Wednesday, May 30, 2012

Comparative Analysis of Voting Systems (2)


Regarding transparency, both perception and reality are of equal importance. Election results should perfectly match the electorate’s expressed preferences, and also be perceived as tamper-free by all constituents. 

The openness to audits, before, during, and after the election will greatly enhance the perceived transparency of any system. Voting systems that provide a dual record of voting, such as DREs (if equipped with voter verified paper audit trail) and OS systems (if they are precinct count), posses quite an advantage over other voting systems. 

A DRE system with Voter Verified Paper Audit Trail (VVPAT) maintains an electronic record, and a paper record of the vote, each recorded at the time that the vote is cast. Random recounts can verify that both totals match each other. In case of an electronic attack on the system, the paper record would be available; and if the paper records were damaged or altered, the electronic record would still be available. This level of redundancy is key to generate a positive perception. 

Likewise, a precinct-based OS system contains an electronic record, recorded when the voter scans the ballot, as well as the paper ballot. Two independent records made at the same time that preclude any possibility of tampering. 

By comparison, an optical scanner system where the ballots are sent to a central site for scanning, or manual paper-based system are all single record approaches. A single record environment does not ensure the same level of voter confidence that a dual record does. 

It is important to mention that advanced systems such as Smartmatic’s Automated Electoral Solution, offer up to seven instances of vote verification. This level of transparency enhances the credibility of the results obtained. 


An accurate technology should truthfully capture, record and count the voter’s intent. 

As we learned from the infamous “butterfly ballot” used in Palm Beach County, Florida (as well as a number of other jurisdictions), during the 2000 Presidential Election, a poorly designed ballot can prevent some voters from accurately recording their intended selection on a ballot. Patterns of undervotes or spikes in voting for obscure, third party candidates provide compelling evidence of votes gone astray. 

A well-designed DRE system avoids these problems entirely. Overvoting is impossible, as the machine will not accept more selections than are valid for a particular contest. Undervoting protections are in place, as the voter is notified of any contests where no selections were made. Illegible write-ins are avoided, as selections are typed into an electronic keyboard. A summary can be presented to the voter to allow a final review and confirmation of choices. 

A precinct-based optical scan system has many of these same benefits. Over votes are rejected, and the voter is required to fill out a new ballot; under votes are identified to the voter. Traditionally OS have had a unacceptable rate of error (ranging from 5% to 15%) reading ballots. The accuracy level of OS is definitely lower that DREs. 

Manual systems are in a serious disadvantage when it comes to accuracy. Humans are bound to make mistakes, and manual systems greatly depend on humans on key stages of the voting process such as tallying, and transmitting results.

Tuesday, May 29, 2012

Comparative Analysis of Different Voting Systems

In our previous post (Basic criteria for choosing voting technology) we covered the five basic principles to consider when choosing a voting system. We will now analyze Manual Voting (MV), Optical Scanners (OS), and Direct-Recording Electronic Voting Machines (DRE), according to the first two principles described before: Accessibility and Security.


The strongest supporters of MV (manual voting and manual counting) proclaim accessibility as one of its most alluring advantages. For the average voter, manual voting is simple and intuitive. It is actually a familiar system that most of us have used in the past to either elect a high school student council member, a graduation song, a advisory board member, etc. 

In theory, OS (manual voting but automated tallying) are also fairly easy to use. The voter marks his/her selection to later feed the ballot into the machine. However, when the election requires different ballots for many posts to be elected, this simplicity fades away. Furthermore, reading errors often reach a problematic 5-15% on Election Day. In this sense, experience demonstrates that OS can be frustrating for the voter, and also for operators and electoral authorities, as they will be held responsible for a considerable number of null votes. 

With the new technological developments, such as touch screens and voice sensitive devices, voting with a DRE system (electronic voting and automated tallying) has evolved to become user friendly. Modern voting machines simplify interaction by providing voters with feedback messages, visual or auditory help, and a confirmation of their choice in any preferred language. It is important to mention that this is the only voting method that allows correcting a wrongly marked selection. 

In recent years, an urge to grant access to a hassle-free voting experience to individuals with special needs has prompted electoral authorities to set new guidelines for voting systems. For instance, the federal law Help America Vote Act (HAVA), established in 2002 new conditions to give citizens with disabilities, including the blind and visually impaired, equal opportunities to participate in elections. 

Electoral authorities have had a difficult time trying to cope with these new standards while complying with universal mandates such as secrecy. For example, assisting an impaired individual can compromise the secrecy of the vote and independence of the voter. MV and OS systems offer little options to improve the voting experience of these citizens. Through the use of touch screens, friendly interfaces, visual and auditory assistance, DRE machines developers have taken ease of use to a whole new level. Their more customizable interface has allowed them to gain an edge over the other technologies in this regard. 


Many of the traditional problems faced in manual elections revolve around security issues. The high levels of human intervention necessary in sensitive areas, such as tallying and transmission of results, and the always latent incentive to alter results, are the root of unintended and malicious errors. Pre marking ballots, or ballot stuffing (the illegal act of one person submitting more ballots than allowed) are some of the usual problems encountered with manual voting. 

In terms of Security, Optical Scanners have had an advantage over MV as they diminish human interaction by substituting human counting and the transmission of results. However, the process of transmitting results is usually made through methods that do not guarantee 100% the safety of the data. Moreover, the typical irregularities often found with paper ballots are present with this technology. 

DRE voting system providers have in Security one of the most compelling selling attributes. By minimizing human interaction with electoral material, increasing levels of auditability, and recording, processing and transmitting the information through safe protocols, a well designed DRE machine impedes any alteration of the voter’s intent. 

An advanced DRE system provides a paper trail for each vote casted, thus creating a double record (physical and electronic) which enhances the transparency of the process. Some advanced Optical Scanners also offer this redundancy of records.

Friday, May 25, 2012

Who’s Ready for E-Voting?

The increasing implementation of automated voting solutions is nowadays a reality that nations with manual or obsolete electoral systems are evaluating in order to guarantee more transparency and confidence to their citizens. What are the factors or standards a nation has to consider in order to determine if it's ready for adopting or modernizing its voting system?

Robert Krimmer and Ronald Schuster, from the Competence Center for E-voting and Participation in Vienna, have proposed an E-voting Readiness Index. They analyzed several countries under different criteria such as their information society context, legal and political circumstances and e-voting application. Krimmer and Schuster expected to find a higher level of inclination in countries where computer and Internet penetration was high and e-government standards had been developed. Based on these decisive factors, results were ranked in an effort to see which criteria would affect each country’s readiness to adopt e-voting.

However, the study failed to serve as a reflection of reality, as the index became a simple group of figures and information that described various aspects of each country’s current situation but had little relation with the actual status of its e-voting experience. For example, even though Venezuela scored low in Krimmer and Schuster’s study, the country’s success in implementing e-voting has been enormous. It has actually set an example for the rest of the region in terms of electoral technology adoption. In Venezuela, both political forces have won elections using the same electoral technology since 2004. Just in February 2012, the Coalition for Unity (formed by opposition parties) held its primary elections with the automated voting system, and the event was proclaimed a definite success. If the score of Venezuela was as low as the researchers suggested in their study, the electoral authority wouldn’t exhibit the 73% of approval and trust it has, according to the respected local polling firm DatanĂ¡lisis (2012). Conscious of the study’s limitations, the researchers concluded that “future work will concentrate on finding significant relationships between contextual factors and the successful deployment of E-Voting.”

We could reckon that there are no formulas to determine which nations are ready to implement a transformation in their systems. A clear willingness to change is the first of several important factors that should be considered. A thorough review of the electoral legislative framework, an evaluation of providers able to adapt to each country’s needs, and clean bidding process to select them, are some of the key aspects that will guarantee a successful implementation of voting technology.

Sunday, May 20, 2012

Irregularities in Zimbabwe expose the vulnerabilities of manual voting systems

This video is a vivid testimony of the numerous problems manual voting often encounters. The picturesque scene depicts an election official communicating voters that they had run out of ballots, and if voting were to be resumed, it would be done by reusing those already cast. Astonished voters expressed their anger, frustration, and determination to vote. 

Incidents such as this one at the local Zanu-PF elections sparked fears of even worse irregularities to come during the future presidential elections to be held this year or next. The worst part is that this was not an isolated incident. In other regions of the country, potential voters were screened for their political affiliation before they were allowed to register to vote.

Zimbabwe is yet another clear example of the need to transition from manual into an integral electronic voting solution. A fully automated and auditable electronic voting system, which includes biometric identification to access the precinct and activate the voting session, can eliminate these types of inconveniences and provide the levels of transparency modern democracies deserve. 

In a world where all the conditions to guarantee the success of an electoral process can be perfectly covered if the right system and provider are selected, “Not enough paper ballots” is an unacceptable excuse for any electoral authority. It’s their duty to provide an electoral kit that covers all citizens expected to exert their right to suffrage.

Monday, May 7, 2012

Basic criteria for choosing voting technology

Throughout the history of elections, a plethora of technologies have been used to register the will of voters. Pencil and paper, lever machines, punch cards, optical scanners, touch screen voting machines, internet voting, are some of the alternatives electoral authorities have at their disposal to identify the people's chosen representatives and to tally their expressed preferences.

With all these options in hand, and the single most important act of democracy at stake, the process of selecting the appropriate alternative to implement becomes crucial.

Electoral authorities need to consider the following five basic principles in order to make a wise and well-informed decision:

1. Accessibility: The voting process should enfranchise all eligible voters and be operated and maintained without the assistance of any specialist. It must be easy to use, and allow people with special needs or disabilities exercise their right to vote.

A voting system must also comply with the accessibility requirements of authorities and operators. The whole voting procedure should be designed as to allow any citizen, once given a minimal training, to operate it and to guide voters.

2. Security: A reliable voting system cannot allow any vote tampering. The different features comprising it should constitute a bulletproof architecture guaranteeing that the voter’s choice remains unchanged. It has to be resistant to malicious or unintentional errors.

Also, a secure system must guarantee total secrecy. A secret ballot is essential to ensure voter autonomy and also avoid vote buying. 

3. Transparency: A voting technology will be transparent as long as it is understandable, verifiable, and accountable to the electorate. Its openness to audits, before, during and after the election will greatly enhance the perceived transparency of the system. 

If citizens are to abide by election outcomes and the laws created by an elected government, they must believe that the voting system reflects their preferences. 

4. Accuracy: Every vote must be counted equally. That is a universal mandate of any modern constitution. An accurate technology should truthfully capture, record and count the voter’s intent.

5. Cost-efficiency: The financial aspect of the different voting technologies must be carefully reviewed as their nature is quite dissimilar. A thorough cost analysis should include not only the initial investment, but also, other associated expenses such as: labor, transportation, guarding, and storage. In order to grasp an all-encompassing view of the economic implications of using each technology, the analysis must include several elections cycles.

The principles mentioned above constitute the basic criteria to start the challenging process of choosing the most appropriate technology for the cornerstone of any modern democracy, elections.