I. In Focus This Week
Working with observers to improve voting system reliability
Election officials must work more closely with election observers
By David Levine
Special to electionlineWeekly
The goals of election observation are enhanced public confidence in the efficiency and integrity of the election process, and more efficient election operations.
Any voting system -- whether an optical scan paper ballot system, a Direct Recording Electronic (DRE) system, or something else -- has to function for the duration of the voting process, and election observers are trained to spot situations in which malfunctions, power outages, lengthy-set up times or other problems prevent voters from casting their votes, discourage them from doing do so, or cause votes already cast to be lost.
Election observers look at how voting machines protect against malfunctions, whether election officials can easily repair basic problems, and whether officials have been trained to deal with problems that arise during the voting process.
This means that one way to improve the reliability of voting systems in the United States is to have election officials work more closely with election observers.
This article discusses three ways to help achieve that goal -- providing guidance to election officials to ensure that all voting systems are observable; improving observer understanding of voting systems; and utilizing observers to evaluate (and improve) voting system reliability.
Taken together, these steps would improve coordination between observers and election officials and give the latter useful feedback on how their systems are working and how to fix flaws in them.
Ensuring that a voting system is observable
As with any electoral process, an integral part of voting system assessment is transparency, which is crucial to verifying (and assuring the integrity of) the electoral process and building public confidence.
Just observing voters and officials operating electronic voting and counting technologies is not meaningful. Observers need additional access to be confident that an election is proceeding properly. They can’t interfere in the process, but they should have full access to documentation about the system, including certification and testing reports.
And observers should not have to sign non-disclosure agreements to gain access to documentation or observe processes -- that jeopardizes their ability to report what they find, and contradicts the whole goal of transparency.
One way to make a voting system observable is to open its underlying processes to observation. Not all aspects of a voting system can be observed (e.g., a component or process may be protected by law from disclosure), but many can be observed and should be open to observers. Examples include election administrator and vendor deployment, set-up, and modifications, and certification, testing and audit activity.
A second way to ensure that a voting system is observable is to facilitate observer access, e.g., by offering observers the opportunity to test a system independently. While opportunities for external testing may have to be limited due to security, logistical and time constraints, allowing such testing is evidence of transparency.
As noted earlier, a third way to ensure that a voting system is observable is by making documentation accessible to observers. This includes identifying unavailable information and explaining why it is not available. Full documentation does not prove the reliability of a voting system, but its absence can indicate problems. The absence of documentation addressing known technical problems, for example, is more telling than the availability of other system documentation.
A fourth way to ensure that a voting system is observable is to make its source code publicly available. Election administration authorities should indicate whether source code has been checked/verified by independent group(s), and indicate how observers can verify that the identified source code is actually used on Election Day. Election observers rarely have the time or capacity to audit/validate source code, but they can determine if others have done a meaningful assessment and evaluate their conclusions.
Finally, a voting system can be made observable by printing result protocols and making them available at the polling place, the jurisdiction’s office, and the state election office, so that the observers can perform the classic audit task of checking the results at polling stations against those that were centrally recorded.
Some of the tasks outlined above call for advance preparation by election officials, but the work required pales in comparison to the downside of denying full access – inaccurate observer conclusions, bad press, and tension between election officials and observers, to name a few.
Improving observer understanding of voting systems
Most US jurisdictions now use electronic voting systems – either DREs or Optical Scan systems. But some continue to use, or have returned to, paper-based voting. And a few have embraced alternatives like mail or Internet voting.
Technologically advanced voting systems pose challenges to election observation. For example, as noted above technology can complicate direct physical observation of some processes.
Another challenge is that advanced voting technologies made not be well understood by the typical observer. Some election observer organizations, like the Organization for Security and Cooperation in Europe (OSCE)/Office for Democratic Institutions and Human Rights (ODIHR), have put out guidance on how to observe newer technologies. But election observers would benefit from current guidance about voting systems from the people who built, certified, purchased, and implemented them.
There are several ways to provide that guidance. Election officials can offer observer training and explain how their systems are protected against foreseeable malfunctions and how they are training workers to deal with problems as they arise.
Election officials can also put together written guidance in the form of Frequently Asked Questions (FAQs) that address issues like
- how a voting system is designed and configured to be user-friendly -- that the physical design of the system helps citizens vote, that the system gives clear feedback and prompts during a voter’s interaction with it, and that it addresses under-votes, over-votes, termination or interruption of the voting process before a ballot is completed, and blank and invalid ballots.
- whether all contestants are presented fairly on the ballot;
- what facilities that have been incorporated to increase access for voters with disabilities or difficulty with English; and
- how/whether a system can be repaired quickly if it malfunctions.
In addition to (or instead of) FAQs, election officials can put together a video that offers the same information.
And here’s a plus – the tools that help observers do their jobs would, in almost all cases, also be useful to poll workers.
Utilize observers to evaluate voting system reliability
Election observation is not a road to riches. People do it because they want to be helpful, and a good observer can be an excellent resource because s/he will often have had prior experience administering elections in another jurisdiction. It makes sense to take advantage of that.
Observers are trained to report on an array of voting machine issues, and their observations can be a source of immediate assistance. Here are some examples:
- problems that arise during voting machine setup at polling stations;
- ensuring that electronic memories do not contain any votes at the start of voting;
- whether voting machines are set-up to protect the secrecy of the vote;
- whether voters appear to understand how devices function;
- whether disabled and elderly voters can vote without assistance (and receive untainted assistance when they need it);
- using alternate languages in the voting process without difficulty;
- proper sealing of elements of a machine that are supposed to be sealed;
- whether polling officials or technicians understand the process, and improperly operate or handle voting machines on election day;
- procedure(s) for processing voters if voting systems become unavailable for more than a few minutes;
- adherence to closing procedures and timely, proper transmission of results to higher levels of election administration;
- immediate audits of the results conducted at a polling station; and
- what happens to recorded votes after election day.
As the use of new voting technologies expands, election observers are increasingly interested in how systems cope with malfunctions or voter errors, conducting pre-election tests, analyzing voter education material, and interviewing relevant stakeholders about their involvement in these activities. To ensure that observers can observe without interfering, it is important that election officials work closely with them. Making voting systems as transparent and accessible as possible, providing guidance/education in new developments, and listening to useful feedback will improve the election process – which is, after all, what it’s all about.
David Levine is an Election Management Consultant who has administered county, state, federal and private sector elections; developed election policy for non-profit organizations; and monitored elections in other countries. His expertise includes voter registration, election administration, poll worker training, outreach, research design and evaluation, voting system standards, logic and accuracy testing, post-election audits, voting accessibility, evaluating proposals and voting technology.
- Next >>