates several safety features including a watermark with a Belgian insignia and a slogan was used for the ballots in manual precincts.
  • All results were published in the website http://verkiezingen2014.belgium.be/nl/
  • At 16.10 p.m., the first official results of the election regarding Belgians who voted in embassies abroad were known.
  • European stock markets moved higher Monday as election results rolled in from the European Union and Ukraine.
  • Despite numerous prior testing and certifications by PwC, a minor inconsistency present in the digital transmission and recording system was not identified until e-day. In consequence, the totalization of the preferential votes in several cantons using Jites voting machines was erroneous, causing the temporary suspension of the results proclamations. Stésud corrected the bug after a few hours.

Friday, May 23, 2014

Europe joins the new wave of election automation

Image: Gozo News
2014 is proving to be an interesting year for modern democracies as more nations -seeking to provide greater transparency and efficiency to their elections- take firm steps to modernize their electoral systems.

Earlier this year, Ecuador implemented an electronic voting pilot during its recent elections. India began attaching printers to their machines to make voting in their machines more transparent. Costa Rica, Panama, Pakistan and Slovakia also made important progress to ensure inclusion and build electoral confidence via election automation.

In Europe, despite some setbacks during the last decade in Ireland, the Netherlands and Germany, the scenario has changed radically and electoral automation is regaining strength.

The European Union Parliamentary elections to being held this weekend will be an excellent opportunity for nations to continue to developing e-voting platforms already in place, or to show its advantages over the outdated and unreliable manual systems in electronic voting pilots.

Estonia is leading the way with an I-voting system that has already yielded excellent results in six national elections. I-voting is optional, as voters can head to poll stations and use the traditional paper and pencil. Since its inception in 2005, casting a ballot online has been gaining popularity in Estonia.

Meanwhile, in Belgium, three elections will take place at the same time: Federal, Regional and European parliamentary authorities are to be chosen. Smartmatic technology (including more than 17,400 voting machines) will be used in polling stations across 153 communes in the Flanders (308 municipalities) and Brussels (19 municipalities) regions. Over 1,000 polling stations in Brussels and Wallonia will be using technology by Jites and Digivote. 


There’s also Switzerland, with its system of electronic voting, known as vote électronique, which allows the electorate to participate in elections and referendums on the Internet without having to go to a polling station.

Bulgaria will have the opportunity to test the benefits of a verifiable voting system.

In Bulgaria, the National Electoral Commission of this country has decided to implement a fully automated system: Citizens of Sofia and three other Bulgarian cities will have the option to use the voting technology platform, which includes touchscreen voting machines, software applications and related services.

Tuesday, May 20, 2014

The Estonian I-voting system in a nutshell

An Estonian voter inserts her ID card for authentication before voting
The immense possibilities that the Internet offers to enhance efficiency in election administration has made Internet voting one of the hottest topics in the world of elections today. And it couldn’t be otherwise, in the midst of the cyber revolution we are living in.

Since 2005, Estonia has been at the world-wide forefront of I-voting. It has carried six national elections (three local, two parliamentary, and one for the European Parliament) in which voters have had the option to either cast a ballot at a polling station using paper and pencil, or to vote remotely using an Internet-based platform.

So far, the success of the system has been evident. No significant or credible hacking or fraud accusations have been made, and the number of I-voters continues to grow sharply as the natural barriers to the adoption of any new technology fade away. In the first elections with optional I-voting (2005), 10,000 people cast a vote remotely. By 2011, that number had grown to 140,000– an approximate 24% of the voting population. Authorities are confident that this technology will continue to gain steam in the near future.

In each election, paper and pencil voting begins 13 days prior to Election Day in certain designated early voting polling stations. Voters can also wait until Election Day and cast a ballot in their neighboring precinct. In addition to the traditional method, casting an I-vote is available for a period of 7 days spanning from the 10th day until the 4th day prior to E-day.

The I-voting process has a similar scheme to that of any traditional voting process. Voter authentication, a crucial first step to avoid voter impersonation, can be done through different means: an ID card (National ID card), a Digital ID (a document identifying a person in electronic environments and involving digital signatures), or a mobile-ID (system based on the SIM card of a phone acting as an ID card and a card reader).

Once the identity of the voter has been authenticated, the voter downloads the voting application from the site www.valimised.ee. After the identity of the voter has been also validated via PIN, the voting process begins. The voter proceeds to choose options and then confirms his/her choices. A notice stating that the vote has been accepted is sent to the voter.

In order to continue improving the service, Estonia is working on a system called Verification of the I-votes. The idea is to detect if any malware has affected the computer used by the voter, and if that malware could have compromised or even changed the vote. The system will also allow voters to verify their I-votes with a smart device (mobile phone or a tablet) equipped with a camera and Internet connection. To learn more about the verification, click here.

Security has always been a concern when it comes to electronic voting. In consequence, different security mechanisms have been put in place to safeguard the I-vote. The vote is encrypted with the most advanced encrypting algorithms available. Also, the voter’s personal data is digitally signed and added to the encrypted vote prior to transmission.

Before processing voting results in the evening of Election Day, the encrypted votes and the digital signatures (i.e. data identifying the voter) are separated to guarantee the secrecy of the vote. Then anonymous I-votes are accepted and counted. The system accepts votes only if they are not connected to personal data. 

A frequently asked question when debates on I-voting begin is -How can you guarantee that the person casting a vote at home is doing so freely?

To mitigate voter coercion, voters are allowed to change their electronic vote by voting as many times as they please. Only their last vote is counted. The voter has the option to log into the electronic voting system again while I-voting is available, or to head to a polling station and cast a vote in person on Election Day.

For sure, Estonia will continue to lead the way in online voting for some time. Its government has been committed to better governance through technology for decades, and in view of the results, it has all the incentives to continue doing so. Paperless cabinet meetings, e-voting, e-health, e-schools, are just some of the initiatives giving this small European nation a very relevant position in the cyber revolution.

Thursday, May 15, 2014

Remote voting civil initiative in Slovakia needs 50k signatures

Source: Google Images
When discussing the technology involved in running an election, most conversations turn to voting machines, for example, or they may talk about the infrastructure involved in recording, tallying and transmitting the ballots as part of an election. By large, the vast majority of ballots cast in an election are done in person in some form or another, but it is just as important to consider the voting rights of citizens who are unable to make a physical appearance at an official polling place on Election Day. And this was a hot topic issue leading up to the presidential elections in Slovakia earlier this year.

After no candidate was able to secure a majority in the first round of voting on March 15, a second round of voting was conducted on March 29 in which Andrej Kiska defeated Robert Fico with 59.38% of the popular vote. Part of the problem with this election was that it effectively did not allow for true remote voting. The current Slovak law says that citizens without permanent residence in the country can vote, but they must be present on Election Day. There is no support for remote voting, since this cohort “still isn't numerous enough to make it worthwhile investing such an amount of money”, according to Slovakia's Interior Minister Robert Kaliňák.

To demonstrate the interest and desire of the Slovakian people in Internet voting and support for remote voting for those living abroad, a civic association of Slovaks living abroad has put together a petition. As it stands, expatriate Slovaks can vote via the postal service, but having an Internet-based system could make it easier for voters to engage and participate. The petition seeks to get at least 50,000 signatures before being delivered to the Slovak parliament.

The two main reasons cited by Kaliňák as to why Internet voting is not yet viable in Slovakia are related to cost and security risks. These are issues that have also been raised in other countries around the world, but they have also been suitably addressed by voting experts like William J. Kelleher, Ph.D. He says that most hacking jobs “are the result of human insiders abusing their positions” and not the fault of the voting technology itself.

Indeed, recent expansions in remote voting technology have already been successfully demonstrated in Australia and the Philippines. However, if the petition is able to get the 50,000 signatures it desires, this would provide a clear illustration to the Slovak government that a large contingent of the Slovak people, particularly those living and working abroad, are calling for an updated electoral system where they are better able to cast a remote vote without having to rely on a more archaic postal-based system.

The 2014 Slovak presidential elections have come and gone, but the civil initiative of those Slovak expatriates still have several months to get the signatures they need. The goal is to collect the 50,000 signatures in 12 months, which would give them a soft deadline of January 2015. 

Tuesday, May 6, 2014

The relationship between partisanship and confidence in election system integrity

Source: Google Images
There are many factors that should come into consideration when voters evaluate their level of confidence in how elections are run. The voter should think about the voting technology being used, for instance, and whether the voting machines have been subject to appropriate pre- and post-election audits for reliability and security. The voter should also think about the infrastructure in place for ensuring that ballots are recorded accurately and tabulated without fault. However, should the results of the election have any bearing on whether the election itself was run fairly?

This notion was brought up by the Pew Center on the States in a February 2012 report titled Election Administration by the Numbers: An Analysis of Available Datasets and How to Use Them. In it, they found that political partisanship appears to have a link to voters' confidence that the ballots are being counted correctly and as intended.

More specifically, the report found data showing “that many voters have more confidence in the election system's integrity when their candidate has won.” This could have to do with the psychological effect that winning and losing may have. If their preferred candidate wins, the voter is then pleased with the election result and thus may also have positive feelings toward the election infrastructure. Conversely, if the preferred candidate loses, the voter is then displeased with the election results and may be – consciously or unconsciously – seeking reasons why the results didn't go the way they had hoped. One of these reasons, in their mind, is a possible lack of integrity in the election system.

Of course, the legitimacy of a government is inherently tied to the confidence that voters have in election results. If the electorate feels that an election is rigged or otherwise compromised, they may feel that the elected government is there unjustly. As the administrative bodies in charge of running an election should be unbiased and objective, they should have no direct connection whatsoever with the party that ends up winning the election. Even so, voter perception can be very powerful.

The Pew Center report looks specifically at the American general elections of 2004, 2006, 2008 and 2010. When George W. Bush, a Republican, won the presidency in 2004, 81.5 percent of Republicans were “very confident” that their vote was counted accurately, but this number fell to 74.2 percent in the 2008 election that saw Democrat Barack Obama win the presidency. Conversely, only 56.9 percent of Democrats were “very confident” that their vote was counted correctly in 2004, rising to 74.8 percent in the 2008 election.

Interestingly, the level of confidence in regards to the nationwide ballots as a whole was recorded as being lower by voters of both parties when compared to the confidence that the voters had in the accurate tabulation of their own votes. As an aggregate, the confidence of voters in the United States that the election as a whole was conducted fairly has hovered at around 50 percent. That's very low, especially when compared to the 89 percent of those polled in Denmark. However, the same partisan effect demonstrated on an individual voter basis also persisted on national perception.

A mere 20.7 percent of Democrats were “very confident” that votes across the country were tabulated accurately in the 2004 election, compared to 75.0 percent of Republicans. When Barack Obama won in 2008, the trend reversed: 53.9 percent of Democrats were very confident, whereas only around 30% of Republicans were very confident in the accurate counting of votes in the 2008 presidential election.

Even though partisanship should not play a role in determining how confident a voter is that ballots are being counted accurately, it is clear that political allegiances do have an impact. Perhaps a greater move toward using even more impartial third parties to administer elections is the solution to this problem, helping to elevate voter confidence with better audits and greater transparency in the electoral system.