Copyright © 2008 jsd
We propose a rating service to evaluate the reliability of procedures used for elections. At present, many jurisdictions use procedures that are in dire need of improvement.
Voters need to take the lead by demanding that their votes be counted properly. However, security demands attention to detail, and individual voters do not have the expertise or the opportunity to attend to all these details. Therefore voters need a ratings service so they can know what needs to be done.
Responsibility for elections is rather spread out:
In particular, voter responsibility does not begin or end on election day. Voters must make sure that proper procedures are in place, and make sure proper procedures are carried out; otherwise there is much too much risk that the votes cast will not be counted properly.
Security requires attention to detail. Individual voters can look after some of the details but not others. It is impractical to expect individual voters to be up-to-speed on what kind of tamper-resistant ink should be used to mark ballots, or what kind of tamper-resistant software should be used for tallying the ballots.
Situations of this sort have arisen many times before, and we know how to deal with them.
By way of analogy: Consumers want to know whether they will be electrocuted by the appliances they buy, but most consumers do not have the technical expertise to judge electrical safety. Therefore they rely on ratings from Underwriters Laboratory and similar organizations.
There are a host of ratings agencies:
We need a comparable service to inspect, evaluate, and certify election processes. See section 3 for the next level of detail.
It is important to distinguish between ultimate responsibility and delegated responsibility. The citizens have ultimate responsibility for election integrity ... but they may reasonably delegate the technical details to a trusted rating service.
The primary activity of the rating service is to inspect, evaluate, and certify election processes.1 (When we refer to election processes, that is not limited to what happens on election day, because many of the most grievous abuses happen before or after election day.)
The ratings service must remain scrupulously impartial and nonpartisan.
The ratings would apply on a jurisdiction-by-jurisdiction basis, for each jurisdiction nationwide.
Voters cannot engage in comparison shopping for election services they way they can for appliances. Therefore one of the primary “deliverables” of the election rating service will be to publish ratings along with specific recommendations for improvement. These will be publicized well in advance of election day, so that voters and other stakeholders will have time to take remedial action.
The service will also publish proactive informational and educational material, including
For example, don’t have everybody in the office login to the tally machine using the “Admin” userid, since this undercuts any notion of responsibility and accountability. Similarly, don’t install high-tech security on four of the five doors to the tally room, while leaving the fifth door unsecured. (These are not made-up examples.) See reference 2.
Copyright © 2008 jsd