Last February, during its sectional elections, Ecuador took an important step towards the modernization of its elections as it carried out three electronic voting pilots in selected provinces. Simultaneously, in all other provinces, the electoral commission used technology to process the votes, from counting to results publishing.
The Union of South American Nations (Unasur) recently released its final report on these elections, praising the results of the binding e-voting test, as well as the organization, training, and logistics of the process. However, it also expressed disapproval of the automated result processing system used in manual precincts.
Unasur’s assessment is based on the fact that, while three provinces used different electoral automation models successfully, the Spanish company Scytl failed to perform its duty in the rest of the nation: digitizing poll counts for their later consolidation and publishing.
According to the report, “the system was ill equipped to process the amount of information generated simultaneously, which made its timely processing difficult and made errors evident.” Besides, it was clear that this was not the first time that the system performed poorly. “The repeated lack of adjustments indicates that the company has not contemplated all the requirements for its use in elections. The provider behaved negligently by insisting on using procedures and mechanisms that are inadequate to this end.” This situation hampered the promise of presenting nationwide results in 78 hours, as their delivery actually took more than a month.
In fact, Unasur insisted in its report that there were various reasons for the failure of the Intermediary Scrutiny Boards (JIEs for its initials in Spanish), “many of which stemmed from failure to comply with the recommendations made by Unasur’s mission in 2013, as well as those from this year’s mission. But the main source of trouble was the system freezing while processing information coming from the JIEs in the BONITA BMP software, which manages the system’s workflow and was provided by the company offering the electoral software suite.”
Moreover, the report highlighted “errors in processing data from the Guayas province’s JIE. The problem that arose in Guayas was made worse by the fact that poll count minutes were printed on both sides and data was uploaded from only one side of the sheets. This generated widespread errors in result aggregation. This error was fixed, but it caused significant delays in the result release process.”
In the other hand, Unasur acknowledged the optimum performance of the three types of electronic voting tested in Ecuador. It specially highlighted the technologies used in the two provinces where the results of the pilots were binding: the one deployed in Santo Domingo de los Tsáchilas, provided by Smartmatic, and the one used in Azuay, from Magic Software Argentina (MSA). The institution expressed its satisfaction over the performance, vote secrecy preservation, contingency resolution, and general fulfillment of the goal of automating voting and safeguarding people’s intent.
After a complex but comprehensive experience showing the benefits that a correctly applied technology offers, Ecuador now has the possibility to transition safely and successfully to vote automation. Unasur recommended carrying out “conceptual re-engineering to redefine all the processes inherent (…) to recount and publish of results,” aspects where the system failed due to the provider’s inability to fulfill requirements.