This pilot is supported by the ‚ÄčMinistry of the Interior and Economic Affairs, and in collaboration with ‚ÄčKL.

How can we determine whether vote-counting systems, manual or technology based, have performed well in specific elections? A risk-limiting post-election audit can ensure that elections are accurate. A risk-limiting audit confirms statistically that the correct electoral outcome is correct. "Electoral outcome" means, for instance, the number of seats each party gets, the candidates who receive seats, or whether a referendum won or lost. Risk-limiting audits involve manually inspecting randomly selected ballots. The audit examines more ballots until the inspection has given strong evidence that the outcome is correct, or--if that does not happen--until there has been a full manual count to check the outcome. Risk-limiting audits are designed to minimize the number of ballots that must be inspected when the outcome is correct, while ensuring a large, pre-specified chance of correcting the outcome if the outcome is wrong. For example, if the original outcome is incorrect, a risk-limiting audit with a risk limit of 10% has a 90% chance of correcting the outcome. It will never alter outcomes that are already correct.

Denmark was trying Risk-Limiting post-election audits for the first time in May 2014, to check its portion of the European Parliamentary Election. We inspected circa 150 ballots from the nation as a whole. Each municipality's share of the work is roughly in proportion to the number of ballots cast in that municipality.

In the "fine count," municipalities sort, count, and bundle the voted ballots. Each bundle is supposed to contain ballots with the same votes, for instance, all ballots in a bundle should show a vote for the same candidate. The election can be audited efficiently if each municipality provides the auditors a list of all its bundles of ballots, together with the number of ballots in each bundle and the name of the candidate, party, or position that the ballots in that bundle are supposed to show. This pilot was supported by the Ministry of the Interior and Economic Affairs, and in collaboration with Local Government Denmark (in Danish: KL).