Best practices in the use of ICTs in elections are still developing. However, drawing on recent experience of election practitioners, the ACE project identifies several guiding principles for the adoption and use of ICTs in an electoral context.21 When an EMB considers introducing and procuring technology, these principles provide a useful starting point. They can help to establish and maintain public confidence in the electoral process.
These guiding principles are:
Technology is just a tool and therefore it is not an end in itself. When technology is applied to any activity related to electoral administration and elections, it is important to consider carefully the electoral context in which the technology is used. For example, if technology is being considered for vote counting, the guiding principles that apply to vote counting also apply to the technology.
It is also good to keep in mind that electoral processes involve more than just operations and technology. They include relevant laws, regulations and guidelines, social and political contexts, the organisational culture of the body implementing the technological solution, the procedures developed to complement the technology and the training of technicians and users.
When a new system is being considered to replace another, an evaluation is needed to assess the impact of the change on all stakeholders. Once the system is adopted, the transition phase needs to be carefully managed to ensure that problems do not occur and that all functions can continue to be effectively carried out. Some strategies that can be used by electoral administrators to minimize the impact of new systems, new technology and changes on electoral processes are:
Transparency – meaning openness and accountability – is a key feature for the credibility of democratic elections. An indicator of transparency can be the access afforded to electoral observers, both domestic and international, to all procedures at every stage of the electoral process.
With manual processes, transparency is relatively straightforward, as the processes are usually visible and it is not difficult to provide meaningful access to observers. By contrast, with the use of some technologies it may be more difficult or even impossible for observers to testify that the outcome is correct. Electronic voting systems are one example where in some cases it can be very difficult to ensure that the vote that is registered is indeed the vote cast by the voter and, therefore, that the resulting vote count is accurate. On the other hand, the use of technology may enhance transparency once the data entry into the respective system is proved to be accurate and large quantities of data can produce meaningful reports with very few errors.
It is useful to compare “micro” and “macro” perspectives when analysing the transparency of manual and electronic systems. With manual processes, transparency is relatively straightforward at a micro level. For example, observers or party agents can easily observe operations in a paper-based voting station, and compare results in that station to those published centrally. However, a macro-level perspective is more difficult. Results are not immediately available and there are no assurances that unobserved voting stations that haven’t been observed directly are presenting accurate results.
In contrast, electronic systems can facilitate transparency at the macro-level. For example, electronic systems can allow for results to be easily generated almost immediately after polling stations close, with results for the entire country tabulated and presented at different regional levels all the way down to individual stations. But one downside of electronic systems is that an observer may not always be confident that the vote cast by the voter is indeed the resultant vote counted at the end of the process.
If a random sample of the equipment is tested and found to be accurate, however, then an observer can reasonably assume that the rest should behave similarly. Whether an electronic system is observed or not it will behave the same way. By contrast some argue that a manual process may be affected by whether or not it is observed. For example, voting station staff may pay more attention to adhering to rules and regulations and accurately counting the votes while they are being observed, and less so if left to their own devices. Knowing that a small percentage of voting stations observed operated correctly does not necessarily extrapolate to the remaining percentage of unobserved stations.
With the use of technologies, transparency may be achieved in different ways depending on the technology. This may involve the use of techniques, such as the creation of audit trails, creation of log files, code verifications, digital signatures and compilation checks, among others. As a result, the skills needed by observers of electoral processes using various technologies, may be completely different than those needed to observe manual processes. To ensure transparency, the election management body may seek the assistance of specialized experts or auditors able to verify the accuracy of their systems and to provide special training to observers.
In addition, auditing the implementation of technology is required. External audits are recommended. Regarding issues of transparency and trust, EMBs are expected to follow appropriate ethics when implementing new technology. For instance, when choosing technology suppliers, EMBs should ensure that the tendering process is fair and open, without favouritism or corruption, and that all government purchasing procedures applicable to the selection of technology are followed.
When buying hardware and software, EMBs should ensure that proper licenses are obtained. Apart from the legal and ethical issues with using unlicensed or unregistered software, users also run the risk of not being notified of known bugs, software fixes or upgrades. Antivirus software should be used for laptops and desktop computers. A complication is that most antivirus software programmes will not update their virus definition files without an internet connection. This poses a serious problem for laptopbased kits deployed to remote regions without internet access.
Computer systems used for elections must include high levels of security. Unauthorised persons must be prevented from accessing, altering or downloading sensitive electoral data. Demonstrable security levels are another way of ensuring that election IT systems are transparent and trustworthy.
Various mechanisms exist to provide for computer security. These include password protection, encryption, verification programmes and physical isolation.
Accuracy is vital to the credibility of an election, particularly for voters’ lists, vote recording, vote counting and the reporting of election results. To ensure that technological systems are trustworthy, there must be ways to test and verify that data is recorded properly and that the manipulation of this data produces accurate outputs.
System accuracy may be tested by randomly entering known data into the system and verifying that the resulting outputs are correct. In addition, for voting systems, a test should be performed to verify that the same set of data processed through several randomly chosen, similar but independent systems produces the same results.
On the whole, it should be expected that accuracy increases as new and improved technologies are adopted.
Electoral computer systems often contain sensitive personal data for large numbers of individuals, which can include names, addresses and other personal details. Many countries have privacy laws and policies that place restrictions on access to personal information, under the general principle that every person has the right to personal privacy. If a person’s privacy is invaded by improper use of electoral data, both the person and the respect accorded to the electoral process may suffer.
Accordingly, security features in electoral computer systems should take account of the need to protect the privacy of personal data. Audit trails can be built into systems containing personal data to track and monitor which individuals have accessed or modified personal data and to prevent any unauthorised invasion of privacy.
Conflicts arise when an electoral IT system needs to ensure that a person that voted is a citizen, has the right to vote, and did not vote more than once. To ensure this security, the system should know the identity of the voter. But the system should not, however, know how the voter voted.
There are solutions for these issues, but EMB’s and voters need to be educated on these solutions. EMBs need proof that the electoral IT system ensures a valid voting process, and the voter needs to be ensured of his voting privacy.
Elections are largely about providing a service for its clients, the voters. When choosing a new technology, the level of client service, including the ”ease of use and voters’ satisfaction”, as well as how the technology is perceived by its users’, needs to be a priority consideration for the EMB in order to ensure that the credibility of the electoral process is not jeopardized.
There are a number of strategies that can be adopted to reduce users’ unrealistic expectations or suspicion, and in order to ensure trust. There should be:
Where a technology with impact on internal and/or external users is being adopted, it is advisable to organize a consultation process with those users or their representatives to ensure that their needs are met and that they are satisfied that the new system is acceptable and reliable. It is important to provide sufficient information to users to enable them to feel included in the process and therefore increase the likelihood that the new technology will be successfully implemented.
Since election technology has the potential to directly affect the political process, it is important to engender a sense of ownership in its users, much more so than might be considered for other government systems.
In addition, access and equity considerations should be taken into account when adopting new technology to ensure that people with special needs are included.
Therefore, before making a commitment to implement new technology, it is important to determine whether it is financially viable, by getting a complete account of all likely costs and savings, namely:
In addition, carrying out the following studies might also prove useful:
These considerations and the results of these studies can help to identify additional functions, benefits and user satisfaction to be provided by the technology under consideration. These considerations can also help to provide a realistic estimate of the expected costs or savings needed for a thorough evaluation of the proposed technology’s cost/efficiency ratio. This, in turn, can help determine whether the technology is worthwhile, viable and affordable.
Relatively minor improvements may not justify the extra cost and by the same token, if a low-cost, low-tech solution is available and it is acceptable for the specific task, there may not be a need to implement a proposed high-cost, high-tech solution. Whenever possible, the expected life-span of the technology should also be determined. In principle, a technology that can be re-used for more than one purpose or for more than one election is more cost-effective than one that can only be used once before it needs to be replaced.
Although it is often expected that a new technology is more efficient than the old one, it might not be the case in practice. It is possible to replace an efficient manual process with an inefficient automated one if the new process is not carefully thought through. When deciding to implement a new technology, it is important to ensure that the new system is more efficient than the one it replaces.
One of the anticipated benefits of technology is the ability to perform accurately large complex tasks in a relatively short time compared to the time needed to undertake the tasks manually. While this may enable EMBs to reduce costs by deploying fewer staff, it may also lead to additional pressure on staff as productivity rises and more is expected of them.
Evaluating sustainability is one of the most important criteria in the adoption of new technologies. In his 2009 report on elections, the Secretary General expressed his concern that some of the world’s poorest countries are adopting some of the most expensive technology for elections. Technology is generally expensive to acquire, but if it is sustainable, and able to be used for more than one purpose or more than one election, it can be more cost-effective in the long term.
New technology may also require the use or availability of other technologies or specific infrastructures, such as communications and electrical energy networks. For example, if a country has an intermittent power supply, implementing a network of personal computers may not be the best choice. High humidity or high levels of dust or sand may also restrict the choice of appropriate technology.
In general, local infrastructure must be capable of supporting a chosen technology. If it is not, it could mean that the technology being considered is not appropriate or that there is a need to upgrade the infrastructure, in which case additional costs and benefits should be carefully assessed.
There is also a need to evaluate the technical and financial capacity of the electoral management body (EMB) to maintain a new system during an agreed upon or required time span. Usually, it is desirable for an adopted technology to be used for several years; therefore funds will need to be secured, not only for the initial implementation, but also for the foreseeable lifespan of the technology. These funds need to be budgeted for and guaranteed before committing to a new technology.
Another aspect related to the sustainability of a new system is the availability of appropriately skilled staff to run and maintain it, either by the EMB or by outside service providers. The costs associated with keeping staff skills up-to-date and the capacity of both EMB and service providers to retain trained staff needs to also be considered.
Another potential pitfall is attempting to do too much too quickly. Instead of implementing a radically new technology (compared to the existing local technology) in one step, it may be more appropriate to gradually implement change over a series of electoral events.
A technology that is flexible is also more likely to be cost-effective and sustainable. It may be desirable to adopt a technology that has the potential for more than one use if this does not compromise the primary need for the technology.
In cases of transitional environments, where the future of election management methods is uncertain, it is desirable to avoid purchasing technology that is highly specialised and inflexible, and to purchase instead more generic products that can facilitate the use of a wide range of applications.
Another of the perceived advantages of using technology for elections is its ability to provide users with greater flexibility and more options. Therefore, procedures should be implemented to ensure that the new technology does not reduce flexibility by limiting the number of options available.
EMBs, Government agencies, donors and UNDP need to ensure that the product of implementation of high-tech IT systems – the databases that contain the registration details of potentially millions of citizens, remain under the control of the EMB and/or Government bodies at all times, and are accessible indefinitely regardless of the interfaces, software platforms, encryption methods or other hosting systems put in place by the vendors/suppliers of the system. EMBs and their partners should never remain permanently tied to one supplier. Scope and opportunity should always exist for the EMBs/Government agencies to “move” supplier in the future, if there is a day to day management component built into the supply contract, should the management and day to day operation of the system lead to implementation difficulties that may lead the national authorities to consider switching the managers of the system. In any case, high-tech solutions, no matter how “cutting edge” today, will eventually require upgrading, and possible re-tendering, at some point in the future.
Legal advice should be employed at the outset, including at all stages of the procurement process and particularly upon contract signing, to ensure that the national authorities have the right to eventually consider “taking” the file databases in the future to a new system manager, without having to request of citizens to return to register all over again, under the system of a new supplier.
