Skip to content
Joseph R. Kiniry edited this page Mar 6, 2013 · 1 revision

= An Executive Summary of the State-of-Affairs in Computer-Based Voting in Denmark = == edited by Joseph Kiniry

What follows is an executive update on electronic elections in Denmark.

  1. Computers have been used in elections in Denmark since the mid-1980s as they are used to create and manage the voter list, send out printed voter cards, create the physical voter lists for election day, print ballots, and compute the final outcome of the election.

  2. Computing the final election result, i.e., determining exactly which persons sit in Parliament, is a non-trivial algorithm. From the mid-1980s until 2011 this computation was performed by a program written by an employee of the Ministry responsible for elections. No one we have spoken with outside of that programmer has any knowledge of how the program works, has never seen the code, etc. Rumor has it that it is tens of thousands of lines of APL. We have been trying, without success thus far, to obtain a copy of that program and all artifacts related to its development. (as of late 2012)

  3. In 2011, when said programmer retired, the Ministry contracted Denmark Statistics, now a branch of said Ministry, to write a new program to compute the final outcome of the election. This program is also not public and has not seen outside review. We are told that it is tens of thousands of lines of PL/SQL. We about about to enter into an agreement with the Denmark Statistics to obtain a snapshot of that program and all related development artifacts so that it can see outside review. (week 51 2012) We have recommended that the system be Open Sourced and put under public oversight and control, if it continues to be used.

  4. Municipalities have been pushing to introduce computers in elections, via kiosk-based evoting in a controlled environment, for around two decades in Denmark. Their reasoning follows that which we have heard elsewhere: it will reduce the costs of elections, it will raise participation, it is the natural, "modern" thing to do, it will permit the disabled to vote independently, it will permit faster and more accurate counts. We believe that the long-term goal of the Municipalities is to push for internet-based elections.

  5. Until 2011 the Ministry and Head of Elections have always said "no" to these informal and formal requests. Previous governments have created Parliamentary working groups, various reports, etc. that all questioned the claims of the Municipalities and argued that Denmark's democracy works very well and why fix what isn't broken?

  6. In 2010 the Danish Technology Board issued a report on evoting, including recommendations for trials in early voting via "voting busses". See http://eptanetwork.org/projects.php?pid=650 and Danish Technology Board E-elections Report for an English translation of key findings in the report.

  7. In 2011 a new government came into power, around a dozen Municipalities made a new formal request to conduct trials in electronic voting, and the new Minister responsible for elections said that the Head of Elections should look into it.

  8. In the 2011 elections a new Digital Voter List (DVL) system was used to replace analog pollbooks in dozens of polling places. The DVL system uses is proprietary, closed source, build and managed by the local IT consultancy monopoly KMD, and the support contract between KMD and the Municipalities coupled payment with the number of voters processed by the system, thereby enormously increasing the cost of elections in said Municipalities (rumor has it, by up to 400%).

  9. The Ministry met in a technical dialog with just over half a dozen evoting systems vendors to get a feeling for the state of the technology, the claims of the vendors wrt functional and non-functional properties of their systems, the estimated cost of elections, etc.

  10. Based upon that technical dialog, the Ministry issued a report to the Municipalities summarizing their (in our view, non-conservative) cost estimates for the trial and for evoting in the long run. They then asked the Municipalities which ones still wanted to move forward with trials, as the cost of trials would be born by said Municipalities. A handful of Municipalities stated that they still wanted to move forward.

  11. Based upon weekly consultation with us and consultation with others, especially via a public workshop on evoting, the Ministry proposed a change in electoral law that would permit binding trials in 2013 and 2014 in said Municipalities. More information on the workshop is available via E-elections in Denmark workshop.

  12. Two weeks ago the proposed change in electoral law was published and a couple of dozen organizations were formally asked to give feedback. We were one such organization.

  13. (2012) The proposed change got a lot of coverage in the technical press in Denmark, as has evoting in general over the past couple of years. The opinion of the technical citizens that engage in such forums has been fairly negative; i.e., they do not want evoting in Denmark because of all of the problems they have seen elsewhere. For one of the most recent articles on such, see http://www.version2.dk/artikel/massiv-kritik-af-lov-om-e-valg-det-bliver-dyrere-og-mindre-sikkert-49461. Google Translate will give you the gist of the tone and content. Comments from politicians seem to indicate that they are effectively already ignoring the input from the experts, are found in http://www.version2.dk/artikel/socialdemokraterne-staar-fast-e-valg-er-bedre-og-sikrere-end-papirvalg-49475.

  14. (weeks 49 and 50, 2012) We have written a strong response to the Ministry. We have also seen the responses of several other parties, including the top political scientists and election/democracy experts in Denmark. All responses we have seen thus far, outside of that written by the main IT trade organization, have been negative, arguing that Denmark should not go down the path of electronic elections. We are translating our response into English now for the world and will post it on our website. It is my understanding that all feedback the government receives becomes public soon.

  15. (2011-2012) In parallel to the above, DemTech has implemented replacement Open Source systems for Digital Voter List management and voter registration systems as well as the final tally system. All of our replacement systems are designed and developed using safety-critical systems standards and tools, thereby aiming at an equivalent to EAL level 7 safety and correctness certification and ISO 27000-series information security certification. We now intend to do the same with a very novel, Open Source kiosk-based election system in early 2013, if the government moves forward with this exercise.

  16. (week 51, 2012) The Ministry now has to synthesize all responses to the proposed change in law and summarize such to the members of Parliament prior to Parliament. The Ministry will also revise the proposed changes in law, perhaps make public their decision-making wrt the input they have received, and send the new revised law text to the Parliament. Alternatively, based upon the input, the Ministry can decide to stop the whole process and kill the proposed change in law. We'll see what happens.

Clone this wiki locally