Electronic Voting, Spring 2020Apr 24 notes and discussion
Part of
the CS:4980:0004 Electronic Voting Notes
|
Early responses to the pandemic have varied widely, from Wisconsin's "we won't deviate from normal rules at all," to Iowa's June primary, where the secretary of state has decided that every Iowa resident should receive an absentee ballot request form by mail.
Planning for the fall is complicated by debates over providing any kind of bailout for the US Postal Service. There appears to be a faction that is willing to allow the Postal Service to go bankrupt.
The authors, a political scientist and a lawyer, have long involvement in questions of election administration. Their suggestion for postal voting is, of course, complicated by recent attacks on the Postal Service. Another issue they don't address is timing. If states that currently rely largely on touch-screen machinery (either DRE voting or ballot-marking devices) switch en-masse to postal voting, the demand for ballot stock and ballot printing will soar. If this decision is made before June, the paper industry and ballot printers may be able to ramp up capaicity and meet the demand. If, on the other hand, the decision is delayed into August, it may be impossible to meet even under normal circumstances. There is a natural tendancy of decision makers to drag their feet, hoping that the problem will go away, and under current circumstances, this response threatens to eliminate viable alternatives.
Groups from across the spectrum have been pushing to use the COVID-19 pandemic as a lever to push for their agenda. In the election domain, there are groups who want a national switch to postal voting, or national adoption of Internet voting, or the complete privatization of postal service. The authors argue strongly to resist all of these, that we should not make long-term decisions in times of crisis, but rather, deal solely with the crisis and await calmer times to think about long-term policy.
When I asked an election official in Miami, back in 2004, "if you were a crook, how would you 'fix' elections in Miami." His eventual response, after objecting to the question, was "I'd go after the absentee ballots, because they are handled by more people than ballots cast in any other way."
The authors of this commentary discuss the downside of an emergency move to postal voting, noting that all of the considerations suggested by the Miami official I questioned invite legal challenges and open a figurative can of worms.
For example, many states have contracted with private vendors to do automatic signature checking on absentee ballot envelopes. These companies use algorithms about wich essentially nothing is known to compare signatures on ballot envelopes with the signature of record (frequently, this is the signature on the drivers license). Such basic information as the false-positive and false-negative rate for their algorithms are not public information! (A false positive is a declaration by the algorithm that two signatures match when they were signed by different people, a false negative is a declaration that two signatures by the same person do not match.)
Many state laws require that ballots be postmarked by election day eve, but the post office has greatly reduced the use of postmarks over the past few decades. Today, to get a postmark, you have to ask for it at the office as you hand your envelope to a clerk. All other mail goes through a sorting machine that just splashes an ink-jet printed bar code across the bottom edge of the envelope. In Iowa, there was a legal fight over ballots received with these bar codes; the court declared that, even though the post office declared that the bar codes on the envelopes in question were proof that the envelopes were in the postal system by the deadline, they were not acceptable proof under Iowa's law -- a law that allowed "intelligent bar codes" on envelopes, whatever that is.
Various non-governmental organizations have entered the fray with their own recommendations. Verified Voting traces its origins to a rebellion against the adoption of DRE voting machines by Santa Clara County, the home of Stanford University. David Dill, a CS professor there, launched a peresonal e-mail list to coordinate the opposition, and Verified Voting grew from that effort.
Note: I am on the Verified Voting Board of Advisors.
The AAAS is another nongovernmental organization, the American Association for the Advancement of Science. In this case, they coordinated the drafting of this letter and collected endorsements from numerous other organizations, including the ACM and CRA (the former is an association of computer professionals, the latter an association of computer science departments and research organizations).
This letter is narrower than the Verified Voting letter, it focuses on what not to do instead of making positive recommendations about changes that can be made.
Note: I am a co-signer of this letter, as well as being a member of at least 3 of the organizations that endorsed it (AAAS, ACM-USTPC and Verified Voting).
Added Apr 28, hot off the press: This is crafted by a broad group including lawyers, computer scientists and political scientists.