S.B. 1051: Too much, too little, too risky

Last week the Government Administration and Elections Committee passed a modified version of S.B. 1051, hailed by the Secretary of the State and ROVAC (Registrars Of Voters Association of Connecticut) as a ‘bipartisan’ compromise.

Yet, all the compromising seems to be the agreement of election officials on a bill that would make registrars jobs easier while adding largely undefined and unchecked powers for the current and future Secretaries of the State.

Two members of the Republican minority voted against the bill primarily because it would give the Secretary sole authority decide to temporarily remove registrars from office for any complaint filed by the Secretary or failing to maintain certification. We agree it goes too far in that provision.  It should and does provide a more objective means for permanently removing registrars.  We fail to see where a provision for the Secretary to temporarily remove registrars would have solved the recent problems noticed in the heat of election days. If that were the only weak and risky provision we might be able to live with the bill and some of its helpful provisions.

We are all in favor of effective training, certification, and fair procedures for removing registrars from office.  The bill has what we suspect will turn out to be relatively weak certification requirements and an alternate procedure for removal by charges from the state’s attorney and any superior court judge.  Even that seems to be a bit weak, requiring only a single judge to rule on removing an elected official from office. Consider:

  • Their is an ‘advisory’ committee to create certification.  In the existing law, never implemented, the committee was not advisory.  Now the current or future Secretary of the State approves the certification program.
  • The committee consists of six members, five appointed by the Secretary.
  • Decertifying a registrar does take concurrence of a majority of the committee.
  • Strengthening the existing law, sitting registrars must be certified within two years of taking office, except perhaps untended,  the law requires registrars who are appointed to fill the remainder of two-year terms to complete certification by the end of the term.
  • We can hope that the actual certification, examination, and continuing education result in relevant, meaningful requirements.

We support professionalization.  Certification in election matters is only part of that.  Additional skills, education, and experience also play a part. We are skeptical that without increased compensation that many highly skilled, organized, and experienced individuals will be attracted to the jobs in small towns.  We wonder how much certification would have prevented the problems seen in recent years in Hartford, West Hartford, and Bridgeport. We support professionalization through regionialization.  That might be the result of another bill passed by the committee, S.B. 1083.

There are other risky, insufficiently defined provisions in the bill associated with closing of the polls and reporting results:

  • One requires quicker reporting of partial results “Once completed, the vote totals produced by the tabulator shall be prepared for transmission to the Secretary of the State”.
  • This is ambiguous.  Yet, according to the Secretary’s testimony on the bill, it seems that the intention is to transmit the results from optical scanners to the central GEMs system for automated calculation of results.
  • To connect our optical scanners to the GEMS requires reversing longstanding security policy implemented by the Bysiewicz administration to keep the scanners sealed from communication that risks infecting the scanners with fraudulent code.
  • We add that the GEMS system is no gem.  It figured prominently in the reporting errors discovered in the Humboldt Project.  We also recall Bev Harris demonstrating to Howard Dean how easy it would be for him to change election results on the GEMS, undetected.
  • Maybe it will turn out OK.  Once again, we are left to hope that in the end, this Secretary and all future secretaries work to maintain security of the scanners, memory cards, and their programming.

Further, the bill gives officials 48 hours after the election to report the rest of the results: hand counted ballots, write-in ballots, and for checkers to sign the pollbooks.

  • We are all for giving officials time to get thing right. Especially in situations like Bridgeport in 2010 where there are huge numbers of unexpected ballots to count by hand.  We wish the media could hold off the pressure for “results, any results”.
  • Yet, these changes seem to lack any security and transparency requirements.  If counting is stopped to continue later, we need convincing, sufficient, enforceable, and enforced security for ballots and checkin lists. We need formal requirements for notification of the public of when counting will resume.
  • When it comes to checkin lists, we see no point in not having checkers total and sign the lists at the polling place on election night — except it they are using electronic pollbooks and the lists are not printed until later by someone else — then we see nt good reason to have them sign printed paper lists that they have not created, from a system they do not understand, and have not held in custody.  Perhaps they or the polling place moderator should have a form to record the number of voters the machine reports as having checked in – signed and submitted on election night.

Finally, we come to electronic auditing.  The bill has this provision near the end:

Notwithstanding   any provision of title 9 of the general statutes, the Secretary of the State, in consultation  and  coordination  with  The  University  of  Connecticut, may  authorize  the  use  of  electronic  equipment  for  the  purpose  of conducting any audit required pursuant to section-320f of the general statutes,  as  amended  by  this  act,  for  any  primary  or  general  election held on or after January 1, 2016, provided (1) the Secretary of the State prescribes  specifications  for  (A)  the  testing,  set-up  and  operation  of such equipment,  and  (B)  the  training  of election officials  in  the  use  of such equipment; and (2) the Secretary of the State and The University of  Connecticut  agree  that  such  equipment  is  sufficient  in  quantity  to accommodate  the  total  number  of audits  to  be conducted.  Nothing  in this  section  shall  preclude  any  candidate  or  elector  from seeking additional remedies pursuant to chapter 149 of the general statutes as a result of any information revealed by such process.

As readers of CTVotersCount know, we have long been supporters of machine assisted auditing.  We are here left to hope that the Secretary and UConn do the right thing i.e. support a method of auditing that is transparent and meets the requirements of evidence based elections, such that the public can verify the results of the audit without depending on officials.  How is that possible? It has been outlined by three leading experts in the field of election auditing and prototyped in CA and CO.

In fact, we provided a bill which included a provision for safe machine assisted auditing this year, S.B. 1041. Even though that bill received wide support and no opposition in testimony, it did not move forward.

If S.B. 1051 moves forward in its current form we are left to hope that the Secretary and UConn will use its provisions to provide safe verifiable auditing.  Yet, left with the concern that they might not, and that some future Secretary and some future UConn scientist or UConn leader collude to disregard science to provide some all but useless, untrustworty version of electronic “black-box” auditing.

Facebooktwittergoogle_plusredditpinterestlinkedintumblrmailFacebooktwittergoogle_plusredditpinterestlinkedintumblrmail

Leave a Reply

You must be logged in to post a comment.