A piece by Sewell Chan in the New York Times notes that Dutch authorities have pulled the plug on computerized ballots and ballot counting for their next national election.
The move was prompted by concerns over the integrity of the election in the face of hacking concerns. In particular, allegations of tampering in the recent US election have caused the Dutch government to re-examine their setup, which was found wanting:
In a report on Monday, the broadcaster RTL concluded that the Dutch election would be “easy to hack,” citing interviews with experts and an in-depth investigation of the vote-tallying software the nation has used since 2009.
So, on March 15, Dutch voters will fill out paper ballots using pens, which will then be counted by hand.
This situation illustrates a progress problem with an interesting twist. Normally, a progress problem involves a dilemma over adoption of a new technology where there is significant uncertainly over how different social constituencies will be affected.
In rough terms, there are two ways to handle such a problem: Precautionary or permissive. Normally, the permissive strategy means adopting a new technology until such time as it is deemed too harmful. The precautionary strategy means holding off on adoption until it is deemed sufficiently safe.
In view of concerns over hacking, though, the permissive strategy would seem to be keeping electronic voting and tallying, whereas the precautionary strategy would be to revert to the pen-and-paper, hand-counting method.
So, the difference between strategies lies not so much in the novelty of a technology but where the main concerns over risk and safety lie.
PS. The recent House of Commons Committee on electoral reform in Canada recommended against proceeding with online voting here out of concern for the integrity of the ballot but made no mention of vote tallying.