[Date Prev][Date Next] [Chronological] [Thread] [Top]

GEMS 1.18.6 Cross Endorsement - Push

I asked for cross endorsement feedback Aug 21, but of course the only feedback was the standard panic RCR a month later.
Working with LHS, we have implemented AVOS cross endorsement in GEMS 1.18.6.  You need a new 194us.abo or 194usma.abo that ships with 1.18.6.  Through some fancy ABasic footwork no AVOS firmware revision is necessary. 
LHS requested that we implement a push variation similar to that requested by New York.  In Connecticut they push to the lower party.  In New York they appear to push to the highest party.  I still don't have good definitions for "low" and "high" for either of these, but the gist is this:  If a voter votes for the same cross endorsed candidate more than once (ie for more than party), than the vote is still cast, but the vote goes to the lowest/highest ranked party only.
For now at least, we have proposed an alternate system.  When a candidate is cross endorsed, a non-partisan entry is included along with the endorsements.  This is done automatically in the race editor.  So, you can have a single endorsed candidate, or have two-or-more cross endorsements for a candidate plus a non-partisan endorsement of the candidate.  That is, one or three+ entries, never two.  The non-partisan entry does not show up on the ballot, but an extra counter for this candidate is created and it shows up on reports.  When voter votes for the same cross endorsed candidate more than once, the vote is assigned to this "unknown" party.  On the AV tape, you will see:
Brian Courtney  DEM   25
Brian Courtney  LIB     30
Brian Courtney  UNK   10
Brian Courtney  TOT    65
Jurisdictions that want the push statistic can just add the UNK value to whatever party they like.  My sincere hope is that this will meet everyone's statutory requirements.  This is, after all, a strict superset of push.  If we were instead to add the votes to the push party implicitly and not store the UNK statistic, you could never tell whether the vote was assigned explicitly by the voter, or implicitly by the push rule.  This way you can get the push stat, plus know how many of those votes were implicit.
This argument appears to have worked with LHS in this election at least, but I would like to see the statutes for each jurisdiction that does cross endorsement before the next panic RCR.  My gut feeling is that there are no steadfast statutes on this, and that the push rules are a holdover from the way that lever machines work, but I need confirmation of this theory from those responsible.